Free ebooks ==> www.Ebook777.

com

www.Ebook777.com
Free ebooks ==> www.Ebook777.com

GENOMIC MEDICINE

www.Ebook777.com
OXFORD MONOGRAPHS ON MEDICAL GENETICS
general editors:
Judith G. Hall
Peter S. Harper
Louanne Hudgkins
Evan Eichler
Charles J. Epstein (deceased 2011)
Arno G. Motulsky (resigned 2011)

1. R. B. McConnell: The Genetics of Gastrointestinal Disorders

2. A. C. Kopéc: The Distribution of the Blood Groups in the United Kingdom
3. E. Slater and V. A. Cowie: The Genetics of Mental Disorders
4. C. O. Carter and T. J. Fairbank: The Genetics of Locomotor Disorders
5. A. E. Mourant, A. C. Kopéc, and K. Domaniewska-Sobezak: The Distribution of the Human Blood Groups and Other Polymorphisms
6. A. E. Mourant, A. C. Kopéc, and K. Domaniewska-Sobezak: Blood Groups and Diseases
7. A. G. Steinbert and C. E. Cook: The Distribution of the Human Immunoglobulin Allotypes
8. D. Tills, A. C. Kopéc, and R. E. Tills: The Distribution of the Human Blood Groups and Other Polymorphisms: Supplement I
10. D. Z. Loesch: Quantitative Dermatoglyphics: Classification, Genetics, and Pathology
11. D. J. Bond and A. C. Chandley: Aneuploidy
12. P. F. Benson and A. H. Fensom: Genetic Biochemical Disorders
13. G. R. Sutherland and F. Hecht: Fragile Sites on Human Chromosomes
14. M. d’A Crawfurd: The Genetics of Renal Tract Disorders
16. C. R. Scriver and B. Child: Garrod’s Inborn Factors in Disease
18. M. Baraitser: The Genetics of Neurological Disorders
19. R. J. Gorlin, M. M. Cohen, Jr., and L. S. Levin: Syndromes of the Head and Neck, Third Edition
21. D. Warburton, J. Byrne, and N. Canki: Chromosome Anomalies and Prenatal Development: An Atlas
22. J. J. Nora, K. Berg, and A. H. Nora: Cardiovascular Disease: Genetics, Epidemiology, and Prevention
24. A. E. H. Emery: Duchenne Muscular Dystrophy, Second Edition
25. E. G. D. Tuddenham and D. N. Cooper: The Molecular Genetics of Haemostasis and Its Inherited Disorders
26. A. Boué: Foetal Medicine
27. R. E. Stevenson, J. G. Hall, and R. M. Goodman: Human Malformations
28. R. J. Gorlin, H. V. Toriello, and M. M. Cohen, Jr.: Hereditary Hearing Loss and Its Syndromes
29. R. J. M. Gardner and G. R. Sutherland: Chromosome Abnormalities and Genetic Counseling, Second Edition
30. A. S. Teebi and T. I. Farag: Genetic Disorders Among Arab Populations
31. M. M. Cohen, Jr.: The Child with Multiple Birth Defects
32. W. W. Weber: Pharmacogenetics
33. V. P. Sybert: Genetic Skin Disorders
34. M. Baraitser: Genetics of Neurological Disorders, Third Edition
35. H. Ostrer: Non-Mendelian Genetics in Humans
36. E. Traboulsi: Genetic Factors in Human Disease
37. G. L. Semenza: Transcription Factors and Human Disease
38. L. Pinsky, R. P. Erickson, and R. N. Schimke: Genetic Disorders of Human Sexual Development
39. R. E. Stevenson, C. E. Schwartz, and R. J. Schroer: X-Linked Mental Retardation
40. M. J. Khoury, W. Burke, and E. J. Thomson: Genetics and Public Health in the 21st Century
41. J. Weil: Psychosocial Genetic Counseling
42. R. J. Gorlin, M. M. Cohen, Jr., and R. C. M. Hennekam: Syndromes of the Head and Neck, Fourth Edition
43. M. M. Cohen, Jr., G. Neri, and R. Weksberg: Overgrowth Syndromes
44. R. A. King, J. I. Rotter, and A. G. Motulsky: The Genetic Basis of Common Diseases, Second Edition
45. G. P. Bates, P. S. Harper, and L. Jones: Huntington’s Disease, THird Edition
46. R. J. M. Gardner and G. R. Sutherland: Chromosome Abnormalities and Genetic Counseling, THird Edition
47. I. J. Holt: Genetics of Mitochondrial Disease
48. F. Flinter, E. Maher, and A. Saggar-Malik: The Genetics of Renal Disease
49. C. J. Epstein, R. P. Erickson, and A. Wynshaw-Boris: Inborn Errors of Development: THe Molecular Basis of Clinical Disorders of Morphogenesis
50. H. V. Toriello, W. Reardon, and R. J. Gorlin: Hereditary Hearing Loss and Its Syndromes, Second Edition
51. P. S. Harper: Landmarks in Medical Genetics
52. R. E. Stevenson and J. G. Hall: Human Malformations and Related Anomalies, Second Edition
53. D. Kumar and S. D. Weatherall: Genomics and Clinical Medicine
54. C. J. Epstein, R. P. Erickson, and A. Wynshaw-Boris: Inborn Errors of Development: THe Molecular Basis of Clinical Disorders of Morphogenesis, Second Edition
55. W. Weber: Pharmacogenetics, Second Edition
56. P. L. Beales, I. S. Farooqi, and S. O’Rahilly: The Genetics of Obesity Syndromes
57. P. S. Harper: A Short History of Medical Genetics
58. R. C. M. Hennekam, I. D. Krantz, and J. E. Allanson: Gorlin’s Syndromes of the Head and Neck, Fifth Edition
59. D. Kumar and P. Elliot: Principles and Practices of Cardiovascular Genetics
60. V. P. Sybert: Genetic Skin Disorders, Second Edition
61. R. J. M. Gardner, G. R. Sutherland, and L. C. Shaffer: Chromosome Abnormalities and Genetic Counseling, Fourth Edition
62. D. Kumar: Genomics and Health in the Developing World
63. P. S. Harper: A Short History of Medical Genetics, Second Edition (online)
64. G. Bates, S. Tabrizi, and L. Jones: Huntington’s Disease, Fourth Edition
65. D. Kumar and C. Eng: Genomic Medicine: Principles and Practice, Second Edition
 GENOMIC MEDICINE:
PRINCIPLES AND PRACTICE

SECOND EDITION

EDITED BY EDITED BY

Dhavendra Kumar MD Charis Eng MD PhD

FRCP FRCPCH FACMG FACP
INSTITUTE OF CANCER & GENETICS SONDR A J. & STEPHEN R . HARDIS ENDOWED
CARDIFF UNIVERSITY SCHOOL OF MEDICINE, CHAIR OF CANCER GENOMIC MEDICINE
A L L WA L E S M E D I C A L G E N E T I C S S E R V I C E DIRECTOR- GENOMIC MEDICINE INSTITUTE &
U N I V E R S I T Y H O S P I TA L O F WA L E S , C A R D I F F CENTER FOR PERSONALIZED GENETIC
& H E A LT H C A R E C L E V E L A N D C L I N I C ,
G E N O M I C P O L I C Y UN IT, AND PROFESSOR AND VICE CHAIR
T H E FA C U LT Y O F L I F E S C I E N C E S & E D U C AT I O N DE PA RTM E N T OF G E N ET I C S & G E N OM E
T H E U N I V E R S I T Y O F S O U T H WA L E S , SCIENCES CASE WESTERN RESERVE
PONTYPRIDD, UK UNIVERSITY SCHOOL OF MEDICINE
CLEVELAND, OHIO, OH

1
 Free ebooks ==> www.Ebook777.com

1
Oxford University Press is a department of the University of
Oxford. It furthers the University’s objective of excellence in research,
scholarship, and education by publishing worldwide.

Oxford New York
Auckland  Cape Town  Dar es Salaam  Hong Kong  Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto

With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam

Oxford is a registered trademark of Oxford University Press

in the UK and certain other countries.

Published in the United States of America by

Oxford University Press
198 Madison Avenue, New York, NY 10016

© Oxford University Press 2015

All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system, or transmitted, in any form or by any means, without the prior
permission in writing of Oxford University Press, or as expressly permitted by law,
by license, or under terms agreed with the appropriate reproduction rights organization.
Inquiries concerning reproduction outside the scope of the above should be sent to the
Rights Department, Oxford University Press, at the address above.

You must not circulate this work in any other form

and you must impose this same condition on any acquirer.

Library of Congress Cataloging-in-Publication Data

Genomics and clinical medicine.
Genomic medicine : principles and practice / edited by Dhavendra Kumar, Charis Eng. — Second edition.
  p. ; cm.
Preceded by Genomics and clinical medicine / edited by Dhavendra Kumar. [First edition]. 2008.
Includes bibliographical references and index.
ISBN 978–0–19–989602–8 (alk. paper)
I.  Kumar, Dhavendra, editor of compilation.  II.  Eng, Charis, 1962– editor of compilation.  III.  Title.
[DNLM: 1.  Genomics.  2.  Genome—physiology.  3.  Pharmacogenetics. QU 58.5]
RB155
616′.042—dc23
2014001647

This material is not intended to be, and should not be considered, a substitute for medical or other
professional advice. Treatment for the conditions described in this material is highly dependent on the
individual circumstances. And, while this material is designed to offer accurate information with respect
to the subject matter covered and to be current as of the time it was written, research and knowledge about
medical and health issues is constantly evolving and dose schedules for medications are being revised
continually, with new side effects recognized and accounted for regularly. Readers must therefore always
check the product information and clinical procedures with the most up-to-date published product
information and data sheets provided by the manufacturers and the most recent codes of conduct and
safety regulation. The publisher and the authors make no representations or warranties to readers, express
or implied, as to the accuracy or completeness of this material. Without limiting the foregoing, the
publisher and the authors make no representations or warranties as to the accuracy or efficacy of the drug
dosages mentioned in the material. The authors and the publisher do not accept, and expressly disclaim,
any responsibility for any liability, loss or risk that may be claimed or incurred as a consequence of the use
and/or application of any of the contents of this material.

1 3 5 7 9 8 6 4 2
Printed in the United States of America
on acid-free paper

www.Ebook777.com
 “To our Parents and Families,

To our Patients and their Families”

 FOR EWOR D
In the foreword of the previous first edition of this book
(see page ix), Dr. Francis Collins had welcomed the reader
to the genome era, and has eloquently introduced the
genome into the medical practice. It is my great pleasure
and honor to be able to comment on the significance of the
genome analyses in medicine.
It has become abundantly clear that there are two
main etiological components when considering health
and disease: the genomic variation, and the environmen-
tal insults. In other words, the majority of disorders result
from the interaction between individual genomic compo-
sition (and history of the subsequent somatic mutations),
and the environmental variables. Thus it is of paramount
importance to study the genomic variation of each indi-
vidual in order to begin to understand his or her disease,
improve the diagnostic possibilities, and introduce intel-
ligent treatments.
The progress in human genomic sciences continues
at a remarkable pace:  the HapMap Project and the 1000
Genomes Project have provided the opportunity to assess
the common and rare variations in genomes from differ-
ent geo-ethnic groups; the ENCODE Project has pro-
vided initial information on the functional elements of
the genome; the genome-wide association studies have
provided genomic signals for functional and diagnostic
studies related to almost all the complex human disorders
and traits with measurable heritability; the work in model
organisms has provided the basis for the functional analysis
of genomic variation by taking advantage of evolutionary
conservation; the discovery of the pathogenic mutations of
a large number of (near)-Mendelian disorders has extended
the important list of rare disorders; and the exploration
of de novo genomic variants points to genes involved in
complex phenotypes and underscores the importance of
the unexplored genetics of sporadic cases. All of the above
advances were possible thanks to the truly extraordinary
and unexpected progress in sequencing technologies, the
availability of bioinformatic tools, and the collaboration of
scientists worldwide.
vii
It is extraordinary that, within only a few years after
the completion of the sequencing of the first genomes, we
have the capacity to determine genomic variability rapidly
and within reasonable cost constraints; however, we are far
from understanding the functional significance of the vast
majority of variants, and we are also far from determining,
monitoring, and understanding the environmental impacts
of the same.
The physician who practices genetic medicine now
has an “organ” or field of expertise similar (or superior?)
to those of other disciplines:  the genome! As the cardi-
ologist is an expert in the cardiovascular system, the neu-
rologist in the nervous system, the endocrinologist in the
endocrine glands, and the ophthalmologist in the eye, the
geneticist is becoming an expert in the genome’s anatomy,
variability, pathogenicity, function, inheritance, history,
and evolution. The geneticist using all this information
and knowledge participates as a primary actor in the diag-
nosis and treatment of individuals, and becomes a notable
voice in the chorus for family and health planning, and a
whole litany of ethical, legal, social, financial, and educa-
tional issues.
Remarkably, our knowledge of the individual genomic
variations, and the access to information through the
Internet and the social media, have made the physician only
one aspect of the health management and have elevated the
patient/consumer to a position of partnership. Medicine is
becoming more participatory, and the physician less pater-
nalistic regarding diagnostic strategies and therapeutic
options. In addition, several other disciplines well versed in
computational biology, information technology, and statis-
tics participate in the diagnostic and therapeutic schemes,
and their expertise is indispensable in the gathering the vast
amounts of data and distilling the relevant information that
guides medical decisions.
Genome sequencing has already become an important
diagnostic tool. Yet, could the reading of each individual’s
genome solve the majority of the diagnostic problems?
Undoubtedly not, even if we achieve a full functional
understanding of each nucleotide. So-called personalized
medicine will need:
(a) the sequence of the genomes of key somatic cells
that have become malignant, or have developed a
recognizable pathological phenotype;
(b) the sequence of the genomes of billions of bacterial
and other microbes that we all host in our bodies
(microbiomes);
(c) the epigenetic modifications in different cell types;
(d) the repeated analysis of transcriptomes of various cell
types; and
(e) the repeated analysis of other “-omics” components.
The list is not exhaustive, and the exact battery of
genome-based tests will be determined on an individual basis.
viii  •   F o r ewo r d
The revised and rich contents of this remarkable book,
edited by our outstanding colleagues, Professors Dhavendra
Kumar and Charis Eng, and written by an equally outstand-
ing roster of authors, deal with all aspects of genomic medi-
cine. The book combines general principles and specific
aspects of clinical practice, all in the light of genome analysis
and the current laboratory methodologies. The genome era
of medicine is well into development, and, as Shakespeare
wrote in The Tempest: “What’s past is prologue.”
Stylianos E. Antonarakis
Professor and Chairman of Genetic Medicine,
University of Geneva
President of Human Genome Organization
Geneva, Switzerland
June 23, 2013
 Free ebooks ==> www.Ebook777.com
FOREWORD TO T HE FIR ST EDIT I ON
A scant twenty years have passed since the word “genom-
ics” was coined by Victor McKusick, Frank Ruddle, and
Tom Roderick to describe a new discipline. The suffix of
the word derives from the Greek ome meaning all, and aptly
conveyed an intention to transition the study of heredity
from a focus on single genes (genetics) to the more global
perspective of all of the hereditary material. A proliferation
of other “omics” disciplines has subsequently erupted—
including proteomics, metabolomics, transcriptomics, gly-
comics, microbiomics, and many more.
But genomics remains the foundation of the rest, reflect-
ing as it does a comprehensive analysis of the DNA instruc-
tion book. The success of the Human Genome Project has
now laid that instruction book wide open. As a result, the
life sciences have been catapulted forward, and biology has
now taken its rightful place alongside physics and chemistry
as a truly digital and quantitative science.
It is the application of genomics to medicine that car-
ries its greatest promise of benefit to humankind. Thus, the
publication of this first textbook of “Genomics and Clinical
Medicine” marks a milestone, a coming of age. Here in the
early years of the third millennium we can see the emerg-
ing outlines of a new synthesis of the noble tradition of the
healing arts with an increasingly precise way of understand-
ing the causes of disease, based on an understanding of the
human genome.
For some in the clinical medicine community, how-
ever, this textbook may come as a surprise. After all, there
are still many practicing physicians who would say they see
no evidence of genetics or genomics as part of their daily
medical practice. Surely, however, that reveals a problem
with the successful communication of rapid new develop-
ments in this field, not the facts of the matter. For in these
forty-two chapters, a vast array of genomic implications for
nearly every condition that affects humankind is laid out in
elegant and comprehensive fashion.
The pace of progress in genomics has been astound-
ing. Over just the last fifteen years, largely as a consequence
of the tools made available through the Human Genome
Project, genes have been identified for more than two
ix
www.Ebook777.com
thousand inherited conditions. With recent rapid advances
in the understanding of human genetic variation, the spe-
cific hereditary contributions to common diseases like dia-
betes, heart disease, cancer, and mental illness are emerging
at an unprecedented rate. The very real possibility of offer-
ing individuals who are currently healthy a personalized
prediction of future risks of illness is no longer a distant
dream. And given that many of the common disorders for
which predictions are becoming possible are associated
with proven means of reducing risk through diet, exercise,
lifestyle change, medical surveillance, or pharmacotherapy,
the real likelihood of widespread individualized programs
of preventive medicine grows by the day. Similarly, the abil-
ity to make predictions about the possibility of a beneficial
or undesirable response to drug therapy, the field of phar-
macogenomics, is advancing rapidly, and will soon require
health care providers to determine the genotype before
writing the prescription, at least for certain drugs. Many of
us predict that the complete genome sequence of an indi-
vidual will become part of that person’s medical record
within about ten years, at a cost of $1000 or less. And the
therapeutics that we use in the future will likely be heav-
ily dependent upon an understanding of the genomic basis
of illness, leading to interventions that are both more accu-
rately targeted to the underlying problem and less likely to
cause side effects.
All of these advances should be welcomed by anyone
interested in the alleviation of human suffering. Yet a num-
ber of major ethical, legal and social challenges lie along
the path if this vision is going to be realized. In the United
States, for example, we still lack effective federal legisla-
tion to prevent discriminatory uses of predictive genetic
information. Major challenges also lie ahead with regard
to ensuring equitable access to new genomic technologies,
especially as our medical care system seems to undervalue
opportunities for preventive medicine, focusing instead on
treating disease once it has already appeared. But perhaps
the greatest barrier, and the one which this book admirably
seeks to address, is an educational one. Most members of
the public are interested in genomics, but relatively unsure
of the details. Seeking advice, they generally turn to their
health care providers, but many of those professionals are
poorly prepared to become practitioners of this new art.
After all, most physicians have had little or no training in
genetics or genomics, and will be hard pressed to quickly
acquire the scientific principles, the medical knowledge, and
the psychosocial skills that will be necessary for the success-
ful introduction of genomic medicine. Busy practitioners
will desperately need an authoritative source of information
x  •   F o r ewo r d to t h e F i r s t E d i t i o n
that includes both principles and specific applications. The
introduction of this textbook, with its distinguished and
authoritative list of contributors, thus arrives in the nick of
time. Welcome to the genome era.
Francis S. Collins, M.D., Ph.D.
National Human Genome Research Institute
National Institutes of Health
Bethesda, MD, USA
 PREFACE TO T HE FIR ST EDIT I ON
Although the science of genetics is only 150  years old,
genetics as inheritance has been a concept discussed since
ancient times. The evolution and natural-selection theories
put forward by Charles Darwin had clear overtones that are
reflected in some of our present-day concepts of the genetic
basis of biological life. Gregor Mendel’s laws of inheritance
and successive discoveries in various aspects of genetics laid
the foundations of a number of disciplines covering differ-
ent areas within the science of genetics. Human genetics
was no exception. However, this was heavily shrouded by
the dark clouds of the so-called eugenics movement (ster-
ilization of the “unfit,” etc.) of the early twentieth century,
when history recorded one of the worst practical applica-
tions of modern science on fellow human beings under the
pretext of scientific research.
It has taken almost sixty years to arrive at our present state
in the science of genetics. The future now appears bright,
opening up many opportunities on the horizon. Clinical
genetics is now a recognized medical specialty among sev-
eral disciplines composing the current spectrum of modern
medicine. The basis of clinical genetics is grounded in the
sound knowledge and understanding of medical genetics
that emerged as a spinoff of “human genetics.”
Fifty years after the discovery of the double-helix struc-
ture of the deoxyribonucleic acid (DNA) molecule (Watson
and Crick, 1953), characterization of the complete sequence
and organization of the human genome was successfully
accomplished (Lander et al., 2001; Venter et al., 2001). This
major scientific achievement laid the foundation of “human
genomics”; the section of the biological sciences that studies
variations, mutations, and functions of genes and control-
ling regions, and their implications for human variations,
health, and disease. This is strengthened by developments
in the other areas of genomics relating to bacteria, vectors,
parasites, animals, and plants.
The identification of all human genes and their regu-
latory regions provides the essential framework for our
understanding of the molecular basis of disease. This
advance has also provided a firm foundation for the future
development of genomic technologies that can be applied
xi
to modern medical science. Rapid developments in global
gene analysis, gene product analysis, medical bioinformat-
ics, and targeted molecular genetic testing are destined to
change the practice of modern medicine. However, many
practicing clinicians perceive developments in genomics as
primarily confined to the research arena, with little clini-
cal applicability. But DNA- and RNA-based methods of
disease-susceptibility screening, molecular-based disease
diagnosis and prognosis, and genomics-based therapeutic
choices and prediction of treatment outcomes are some of
the key areas that are likely to influence the practice of mod-
ern clinical medicine.
Undoubtedly the science of genomics holds tremen-
dous potential for improving human health. The World
Health Organization (WHO) has made several recommen-
dations on the scope and application of genomics on global
health. It is acknowledged that the information generated
by genomics will provide major benefits in the prevention,
diagnosis, and management of communicable and genetic
diseases as well as other common medical diseases, includ-
ing cardiovascular diseases, cancer, diabetes, and mental
illnesses (Cardon and Bell, 2001). Together, these consti-
tute the major global health burden, as reflected in chronic
ill-health and mortality. In addition, a number of infectious
diseases are associated with genomic mutations, manifest-
ing in the form of increased susceptibility, clinical sever-
ity, favorable or unfavorable responses to anti-microbial
therapy—or in conferring protection. It is possible that the
protective effect of a microbial vaccine might be influenced
by genomic variation.
The sequence of the entire human genome is now
complete—but each person carries a distinct sequence.
The variation among all humans is reflected in varia-
tion within the human genome. The genomic variation
between individuals, together with environmental factors,
probably determines each person’s disease susceptibil-
ity, and is important in drug efficacy and side effects for
that person (Holden, 2000; Chakravati, 2000). The key to
genomic variation lies in finding single-nucleotide poly-
morphisms (SNPs) and their use in disease-association
studies (Stephens et  al., 2001). The positional cloning
(identifying the gene by location, followed by functional
analysis) of the disease susceptibility loci will depend on
the successful application of haplotype associations. In
addition, these will be important in clinical studies to find
individuals in whom a drug is likely to be efficacious. The
use of SNPs in pharmacogenetics is currently restricted
to studying genes for drug-metabolizing enzymes, such as
P450s, and variations in genes that target drug receptors.
The newly emerging dynamic field of pharmacogenomics is
an exciting application of genomic variation in drug dis-
covery and drug development.
The recent cloning of real disease-susceptibility genes
for multifactorial diseases is encouraging: for example, the
identification of NOD2 as a susceptibility gene for Crohn’s
disease (Hugot et  al., 2001; Ogura et  al., 2001). This is a
major development in understanding the pathophysiology
of inflammatory bowel disease. Similar studies are likely to
unravel the genetic mechanisms in other complex medical
diseases. A comprehensive SNP map will allow the cloning
of other susceptibility alleles. However, this will depend
upon population sample and size, the method employed,
linkage disequilibrium, or association studies, rather than
on the technology used (Cardon and Bell, 2001). Some
of the best genetic studies of this kind include studies of
susceptibility to infectious disease; for example, of an asso-
ciation between chemokine receptors (CCR5) and HIV
susceptibility, and between the bacterial transporter pro-
tein Nramp and resistance to macrophage-infecting bac-
teria such as Mycobacterium tuberculosis. Similarly, various
alleles at the G6PDH locus determine malaria susceptibil-
ity (Tishkoff et al., 2001).
These kinds of studies, and clinical applications of the
resulting outcomes, are not without ethical concerns. Some
of the questions and concerns are related to “ownership” of
the genes and the freedom to use collected DNA for such
studies. These are complex and emotional issues, especially
when we are dealing with populations who may have been
exploited or are perceived to have been exploited. These
issues should always be dealt with carefully under the statu-
tory requirements and rules.
There has been a tremendous surge in various subspe-
cialties and technologies with names ending in -omics.
We are rapidly moving into the “omics” era. In addition
to genomics, several new specialist fields with an “-omics”
suffix have recently appeared; for example, pharmacoge-
nomics, nutrigenomics, metabonomics, transcriptomics,
proteomics, microbiomics, glycomics, toxicogenomics,
xii  •   P r e fac e to t h e F i r s t E d i t i o n
and many more. Some of these areas are included in
this book. Whatever the basis of distinction might be,
the driver of all these terms is GENOMICS—the study of
genomes in its entirety.
Genomics is not just about genome sequencing. Apart
from full-length cDNAs and their sequences, copies of
mRNAs that actually exist and code for different proteins
are probably more important. The study of proteins thus
derived falls within the broad field of proteomics, a likely out-
come of functional genomics and probably a true compan-
ion to genomics. It is likely that proteomics will eventually
have more practical applications in clinical medicine. This is
rapidly moving ahead with the completion of the HapMap
Project (Nature, 2005)  and the future “functional-variant
database,” a natural outcome of the HapMap Project (Gibbs,
2005).
It is vital that existing gaps in our knowledge about
various “omics” disciplines be filled to ensure efficient use
of the valuable information emerging from research. It is
also important that the gap between “genetic profession-
als” and the primary-care community, as well as the “pub-
lic health community,” be narrowed (Khoury et al., 2003).
Integration of this knowledge into the medical education
curriculum and the continued professional education pro-
grams is urgently required to ensure applications of genom-
ics in the provision of health care.
During the last two decades, the practice of medical
genetics or clinical genetics has found its niche within the
broad purview of clinical medicine. Genetic services now
constitute a small, albeit important, component of mod-
ern medical practice and public health. Currently, genetic
services focus on providing information on chromosomal
and single-gene diseases, with limited contributions to
multifactorial/polygenic diseases. How would this then
be different from genomics? Already there is tremendous
enthusiasm for the recently introduced term of “genomic
medicine.” In a primer on genomic medicine, Guttmacher
and Collins (2002) viewed “genetics as the study of single
genes and their effects” and genomics as “the study not
just of single genes, but of the functions and interactions
of all the genes in the genome.” In simple terms, there is a
quantitative difference between the two fields—the study
of multiple genes as opposed to one gene. Thus genetics can
be seen as part of genomics. However, there is a qualitative
difference between genetics and genomics in medical and
health applications, ranging from the concept of disease in
genetics to the concept of information in genomics (Khoury
et al., 2003).
The practice of medical genetics has traditionally
focused on conditions that result from specific alterations
or mutations in single genes (e.g., inborn errors of metab-
olism, Duchenne muscular dystrophy, and Huntington’s
disease); in parts of, or whole, chromosomes (e.g., trisomy
21 in Down syndrome); or associated with congenital
malformations and developmental disabilities. The exist-
ing model of medical genetic services for these conditions
includes laboratory diagnosis and genetic counseling and
management. This is supported by public health mea-
sures to ensure the delivery of genetic services and genetic
screening (e.g., newborn screening or screening the
high-risk population). On the other hand, the practice
of genomics in medicine and public health will focus on
information resulting from variations at one or multiple
loci and strong interactions with environmental factors;
for example, diet, drugs, infectious agents, chemicals,
physical agents, and behavioral factors (Khoury et  al.,
2003).
What medical and public health applications could one
foresee following the completion of the human genome
sequence in 2003? How could these be applied and deliv-
ered to the 95% of human diseases that do not fall under
the rubric of “genetic disorders”? These are some of the
likely questions related to genomic medicine. Medical
and public health professionals urgently need to make
the changes necessary to accommodate rapid identifica-
tion and characterization of the numerous genomic vari-
ants at multiple loci that increase or decrease the risks for
various diseases, singly or in combination with other genes,
and with various chemical, physical, infectious, pharma-
cological, and social factors (Khoury, 1999). This genetic
and genomic information is crucial in assessing the disease
susceptibility of healthy individuals, and in personalized
primary- and secondary-prevention planning. Collins and
McKusick (2001) stated that,
By the year 2010, it is expected that predictive
genetic tests will be available for as many as a dozen
common conditions, allowing individuals who wish
to know this information to learn their risks for
which interventions are or will be available. Such
interventions could take the form of medical surveil-
lance, lifestyle modifications, diet, or drug therapy.
Identification of persons at highest risk for colon
cancer, for example, could lead to targeted efforts to
provide colonoscopic screening to those individuals,
with [the] likelihood of preventing many premature
deaths.
P r e fac e to t h e F i r s t E d i t i o n   •   xiii
Personalized medicine will not only encompass com-
mon medical diseases, but could also include a wide
range of preventable diseases. Genetic testing for future
disease-susceptibility using multiple genomic variants
will be possible and affordable with the application of
“high-throughput” microarray-based genetic testing.
A wealth of information on genomics is rapidly being
acquired, with the potential for a major impact on human
health. However, these data and this information are scat-
tered throughout several scientific journals, reviews, and
state-sponsored reports and bulletins. A clinician or health
professional often has difficulty in accessing and assimilat-
ing this information for application in her or his medical
and public health practice. More importantly, an inability
to assimilate and interpret this information can lead to
frustration, and therefore avoidance of potentially useful
information.
In view of the above developments and the rapidly
increasing gulf between the practitioners and the avail-
able literature resources, the need for a dedicated book on
genomic medicine was appreciated. Writing such a book
was obviously a nearly impossible task for a single author.
Several leading experts in different fields of the genome
science and technology therefore offered to contribute.
The views and opinions reflected in their individual chap-
ters are largely influenced by each author’s experience,
perception, and interpretation of the available data and
information.
This book provides a wide-ranging coverage of the
subject, from the historical progress to general aspects of
genomics, and describes in some detail the medical and
health applications. Generally, all chapters follow the
same format and are written by experts in their respective
fields of research and clinical expertise. Each chapter pro-
vides a detailed and comprehensive account of its subject.
However, it is likely that some gaps might exist, due to the
inevitable time-lag between the time of writing and appear-
ing in print. This is due to rapid developments in each field.
However, all efforts have been made to provide the reader
core information on the basic principles, scientific facts,
current and likely future applications, useful relevant ref-
erences, and information on Internet-based resources that
should be helpful in exploring the subject further.
It is hoped that this book will facilitate acquiring fac-
tual information on genomics, developing concepts about
the genomic basis of human disease, and provide a practical
base for any interested clinicians and health professionals
to develop an understanding of applications of genom-
ics in clinical medicine and health. It is aimed at a wide
range of scientists, clinicians, and health professionals who
are engaged in research, teaching, and training in medi-
cal and health applications of genome-based science and
technology.
Finally, the practice of medicine is an art based on sound
scientific principles. It would be appropriate to quote Sir
William Osler’s remark, “If there were no individual vari-
ability, medicine would have been science, not an art.”
Genomics in this context provides the basis of individual
variability, and the modern post-genomic clinician will
need to ensure that this is applied as an art.
Dhavendra Kumar
Institute of Medical Genetics
Cardiff University, Wales
United Kingdom
RE F ERENCES
Cardon LR, Bell, JI (2001). Association study designs for complex dis-
eases. Nat Rev Genet. 2:91–99.
Chakravati A (2000). To a future of genetic medicine. Nature.
409:822–823.
Collins FS, Guttmacher AE (2001). Genetics moves into medical main-
stream. JAMA. 286:2322–2324.
Collins FS, McKusick VA (2001). Implications of the Human Genome
Project for medical science. JAMA. 285:540–544.
xiv  •   P r e fac e to t h e F i r s t E d i t i o n
Genovations—the advent of truly personalized healthcare. Available at
http://www.genovariations.com.
Gibbs R (2005). Deeper into the genome. Nature. 437:1233–1234.
Guttmacher AE, Collins FS (2002). Genomic medicine:  a primer. N
Engl J Med. 347:1512–1520.
Holden AL (2000). The SNP consortium: a case study in large pharma-
ceutical company research and development collaboration. J Com
Biotech. 6:320–324.
Hugot JP et al. (2001). Association of NOD2 leucine-rich variants with
susceptibility to Crohn’s disease. Nature. 411:599–603.
Khoury MJ (1999). Human genome epidemiology: translating advances
in human genetics into population-based data for medicine and
public health. Genet Med. 1:71–73.
Khoury MJ, McCabe LL, McCabe ER (2003). Population screening in
the age of genomic medicine. N Engl J Med. 348:50–58.
Lander ES et  al. (2001). Initial sequencing and analysis of the human
genome. International Human Genome Sequencing Consortium.
Nature. 409:860–921.
Nature (2005). A haplotype map of the human genome—report from
the International HapMap Consortium. Nature. 437:1299–1320.
Ogura Y et al. (2001). A frameshift in NOD2 associated with susceptibil-
ity to Crohn’s disease. Nature. 411:603–606.
Stephens C et al. (2001). Haplotype variation and linkage disequilibrium
in 313 human genes. Science. 293:489–493.
Tishkoff SA et  al. (2001). Haplotype diversity and linkage disequilib-
rium at the human G6PDH:  recent origin of alleles that confer
malarial resistance. Science. 293:455–461.
Venter JC et  al. (2001). The sequence of the human genome. Science.
291:1304–1351.
Watson JD, Crick FHC (1953). Molecular structure of nucleic acids.
Nature. 171:737–738.
World Health Organization (2002). Genomics and world health—report
from the Advisory Committee on Health Research. Geneva: WHO.
 PR EFACE
Since the publication of Kumar and Weatherall’s Genomics
and Clinical Medicine (2008), the science of genomics
has made tremendous progress (Kumar, 2008; Hamburg
and Collins, 2010). The preface for the first edition (see
pages. . . .) set out a challenging and visionary goal for
authors and editors. This was successfully accomplished, as
reflected in positive and constructive reviews, personal or
published (Feero, Guttmacher et  al., 2010). Exciting new
developments in biotechnology and bioinformatics have
opened horizons that were inconceivable only a few years
ago. The chatter about next-generation sequencing is not
restricted to post-doctoral trainees and young investiga-
tors. It is evident everywhere and is now firmly ingrained
in the minds and souls of genetic and genomic researchers
and healthcare professionals (Berg, Khoury et  al., 2011).
Indeed, even as we write this second edition of (probably)
the first book on genomic medicine, many tertiary genetics
centers are utilizing whole-exome sequencing for routine
clinical care (Lupski, Reid et  al., 2010; Green and Guyer,
2011). Unravelling the complexities of the RNA molecules
has made a huge impact in molecular and experimental
biology. Challenging and controversial stem cell genomic
research captured the headlines and was applauded by the
awarding of the 2012 Nobel Prize. This is truly the begin-
ning of a promising phase for applied and translational
genomics. The sky is the limit.
Enormous genomic data and information generated
by genome-wide association studies (GWAS), deciphering
the complex phenotypes by copy-number variations and
single-nucleotide polymorphisms, and applying knowledge
gained from genetic and genomic analysis in the so-called
rare Mendelian disorders, which affect more than 18 million
Americans alone, have offered fine molecular understanding
of the underlying pathogenic mechanisms, as well as imple-
mentation in clinical care of molecular diagnostics, risk
assessment, genetic counseling, management, and predic-
tive testing of at-risk relatives (Chin, Andersen et al., 2011).
There is a lot of enthusiasm for applying next-generation
sequencing methods (alongside Sanger sequencing, given
the necessity of validating of next-generation data) in new
xv
gene discoveries, unravelling novel molecular mechanisms,
and identifying critical focal points in molecular pathways
in designing and developing targeted molecular therapy
models. Inevitably, this has led to intense debate on practi-
cal matters related to disclosure and applications of perti-
nent and incidental findings (Green, Berg et al., 2013).
Several major global initiatives are being pursued to
curate and annotate the enormous amount of genomic data
and information from new genomic technological advances.
The common theme is genotype–phenotype correla-
tion, which is vital for clinical care. Leaders in this type of
multi-institutional approach include the Human Variome
Project (www.humanvariomeproject.org), the GenPhen
project (www.Gen2Phen.Org), and the recently launched
Global Genome Alliance (http://www.ebi.ac.uk/about/
news/press-releases/Global-Alliance). Successful outcomes
of these projects might offer clarification and evidence that
could be applied to personalizing healthcare and wellness.
However, there is sufficient evidence supporting the argu-
ment for genomic applications for enhancing the diagnostic
and probably the prognostic potential of genomic medicine
and health (Khoury, Gwinn et  al., 2007). Promising new
therapeutic developments have followed, particularly the
discovery and development of new drugs and the phar-
macogenetic/pharmacogenomic evidence necessary for
personalizing pharmacotherapy; or at least we have made a
good start (Chin, Andersen et al., 2011).
So how do genomics and all the related genome tech-
nologies affect medicine and health? Do we have enough
data, understanding, and robust evidence, especially of clin-
ical outcomes, to apply and translate into practicing effec-
tive and efficient clinical medicine? It is probably safe for
us to gently move into the next phase of genomic medicine
and personalized healthcare.
The thoroughly revised and practically wholly new text
in this second edition aims to address the above questions
and dilemmas. We are pleased to present this edition under a
new title, Genomic Medicine: Principles and Practice. Several
of our colleagues and professionals in related networks
might agree on this ambitious title and probably share the
view that genetics and genomics knowledge is ripe for judi-
cious clinical practice, but determining the practical clinical
implementation remains the challenge. Part of the challenge
is bringing all clinicians some minimum of practical knowl-
edge of genomic advances so that genomics-enabled clini-
cal practice can be understood, embraced, and leveraged
for the delivery of value-based healthcare. Our second edi-
tion seeks to reach this lofty goal. The views and opinions
expressed herein reflect each individual author’s content
expertise and knowledge, interpretation, and determina-
tion for puutting forward their views and opinions on the
future of medicine in the genome era. The editors have sim-
ply facilitated the process to present the material in the best
possible and deliverable manner. Naturally, all those who
worked in developing and producing the second edition of
this unique genomic medicine textbook will be anxious to
know how medical students, post-doctoral fellows, genome
scientists, and genetic/genomic physicians would rank the
new edition.
Even with the genome in our palm, we remain humbly
aware that we continue to strive to be better healers, as we
have for 4,600 years:
Superior doctors prevent the disease;
Mediocre doctors treat the disease before evident;
Inferior doctors treat the full-blown disease.
—Attr. Nai-Ching (first Chinese medical text), 2600 BC
Dhavendra Kumar
and
Charis Eng
May 2014
xvi  •   P r e fac e
RE F ERENCES
Berg JS, Khoury MJ, et al. (2011). Deploying whole genome sequencing
in clinical practice and public health: meeting the challenge one bin
at a time. Genet Med. 13(6):499–504.
Chin L, Andersen JN, et al. (2011). Cancer genomics: from discovery
science to personalized medicine. Nature Med. 17(3):297–303.
Feero WG, Guttmacher AE, et  al. (2010). Genomic medicine—an
updated primer. N Engl J Med. 362(21):2001–2011.
Green ED, Guyer MS (2011). Charting a course for genomic medicine
from base pairs to bedside. Nature. 470(7333):204–213.
Green RC, Berg JS, et  al. (2013). The American College of Medical
Genetics (ACMG) recommendations for reporting of incidental
findings in clinical exome and genome sequencing. Genet Med.
15(7):565–74; doi:10.1038/gim.2013.73.
Hamburg MA, Collins FS (2010). The path to personalized medicine.
N Engl J Med. 363(4):301–304.
Khoury MJ, Gwinn M, et  al. (2007). The continuum of translation
research in genomic medicine: How can we accelerate the appropri-
ate integration of human genome discoveries into health care and
disease prevention? Genet Med. 9(10):665–674.
Kumar D (2008). Clinical medicine in the genome era: an introduction.
Genomics Clin Med. (53):145.
Kumar D, Weatherall DJ (2008). Genomics and Clinical Medicine.
Oxford, UK: Oxford University Press. New York
Lupski JR, Reid JG, et  al. (2010). Whole-genome sequencing in a
patient with Charcot-Marie-Tooth neuropathy. N Engl J Med.
362(13):1181–1191.
 ACK NOWLED G MEN TS
Nelson Mandela once said, “It always seems impossible
until it’s done.” This probably applied to this mammoth
book project. On a number of occasions, it did not seem
possible that the book would ever come to be. Nevertheless,
we are proud and very pleased to present the second edi-
tion of Kumar and Weatherall’s Genomics and Clinical
Medicine largely rewritten, extensively revised, presented in
an entirely different style and format, and wrapped with a
new title of “Genomic Medicine: Principles and Practice.”
Developing and working on the second and revised edi-
tion of the Oxford genomic medicine textbook has been
a rewarding and learning experience. The project had the
blessings of Sir David Weatherall, who inspired and guided
the original version of this book (Genomics and Clinical
Medicine, Oxford University Press, 2008) and continued to
advise and mentor us on this entirely new version from con-
ception to completion. We are fortunate to have the support
of a fresh team of dedicated experts and authors who have
produced the finest possible text, presented in a number of
chapters. Whilst we have retained a few selected authors
and contributors from the first edition, several of our cur-
rent authors are entirely new to the emerging and chal-
lenging field of genomic medicine. We will always remain
indebted and grateful for their support and contribution.
A few names deserve special thanks:  notably Andrew
Read (Manchester, England), Stephen Pennington (Dublin,
Ireland), Richard Festenstein (London), Kevin White
(Chicago, Illinois), Patrick Stover (Cornell, New  York),
Rino Rappuoli (Novartis, Italy), Teri Monolio (National
Institutes of Health, Bethesda, Maryland), Angus Clarke
xvii
(Cardiff, Wales), Reed Pyeritz (Philadelphia, Pennsylvania),
Michael Parker (Oxford, England), Jane Kaye (Oxford,
England), Dan Arking (Baltimore, Maryland), Kenneth
Mills (Belfast, Northern Ireland), Bill Cookson (London),
Sarra Jamieson (Perth, Australia), Graeme Black
(Manchester, England), Karen Avraham (Tel Aviv, Israel),
Eugene Healy (Southampton, England), Julian Sampson
(Cardiff, Wales), Ben Lim (Singapore), Neil Robertson
(Cardiff, Wales), Julian Knight (Oxford, England), and
Stylianos Antonorakis (Geneva, Switzerland).
Apart from having a team of world-class authors, the
publishing team at the Oxford University Press (OUP)
in New  York worked hard and demonstrated unmatched
patience for managing delays, handling few unacceptably
large and poorly presented manuscripts, and applying their
superb editorial and technical skills in the production of
this new edition. The OUP team, led by Catherine Barnes
and ground managed by Chad Zimmerman and Meredith
Keller, deserve praise and gratitude for bringing this excep-
tional book to reality and allowing it a place in the presti-
gious series, Oxford Monographs on Medical Genetics.
All clinicians work and live for serving patients and their
families—this book is dedicated to all our patients. Finally,
no small or large project could be completed without the
blessings and support of the family, particularly those close
to our heart and soul. We are deeply indebted and grateful
to our parents and families for their untiring and infinite
support in the completion of this book.
Dhavendra Kumar and Charis Eng
 Free ebooks ==> www.Ebook777.com
CON T EN TS
Foreword vii
Stylianos Antonarakis
Foreword to the First Edition ix
Francis S. Collins
Preface to the First Edition xi
Dhavendra Kumar
Preface xv
Dhavendra Kumar and Charis Eng
Acknowledgements xvii
Dhavendra Kumar and Charis Eng
PA RT I
PRINCIPLES OF GENOMIC MEDICINE
1. Genes, genetics, and human genomics 3
Dhavendra Kumar
2. The human genome—structure and organization 13
Andrew P. Read
3. Human proteomics 27
Brian Morrissey, Lisa Staunton,
and Stephen R. Pennington
4. Epigenetics, epigenomics, and human disease 39
Aravind Ramesh, Cihangir Yandim, Theona
Natisvili, Marta Mauri, Piu Pik Law, Jackson P. K.
Chan, Santiago Uribe Lewis and Richard Festenstein
5. Genes, genome, and developmental malformations 61
Dhavendra Kumar
6. Bioinformatics, systems biology, and systems
medicine 83
Binay Panda and Neeraja M. Krishnan
7. Pharmacogenomics—critical component of
genomic medicine 97
Wolfgang Sadee
8. New drug development, drug response,
and precision medicines 114
Michelle Penny and Duncan McHale
9. Mitochondrial genetics and genomics in clinical
medicine 131
Agnès Rötig and Dhavendra Kumar
xix
www.Ebook777.com
10. Genomics technology in clinical diagnostics 148
Kevin White and Jeremy Segal
11. Microbial genomics: targeted antimicrobial therapy
and genome vaccines 167
Immaculada Margarit and Rino Rappuoli
12. Nutritional genomics 180
Zhenglong Gu, Kaixiong Ye, and Patrick J. Stover
13. Genomics in public and population health 210
Anastasia L. Wise and Teri A. Manolio
14. Genetic testing and genomic screening 218
Angus John Clarke
15. Biobanking for genomics-based translational medicine 235
Steven J. Madore
16. Genetics and genomics education: the path from
helix to health 242
Reed E. Pyeritz
17. Ethical, legal, and social issues in clinical genomics 250
Caroline F. Wright, Anna Middleton,
and Michael Parker
18. The regulation of human genomics research 259
Jane Kaye
PA RT II
GENOMIC S IN CLINICAL PRACTICE
19. Genetic and genomic approaches to clinical medicine 269
Dhavendra Kumar
20. Genetic and genomic taxonomy of human disease 294
Dhavendra Kumar
21. Genomics of complex cardiovascular disease 316
Foram N. Ashar and Dan E. Arking
22. Genomics of type 2 diabetes mellitus and obesity 337
Venkatesan Radha and Viswanathan Mohan
23. Genetics and genomics of osteoporosis and related
disorders 352
Yoshiji Yamada
24. Genetics and genomics of chronic kidney disease 369
Albert C. M. Ong and Alexander P. Maxwell
 25. Genetics and genomics in clinical hematology, I:
Hemostasis and thrombosis 393
John H. McVey
26. Genetics and genomics in clinical hematology, II:  Vinod Cherian Varghese, Sian Morgan,
Inherited disorders of hemoglobin 404
Sir David J. Weatherall
27. Genetic and genomics in clinical hematology, III:  The spectrum of genetic eye disease 623
Acute leukemias 412
Kenneth Mills
28. Genetics and genomics of chronic inflammatory
disorders, I: Inflammatory bowel disease 431
Saad Pathan and Derek P. Jewell
29. Genetics and genomics of chronic inflammatory
disorders, II: Rheumatoid arthritis and related
arthropathies 448
Kate McAllister and Stephen Eyre
30. Genetics and genomics of chronic inflammatory
disorders, III: Bronchial asthma 462
William Cookson
31. Genetics and genomics of neuro-psychiatric diseases,
I: Seizure disorders 473
William Owen Pickrell
32. Genetics and genomics of neuro-psychiatric diseases,
II: Multiple sclerosis 487
Katharine Harding and Neil Robertson
33. Genetics and genomics of neuro-psychiatric diseases,
III: The common dementias 508
Amy Gerrish, Rebecca Sims, and Julie Williams
34. Genetics and genomics of neuro-psychiatric diseases,
IV: Schizophrenia and bipolar disorder 521
Jinbo Fan
35. Genetics and genomics of neuro-psychiatric diseases,
V: Learning and behavioral disorders 530
F. Lucy Raymond
36. Clinical cancer genomics 541
Joanne Ngeow and Charis Eng
37. Genomics and infectious diseases: susceptibility,
resistance, response, and antimicrobial therapy 565
Michaela Fakiola, Wei Lu, Sarra E. Jamieson,
and Christopher S. Peacock
xx  •   C o n t e n t s
38. Genomic perspectives of clinical immunology 591
Cornelius L. Verweij
39. Genomic applications in clinical pediatrics 603
and Ian Frayling
40. Genetics and genomics in clinical ophthalmology, I:
Graeme Charles M. Black
41. Genetics and genomics in clinical ophthalmology,
II: Glaucoma 636
Roshanak Sharafieh, Anne H. Child, and
Mansoor Sarfarazi
42. Genetics and genomics in clinical ophthalmology,
III: Age-related macular degeneration 652
Mark E. Kleinman and Jayakrishna Ambati
43. Genomic applications in audiological medicine 663
Daphne Karfunkel-Doron, Zippora Brownstein, and
Karen B. Avraham
44. Genetics and genomics of skin diseases I: Atopic
dermatitis and other skin complex diseases 683
Nilesh Morar
45. Genetics and genomics of skin diseases, II: Genomics
of pigmentation and skin cancer 696
Eugene Healy
46. The genetic and genomic practice of reproductive
medicine 713
Dhavendra Kumar
47. Stem cell genomics: Developmental competence 741
Kyle M. Loh, Bing Lim, and Lay Teng Ang
48. Genomic applications in critical care medicine 766
Matthew C. Frise, Charles Hinds, and
Julian C. Knight
49. Molecularly targeted therapy for Mendelian disorders 781
Mark Davies and Julian Roy Sampson
50. Glossary for genetic and genomic medicine 788
Dhavendra Kumar
Index 801
 CON T R IBU TOR S

Ambati, Jayakrishna, MD Chan, Jackson, PK

Departments of Ophthalmology and Visual Sciences Gene Control Mechanisms and Disease Group
and Physiology Department of Medicine
University of Kentucky Imperial College
Lexington, Kentucky, KY Hammersmith Hospital
Ang, Lay Teng, PhD London, England, UK
Genome Institute of Singapore Child, Anne H., MD
Stem Cell and Developmental Biology Group Department of Cardiovascular Medicine
Singapore St. George’s Hospital Medical School and Hospital
Arking, Dan E., PhD London, England, UK
McKusick-Nathans Institute of Genetic Medicine Clarke, Angus John, DM, FRCPCH
Johns Hopkins University School of Medicine Institute of Cancer & Genetics (formerly Institute
Baltimore, Maryland, MD of Medical Genetics)
Ashar, Foram N., BS University Hospital of Wales
McKusick-Nathans Institute of Genetic Medicine Cardiff University School of Medicine
Johns Hopkins University School of Medicine Cardiff, Wales, UK
Baltimore, Maryland, MD Cookson, William, FRCP, DPhil
Avraham, Karen B., PhD Asmarley Centre for Genomic Medicine
Department of Human Molecular Genetics and The National Heart and Lung Institute
Biochemistry Hammersmith Hospital
Sackler Faculty of Medicine Imperial College
Sagol School of Neuroscience London
Tel Aviv University England, UK
Tel Aviv, Israel Davies, Mark, MBBS, PhD
Black, Graeme Charles M., PhD, FRCOphth Institute of Cancer & Genetics
Manchester Centre for Genomic Medicine, Institute of Cardiff University School of Medicine
Human Development, University of Manchester and Central University Hospital of Wales
Manchester University Hospitals NHS Foundation Trust, Cardiff,
Manchester Academic Health Science Centre (MAHSC) Wales, UK
St. Mary’s Hospital
Eng, Charis, MD PhD FACP
Manchester, England, UK
Sondra J. & Stephen R. Hardis Endowed Chair of
Brownstein, Zippora, PhD Cancer Genomic Medicine Director- Genomic
Department of Human Molecular Genetics Medicine Institute & Center for Personalized
and Biochemistry Genetic Healthcare Cleveland Clinic,
Sackler Faculty of Medicine and Professor and Vice Chair Department of Genetics &
Sagol School of Neuroscience Genome Sciences Case Western Reserve University
Tel Aviv University School of Medicine Cleveland, Ohio, OH
Tel Aviv, Israel

xxi
Eyre, Stephen Harding, Katharine, BM, BCh, MRCP
Arthritis Research UK Epidemiology Unit, Centre for Institute of Psychological Medicine and Clinical Lecturer
Musculoskeletal Research, Institute of Inflammation in Neurology
and Repair, and NIHR Manchester Musculoskeletal Cardiff University School of Medicine
Biomedical Research Unit, MAHSC, Stopford Building University Hospital of Wales
The University of Manchester Cardiff, Wales, UK
Manchester, England, UK
Healy, Eugene, PhD FRCP
Fakiola, Michaela, PhD Department of Dermatology
Cambridge Institute for Medical Research Southampton University Hospitals NHS Trust
Addenbrooke’s Hospital University of Southampton
Cambridge, England, UK Southampton
England, UK
Fan, Jinbo, PhD
Departments of Epidemiology and Biostatistics Hinds, Charles, MBBS, FRCP, FRCA
and Psychiatry Professor of Intensive Care Medicine
Case Western Reserve University School of Medicine West Smithfield
Cleveland, Ohio, OH London, England, UK
Festenstein, Richard, MD Jamieson, Saara E., PhD
Gene Control Mechanisms and Disease Group Telethon Institute for Child Health Research
Department of Medicine Centre for Child Health Research
Imperial College University of Western Australia
Hammersmith Hospital Subiaco, Australia
London, England
Jewell, Derek P., FRCP, Dphil
Frise, Matthew C., BM, BCh, MRCP Nuffield Department of Clinical Medicine
Specialty Registrar in General Internal Medicine John Radcliffe Hospital,
and Intensive Care Medicine The University of Oxford
Oxford University Hospitals NHS Trust Oxford, England, UK
John Radcliffe Hospital
Karfunkel-Doron, Daphne, PhD
Oxford, England, UK
Department of Human Molecular Genetics
Frayling, Ian, PhD FRCPath and Biochemistry
All Wales Genetics Laboratory Service Sackler Faculty of Medicine
Institute of Medical Genetics Sagol School of Neuroscience
University Hospital of Wales Tel Aviv University
Cardiff, Wales, UK Tel Aviv, Israel
Gerrish, Amy, PhD Kaye, Jane, DPhil
Research Fellow HeLEX—Centre for Health, Law and Emerging
Medical Research Council (MRC) Centre for Technologies at Oxford
Neuropsychiatric Genetics and Genomics Nuffield Department of Population Health
Cardiff University School of Medicine University of Oxford
Cardiff, Wales, UK Oxford, England, UK
Gu, Zhenglong, PhD Kleinman, Mark E., MD
Division of Nutritional Sciences Departments of Ophthalmology and Visual Sciences
Cornell University and Physiology
Ithaca, New York, NY University of Kentucky
Lexington, Kentucky, KY

xxii  •   C o n t r i b u to r s
Knight, Julian C., DPhil, FRCP Lu, Wei, PhD
Senior Clinical Research Fellow in Genomic Medicine School of Pathology and Laboratory Medicine
Wellcome Trust Centre for Human Genetics Faculty of Medicine and Dentistry
University of Oxford The University of Western Australia
Oxford, England, UK Nedlands, Australia
Krishnan, Neeraja M., PhD McAllister, Kate, PhD
Strand Life Sciences Arthritis Research UK Epidemiology Unit, Centre for
Bangalore, India Musculoskeletal Research, Institute of Inflammation
and Repair, and NIHR Manchester Musculoskeletal
Kumar, Dhavendra, MD FRCP FACMG
Biomedical Research Unit, MAHSC
Institute of Cancer & Genetics
The University of Manchester
Cardiff University School of Medicine,
Manchester, England, UK
All Wales Medical Genetics Service
University Hospital of Wales, Cardiff McHale, Duncan
& Global Exploratory Development
Genomic Policy Unit, UCB Pharma
The Faculty of Life Sciences & Education Stockport
The University of South Wales, England, UK
Pontypridd, UK
McVey, John H., PhD, FRCP
Law, Piu Pik MRC Centre for Transplantation
Gene Control Mechanisms and Disease Group Innate Immunity Section
Department of Medicine King’s College London, Guy’s Hospital
Imperial College London, England, UK
Hammersmith Hospital
Madore, Steven J., PhD
London, England, UK
Coriell Institute for Medical Research
Lewis, Santiago Uribe Camden, New Jersey, NJ
Gene Control Mechanisms and Disease Group
Manolio, Teri A., MD, PhD
Department of Medicine
National Human Genome Research Institute
Imperial College
National Institutes of Health
Hammersmith Hospital
Bethesda, Maryland, MD
London, England, UK
Margarit, Immaculada, PhD
Lim, Bing, PhD
Novartis Vaccines and Diagnostics
Genome Institute of Singapore, Stem Cell
Siena, Italy
and Developmental Biology Group
Singapore Mauri, Marta
Harvard Medical School, Department of Medicine, and Gene Control Mechanisms and Disease Group
the Beth Israel Deaconess Medical Center, Department Department of Medicine
of Hematology/Oncology Imperial College
Boston, Massachusetts, MA Hammersmith Hospital
London, England, UK
Loh, Kyle M., PhD
Stanford University School of Medicine Maxwell, Alexander P., PhD, FRCP
Department of Developmental Biology Centre for Public Health
Stanford Institute for Stem Cell Biology School of Medicine Queens University
and Regenerative Medicine Belfast, Northern Ireland, UK
Stanford, California, CA Middleton, Anna, PhD
Wellcome Trust Sanger Institute
Wellcome Trust Genome Campus
Cambridge, England, UK

C o n t r i b u to r s   •   xxiii
Mills, Kenneth, PhD Panda, Binay, PhD
Professor of Experimental Haematology Ganit Labs, Bio-IT Centre
Centre for Cancer Research and Cell Biology (CCRCB) Institute of Bioinformatics and Applied Biotechnology
Queen’s University Belfast Bangalore, India
Belfast, Northern Ireland, UK
Parker, Michael, DPhil
Mohan, Viswanathan, MD, DSc The Ethox Centre, Department of Public Health
Diabetes Specialties Centre University of Oxford
Madras Diabetes Research Foundation Oxford, England, UK
ICMR Centre for Advanced Diabetes Research
Pathan, Saad, MD
Dr. M.G.R. Medical University
Senior Consultant
University of Madras
Qatar Biomedical Research Institute
Chennai, Tamilnadu
Qatar Foundation
India
Qatar
Morar, Nilesh, FRCP, DPhil
Peacock, Christopher S., PhD
Department of Dermatology
School of Pathology and Laboratory Medicine
Chelsea and Westminster Hospital
Faculty of Medicine and Dentistry
Fulham Road
The University of Western Australia
London
Nedlands; Telethon Kids Institute
England, UK
The University of Western Australia
Morgan, Sian, BSc (Hons), MRCPath Subiaco, Australia
All Wales Genetics Laboratory Service
Pennington, Stephen R.
Institute of Medical Genetics
University College Dublin Conway Institute
University Hospital of Wales
School of Medicine and Medical Science
Cardiff, Wales, UK
Dublin, Republic of Ireland
Morrissey, Brian
Penny, Michelle, Phd
UCD Conway Institute, School of Medicine
Translational Medicine Unit
and Medical Science
Eli Lilly and Company
University College Dublin
Indianapolis
Dublin, Republic of Ireland
Indiana
Natisvili, Theona USA
Gene Control Mechanisms and Disease Group
Pickrell, William Owen, MRCP, MEng
Department of Medicine
Clinical Research Fellow
Imperial College
Neurology and Molecular Neuroscience
Hammersmith Hospital
College of Medicine, Institute of Life Science
London, England, UK
Swansea University
Ngeow, Joanne, MBBS, MRCP Swansea, Wales, UK
Genomic Medicine Institute
Pyeritz, Reed E., MD, PhD
Cleveland Clinic, Cleveland, Ohio, USA
Departments of Medicine and Genetics
Oncology Academic Program, Duke-NUS Graduate
Perelman School of Medicine
Medical School and National Cancer Centre,
The University of Pennsylvania
Singapore
Philadelphia, Pennsylvania, PA
Ong, Albert C. M., DM, FRCP
Academic Nephrology Unit and Bateson Centre
University of Sheffield Medical School
Sheffield, England, UK

xxiv  •   C o n t r i b u to r s
Venkatesan Radha Sarfarazi, Mansoor, PhD
Diabetes Specialties Centre Molecular Ophthalmic Genetics Laboratory
Madras Diabetes Research Foundation University of Connecticut Health Center
ICMR Centre for Advanced Diabetes Research Farmington, Connecticut, USA
Dr. M.G.R. Medical University
Segal , Jeremy, MD, PhD
University of Madras
Division of Genomic and Molecular Pathology
Chennai, Tamilnadu
Department of Pathology
India
The University of Chicago
Ramesh, Aravind S. Maryland Ave.
Gene Control Mechanisms and Disease Group Chicago, IL 60637
Department of Medicine USA
Imperial College
Sharafieh, Roshanak, PhD
Hammersmith Hospital
Molecular Ophthalmic Genetics Laboratory
London, England, UK
University of Connecticut Health Center
Rappuoli, Rino, PhD Farmington, Connecticut, CT
Novartis Vaccines and Diagnostics
Sims, Rebecca, PhD
Siena, Italy
Research Fellow
Raymond, F. Lucy, FRCP, PhD MRC Centre for Neuropsychiatric Genetics
Department of Medical Genetics and Genomics
Addenbrooke’s Hospital Cardiff University School of Medicine
Cambridge, England, UK Cardiff, Wales, UK
Read, Andrew P., PhD, FMedSci Staunton, Lisa
Department of Genetic Medicine University College Dublin Conway Institute
University of Manchester School of Medicine and Medical Science
St. Mary’s Hospital Dublin, Republic of Ireland
Manchester, England, UK
Stover, Patrick J., PhD
Robertson, Neil, DM, FRCP Cornell University
Institute of Psychological Medicine and Clinical Division of Nutritional Sciences
Neuroscience Ithaca, New York, NY
Cardiff University School of Medicine
Varghese, Vinod Cherian, MD, MRCPCH
University Hospital of Wales
All Wales Medical Genetics Service
Cardiff, Wales, UK
Institute of Medical Genetics
Rötig, Agnés University Hospital of Wales,
INSERM U781 Cardiff, Wales, UK
Hôpital Necker–Enfants Malades
Verweij, Cornelius L., PhD
Université Paris Descartes Paris, France
Department of Immunology
Sadee, Wolfgang, Dr.rer.nat. Academic Medical Centre
Center for Pharmacogenomics University of Amsterdam,
College of Medicine The Netherlands
Ohio State University
Weatherall, Sir David J., DM, FRCP
Columbus, Ohio, OH
Weatherall Institute of Molecular Medicine
Sampson, Julian Roy, DM FRCP University of Oxford
Institute of Cancer & Genetics John Radcliffe Hospital
Cardiff University School of Medicine Oxford, England, UK
University Hospital of Wales
Cardiff
Wales, UK

C o n t r i b u to r s   •   xxv
White, Kevin, PhD Yamada, Yoshiji, MD, PhD
Department of Human Genetics Department of Human Functional Genomics
Department of Ecology & Evolution Life Science Research Center
Institute for Genomics and Systems Biology Mie University
The Pritzker School of Medicine Tsu, Japan
The University of Chicago, IL
Yandim, Cihangir
USA
Gene Control Mechanisms and Disease Group
Williams, Julie, PhD Department of Medicine
MRC Centre for Neuropsychiatric Genetics Imperial College
and Genomics Hammersmith Hospital
Cardiff University School of Medicine London, England, UK
Cardiff, Wales, UK
Ye, Kaixiong
Wise, Anastasia L., PhD Cornell University
National Human Genome Research Institute Division of Nutritional Sciences
National Institutes of Health Ithaca, New York, NY
Bethesda, Maryland, MD
Wright, Caroline F., PhD
Wellcome Trust Sanger Institute
Wellcome Trust Genome Campus
Cambridge; The Ethox Centre,
Department of Public Health
University of Oxford
Oxford, England, UK

xxvi  •   C o n t r i b u to r s
GENOMIC MEDICINE
 PA RT I

PRINCIPLES OF GENOMIC MEDICINE

 1.
GENES, GENETICS, AND HUMAN GENOMICS
Dhavendra Kumar
INTRODUCTION
In the nineteenth and twentieth centuries, major discov-
eries and scientific advances in physical and biological sci-
ences led to the Industrial Revolution and the massive
transformation of the global landscape and standards of liv-
ing. Towards the end of the last millennium, tremendous
growth in the sophistication of the biological sciences was
harnessed in medicine, the food industry, and related indus-
tries. This was aided by major applications and translations
of physical sciences, particularly in the fields of computing
and information technology.
New discoveries and innovations in biological sciences
during the five decades preceding the twenty-first century
have centered on genetics and genomics. It took just over
50  years following the unraveling of the structure of the
molecule of nucleic acids, the key unit of the biological life,
for scientists to embark on sequencing of major organisms’
entire genetic constitution or genome. The word genome
includes gene and ome, implying complete knowledge
of all genes and related elements in any single organism.
Inevitably, this led to enthusiastic expansion of the whole
science and thence to the emergence of genomics.1 The suffix
-omics, derived from the ancient Greek, refers to in-depth
knowledge. Not surprisingly, genomics was followed by a
plethora of related -omics; for example, proteomics, metab-
olomics, transcriptomics, and so on.2 Currently, we have
over 30 such disciplines with the -omics suffix.
The ultimate goal of any scientific discipline is its trans-
lation for the benefit of all humans, crossing all possible
barriers and boundaries. Major advances in medicine and
health were only possible through understanding basic prin-
ciples and mechanisms underlying disease processes. This
was facilitated by rapid applications of physical and chemi-
cal sciences in medicine and health; for example, X-ray
diagnostics, ultrasound diagnosis, microbiology diagnosis,
immune-histochemical diagnosis, and finally, molecular
diagnosis. Developments and advances in genetics have led
3
to a better understanding of the principles governing hered-
ity and the familial transmission of physical characteristics
and diseases, better understanding of the pathophysiology
of diseases, the development of new methods of clinical and
laboratory diagnosis, and innovative approaches to mak-
ing early diagnoses (e.g., prenatal diagnoses and newborn
screening) and offering reproductive choices, including
pre-implantation genetic diagnoses. All these developments
are now accepted within the broad fields of human genet-
ics, medical genetics, clinical genetics, and (lately) genetic
medicine. Not surprisingly, the field remains wide open,
encompassing the massive field of human genomics appro-
priately named genomic medicine.3
This chapter leads the section titled “Principles of
Genomic Medicine.” It is anticipated that the reader, prob-
ably less informed than the specialists herein (or even unin-
formed), might find this helpful in grasping the concepts of
heredity, genes, genetics, and genomics. It is expected that
the reader will proceed to further chapters in this section
and the second section, “Practice of Genomic Medicine,”
better equipped with the introduction to genetic/genome
sciences, genetic diseases, genetics and genomics in medi-
cine, applications in public health, and specific issues related
to society, ethics, and law. The reader will find a major
change in the organization and presentation of material in
this book from that of the previous edition.4
THE HEREDIT Y
The concept of heredity dates back several hundreds and
even thousands of years. It is evident in all forms of bio-
logical life and evolution. Evolutionary scientists, philoso-
phers, and biologists have used “heredity” to put forward
their views on procreation, development, adaptation, and
the transmission of species-specific traits. The popular
Darwinian theory of natural selection rests on the core
concept of the transmission of hereditary factors.5 For several
thousand years, various descriptions and explanations have
been put forward to define the physical shape and func-
tional nature of hereditary factors. From ancient times, and
in almost every civilization, intense debate and arguments
have failed to arrive at a consensus. Most arguments focused
on whether the hereditary factor was a creation by God, a
new product fresh from the soil and water, or something in
the blood and in the semen. The symbolic representation of
the phallus in ancient sculptures and paintings of the Indian
subcontinent is an example of the concept that the phallus,
and thus semen, is a key factor in the creation and transmis-
sion of individuals’ (including families’) physical traits and
behavior characteristics.
In the historical context, the concept of the gene was
introduced only recently as the most acceptable answer to
explain one of the hereditary factors. It is unclear when and
by whom this term was first introduced. It does not matter,
as the term gene (from the Greek genos, race) is now uni-
versally accepted and used in the context of understanding
heredity, and is probably the single most important biologi-
cal factor regulating biological life, ranging from single-cell
organisms to multicellular mammals. Rapid and extraordi-
nary scientific progress made during nineteenth and twen-
tieth centuries has led to the development of genetics, the
science of heredity. This has now been transformed into
the broader field of genomics that includes all genes with all
possible biologically active, heritable, regulatory and evolu-
tionary genetic elements, whether recent or extending back
through several thousand years of life on our planet.
In biological terms, genes, genetics, and genomics are
keys to procreation, development, growth, function, and
survival. The health of any living organism is judged by its
physical and functional well-being. Thus, genes, genetics,
and genomics are central to all forms of biological health,
including that of humans. Human health depends not only
on its own genetic or genomic constitution, but on that of
other organisms whose well-being is also essential to human
health—for example, food (plants, fish, and animals), shel-
ter (homes made of wood from trees), the environment
(water, trees, and plants), protection (clothes from cotton
and animal skin), and transportation (animals and vehicles
made of wood from trees). From a medical perspective, the
science of genetics or genomics offers deep insight into and
evidence for a number of human diseases, including infec-
tious diseases resulting from either lack of protection and/
or failure in controlling the spread of microbial infections
or parasitic infestations. This chapter introduces the reader
to some of the basic facts about genes, genetics, and genom-
ics, and discusses how these impact human health and that
of the plants, crops, and animals necessary for human health
4   •   principles of G enomic M edicine
and survival. This is obviously more relevant to millions
of people in the developing and less-developed countries,
where limited resources and lack of infrastructure limit the
optimal use of the science of genetics and genomics in appli-
cations to eradicate poverty and ensure optimal health. The
reader will find cross-references to separate chapters in the
book containing detailed information and further discus-
sion of each subject.
G E N E S , G E N ET I C S , A N D G E N O M I C S
A detailed description of the basic principles of genetics and
human genetic diseases is beyond the scope of this chapter.
These facts are explained in subsequent chapters and vari-
ous other information resources on basic genetics and medi-
cal genetics (see “Further Reading”). However, some basic
principles and relevant information are outlined in this sec-
tion to assist the reader with limited understanding of basic
genetics.
Living organisms are divided into two large classes—
the eukaryotes and prokaryotes. The cells of the eukaryotes
have a complex compartmentalized internal structure, the
nucleus; these include algae, fungi, plants, and animals.
Prokaryotes, on the other hand, are single-celled micro-
organisms without any specific part harboring the genetic
material or genome; examples include bacteria and other
related microorganisms. The other types of living organisms
are viruses, which are intracellular obligate parasites living
in both eukaryotes and prokaryotes, and composed of short
dispersed nucleic acid (DNA or RNA) sequences.
Genetic information is transferred from one generation
to the next by small sections of the nucleic acid, deoxyri-
bonucleic acid (DNA), which is tightly packaged into sub-
cellular structures called chromosomes. Prokaryotes usually
have a single circular chromosome, while most eukaryotes
have more than two, and in some cases up to several hun-
dred. In humans, there are 46 chromosomes arranged in
23 pairs, with one of each pair inherited from each parent
(Figure 1.1, a & b). Twenty-two pairs are called autosomes,
and one pair is called sex chromosomes, designated as X and
Y; females have two X chromosomes (46, XX) and males
have an X and a Y (46, XY).
A chromosome consists of a tightly coiled length of
DNA and the proteins (e.g., chromatins) that help define
its structure and level of activity. DNA consists of two long
strands of nucleotide bases wrapped round each other along
a central spine made up of phosphate and sugar (Figure 1.2).
There are four bases:  adenine (A), guanine (G)  cytosine
(C), and thymine (T). Pairing of these bases follows strict
 (A) (B)
Human chromosomes!

Centromere 1 2 3 4 5

1 2 3 4 5 6 7 8 9

6 7 8 9 10 11 12

a
13 14 15 16 17 18
10 11 12 13 14 15 16 17 18

19 20 21 22 X 23Y or
Chromatid Y
b 19 20 21 22 X X X

Figure 1.1  Human chromosomes: a) Diploid set in a male (46, XY); b) Complete set of human chromosomes map.

Central Axis arranged in sets of three, referred to as codons. Each codon

Base “codes” for a specific amino acid; hence the term genetic
O
H 5’
C A T
code. Codons are located in exons, which contain the cod-
H
4’
H H
1’
G C
ing sequences. A  gene may consist of several such coding
H H DNA segments. Exons are separated from each other by
3’ 2’ T A
O H non-coding sequences of DNA, called introns. Although
C G
–O P = O they are not yet known to be associated with any specific
Base
O O
A T
function, it is likely that some of these introns might be of
H 5’
H
4’
C
1’
evolutionary significance, or associated with other funda-
G C
H
H H
mental biological functions. During the transcription of
H T A
3’ 2’ DNA, the introns are spliced out, and the exons then attach
C G
O
–O P = O
H
to mitochondrial RNA (mRNA) to start the process of pro-
A T
Base tein synthesis.
O O
H 5’
C
G C
Proteins are one of the major constituents of the body’s
H
4’
H H
1’ chemistry. These are remarkably variable in their structure,
T A
H
H ranging from tough collagen that forms connective tissue
3’ 2’ C G
O H and bone, through the fluid hemoglobin that transports
A T
–O P = O oxygen, to thousands of enzymes, hormones, and other
Base
O O
G C
biological effectors and their receptors that are essential for
H 5’
H
C T A the structures and functions of the body. Each protein is
4’ 1’
H
H H
made up of one or more peptide chains consisting of series
H G C
3’ 2’ of amino acids, only of which 20 occur in living organisms.
O H
The different structures and functions of proteins depend
Figure 1.2 
The Watson-Crick model of the double helix struture of the on the order of amino acids as determined by the genetic
nucleic acid molecule (Turnpenny and Ellard, 2011). code.
DNA has the remarkable property of self-replication.
rules: A always pairs with T, and C with G. Two strands are, The two strands of a DNA molecule separate as chromo-
therefore, complementary to each other. somes divide during cell division. There are two types of cell
Genes are made up of specific lengths of DNA that division; mitosis in all body cells, and meiosis, which is spe-
encode the information to make a protein, or ribonucleic cifically confined to the gonads in making sperm and eggs
acid (RNA) product. RNA differs from DNA in that the (Figure 1.4). During mitosis, no reduction of the number
base thymine (T) is replaced by uracil (U), and the sugar of chromosomes takes place (diploid, or 2n), while meiosis
is ribose. It acts as a template to take the coded informa- results in half the number of chromosomes (haploid, or 1n).
tion across to ribosomes for final assembly of amino acids The new pairs of DNA are identical to those from which
into the protein peptide chain (Figure 1.3). The bases are they were synthesized. However, sometimes mistakes or

G enes , G enetics , and H uman G enomics   •   5

 Free ebooks ==> www.Ebook777.com

C C
A C T
C A A A TA
C A T T AA
C T A G AA
FLANKING IVS IVS 2 Gene
NC GT AG GT AG NC

5’ 3’ mRNA Precursor

5’ CAP Excision of Introns

AAAA A
Splicing of Exons
AAAA A Processed mRNA
Nucleus
Cytoplasm

AAAA A Translation
Ribosome UAA

Transfer RNA

Amino
Acid Growing
Chain
Finished
Chain

Figure 1.3  The synthesis of the peptide chain from the coding sequences in the exon (Turnpenny and Ellard, 2011).

mutations occur. These usually result from substitution of a

Mitosis Meiosis
different base, or are due to extensive structural changes to
genes. In other words, any “spelling mistake” in the letters
A-T or C-G could result in either absence of coded infor-
DNA replication
DNA replication mation (nonsense mutation) or a different message (missense
mutation). However, not all mutations or spelling mistakes
have an adverse effect (neutral mutations). Conversely, some
changes in the genes might result in a favorable property;
Homologous paring
for example, resistance to disease or other environmental
hazard. This is the basis for the gradual changes in species
Line up on spindle
Line up on spindle over millions of years of evolution. On the other hand,
mutations may result in defective gene functions, leading
to a disease, or susceptibility to a disease, due to qualita-
Division 1
tive or quantitative changes in the gene product, the pep-
tide chain. However, these changes may also result from
epigenetic mechanisms, abnormal RNA molecules, and
Division post-translational modifications (see Glossary). A  brief
introduction to these molecular processes is provided
Division 2
elsewhere in this chapter; interested readers are advised to
consult dedicated texts on cell and molecular biology (see
“Further Reading”).
Studies on human genomic variations in different
population groups and the resemblance of several genome
sequences to other genomes (comparative genomics) have
Recombination offered wide-ranging evidence to support the follow-
ers of Charles Darwin. Apart from reproduction, genes,
Figure 1.4 
Steps in mitosis and meiosis during a eukaryotic cell division;
note (bottom) exchange of the genetic material (recombination) through gene-sequence variation, genomic variation, and epigenetic
homologous pairing (Turnpenny and Ellard, 2011). factors are important in growth, development, aging, and

6   •   principles of G enomic M edicine

www.Ebook777.com
senescence. Some of these may be evolutionarily conserved
across species, but relevant to human health. Mutations and
alterations in several of these genomic elements are linked
to a broad range of medical conditions.
THE HUM AN GENOME PROJECT
The advent of recombinant DNA technology in the 1970s
revolutionized our ability to characterize and capitalize
on the molecular basis of human genetic disease. This laid
the foundation of eventually mapping and deciphering the
DNA sequence of all the structural and functional genes of
the human genome. The Human Genome Project (HGP)
was, therefore, a natural progression from all previous
developments in the field of human genetics. Such a mam-
moth task could not have been accomplished without the
international collective efforts supported by generous fund-
ing from governmental and nongovernmental sources.6
The project (HGP) has helped map and provide nucle-
otide sequences of around 23,000 nuclear genes, which,
along with a number of other sequence variations, compose
the whole human genome (see Chapter  2). Although a
large number of the nuclear genes have been assigned with a
structural or functional link, the precise roles of other parts
of the genome are not yet fully understood. However, HGP
provides the basis for “functional genomics” to explore fur-
ther the genome’s functional role, and understand the com-
plex mechanisms through which genes and their products
interact to affect biological function and influence disease
processes. The development of new therapeutic agents is
now possible on the basis of genomic arrangement and its
designated functional role. This approach also helps char-
acterize the genomes of various pathogens and other organ-
isms, an invaluable tool in realizing the full potential of this
field to improve human health.7
H U M A N G E N O M E VA R I AT I O N
AND HUM AN DISE ASE
M E A S U R I N G G E N ET I C A N D G E N O M I C
VA R I AT I O N
The most direct way to measure genetic differences, or
genetic variation, is to estimate how often two individu-
als differ at a specific site in their DNA sequences—that
is, whether they have a different nucleotide base pair at a
specific location in their DNA. First, DNA sequences are
obtained from a sample of individuals. The sequences of all
G enes , G enetics , and H uman G enomics   •   7
possible pairs of individuals are then compared to see how
often each nucleotide differs. When this is done for a sam-
ple of humans, the result is that individuals differ, on aver-
age, at only about one in 1,300 DNA base pairs. In other
words, any two humans are about 99.9% identical in terms
of their DNA sequences (see Chapter 2).
During the past several years, a new type of genetic vari-
ation has been studied extensively in humans: copy-number
variants (CNVs) —DNA sequences of 1,000 base pairs or
larger are fairly distributed across the genome.8 In some
instances, CNVs could be deleted, duplicated, or inverted
in some individuals with mild phenotypical effects. Several
thousand CNVs have been discovered in humans, indicat-
ing that at least 4 million nucleotides of the human genome
(and perhaps several times more) vary in copy-number
among individuals. CNVs thus are another important class
of genetic variation and contribute to at least an additional
0.1% difference, on average, between individuals. Despite
significant progress, the medical and health implications of
CNVs are not entirely clear.9
Comparisons of DNA sequences can be done for pairs
of individuals from the same population or for pairs of
individuals from different populations. Populations can be
defined in various ways; one common way is to group indi-
viduals into populations according to a continent of origin.
Using this definition, individuals from different popula-
tions have roughly 10% to 15% more sequence differences
than do individuals from the same population (this estimate
is approximately the same for both SNPs—see below—and
CNVs). In other words, people from different populations
are slightly more different at the DNA level than are people
from the same population. The slightness of this difference
supports the conclusion that all humans are genetically
quite similar to one another, irrespective of their geographic
ancestry.10
Because it is still fairly expensive to assess DNA sequences
on a large scale, investigators often study genetic variations
at specific sites that are known to vary among individuals.
Suppose that a specific site in the DNA sequence harbors
an A in some individuals’ DNA sequences, and a G in oth-
ers. This is a single nucleotide polymorphism (SNP), where
polymorphism refers to a genetic site that exists in multiple
forms. The proportion of individuals who have an A  and
the proportion with a G give the frequency of each form,
or allele, and this frequency can be estimated for a sample
of individuals from a population. If the frequencies of A in
three different populations are .10,.20, and .50, the genetic
distance between the first two populations is smaller than
that between the third population and the first two. On
the basis of this assessment, the first two populations are
genetically more similar than either is to the third. To get
a more accurate picture of genetic differences, hundreds or
thousands of SNP frequencies would be assessed to yield
the average genetic difference among pairs of populations.11
G E N O M E VA R I AT I O N A N D HUM A N
DISEASE
Nearly all human diseases are influenced by genes. Because
individuals have different variants of genes, it follows
that the risk of developing various diseases will also differ
among individuals. Consider a simple example. Jim Fixx, a
well-known runner and fitness enthusiast, died of a heart
attack at the age of 52. Sir Winston Churchill, who was
renowned for his abhorrence of exercise and his love of
food, drink, and tobacco, lived to the age of 90. It is plau-
sible that genetic differences between Fixx and Churchill
were responsible, at least in part, for the paradoxical differ-
ence in their life spans. (Indeed, Jim Fixx’s father had a heart
attack at the age of 35, and died of a second heart attack at
the age of 43.)
Because genes are passed down from parents to off-
spring, diseases tend to “cluster” in families. For example, if
an individual has had a heart attack, the risk that his or her
close relatives, offspring, or siblings will have a heart attack
is two to three times higher than that of the general popula-
tion. Similar levels of increased risk among family members
are seen for colon cancer, breast cancer, prostate cancer, type
2 diabetes mellitus, and many other diseases. This clustering
in families is partly the result of shared non-genetic factors
(e.g., families tend to be similar in terms of their dietary
and exercise habits), and partly the result of shared genes.
As we have seen, populations differ somewhat in their
genetic backgrounds. It is thus possible that genetic differ-
ences could be partly responsible for differences in disease
prevalence. For many disorders caused by genetic changes
in single genes, these differences are readily apparent.
Cystic fibrosis, for example, is seen in about one in 2,500
Europeans, but only in one in 90,000 Asians. Sickle-cell dis-
ease is much more common in individuals of African and
Mediterranean descent than in others, although it is found
in lower frequency in many other populations due to migra-
tion and intermarriage.
These differences in prevalence can be attributed to the
evolutionary factors that influence genetic variation in gen-
eral. Mutation is the ultimate source of all genetic variation.
In some cases, such as hemochromatosis in Europeans and
sickle-cell disease in Africans, the responsible mutations
have arisen within the last few thousand years, helping to
account for a fairly restricted distribution of the disease.
8   •   principles of G enomic M edicine
Natural selection also plays a role in population differences
in some genetic diseases. For sickle-cell disease and related
diseases known as the thalassemias, heterozygotes (those
who carry a single copy of a disease-causing mutation) are
relatively resistant to the malaria parasite. Cystic fibrosis
heterozygotes are resistant to typhoid fever, and hemochro-
matosis heterozygotes absorb iron more readily, perhaps
protecting them against anemia. Also, the process of genetic
drift, which is accentuated in small populations, can raise
the frequencies of disease-causing mutation quickly just by
chance (e.g., Ellis-van Creveld disease, a reduced-stature dis-
order, is unusually common among the Old Order Amish of
Pennsylvania).12 In contrast to the effects of natural selection
and genetic drift, which tend to promote population differ-
ences in disease prevalence, gene flow (the exchange of DNA
among populations) tends to decrease differences among
populations. With the enhanced mobility of populations
worldwide, gene flow is thought to be increasing steadily.
These same factors can affect common diseases such
as cancer, diabetes, hypertension, and heart disease, but
the picture is more complex, because these diseases are
influenced by multiple genetic and non-genetic factors.
Common diseases do vary in frequency among popula-
tions:  hypertension occurs more frequently in African
Americans than European Americans, and type 2 diabetes
mellitus (T2DM) is especially common among Hispanic
and Native American populations.13 Although genes clearly
play a role in causing common diseases, it is less clear that
genetic differences between populations play a signifi-
cant role in causing differences in prevalence rates among
populations. Consider another example:  the Pima Native
American population in the southwestern United States
now has one of the highest known rates of type 2 diabe-
tes in the world. About half of adult Pimas are affected.
Yet this disease was virtually unknown in this population
prior to World War II. Obviously, the Pimas’ genes have not
changed much during the past 50 or so years. Their environ-
ment, however, has changed dramatically with the adoption
of a “Western” high-calorie, high-fat diet, and a decrease in
physical exercise. In this case, it is almost certain that the
rapid increase in type 2 diabetes prevalence has much more
to do with non-genetic than genetic causes.14
But why does a Western diet seem to have a greater
effect on some populations than others? Perhaps differ-
ences in genetic background, interacting with dietary and
other lifestyle changes, help account for this variation. As
additional genes that influence susceptibility to common
diseases are discovered, and as the roles of non-genetic fac-
tors are also taken into account, it is likely that this picture
will become clearer.
 FUNCTIONAL GENOMICS
A N D   P R OT E O M I C S
Functional genomics is a systematic effort to understand the
function of genes and gene products by high-throughput
analysis of gene transcripts in a biological system (cell, tis-
sue, or organism) with the use of automated procedures that
allow scale-up of experiments classically performed with
single genes.15 Functional genomics can be conceptually
divided into gene-driven and phenotype-driven approaches.
Gene-driven approaches rely on genomic information to
identify, clone, and express genes, as well as to characterize
them at the molecular level. Phenotype-driven approaches
rely on phenotypes, either identified from random mutation
screens or associated with naturally occurring gene variants,
such as those responsible for mouse mutants or human dis-
eases, to identify and clone the responsible genes without
prior knowledge of the underlying molecular mechanisms.15
The tools of functional genomics have enabled the devel-
opment of systematic approaches to obtaining basic infor-
mation for most genes in a genome, including when and
where a gene is expressed and what phenotype results if it
is mutated, as well as the identification of the gene product
and the identity of other proteins with which it interacts.16
Functional genomics aspires to answer such questions sys-
tematically for all genes in a genome, in contrast to conven-
tional approaches that address one gene at a time.
Analysis and applications of the rapid accumulation
of highly sophisticated genome and proteome data neces-
sitated development of powerful computational pro-
grams and relevant hardware tools. Storage, retrieval, and
assimilation of enormous amounts of data require fast and
accurate computational skills. Bioinformatics deals with
these requirements within the broad biomedical and bio-
technology sectors. There are several literature and online
resources with detailed descriptions of the role and scope of
bioinformatics.17
A number of biomedical and biotechnology disci-
plines have emerged during the last two decades, all ending
with the suffix -omics. -Omics is derived from ome (Greek,
omoyous), which refers to complete knowledge. The ancient
language Sanskrit has a similar word, ohm, with similar
meaning and expression. A  number of these “omics” have
direct or indirect links to the fundamentals of genome sci-
ence and technology. A number of biological models have
been developed and tested using genomic, transcriptomic,
proteomic, and metabolomic approaches (Figure 1.5).
Systems biology refers to developing and testing biological
models based on -omic sciences.18 The central dogma is the
computational analysis of complex and enormous data at
G enes , G enetics , and H uman G enomics   •   9
The “Omics” Information Universe
Proteomics
Transcriptomics - Identification
Gene Expression - Expression
iRNA - LC-MS driven
Discovery
Biomarkers
Genomics
DNA Sequencing
Metabolomics
Mutations
Polymorphisms
The current ‘omics’ paradigm Small Molecules,
inefficiently delivers Metabolites
Epigenomics
biomarkers productive to
Biopharma
Figure 1.5  The “OMICS” paradigm, showing four major branches.
all biological levels—gene, molecule, cell, tissue, organ, and
whole body.
T R A N S L AT I O N A L G E N O M E R E S E A R C H
AND GENOMIC MEDICINE
The potential of applications of genome science and tech-
nology in medicine and health has led to the emergence
of genomic medicine, a natural outcome of the tremen-
dous progress made in medical genetics and genomics.19
However, final endpoints in genomic medicine will largely
depend upon judicious and efficacious application and
utilization of the diagnostic and therapeutic potential of
genome-based technologies; for example, clinical applica-
tions of microarray technology. This process requires multi-
faceted systematic and analytical research efforts to translate
the basic scientific information into practical and pragmatic
applications following the principles of good medical prac-
tice. There is no disagreement that this translational genome
research is vital for the successful and efficient delivery of
promises made by researchers and physicians behind the
genomic medicine movement.
The process for translational genome research includes
the participation of several researchers drawn from different
disciplines. The multidisciplinary model for translational
genome research is widely accepted, and includes several key
elements. Informatics and computational networks remain
the central dogma for translational genomics research and
systems biology (Figure 1.6).20 A framework for the con-
tinuum of multidisciplinary translation research is recom-
mended to utilize previous research outcomes in genomics
and related areas of health and prevention.21 The whole
process includes four phases and revolves around the devel-
opment of evidence-based guidelines. Phase 1 translation
(T1) research seeks to move a basic genome-based discovery
into a candidate health application, such as a genetic test or
intervention. Phase 2 translation (T2) research assesses the
value of genomic applications for health practice, leading
to the development of evidence-based guidelines. Phase 3
translation (T3) research attempts to move evidence-based
guidelines into health practice through delivery, dissemi-
nation, and diffusion research. Phase 4 translation (T4)
research seeks to evaluate the “real world” health outcomes
of a genomic application in practice. It is important to
appreciate that the whole process of translation research
leading to evidence-based guidelines is a dynamic one, with
considerable overlap between the different stages. The pro-
cess should be able to accommodate new knowledge that
will inevitably arrive during translation research.
The role of translational genome research, including
that of clinical trials, is crucial in developing evidence-based
good-practice guidelines.22 The aim should be to obtain
vital genetic and genomic information, including laboratory
material for research, from the patient, family, and commu-
nity, and then use this scientific data and information for
clarification and ratification. The outcomes of translational
gene research should be valid and deliverable in the clinic
for diagnostic and therapeutic applications.
HUM AN GENOMICS FOR SOCIO-
E C O N O M I C D E VE L O PM E N T
During the last decade, rapid progress has been made in har-
nessing the huge potential of genome science and technology
Biological System Analysis
And Model Formation
“Dry” Experiments
(Simulation)
Data and Hypothesis
Driven Modeling
Data Analysis
synthesis
Experimental
Data Analysis
Figure 1.6  Informatics as the central dogma for systems biology and genome sciences.
1 0   •   principles of G enomic M edicine
for its economic and health benefits globally, in particular in
less- and least-developed nations.23 Apart from the World
Health Organization (WHO), other international and
national institutions engaged in this endeavor include the
Human Genome Organization (HUGO), Organization
for Economic Cooperation and Development (OECD),
the McLaughlin-Rotman Center for Global Health (The
University of Toronto, Canada), the Mexican Health
Foundation, the Beijing Genomics Institute, the Department
of Science and Technology (Government of India), and many
more. All these institutions are focused on supporting and
exploiting the huge potential of genomic technologies and
related bioinformatics developments on the global economy
and on health.24 The impact of genome sciences and tech-
nologies will manifest in the following wide-ranging areas:
• Personalized medicine and health approaches that will
help people and societies shift the focus from “sick-care”
to “well-care and prevention.”
• Biotechnology methods to produce environmentally
clean and efficient fuel and chemicals to accelerate tran-
sition from petroleum-based economies
• Genome-driven plant- and crop-growing methods for
producing affordable food for less- and least-developed
economies
• Promoting genomic science and technology in animal
breeding and livestock improvement
• Supporting genome research for new drug discovery and
drug development for enhancing pharmaceutical efficacy
Prediction And
Hypothesis Refinement
Experimental
Model Design
Experiment Data Acquisition
Genome
Transcriptome
Proteome
Interactome
Metabolome
 Free ebooks ==> www.Ebook777.com
• Applications of genomic biotechnologies in the study
and monitoring of environmental health
C O N C LU S I O N S
Developments in genetics and the subsequent sequencing
of the human and other genomes have provided us with an
opportunity to review the role of genes and genomes in all
aspects of health and disease. Human health, including cau-
sation of disease, is not exclusively dependent on the human
genes and genome. Evolutionary links with other genomes
and ecologically relevant and beneficial parts of other
genomes play crucial roles in the maintenance of human
health and, to some extent, in morbidity and mortality.
Understanding genomes of microbes, parasites, animals,
plants, and crops is an acknowledged priority of current
biomedical and biotechnology research.
Conventionally, the causation of human disease
includes malformations, trauma, infection, immune dys-
function, metabolic abnormality, malignancy, and degen-
erative conditions associated with aging. Genetic factors
have long been recognized in all of these disease groups.
The traditional genetic categories of diseases include chro-
mosomal disorders, single-gene or Mendelian diseases,
and several forms of multifactorial/polygenic conditions.
In addition, somatic genetic changes and mutations of the
mitochondrial genome probably account for a small, albeit
important, number of diseases. These groups of disorders
are well recognized and have an established place in the
classification of human disease. Recent developments in
genome research have provided vast data indicating differ-
ent genomic mechanisms to explain complex pathogenesis
in some disorders. The spectrum of these disorders is wide
and includes both acute and chronic medical and surgical
diseases. Perhaps it is reasonable to identify these disorders
on the basis of underlying molecular pathology, including
genomic imprinting, genomic rearrangements, and gene–
environment interactions involving multiple genes and
genomic polymorphisms.
This chapter has reviewed the genetic and genomic
approaches to human health and disease. The genomic
approaches to understanding and managing human disease
are rapidly being incorporated in the practice of clinical
medicine. In addition, applications of genome science and
technology are also reforming biotechnologies in a number
of industries, including pharmaceutical, agricultural, and
ecological bioengineering. The enormous impact of genome
sciences and technologies on the economy of the developing
G enes , G enetics , and H uman G enomics   •   1 1
www.Ebook777.com
world will be judged on applications in a number of areas,
including bio-fuels, accelerated breeding of crops and live-
stock, personalized health products, pharmaceutical effi-
cacy, and genomic monitoring of environmental health.
F U RT H E R R E A D I N G
Readers who wish to enhance their knowledge or seek more
information are advised to consult the following books.
Turnpenny and Ellard, Emery’s Elements of Medical Genetics,
Churchill Livingstone, 2011.25
Genomics and World Health: Report of the Advisory Committee
on Health Research 2002, World Health Organization.26
Harper, A Short History of Medical Genetics, Oxford University
Press, New York, 2008.27
REFERENCES
1. Temple LK, et  al. Defining disease in the genomics era. Science.
2001;293(5531):807–808.
2. Lay JO Jr., et  al. Problems with the “omics.” TrAC.
2006;25(11):1046–1056.
3. Feero WG, et al. Genomic medicine—an updated primer. N Engl J
Med. 2010;362(21):2001–2011.
4. Kumar D, Weatherall DJ. Genomics Clin Med. 2008:  Oxford
University Press, Oxford.
5. Fisher RA. The Genetical Theory of Natural Selection:  A  Complete
Variorum Edition. 1999: Oxford University Press, Oxford.
6. Collins FS, Morgan M, Patrinos A. The Human Genome Project: les-
sons from large-scale biology. Science. 2003;300(5617):286–290.
7. Collins FS, McKusick VA. Implications of the Human Genome
Project for medical science. JAMA. 2001;285(5):540–544.
8. Conrad DF, et  al. Origins and functional impact of copy number
variation in the human genome. Nature. 2009;464(7289):704–712.
9. Pinto D, et al. Functional impact of global rare copy number variation
in autism spectrum disorders. Nature. 2010;466(7304):368–372.
10. Freeman JL, et al. Copy number variation: new insights in genome
diversity. Genome Res. 2006;16(8):949–961.
11. Akey JM, et al. Interrogating a high-density SNP map for signatures
of natural selection. Genome Res. 2002;12(12):1805–1814.
12. Puffenberger E. Genetic heritage of the Old Order Mennonites of
southeastern Pennsylvania. In Am J Med Genet C: Sem Med Genet.
2003: Wiley Online Library.
13. Harris MI, et  al. Is the risk of diabetic retinopathy greater in
non-Hispanic blacks and Mexican Americans than in non-Hispanic
whites with type 2 diabetes? A US population study. Diabetes Care.
1998;21(8):1230–1235.
14. Harris MI, et al. Racial and ethnic differences in glycemic control of
adults with type 2 diabetes. Diabetes Care. 1999;22(3):403–408.
15. Yaspo, M.-L. Taking a functional genomics approach in molecular
medicine. Trends MolMed. 2001;7(11):494–501.
16. Steinmetz LM, Davis RW. Maximizing the potential of functional
genomics. Nature Rev Genet. 2004;5(3):190–201.
17. Rehm B. Bioinformatic tools for DNA/protein sequence analy-
sis, functional assignment of genes and protein classification. Appl
Microbiol Biotech. 2001;57(5–6):579–592.
18. Gehlenborg N, et al. Visualization of omics data for systems biology.
Nature Methods. 2010;7:S56–S68.
 19. Burke W, et al. Translational genomics: seeking a shared vision of
benefit. Am J Bioethics. 2008;8(3):54–56.
20. Karp PD, et  al. Pathway Tools version 13.0:  integrated software
for pathway/genome informatics and systems biology. Briefings in
Bioinformatics. 2010;11(1):40–79.
21. Khoury MJ, et  al. Population sciences, translational research, and
the opportunities and challenges for genomics to reduce the burden
of cancer in the 21st century. Cancer Epidemiol Biomarkers Prev.
2011;20(10):2105–2114.
22. Kumar D. Clinical medicine in the genome era:  an introduction.
Genomics Clin Med. 2008(53):145.
1 2   •   principles of G enomic M edicine
23. Singer PA, Daar AS. Harnessing genomics and biotechnology to
improve global health equity. Science. 2001;294(5540):87–89.
24. Kumar D. Genomics and Health in the Developing World.
2012: Oxford University Press, New York.
25. Turnpenny PD, Ellard S. Emery’s Elements of Medical Genetics.
2011: Churchill Livingstone, Edinburgh.
26. WHO Advisory Committee on Health Research. Genomics and
World Health: Report of the Advisory Committee on Health Research
2002. 2002: World Health Organization, Geneva, Switzerland.
27. Harper PS, A Short History of Medical Genetics. 2008:  Oxford
University Press, New York.
 2.
THE HUMAN GENOME—STRUCTURE
AND ORGANIZATION
Andrew P. Read

INTRODUCTION chromosomes to a limited digestion with trypsin and then

staining with Giemsa reagent. The bands so produced are
Humans have two genomes, nuclear and mitochondrial. called G-bands. A  standard nomenclature of the G-bands
Normal diploid cells contain two copies of the nuclear is used to identify chromosomal locations. Locations are
genome and a much larger but variable number of cop- described by the chromosome number followed by p (short
ies of the mitochondrial genome. The nuclear genome is arm) or q (long arm), then the major band, counting out-
approximately 2 x 105 times larger than the mitochondrial wards from the centromere, and sub-band (Figure 2.2). For
genome (3 x 109 vs. 16,569 bp), and contains more than example, “17q24” means chromosome 17, long arm, main
1,500 times the number of protein-coding genes (approxi- band 2, sub-band 4. The pronunciation is “17q two four,”
mately 21,000 vs. 13), including many required for mito- not “twenty-four.”
chondrial functions. Understandably, the phrase “the G-banding reflects small differences in the guanine-
human genome” normally refers to the nuclear genome, cytosine (GC) content of the DNA at different chromo-
and it is used in that sense here. The structure and orga- somal locations. The genome-wide GC content is 41%.
nization of the mitochondrial genome are described sepa- Dark staining G-bands contain DNA with a slightly lower
rately below. GC content (average 37%), whereas the light bands average
45%.1 These differences correlate with systematic variations
in the distribution of genes and various classes of repeat
THE HUM AN GENOME AS elements across the genome (see below). It is intriguing to
CHROMOSOMES speculate why such regularities across 1–5-Mb regions have
persisted through evolution.
The nuclear genome is most readily seen when it is tightly
packaged at the metaphase stage of mitotic cell division
(Figure 2.1). It is important to remember that this rep-
resents a highly abnormal state of the genome. At this
stage in the life of the cell, the DNA has already been
replicated, so that each chromosome consists of two iden-
tical sister chromatids, joined at the centromere. Thus, a 1 2 3 4 5
mitotic cell contains four copies of the nuclear genome,
and the tightly packaged DNA is largely inactive. In a
more typical cell, each chromosome consists of a single 6 7 8 9 10 11 12
highly extended chromatid, with regions of active and
inactive DNA. 13 14 15 16 17 18
The 24 different chromosomes (22 autosomes, one X
and one Y chromosome) can be recognized by their size,
the position of the centromere, and the pattern of dark and 19 20 21 22 X Y

light bands produced by laboratory manipulation. Most Figure 2.1 

Normal male karyotype (46,XY), G-banded. (Courtesy of
commonly, this manipulation consists of subjecting the Dr. Lorraine Gaunt, Manchester, United Kingdom).

13
 3
2 2
p
2 2
1 1 1 1
1 1
1 1
1 1
1 1
2
2 2
q 2 2
3 2

4 3 3 3

1 2 3 4 5 6

2
2 2
p 1
1 1 1
1 1

1 1 1 1
1 1
2 2
q
2 2 2
3 3 2

7 8 9 10 11 12

1 1 1
p 1 1 1
1 1 1 1
1 1
2 2 2
q 2 2
2
3 3

13 14 15 16 17 18

p 1 1 1 1
1 1
1 1
q 1 1 2 1 1

19 20 21 22 Y X

Figure 2.2  Standard cytogenetic nomenclature of human G-banded chromosomes.

Chromosomes have two general functions. On a large
scale, they are necessary to ensure accurate partitioning of
the replicated DNA between daughter cells at mitosis, and
to allow the more complicated events of meiosis. For this
purpose, the gene content and DNA sequence are irrelevant.
All that matters is that each chromosome should be a stable
DNA package with a single centromere and protected ends.
On the micro scale, the way the naked DNA is packaged by
a variety of proteins and small RNA molecules is a crucial
determinant of its function, as will be described below.

1 4   •   P rincip l es o f G enomic M edicine

 T H E S T RU C T U R E O F C H RO MO S O M E S
The stunning success of the Human Genome Project has
had one small negative effect. By focusing attention on
DNA sequences written as extended lines of A, G, C, and
T letters, it encourages the unwary seeker to forget that our
genome functions, not as extended naked DNA, but as
highly folded coils and loops of chromatin, a DNA–protein
complex. Looping brings together DNA sequences that
appear widely separated in the linear DNA sequence, allow-
ing them and their bound proteins to interact.
Watson and Crick showed that successive base pairs
along the DNA double helix are 0.34 nm apart. This allows
us to calculate that, at metaphase, 1 meter (3 × 109 bp) of
DNA is packed into 23 chromosomes with a total length
of approximately 100  μm. During interphase, the DNA
is more extended, but still highly organized, with loops
of packaged DNA occupying defined territories whose
location (central vs. peripheral) and proximity to one
another are believed to be important factors in control-
ling gene expression. Proteins are the main packaging
agents of chromatin, with some involvement of small
RNA molecules.
The basic level of packaging is into a string of nucleo-
somes. One hundred and forty-seven base pairs of naked
DNA wrap around an octamer of histone proteins (two
molecules each of H2A, H2B, H3, and H4) to form a
nucleosome. Successive nucleosomes are separated by
10–80 base pairs of spacer DNA. Nucleosomes are rela-
tively stable structures that nevertheless must permit poly-
merases and other progressive enzymes to move along a
DNA strand. Adenosine triphosphate (ATP)-powered
chromatin remodeling complexes of proteins assist in this
process, while the DNA of active gene-regulatory sequences
is often relatively devoid of nucleosomes.
Adjacent nucleosomes are linked by H1 histones. They
can associate tightly or loosely to form “open” or “closed”
chromatin. Large tracts of closed, genetically inactive chro-
matin are called heterochromatin, whereas open, poten-
tially active chromatin is euchromatin. This level of packing
is determined by chemical modification of the histones and
DNA. Mammalian DNA is modified by methylation of
the 5-position of cytosine bases. Histones are subject to a
great variety of modifications, including methylation, acet-
ylation, phosphorylation, and ubiquitination of specific
residues. These modifications of the DNA and histones
constitute the epigenetic marks discussed below. Different
combinations of epigenetic marks produce a repertoire of
“chromatin flavors” that define different functional regions
of DNA.
T he H uman G enome —S tructure and O rgani z ation   •   1 5
C E N T RO M E R E S
A functioning chromosome must have one, and only one,
centromere. During cell division, spindle fibers pull chroma-
tids apart, and centromeres nucleate the formation of kinet-
ochores, the structures to which the spindle fibers attach.
Thus chromosomes without centromeres cannot move to the
daughter cell nuclei, while chromosomes with two or more
centromeres risk being pulled apart by conflicting fibers.
Human centromeres are marked by megabase-sized tracts of
repetitive DNA consisting of tandem arrays of highly similar
171 bp α-satellite sequences.2 However, the defining feature
of centromeres is not the DNA sequence but a modification
of the nucleosomes. Histone H3 is replaced by a variant,
Centromere protein A (CENPA), and this is both necessary
and sufficient to form a functional centromere. Sequencing
long, highly repetitive tracts of DNA is extraordinarily dif-
ficult, because it can be nearly impossible to work out the
correct assembly of the individual nearly identical clones.
Because of this difficulty, and because they are believed not
to contain any genes, centromeres are not included in the
current human genome sequence.
TELOMERES
The ends of chromosomes require a special structure for
two reasons:
• Because of the detailed enzymology of DNA replication,
it is not possible to replicate the extreme 3′-end of a DNA
strand. Each round of replication shortens a chromosome
by 50–100 nucleotides (nt). Within the life of a multicel-
lular organism, that would be tolerable, provided no vital
gene is located near the end of any chromosome. However,
over evolutionary time, it would be disastrous. Some spe-
cial mechanism is needed to restore chromosome ends, at
least once every generation of the whole organism.
• As part of their mechanism for repairing DNA damage,
cells check for loose DNA ends and join together any
that they find. Chromosome ends need protection from
this mechanism.
Telomeres of human chromosomes carry long arrays of
tandemly repeated (TTAGGG)n sequence. These contract
during successive rounds of somatic cell replication. Germ
cells have a special RNA–protein complex, telomerase, that
is able to restore telomeres to full length by non-templated
addition of the telomeric repeat.3 Specific proteins bound
to telomeres protect the DNA end, which is formed into
a non-standard looped structure that does not trigger the
DNA damage response.
 Free ebooks ==> www.Ebook777.com

THE HUM AN GENOME AS GENES P ROT E I N- C O D I N G G E N E S

These are the classical genes that are transcribed by RNA

The “finished” human genome sequence was published in
polymerase II to make messenger RNA (mRNA), which in
2004.4 Table 2.1 shows the current best estimates of the size
turn is translated by ribosomes according to the genetic code.
and gene content of each chromosome. These figures are for
As usual in higher organisms, the coding sequence of most
the Human Reference Genome. They do not correspond
genes is split into segments (exons) separated by non-coding
precisely to the genome of any actual individual, because
introns. Thus the structure of a typical human protein-coding
the genomes of healthy normal individuals vary somewhat
gene would be as shown in Figure 2.3. Table 2.2 shows aver-
in chromosome sizes and numbers of genes, as described
age numbers and sizes of exons and introns, but these vary
below. Nor is the Reference Genome in any sense an “ideal”
very widely between genes with no obvious rationale. Some
human genome. It is simply an arbitrary and reasonably
genes have no introns, while at the other extreme, the gene
typical reference point for comparing human genome
encoding the muscle protein titin has over 300 exons.
sequences. Uncertainties in the figures relate primarily to
RNA polymerase produces a primary transcript that
the highly repetitive DNA of centromeres and telomeres
includes all the exons and introns, but this must undergo
and to the number of RNA genes, which are difficult to
several stages of processing to produce the mature mRNA
identify from sequence data.

Table 2.1  DNA AND GENE CONTENT OF THE REFERENCE HUMAN GENOME (GRCH37, FEB. 2009)

CHROMOSOME LENGTH (BP) PROTEIN-CODING GENES PSEUDOGENES RNA GENES GENES PER
(KNOWN + NOVEL) MB
1 249,250,621 2037 1131 672 8.17
2 243,199,373 1259 947 526 5.18
3 198,022,430 1066 719 430 5.38
4 191,154,276 758 698 363 3.97
5 180,915,260 874 675 343 4.83
6 171,115,067 1042 726 358 6.09
7 159,138,663 907 800 350 5.70
8 146,364,022 731 568 288 4.99
9 141,213,431 803 714 260 5.69
10 135,534,747 762 498 295 5.62
11 135,006,516 1320 774 290 9.78
12 133,851,895 1051 582 336 7.85
13 115,169,878 326 323 173 2.83
14 107,349,540 652 472 310 6.07
15 102,531,392 605 471 329 5.90
16 90,354,753 867 384 229 9.60
17 81,195,210 1197 255 273 14.74
18 78,077,248 277 56 157 3.55
19 59,128,983 1418 180 198 23.98
20 63,025,520 546 213 189 8.66
21 48,129,895 233 150 69 4.84
22 51,304,566 455 308 105 8.87
X 155,270,560 836 780 351 5.38
Y 59,373,566 53 327 44 0.89
Total 3,286,906,385 21,099 15,520 11,960 6.42

The overall totals are derived from a slightly different analysis from the individual chromosome totals, so the figures do not exactly add up.
SOURCE: Data from Ensemble Release 66, March 20, 2012.

1 6   •   P rincip l es o f G enomic M edicine

www.Ebook777.com
 Primary transcript

ATG TAA
5’UT 3’UT
5’ Intron 1 Intron 2 Intron 3 3’

Exon 1 Exon 2 Exon 3 Exon 4

Figure 2.3 
Structure of a typical gene. This gene has four exons. The positions of the translation start (ATG in DNA, AUG in the mRNA) and stop
(TAA in DNA, UAA in the mRNA) signals are shown. The part of exon 1 upstream of the translation start is the 5′ untranslated region (5′UT),
and the part of exon 4 downstream of the translation stop is the 3′ untranslated region (3′UT). The primary transcript includes all the exons and
introns; in the mature mRNA, the introns are removed and the exons spliced together. In a real gene, the introns would probably be considerably
larger than the exons.

Table 2.2  HUMAN GENE STATISTICS downstream of the 5′-end (the genome-wide aver-
age is 300 nt). The 5′ untranslated region binds the
Average exon number 9
ribosomes, which slide along until they encounter
Average exon size 145 bpa
an AUG codon embedded in a suitable context (a
Average intron size 3365 bp Kozak sequence, consensus GCCRCCAUGG, where
Average size of transcription unit 27,000 bp R means a purine [A or G] and the initiator codon
Average size of mRNA 2600 bp is underlined). At the 3′ end, the stop codon (UAA,
a
This is the average size of internal exons. The 3′ exon of a gene is often consider- UAG, or UGA) is usually several hundred bases or even
ably larger than the internal exons. kilobases upstream of the physical end of the mRNA.
SOURCE: Data from International Human Genome Sequencing Consortium 3′ untranslated sequences contain important elements
(2001).1
that regulate the activity and turnover of mRNAs. Note
that unlike introns and other regulatory elements (see
that is exported to the cytoplasm. A large multimolecular below), the 5′ and 3′ untranslated regions form part of
machine, the spliceosome, cuts out the introns and splices the mature mRNA.
together the exons. A special nucleotide structure, the cap,
• Introns—introns are usually considerably larger than the
is added to the 5′ end of the RNA, and a string of a few hun-
dred A nucleotides (the polyA tail) is attached to the 3′ end. exons they separate.
• Gene regulatory sequences—one reason for the increase
THE EXTENDED GENE in genome size as we move up the evolutionary scale is
the greater complexity and sophistication of gene regula-
Human genes are much larger than their coding sequences. tion in higher organisms. Table 2.3 lists some of the play-
If we define a gene as a functional unit of DNA, extra ele- ers in human gene regulation.
ments include:
The ENCODE consortium of laboratories5 is pur-
• 5′ and 3′ untranslated regions—the AUG ini- suing a large-scale effort to define the nature and action
tiation codon of an mRNA is located some distance of these regulatory elements across the entire human

Table 2.3  REGULATORY ELEMENTS IN THE HUMAN GENOME

ELEMENT COMMENTS
Promoter The 500 bp of DNA immediately upstream (5′) of a gene usually includes a number of different motifs that
attract and bind the transcription factor proteins that recruit and assemble an RNA polymerase complex.
Enhancer A sequence that increases transcription of the gene by binding proteins that help attach or activate the
RNA polymerase. Enhancers are defined by histone modifications closely resembling those of promoters.
They may be upstream or downstream of the promoter, and may be a considerable distance away from
the gene they regulate. Although present in every cell, they usually act only in specific tissues, presumably
because the protein(s) they bind are present only in that tissue.
Silencer Similar to an enhancer, but with a negative action.
Insulator (boundary element) Sequences that limit the extent of influence of a regulatory element. Insulators have to be located between
the regulatory element and the DNA that they are protecting. They act by preventing the spread of chro-
matin modifications along the DNA.

T he H uman G enome —S tructure and O rgani z ation   •   1 7

genome and in two model organisms, the Drosophila fly
and Caenorhabditis nematode worm. Some genes that
play critical roles in development are located in “gene des-
erts” on chromosomes (defined as regions of 500 kb or
greater containing no genes). Their regulation evidently
requires complex arrays of enhancers in a large region free
of competing or interfering transcription units. This hints
at a function for some of the large amount of intergenic
non-coding DNA.
G E N E S E N C O D I N G F U N C T I O NA L R NA S
Not all RNA molecules in a cell are mRNA. Ribosomal
RNA and transfer RNA have long been known, but there
are many other non-coding RNAs (ncRNAs).6 These can
be divided into “classical” ncRNAs and long intergenic
ncRNAs (lincRNAs). The classical ncRNAs are small mole-
cules, typically 16–30 nt, derived by processing much longer
precursors. There has been an explosion in our knowledge
of the numbers and classes of these molecules, and this is
still a very active research area. The main well-established
classes are:
• snRNAs (small nuclear RNAs) form part of the spliceo-
somal machinery.
• snoRNAs (small nucleolar RNAs) act as
sequence-specific guides for enzymes that chemically
modify specific bases in ribosomal and other RNAs.
• miRNAs (microRNAs) control translation of many
mRNAs by binding to sequences in the 3′ untranslated
region.
• piRNAs (piwi-associated RNAs) act in gametes to
ensure stability of the genome. There appear to be many
thousands of piRNA genes, grouped in around 100
clusters.
Table 2.4 lists the numbers of genes encoding these mol-
ecules—but these are subject to major revision because it
is very difficult to identify functional ncRNAs and distin-
guish them from the large number of nonfunctional vari-
ants present in the genome.
Long ncRNAs closely resemble messenger RNAs. They
are often spliced, capped, and polyadenylated just like
mRNAs. Some are transcribed from the antisense strands
of genes; others from intergenic sequences. Some of them
have well-documented roles in controlling expression of
their cognate genes,7,2 but for the majority, the function, if
any, is unknown.9
1 8   •   P rincip l es o f G enomic M edicine
Table 2.4  RNA GENES IN THE HUMAN GENOME
RNA SPECIES NUMBER OF FUNCTIONAL GENES
Ribosomal RNA 150–200
Transfer RNA 496
snRNA 91
snoRNA 375
miRNA 1733
piRNA 114 clusters
SOURCE: Griffiths-Jones (2007),34 miRNAblog.com, piRNAbank.ilab.ac.in.
P S EU D O G E N E S
Researchers analyzing genome sequences to identify genes
often come across sequences that at first sight appear to be
functional genes, but that on closer examination are seen
to harbor changes that make them nonfunctional. These
are pseudogenes. They are often the result of evolutionary
duplication of a gene: when there are two copies, one is free
to acquire mutations without affecting the function of the
other. Large gene families such as the olfactory receptor or
piRNA families are especially rich in pseudogenes.
THE REST OF OUR DNA
The genetic code uses three nucleotides to encode each
amino acid. Thus, we have enough DNA to encode 109
amino acids or several million proteins—vastly more than
the actual number of genes. The gene counts in Table 2.1 are
provisional, but no future adjustments can alter the fact that
very little of our genome codes for protein—probably little
over 1%. Around another 27% comprises the non-coding
portions of genes (the introns and regulatory elements),
some of the rest codes for functional RNAs, and some is
used to form the centromeres and telomeres of chromo-
somes. When all these have been taken into consideration,
they leave at least half the human genome unaccounted
for. This DNA can be classified into unique sequence and
repetitive DNA.
About 5% of all the unique sequence DNA in the human
genome is conserved across species. This implies that it has
some sequence-dependent function such that natural selec-
tion removes the variants that must from time to time arise.
Since only 1–1.5% of the conserved sequence is coding, the
rest presumably has some regulatory function. Bejerano
et al.10 identified 481 “ultraconserved” segments, more than
200 bp long, that are 100% conserved between human, rat,
and mouse. Most were also highly conserved in the chicken
and puffer fish. They included 100% conserved sequences
of 770 and 732 bp on the X chromosome in introns of
the POLA gene, and a nearby 1046 bp region with only a
single nucleotide change. These highly conserved sequences
are supposed to contain complex arrays of enhancers—
although they are remarkable because no known cellular
process requires a sequence as long as a kilobase to be totally
conserved.
R E P ET I T I VE D NA
At least half the human genome comprises sequences that
are present in more than one copy. Repetitive DNA can
be classified in many different ways; one basic distinc-
tion is between tandem and interspersed repeats. In tan-
dem repeats, the same sequence occurs a number of times,
one after another, at a particular chromosomal location.
Interspersed repeats are present in a number of copies at
different locations in the genome. The great majority of
repetitive DNA is non-coding. Exceptions to this include
the tandemly repeated arrays of genes encoding ribosomal
RNA, some small RNAs and olfactory receptor proteins,
and a few interspersed or clustered multicopy genes such as
the histone or ubiquitin gene families.
TA N D E M R E P E ATS
The repeat unit in tandem repeats varies from a single
nucleotide (poly[A]‌runs are particularly frequent) through
short units such as the hexanucleotide telomeric repeat, to
long units like the 171 bp α-satellite present at chromo-
somal centromeres. As mentioned in the following section,
tandem repeat arrays are often polymorphic as regards the
number of repeat units.
•
Satellite DNA was so named because in early experi-
ments using density gradient centrifugation of cellular
DNA, it formed a satellite to the main peak of bulk
DNA. There are several families of satellite DNA (α, β,
satellite 1, etc.). All are mainly located at centromeres
and heterochromatic regions of chromosomes, and
comprise large arrays (up to several megabase pairs) of
tandem repeats.
• Minisatellite DNA comprises tandem arrays, typically
1–20 kb long, of repeating units of a dozen or so nucleo-
tides. Although distributed through the genome, they
are particularly found near chromosome ends, proximal
to the telomeric repeats. Minisatellites with the same
repeat unit are often present at a number of different
chromosomal locations—these are the basis of the origi-
nal DNA fingerprinting technique of Jeffreys.11
T he H uman G enome —S tructure and O rgani z ation   •   1 9
• Microsatellites are short arrays, typically less than 100
bp, with repeat units mostly 1–5 bp. Approximately 3%
of our genome consists of microsatellites, distributed
randomly across all chromosomes.
I N T E R S P E R S E D R E P E ATS
Most interspersed repeats, comprising almost half the entire
human genome, belong to four classes of semi-autonomous
elements that have, or once had, the ability to propagate
themselves within a genome.1 These transposon-derived
repeats can be seen as a sort of intracellular parasite.
Whether they have any function useful to the host cell is
much debated. The four classes are:
• Long interspersed nuclear elements (LINEs). There are
several families, of which the L1 repeats are the most
numerous. Full-length LINEs are 6.5 kb long and have
the ability to be transcribed and to make DNA copies of
the transcript, which can insert at new locations in the
genome (retrotransposition). The human genome con-
tains an estimated 850,000 LINEs, comprising 21% of
our total DNA. Most L1 copies are truncated—the aver-
age LINE is only 900 bp long. Probably only 80–100
copies are fully functional, and these are normally held
in a non-transcribed repressed state. There is evidence
that LINEs may be remobilized in the developing brain,
and this may contribute to the genetic diversity of
neurons.12
• Short interspersed nuclear elements (SINEs). These
are 300 bp long. They encode no proteins, but can be
propagated by the LINE enzymes. Human SINEs can
be grouped into the Alu, MIR, and Ther2 families.
Approximately 1.5 million complete or partial copies are
present, comprising 13% of our genome.
• Long terminal repeat retroposons are closely related to
retroviruses, though lacking the envelope gene necessary
for extracellular existence. Some 450,000 copies make up
8% of our genome. Again, many copies are defective.
• DNA transposons are virus-like entities that propagate
by a cut-and-paste mechanism. At least seven major fam-
ilies are distinguishable in the human genome, totaling
approximately 300,000 copies and 3% of the genome.
The distribution of LINEs and SINEs in the human
genome is interesting. LINEs are four-fold more frequent in
the AT-rich DNA that forms the gene-poor dark G-bands
on chromosomes, whereas SINEs show the opposite distri-
bution. It is difficult to explain this without ascribing some
beneficial function to the SINEs. Transposable elements are
dangerous, disrupting the host sequence, and one would
expect there to be selection against elements’ transposing
into gene-rich regions.
S EG M E N TA L D U P L I C AT I O NS
Approximately 5% of the human genome consists of long
sequences, 1–400 kb in size, that are present in two or more
nearly identical copies.13 Such duplicons, with more than
90% sequence identity occur on every chromosome, mainly
in the pericentromeric and subtelomeric regions, but also
at other positions (Figure 2.4). The copies may be repeated
tandemly or spaced apart on the same chromosome or on
different chromosomes. Mispaired recombination between
syntenic duplicons that lie close together but not adjacent is
a frequent cause of pathogenic chromosome abnormalities;
common copy-number polymorphisms (see below) are also
associated with segmental duplications.
T H E M I TO C H O N D R I A L G E N O M E
The mitochondrial genome is very different from the nuclear
genome (Figure 2.5; Table 2.5). In many respects, it has
more in common with bacterial genomes than the eukary-
otic nuclear genome. This is consistent with the idea that
mitochondria originated as endosymbiotic bacteria within
some ancestral eukaryotic cell. If this theory is correct, then
over the years, the mitochondria have gradually transferred
more and more of their functions to the nucleus. The great
majority of mitochondrial proteins are now encoded by
nuclear genes. Cells contain many mitochondria (typically
100–1000; maybe 100,000 in an oocyte), so mitochon-
drial DNA (mtDNA) might be formally classified among
the repetitive DNA in a cell. Although the mitochondrial
genome is very small compared to its nuclear counterpart,
1 3
9 11
Figure 2.4 
Examples of segmental duplications in the human genome. The central line represents chromosome 7, with the small rectangle representing
the centromere. Duplicated regions at least 10 kb long and 98% identical are connected by lines, either within chromosome 7 or linked to other
chromosomes (nos. 1, 3, 5, 6, 9, 11, 16, and 19). Adapted with permission from http://humanparalogy.gs.washington.edu/build37/ figures/blowup/starburst.S10000.P0.98/chr7_10kb_98perc.pdf.
2 0   •   P rincip l es o f G enomic M edicine
because there are many copies, mtDNA often makes up 1%
or so of total cellular DNA.
As in bacteria, the mitochondrial genome is circular and
closely packed with genes. There are no introns and little
intergenic non-coding DNA. Some genes even overlap. In the
nuclear genome, it is not uncommon for genes on opposite
strands to overlap—Nussbaum et al.14 recorded 59 such pairs
on chromosome 18—but in this case, genes on the same strand
overlap, using the same template but read in different reading
frames. Twenty-four of the 37 genes specify functional RNAs
(two ribosomal RNAs and 22 tRNAs); the other 13 genes
encode components of the electron transport pathway.
A short segment of the mitochondrial genome is
triple-stranded. This D-loop (displacement loop) is pro-
duced by replication forks overlapping as they travel in
opposite directions around the circular DNA. The D-loop
contains the only significant amount of non-coding DNA
in the mitochondrial genome. Perhaps because of this,
it is the location of many of the DNA polymorphisms
that are such useful tools for anthropologists research-
ing the origins of human populations. Because there is no
recombination among mtDNAs, complete haplotypes of
polymorphisms are transmitted through the generations,
modified only by recurrent mutation. This makes mtDNA
a highly informative marker of ancestry, at least along the
maternal line.
mtDNA replication and transcription use nuclear-
encoded polymerases. Transcription proceeds in both
directions round the circle. The initial products are two
large multicistronic RNAs, which are subsequently cleaved
to make the individual mRNAs. All the protein compo-
nents of the translation machinery are nuclear-encoded,
but the tRNAs are exclusively mitochondrially encoded,
and these use a coding scheme slightly different from the
otherwise universal code. There are four stop codons—
UAG, UAA, AGG, and AGA: UGA encodes tryptophan,
and AUA specifies isoleucine, rather than arginine as
5 6
16 19
 Free ebooks ==> www.Ebook777.com

OH PH

D-loop

NA Phe 16S
7S D Val
Thr 23S
CYB
Pro Leu
H
STRAND PL ND1
Glu lie
ND6 Gln f-Met

ND2
ND5 Ala
OL Asn
Cys Trp
Tyr
Leu
Ser
His CO1
Ser
ND4

Asp
L ND4L
STRAND CO2
ND3
CO3 Lys
Arg ATPase 8
Gly
ATPase 6

Figure 2.5 
The mitochondrial genome. The heavy (H) and light (L) strands of the circular 16,659 bp double helix are shown. Protein-coding genes are
shaded; transfer RNAs genes are shown as short lines with the name of the amino acid. There are no introns. OR, OL, and the heavy arrows indicate
the origins and directions of replication of the two strands. PR, PL, and the light arrows show the promoters and the direction of transcription
of the two multicistronic transcripts that are subsequently cleaved into individual mRNAs. Adapted with permission from Figure 9.1 of Strachan T, Read A (2004), Human
Molecular Genetics (3rd ed.), Garland.

normally. Presumably, with only 13 protein-coding genes,
the mitochondrial system could tolerate mutations that
modified the coding scheme in a way the main genome
could not.
Mutations in mtDNA are important causes of disease,
and perhaps also of aging.15 Phenotypes caused by variation in
mtDNA are transmitted exclusively down the maternal line
(matrilineal inheritance), but most genetic diseases where
there is mitochondrial dysfunction are caused by mutations
in nuclear-encoded genes, and so follow normal Mendelian
patterns. As cells contain many copies of the mitochondrial
genome, they can be heteroplasmic, containing a mix of dif-
ferent sequences. Unlike mosaicism for nuclear variants, het-
eroplasmy can be transmitted by a mother to her children.

Table 2.5  COMPARISON OF THE HUMAN NUCLEAR AND MITOCHONDRIAL GENOMES

NUCLEAR GENOME MITOCHONDRIAL GENOME

Size 3 × 10 bp
9
16,659 bp
Topology 23 linear molecules 1 circular molecule
Number of genes Approximately 21,000 37
% coding sequence (incl. genes for functional RNAs) Approximately 1.4% 93%
Average gene density Approximately 1 per 125 kb (variable) 1 per 0.45 kb
Introns Average 8 per gene (variable) None
Repetitive DNA Approximately 50% None

T he H uman G enome —S tructure and O rgani z ation   •   2 1

www.Ebook777.com
 THE FUNCTIONAL GENOME: THE
E P I G E N O M E , T R A N S C R I P TO M E ,
A N D   P R OT E O M E
The genome is fixed and the same in every cell, apart from
the effects of mitotic errors and somatic mutations (espe-
cially important in cancer) and the special rearrangements
of immunoglobulin and T-cell receptor genes in lympho-
cytes. But all these other “-omes” are variable, specific to a
tissue, a cell type, and a point in time. They are responsible
for development, differentiation, and the response to exter-
nal changes.
ONE GENOME, MANY EPIGENOMES
Patterns of cell- and tissue-specific gene expression are estab-
lished and maintained by the patterns of epigenetic marks
on the genome. As mentioned above, these consist of DNA
methylation and a variety of specific covalent modifications
of histones. The marks are established by a large series of
“writers”:  DNA methyltransferases, histone methyltrans-
ferases and demethylases, histone acetyltransferases and
deacetylases, histone kinases and phosphatases, and so on.
In some cases, small RNA molecules help ensure sequence
specificity. The effects on gene expression are mediated by
“readers,” which include methylated DNA-binding pro-
teins, chromodomain and bromodomain proteins that bind
methylated and acetylated histones respectively, and a large
number of other proteins.
As a result of these modifications, chromatin exists in
a variety of epigenetic “flavors.” The basic distinction is
between heterochromatin (inactive, repressed) and euchro-
matin (potentially active), but subtypes define transcrip-
tional activity and regulatory elements such as promoters,
enhancers, and insulators. The flavor depends on a combi-
nation of types and relative quantities of marks rather than
a simple histone code. For example, in an analysis of nine
different epigenetic marks in nine human cell types, Ernst
et al.16 characterized 15 different flavors.
CPG ISLANDS
A striking feature of human DNA is the scarcity of G
nucleotides directly downstream (3′) of a C nucleotide.
The overall GC content is 41%, so we might expect 4.2%
(0.205 x 0.205) of all dinucleotides to be CG. The observed
frequency is one-fifth of this. The explanation lies in DNA
methylation. DNA methyltransferases specifically attach a
methyl group to the 5-position of cytosines in CG (tradi-
tionally written CpG) sequences. Such methylated cytosines
2 2   •   P rincip l es o f G enomic M edicine
are mutational hot spots. All cytosines, methylated or not,
have a tendency to deaminate spontaneously. Deamination
of unmethylated cytosine produces uracil. Cells recognize
this as an unnatural base in DNA and repair the damage
by replacing uracils with cytosine. However, deamination
of 5-methyl cytosine produces the natural DNA base thy-
mine. Cells are therefore unable to recognize such events.
Over evolutionary time, the majority of methylatable cyto-
sines have mutated to thymine, hence the scarcity of CpG
sequences in the human genome.
Although most of the human genome is depleted in CpG
sequences, approximately 1% consists of “CpG islands” where
the cytosines are usually unmethylated, and so have not been
lost. CpG islands are associated with the 5′ ends of approxi-
mately 60% of human protein-coding genes, particularly
the ubiquitously expressed housekeeping genes. Typically,
islands are a few hundred base pairs long, and lie immediately
upstream of the gene, or overlap the first exon. Methylation
of the island is associated with pathological gene silencing.
T H E T R A N S C R I P TO M E
The transcriptome is the totality of RNA transcripts pres-
ent in a cell at a given time. Next-generation sequencing of
bulk cDNAs has allowed detailed studies of transcriptomes.
All transcripts can be catalogued, and the number of times a
given transcript is sequenced is a measure of its abundance.
Transcription of a gene depends on assembling an initia-
tion complex upstream of the gene. This includes the RNA
polymerase, but also a whole suite of transcription factors
and co-activators that provide the specificity and control
of transcription. Sequences to be transcribed are identified
in the DNA both by the chromatin flavor and by specific
small-sequence motifs that bind transcription-factor pro-
teins. Whether such sequences are actively transcribed or
not depends on the availability of the necessary proteins,
and the absence of inhibitory proteins, in the particular cell
at that particular time. Transcripts often vary considerably
in abundance between people, and much of this variation
is heritable.17 Presumably, this variation explains much of
human individuality.
Both the human transcriptome and the proteome are
considerably larger than our total of 21,000 genes would
suggest. Contrary to earlier views, it turns out that the great
majority of all our DNA is transcribed, at least in some cells
and at some times.5 Often transcripts are found from both
strands of the double helix. Thus cells are awash with RNA
molecules of unknown function. How much of this perva-
sive transcription is functional, and how much is just “tran-
scriptional noise,” is unknown.
 A LT E R NAT I VE S P L I C I N G A N D T H E
P ROT EO M E
Proteomes are best studied by mass spectroscopy. Totaled
over all cell types and all human development, humans have
far more different proteins than different genes. Five main
mechanisms account for this:
• A majority of all human genes produce more than one
mature mRNA by alternative splicing of exons.18 The
ENCODE project5 reported an average of 5.4 tran-
scripts per gene. Common mechanisms involve dif-
ferential incorporation or skipping of exons and use of
alternative splice sites within an exon. An ambitious
attempt to identify the determinants of tissue-specific
alternative splicing19 concluded that it depended on
the interplay of a very large number of sequence motifs
and external molecules. Splice sites are not all the same.
They are more or less active depending on the sequence
­context surrounding the invariant GT. . . AG splice
­signals, on the presence or absence nearby of enhancers
or silencers of splicing (sequences that bind proteins that
help or hinder deployment of the spliceosome machine),
and on the repertoire of the binding proteins (SR
­proteins) available at the time. Some alternative splicing
is unquestionably functional, generating two or more
­functional proteins from a single gene in a controlled
way. How much of the total is of this type, and how
much reflects loose control of splicing, is not clear.
• Genes often have two or more alternative promoters.
These may be differentially regulated, and produce tran-
scripts with alternative first exons joined to common
downstream exons. For example, the dystrophin gene
has eight promoters, producing tissue-specific variants of
the protein. The CDKN2A gene has two promoters; the
5′ exons produced from them are spliced to the down-
stream exons in different reading frames, so that this
single gene encodes two proteins that have totally differ-
ent amino acid sequences.
• Around 1,000 human mRNAs are subject to editing by
special enzymes that change specific nucleotides, so that
the mRNA sequence is no longer an accurate reflection
of the genomic DNA sequence.20 Edited and unedited
copies of the same mRNA can then encode subtly dif-
ferent proteins. RNA editing is particularly important
in the central nervous system. A remarkable paper by Li
et al.21 suggests that there may be much more widespread
systematic differences between genome and transcrip-
tome sequences. If confirmed, this points to unknown
mechanisms generating diversity.
T he H uman G enome —S tructure and O rgani z ation   •   2 3
• After assembly on the ribosomes, polypeptide chains are
often extensively modified to make the functional pro-
tein. Sometimes proteolysis produces several functional
polypeptides (for example, some peptide hormones).
Many modifications are reversible—for example, revers-
ible phosphorylation controls the activity of many
signaling proteins—so that cells can contain multiple
­differentially active versions of the same protein.
• The genes encoding immunoglobulins and T-cell recep-
tors use special mechanisms of DNA splicing, recombi-
nation, and mutagenesis to produce a potentially infinite
variety of antibodies and receptors (reviewed in ref. 22).
VA R I AT I O N W I T H I N T H E H U M A N
GENOME
It is usual to divide variation in the human genome into
large-scale and small-scale variations. Large-scale variants are
often defined as variants greater than 1 kb in size, although
there is no natural boundary between variants so described
and insertion/deletion variants (indels) up to 1 kb in size.
L A RG E -S C A L E VA R I AT I O N
Until recently, it was generally supposed that most human
genetic variation was on the 1–100 bp scale, and that the
large-scale structure of the human genome was fixed. New
techniques, such as comparative genomic hybridization23
and next-generation paired-end sequencing,24,25 have radi-
cally changed this view. It is now apparent that structural
variants are a major source of human genetic variation,
involving in total more of our genome than single nucleo-
tide polymorphisms (SNPs).
Structural variants include insertions and deletions
(copy number variations, CNVs) and balanced inversions
and translocations. Copy number variants often overlap
the segmental duplications described above. A large study
by Conrad et  al.26 identified 11,700 copy number vari-
ants in a survey of just 41 ostensibly healthy individuals.
Comparing any two genomes revealed an average of 1,098
CNVs totaling 24 Mb or 0.78% of the genome. Like all
earlier studies, Conrad et al. used hybridization assays that
are better at detecting large variants than small ones; their
smallest reported CNV involved 443 bp of DNA. The 1000
Genomes study used high-throughput sequencing, which
can identify smaller structural variants. The pilot study of
179 individuals reported over 20,000 variants larger than
50 bp, over half of which were previously unreported.27
 Databases of variants have been set up. The Database
of Genomic Structural Variation (dbVar; http://www.
ncbi.nlm.nih.gov/dbvar) aims to be comprehensive. The
Database of Genomic Variants (http://projects.tcag.ca/
variation/) catalogs variants seen in healthy individuals. As
of November 2010, it listed 66,741 CNVs at 15,963 loci,
34,229 indels sized 100 bp–1 kb, and 953 inversions. The
Decipher database (decipher.sanger.ac.uk) lists variants
found by clinical services in patients, together with clinical
details. As cases accumulate, it should become easier to dis-
tinguish pathogenic from incidental variants, although this
is complicated by evidence that some variants act as suscep-
tibility factors, increasing the risk of a condition but being
neither necessary nor sufficient for the condition to occur.
S M A L L-S C A L E VA R I AT I O N
The majority of nucleotide positions in the genome are rela-
tively invariant, with variants occurring at low frequencies
across populations. However, about one in 300 nucleotides
has a common variant. These are the single nucleotide poly-
morphisms. SNPs usually have only two alleles (that is, at
a given location there is a choice of two alternative nucleo-
tides, but not three or four), and can be characterized by the
minor allele frequency (MAF). This pattern does not arise
because those positions are intrinsically mutable; rather, it
is a product of the evolutionary history of our species, as
explained below. Over 10 million SNPs are catalogued in
the dbSNP database, which, however, includes also small
insertion-deletion polymorphisms (indels).
The 1,000 Genomes and other sequencing projects have
shown how the genomes of individuals vary.27 Typically a
person’s genome might differ from the Reference Human
Genome by 3,000,000 SNPs, 100,000–500,000 indels,
and 300–1,000 larger structural variants. Variation in
normal healthy individuals is by no means limited to the
non-coding DNA. Analysis of 1,000 Genomes data suggests
Table 2.6  HAPLOTYPE BLOCK STATISTICS
POPULATION WHITE N. EUROPEAN
ORIGIN, UTAH
Average size of blocks, kb 16.3
Average no. of SNPs per block 70.1
Average no. of haplotypes (MAF > 0.05) per block 4.66
Fraction of genome spanned by blocks 87%
Fraction of chromosomes due to haplotypes with 93%
MAF > 0.05
SOURCE: International HapMap Consortium (2005).30
2 4   •   P rincip l es o f G enomic M edicine
that the average healthy individual has loss of function vari-
ants in around 100 genes, including cases of homozygous
loss of function.28 This highlights the fact that not all genes
are essential. Dispensable genes are most likely to be mem-
bers of large families of closely related genes, such as the
800-strong olfactory receptor gene family. People can vary
quite widely in the number of such genes.29
T H E PAT T E R N O F VA R I AT I O N AC RO S S
T H E G E N O M E
The reports of the HapMap Consortium30–32 show the pat-
tern of small-scale variation across the genome. In Phase I of
the project 1,007,329 SNPs were typed in 269 DNA sam-
ples from people from four locations: Nigeria (Ibadan), the
United States (Utah), China (Beijing), and Japan (Tokyo).
In Phase II, this was extended to a further 4.6 million SNPs,
and in Phase III, to additional populations. Ninety percent
of heterozygous sites in each individual were due to com-
mon SNPs (MAF > 0.05).
When data on adjacent SNPs are combined, it becomes
apparent that chromosomes are a mosaic of haplotype
blocks (Table 2.6). Blocks are statistical concepts, whose
exact properties depend on the statistical criteria used,
but they reflect a real and important structural feature of
human genomes. If a block contains 50 biallelic SNPs, there
are 250 possible haplotypes. However, in reality, at any given
location, the range is far more limited. As Table 2.6 shows,
the average number of haplotypes per block in the Phase
I data ranged from 4.0 in the Chinese and Japanese sample
to 5.6 in the Nigerians.
At most locations, more than 90% of all haplotypes are
one of four to six common alternatives; the exceptions are
likely to be the result of recent mutations that have found
their way into only a small number of descendant chromo-
somes. The block pattern can be explained if each block
haplotype is derived from a single common ancestor. Thus,
SUB-SAHARAN AFRICAN, EAST ASIAN,BEIJING &
IBADAN, NIGERIA TOKYO
7.3 13.2
30.3 54.4
5.57 4.01
67% 81%
94% 95%
most humans inherit their DNA at any particular location
from just a tiny handful of ancestors. This does not mean
that we are all descended from four or six cavemen. The four
or six haplotypes in one block probably came from a differ-
ent set of ancestors than did the four or six haplotypes in
the next block. The size of the blocks is a function of the
number of generations between us and the common ances-
tor. With each round of sexual reproduction, recombina-
tion during meiosis fragments the chromosomes. African
populations are older than others are, so there has been
more fragmentation, and haplotype blocks are smaller. The
mtDNA and the Y chromosome (except for its tip) are not
subject to recombination, and so each is a single large hap-
lotype block.
Much of the research described in the rest of this book
starts from this picture of our genomes as a mosaic of lim-
ited numbers of ancestral chromosomal segments. The
common disease–common variant hypothesis assumes that
most of the genetic determinants of susceptibility to com-
mon disease are ancient polymorphisms. At the relevant
chromosomal location, susceptibility or resistance alleles
should segregate with particular blocks. Thus, identifying
them comes down to identifying blocks that are associated
at the population level with susceptibility or resistance.
It is not necessary to type every SNP in a block to distin-
guish the four or six alternatives at a particular location.
Blocks can be defined by a small number of SNPs (tag-
ging SNPs). The research protocol therefore comes down
to typing individuals for sufficient tagging SNPs to define
every block, and looking for associations. If blocks average
10 kb, and require three tagging SNPs each, this means typ-
ing for approximately 1 million SNPs in a sufficiently large
case-control study.
The Wellcome Trust Case Control Consortium33 pro-
vided the first major demonstration that such genome-wide
association studies (GWAS) could reliably identify suscep-
tibility factors for common diseases. Thousands of such fac-
tors have now been identified for a large number of complex
diseases. However, the proportion of the overall heritability
identified by GWAS is low for most diseases, and genotyp-
ing individuals for these factors does not generally lead to
clinically useful predictions of susceptibility. Where to find
the “missing heritability” is a matter of much debate.
C O N C LU S I O N
The successful completion of the Human Genome Project
ushered in a new era in human genetics. The finished human
sequence is complemented by a rapidly increasing number of
T he H uman G enome —S tructure and O rgani z ation   •   2 5
genome sequences from other species. A major surprise has
been the relatively low number of genes in our genome—not
vastly more than that in the Drosophila fruit fly (approximately
13,000) or the Caenorhabditis elegans worm (approximately
19,000). The greater complexity of mammals compared with
these organisms has been accompanied neither by a great
increase in the number of genes nor by a significant increase
in the number of different proteins produced per gene by
alternative splicing, but by a vast increase in the amount of
non-coding DNA with no known function. This hints at the
existence of much more sophisticated systems for regulating
gene expression, probably mediated by combinatorial bind-
ing of numerous proteins and small RNA molecules to some
of the non-coding DNA, controlled by the local structure
and organization of the chromatin.
A major interest is in how the genome varies between
people. Now that thousands of individual human genomes
have been fully sequenced, we have a much better picture
of the range of normal variation and the evolutionary pro-
cesses that produced that variation. The HapMap project has
detailed how our chromosomes are mosaics of short ances-
tral blocks, with most humans having only a tiny handful of
ancestors for any particular block. Common SNPs (MAF >
0.05) tend to be found in all ethnic groups, albeit at differing
frequencies, while infrequent and rare SNPs are more likely
to be specific to particular ethnic groups or extended families.
Structural variants are numerous and often encompass genes.
The average healthy person carries loss-of-function variants
in around 100 genes, showing that not all our genes are essen-
tial. As clinical genetics services move more and more to
sequencing as the default procedure, a major preoccupation
is distinguishing pathogenic from normal variants.
REFERENCES
1. International Human Genome Sequencing Consortium. Initial
sequencing and analysis of the human genome. Nature. 2001;409:
860–921.
2. Rudd MK, Willard HF. Analysis of the centromeric regions of the
human genome assembly. Trends Genet. 2004;20:529–533.
3. Hahn WC. Telomere and telomerase dynamics in human cells. Curr
Mol Med. 2005;5:227–231.
4. International Human Genome Sequencing Consortium. Finishing
the euchromatic sequence of the human genome. Nature. 2004;431:
931–945.
5. ENCODE Project Consortium. Identification and analysis of func-
tional elements in 2007;1% of the human genome by the ENCODE
pilot project. Nature. 447:799–816.
6. Mattick JS. The genetic signatures of noncoding RNAs. PLoS Genet.
2009;5:e1000459.
7. Cabili MN, Trapnell C, Goff L, et  al. Integrative annotation of
human large intergenic noncoding RNAs reveals global properties
and specific subclasses. Gene Dev. 2011;25:1915–1927.
 8. Guttman M, Rinn JL. Modular regulatory principles of large
non-coding RNAs. Nature. 2012;482:339–346.
9. Sotillo E, Thomas-Tikhonenko A. The long reach of noncoding
RNAs. Nat Genet. 2011;43:616–617.
10. Bejerano G, Pheasant M, Makunun I, et al. Ultraconserved elements
in the human genome. Science. 2004;304:1321–1325.
11. Jeffreys AJ, Wilson V, Thein SL. Individual-specific fingerprints of
human DNA. Nature. 1985;314:67–73.
12. Baillie JK, Barnett MW, Upton KR, et  al. Somatic retrotranspo-
sition alters the genetic landscape of the human brain. Nature.
2011;479:534–537.
13. Bailey JA, Gu Z, Clark RA, et al. Recent segmental duplications in
the human genome. Science. 2002;297:1003–1007.
14. Nusbaum C, Zody MC, Borowsky ML, et al. DNA sequence and
analysis of human chromosome 2005;18. Nature. 437:551–555.
15. Trifunovic A, Wredenberg A, Falkenberg M, et al. Premature age-
ing in mice expressing defective mitochondrial DNA polymerase.
Nature. 2004;429:417–423.
16. Ernst J, Kheradpour P, Mikkelsen TS, et  al. Mapping and analy-
sis of chromatin state dynamics in nine human cell types. Nature.
2011;473:43–49.
17. Deutsch S, Lyle R, Dermitzakis E, et al. Gene expression variation
and expression quantitative trait mapping of human chromosome
2005;21 genes. Hum Mol Genet. 14:3741–3749.
18. Maniatis T, Tasic B. Alternative pre-mRNA splicing and proteome
expansion in metazoans. Nature. 2002;418:236–243.
19. Yoseph Barash, John A. Calarco, Weijun Gao, Qun Pan, et al.
Deciphering the splicing code, Nature. 465:53–59; 2010
20. Wulff B-E, Sakurai M, Nishikura K. Elucidating the inosin-
ome: global approaches to adenosine-to-inosine RNA editing. Nat
Rev Genet. 2011;12:81–85.
21. Li M, Wang IX, Li Y, et al. Widespread RNA and DNA sequence
differences in the human transcriptome. Science. 2011;333:53–58.
2 6   •   P rincip l es o f G enomic M edicine
22. Gellert M. V(D)J recombination: RAG proteins, repair factors and
regulation. Annu Rev Biochem. 2002;71:101–132.
23. Speicher MR, Carter NP. The new cytogenetics: blurring the bound-
aries with molecular biology. Nat Rev Genet. 2005;6:782–792.
24. Korbel JO, Urban AE, Affourtit JP, et  al. Paired-end mapping
reveals extensive structural variation in the human genome. Science.
2007;318:420–426.
25. Mills RE, Walter K, Stewart C, et al. Mapping copy number variation
by population-scale genome sequencing. Nature. 2011;470:59–65.
26. Conrad DF, Pinto D, Redon R, et  al. Origins and functional
impact of copy number variation in the human genome. Nature.
2010;464:704–712.
27. 1000 Genomes Project Consortium. A map of human genome
variation from population-scale sequencing. Nature. 2010;467:
1061–1073.
28. MacArthur DG, Balasubramanian S, Frankish A, et al. A systematic
survey of loss-of-function variants in human protein-coding genes.
Science. 2012;335:823–828.
29. Sudmant PH, Kitzman JO, Antonacci F, et al. Diversity of human
copy number variation and multicopy genes. Science. 2010;330:
641–646.
30. International HapMap Consortium. A haplotype map of the human
genome. Nature. 2005;437:1299–1320.
31. International HapMap Consortium. A second generation human
haplotype map of over 3.1 million SNPs. Nature. 2007;449:851–861.
32. International HapMap 3 Consortium. Integrating common and rare
genetic variation in diverse human populations. Nature. 2010;467:
52–58.
33. Wellcome Trust Case Control Consortium. Genome-wide associa-
tion study of 14,000 cases of seven common diseases and 3,000 con-
trols. Nature. 2007;447:661–683.
34. Griffiths-Jones S. Annotating non-coding RNA genes. Annu Rev
Genomics Hum Genet 2007;8:279–298.
 3.
HUMAN PROTEOMICS
Brian Morrissey, Lisa Staunton, and Stephen R. Pennington
INTRODUCTION
It has long been appreciated that proteins are key ­biological
effectors of life, and while genomes provide a “blueprint,”
it is the proteins the genomes encode that are the essential
building blocks of biology. This pivotal role has ensured
that proteins remain at the forefront of scientific scrutiny
when scientists are trying to unravel the processes driv-
ing human development and disease progression. It was
therefore somewhat inevitable that as soon as the human
genome was sequenced, a new ambition would be the
attempt to replicate the same accomplishment with the
human proteome. Moreover, it was no coincidence that
the Human Proteome Organization (HUPO;www.hupo.
org) was launched on the ninth of February 2001, just one
week before the anticipated release of the landmark papers
describing the human genome sequence. To date, this goal
of mapping and sequencing the human proteome has yet to
be realized, but it is still actively pursued with the establish-
ment of the Human Proteome Project (a HUPO initiative)
in 2009 to map the entire human protein set.1,1a,1b However,
the difficulty in achieving such a goal is widely acknowl-
edged, including recently by Nilsson and colleagues, who
noted that, “Sequencing the human genome was perhaps
the easy part; now making sense of the constantly moving
and changing picture of the proteome will require a lot
more time effort and creativity.”2 Of course, the domain of
genome sequencing has not stood still, and researchers con-
tinue to make significant advances that have embraced new
technologies and transformed the speed, efficiency, and
cost of sequencing. The new opportunities this has afforded
include massive genome-wide disease-association studies
and the prospect of widespread genome sequencing of indi-
viduals to support personalized medicine. Technological
advances in proteomics, while seemingly more diverse, have
perhaps been less dramatic—they have included develop-
ments in mass spectrometry, the imaging of proteins, pro-
tein and antibody arrays, and chemical proteomics.
27
At its most straightforward, proteomics may be viewed
as the analysis of protein expression or steady-state protein
abundances—this reflects the overall balance between tran-
scription, translation, and post-translational processes. In
recent years, the connection between the genome and pro-
teome has been probed by trying to establish the mecha-
nisms underlying the regulation of gene expression and
turnover or homeostasis of proteins.3,4 Highly complex as
it is, this interrelationship encompasses the control of tran-
scription, of mRNA processing and degradation, and the
regulation of protein translation, post-translational modi-
fication, localization, and degradation. Notably, recent
improvements in mass spectrometry for proteomics and
next-generation sequencing for genomics are together sup-
porting more integrated large-scale surveys of genomes,
transciptomes, and proteomes. Thus, emerging studies using
next-generation sequencing-derived data to complement
proteomics data offer the opportunity to better understand
the regulation of protein expression and hence biological
function at a systems level.5,6
In less than two decades, the field of proteomics has
undergone a number of significant developments, many of
which have been underpinned by the growing availability
of DNA sequence data, and for human proteomics, were
spurred on by the initial release of the draft human genome
sequence and its subsequent “completion” in 2003.7–9 Prior to
this genomic era, scientists intent on studying protein expres-
sion at a “proteome” level relied heavily on two-dimensional
gel electrophoresis (2-DE), Edman sequencing for protein
identification, and antibodies for targeted protein measure-
ments. These approaches, while used very effectively by a few
practitioners who excelled in their application,10–12 are labo-
rious, technically demanding, and present significant practi-
cal obstacles. It was not surprising, therefore, that alternative
techniques were sought,13,14 and in the early 1990s it became
apparent that mass spectrometry (MS) was set to play a signif-
icant part in the establishment of proteomics capabilities.15,16
Specifically, the development of two Nobel Prize–winning
techniques, Matrix Assisted Laser Desorption Ionization
(MALDI) and Electrospray Ionization (ESI), enabled ioniza-
tion of bio-macromolecules for subsequent analysis by MS,
which until this time had been confined to small molecule/
chemical analysis.17,18 While this accelerated the development
of proteomics, its true potential was not fully realized until
the genome-wide sequences (human, rat, mouse, and various
human pathogens) became available. This genome sequence
data in the form of extensive and readily accessible sequence
databases provided the template that protein analysis by MS
could use to match against experimentally generated peptide
mass spectra to identify their constitutive proteins.19,20 This
development revolutionized proteomics, paving the way for
a new era in the study of human disease. Although numerous
techniques, including 2-DE and protein arrays, are used and
used very effectively, MS-based proteomics has so far led the
way in the emergence of proteomics. MS instrumentation
has been developed to incorporate new mass analyzers and
hybrid instruments designed to tackle the challenges associ-
ated with protein analysis. These developments have taken
advantage of improvements in mass accuracy achieved with
Fourier transform ion cyclotron resonance MS instruments,
in the resolving power of time-of-flight instruments, and in
the sensitivity and dynamic range of triple quadrupoles.13
This diversity of instrumentation and post-source manipula-
tion of ions has ensured that MS remains the current pro-
teomic technique of choice and is quite likely to continue to
do so for the foreseeable future. However, there is much that
MS as an analytical tool cannot as yet achieve, and approaches
that complement or perhaps even surpass its capabilities
remain highly desirable. As an illustration: a key limitation
of current proteomics is that the widely used “bottom-up”
approach whereby proteins are cleaved to peptides prior to
MS analysis rarely gives 100% coverage of individual pro-
teins (compare this with the redundancy encountered in
genome sequencing), leaving potentially important “parts”
of the sequence of individual proteins refractory to analysis.
Top-down approaches, in which intact proteins are analyzed
and fragmented in the MS itself, are time consuming, require
significant skill, lack sensitivity, and are not readily applied
to complex mixtures of proteins. Technologies that could
address these and other important limitations are still sought.
T WO -D I ME N S I O N A L G EL
ELE C T R O P H O R ES I S
First described in 1975, 2-DE was the answer to a widely
recognized need for greater resolution in protein separa-
tion.11,21 By coupling two analytical methods of protein
2 8   •   P rincip l e s o f G e no m ic M e dicin e
separation—isoelectric focusing (IEF) and SDS-PAGE—
this technique provided the capacity of separating up to
1000 proteins.21 In the first dimension, proteins are sepa-
rated according to their isoelectric point (pI) by IEF along a
pH gradient, which is subsequently followed by SDS-PAGE,
which separates proteins according to their molecular mass.
After separation, proteins are visualized by staining, com-
monly Coomassie Brilliant Blue or the more sensitive silver
stain. Proteins resolved by 2-DE are then analyzed, digested,
and identified using MS. This classical proteomic meth-
odology saw widespread application but was hampered
by experimental drawbacks such as poor reproducibility
and sensitivity. The introduction of commercially available
immobilized pH gradient gels for IEF aimed to overcome
the lack of reproducibility seen with 2-D maps. In 1997, in
an attempt to improve the reproducibility and sensitivity of
the original 2-DE protocol, a new technique, 2-dimensional
difference in gel electrophoresis (2-D DIGE), was estab-
lished.22–24 This technique, which uses fluorescent cya-
nine dyes for protein labelling prior to protein separation,
revolutionized quantitative proteomic analysis, with over
2,700 citations, including primary research papers and
reviews (GE Healthcare, 2011). The advantage of using cya-
nine dyes is that they are size- and charge-matched to the
amino acid lysine and so ensure negligible shift in pI during
first-dimension separation, thus allowing multiple experi-
mental samples to be separated on a single gel, thus reducing
the number of gels to be run in the experiment and allowing
the user to include an internal standard; that is, equal quan-
tities of all experimental samples into one pooled internal
standard sample. The internal standard allows the user to
apply a normalization factor to correct for experimental gen-
erated differences. The generation of images following the
CyeDye TM DIGE flours requires a laser scanner capable of
fluorescence excitation of the dyes, which can pose a finan-
cial disadvantage of the 2-D DIGE approach. The resulting
improvement in experimental reproducibility is particularly
advantageous for clinical studies, where sample numbers can
be much higher than in animal or cell model studies.
Following the generation of the gel images, specialist
software/bioinformatic packages are required to identify
protein spots deemed to be differentially expressed. Such
software is readily available and can come as a package with
laser scanners, such as Quant XL, which is available with
the Typhoon laser scanner series; however, specialist bioin-
formatic packages dedicated to gel imaging and statistical
analysis such as Progenesis SameSpots from Non Linear
Dynamics may also assist in differentially expressed pro-
teins. Protein spots of interest are then excised and digested
for protein identification via MS.
 However, a number of well-publicized limitations have
seen the 2-DE technique fall out of favor with scientists
in recent years.25 The technique in itself is laborious, with
reproduction still an arduous task, although automation
has been attempted. Its separation capability is still limited
when analyzing complex protein species, with both dynamic
range and total numbers of proteins still a major challenge.
Moreover, the loss of hydrophobic protein species during
the IEF step and the inability of SDS-PAGE to resolve pro-
teins with a pI at the extremities of the pH scale have led
to much criticism of the technique.26 While 2-DE has now
largely been replaced by MS techniques, its utility in pro-
teomics should not be ignored. The ability to isolate protein
isoforms and post-translational modifications means it still
provides a significant advantage over alternative proteomic
strategies.27 This advantage was recently highlighted by
Angelica Gorg, one of the pioneers of 2-DE, who wrote,
“Due to the wide diversity of protein abundance and prop-
erties in complex proteomes it is anticipated that no single
proteome analysis will be able to effectively address all the
proteome analysis required.”
P ROT E I N S P OT I D E N T I FI C AT I O N
BY M A S S S P EC T RO M ET RY
MS identification of the protein gel spots consists of excis-
ing the spots of interest followed by in-gel digestion.28 The
reduced complexity of protein gel spots enables protein
identification by MALDI and LC-ESI-based instruments,
with LC-ESI shown to provide a greater protein sequence
coverage.29
M A SS S P E C T R O MET RY
Although in the early days, MS-based proteomics largely
provided a supporting role for 2-DE experiments, the tech-
nology quickly moved to provide an independent means of
analysis for complex samples. Initially providing qualitative
data, new methods for quantification of complex mixtures
and targeted approaches for individual proteins have con-
tinued to develop and improve in recent years.30 This con-
tinued evolution in MS has resulted in a plethora of mass
spectrometers run as single units and coupled, known as
tandem mass spectrometry (MS/MS), which greatly vary in
analytical performance and experimental capabilities.31 The
rapid development of MS has in turn forced the develop-
ment of supportive bioinformatics tools to analyses the
increasingly complex data, a process that now requires as
much consideration as the MS analysis itself.32–34
H u m an P rot e o m ic s   •   2 9
G L O BA L P ROT EO M I C A NA LY S I S
The global analysis of complex protein samples, frequently
referred to as “shotgun” proteomics, can conceptually be
undertaken by three core experimental steps as shown in
Figure 3.1. They consist of 1)  protein extraction (which
can be optionally followed by protein fractionation prior
to enzymatic digestion to reduce sample complexity);
2) enzymatic digestion of protein into peptides (routinely
using trypsin), followed by upfront separation of peptides
by liquid chromatography (LC) (offline peptide separation
can also be considered to further reduce sample complex-
ity); and 3) peptide/protein quantification and identifica-
tion determined from the generation of MS1 or MS/MS
data respectively. MS1 data are generated when peptides
eluting from the LC system “fly” through the mass spec-
trometer intact, with the resulting peptide ion signal pro-
portional to the peptide abundance in the sample. MS/
MS data are generated when peptide ions are fragmented
in a collision cell within the mass spectrometer, generating
a series of fragmentation ions that are used to build MS/
MS spectra. Peptides are then identified by matching these
spectra against a theoretical spectral database generated
from a genome sequence database. Identified peptides are
subsequently rolled up to generate a list of proteins identi-
fications with this process stemming the term “bottom–up
proteomics.” By matching peptides ion (MS1) with their
fragmentation ions (MS/MS), proteins are quantified and
identified. This process is conceptually undertaken in most
LC-MS/MS workflows with the exception of iTRAQ label-
ling and spectral counting, where protein quantification is
determined from the MS/MS data.
While the choice of mass spectrometer is determined
by a variety of parameters, the popularity of liquid chroma-
tography coupled to a mass spectrometer via ESI has largely
overtaken MALDI-based approaches due to its ability to
seamlessly connect the LC and MS, providing peptide sepa-
ration in an automated fashion whereby 2500 proteins can
be identified in a mammalian cell over a 90-minute analy-
sis window.35 In 2011, Nagaraj and colleagues detailed the
in-depth proteomic and transcriptomic profiling (RNA-
seq) of the Hela human cervical cell line to determine
what depth of proteome coverage could be achieved.36
This experiment consisted of 288 hours of analysis time in
a LTQ-Orbitrap mass spectrometer, resulting in the iden-
tification of 10,255 proteins encoded by 9,207 genes from
the total 11,936 estimated genes in total. This undoubtedly
demonstrates the deepest level of coverage of any human
cell to date. However, this level of analysis may still be out-
side the capabilities of most proteomic research centers, and

Proteome
Protein Peptide
Extract Extract
Protein Peptide
Fractionation Fractionation
Figure 3.1  Schematic workflow of “shotgun” proteomic-based approach.
it is extremely time consuming. Moreover, this experiment
was limited to generating a list of protein identifications
for a single cell type. The ability to quantify the proteins
either for an individual cell type or a more complex samples
such as tissue lysates or body fluids would be a much greater
challenge.
Figure 3.1 outlines the core experimental steps for global
analysis of complex protein samples (shotgun proteomics),
including: protein extraction, enzymatic digestion followed
by separation of peptides by liquid chromatography (LC),
and peptide/protein quantification and identification
determined from the generation of MS1 or MS/MS data
respectively.
L A B E L O R L A B E L-FR E E?
The process of generating qualitative and/or quantitative
peptide data is by no means trivial, and one of the first con-
siderations before analysis can take place is to label or not
to label. Both provide relative quantification of peptides
across samples with the choice of employing a labelled or
label-free approach, determined by cost, sample numbers,
sample type, and the degree of required accuracy.37
Labelling approaches utilize the incorporation of a
metabolic, enzymatic, or chemical isotopic tags that change
the mass of the peptides without changing its biochemical
properties, in a process referred to as differential stable iso-
tope labelling. By combining labelled and unlabeled samples
in a single run on a mass spectrometer, both labelled and
unlabeled peptides can be identified and differentiated.
The measured signal coming from both peptides can then
be used to determine the relative quantitative differences
between the two samples.38 Numerous labelling procedures
have been developed that support shotgun proteomics,
3 0   •   P rincip l e s o f G e no m ic M e dicin e
MS spectra
Peak
Quantification
Abundance
LC-MS+
LC-MS/MS Peptide/Protein
Identification
MS/MS
spectra
Database
search
m/z
including O18 labelling, isotope-coded affinity tags (ICAT),
isobaric tags for relative, absolute quantification (iTRAQ)
and stable isotope labelling by amino acids in culture
(SILAC).39–42
O18 labelling is one of the simplest labelling approaches
available for shotgun proteomics. In this enzyme-catalyzed
process, O16 is replaced with O18 at the C-terminal of the
carboxyl group of proteolytical peptides, resulting in a 4Da
mass shift with subsequent relative quantification via the
parent ion peak height/area.42 As labelling is performed
during protein digestion, this technique is particularly use-
ful where metabolic labelling such as SILAC is not possible,
such as human tissue or serum providing a suitable labelling
approach for clinical studies.43,44 However, a number of lim-
itations have been reported, including a limited dynamic
range and variability of uptake among different peptides.45
iTRAQ, first described by Ross et al., consists of a set
of isobaric labels that are isotopically incorporated at the N
termini and lysine side chain peptides in a digest mixture.41
These isobaric tags are indistinguishable at the MS level;
however, following MS/MS fragmentation, the reporter
groups that contain variable mass (114-117 Da or 113-121
Da) are released, with the peak area of the reporter ions
used to determine the relative abundance of the peptides at
the MS/MS level.38 Initially released as a 4-plex, and later
8-plex allowing analysis of up to eight samples, the iTRAQ
procedure has been applied to numerous biological stud-
ies of different samples, such as human saliva, fibroblasts,
and mammary epithelial cells.46 However, given the limita-
tion of eight labels for analysis, this technique is incompat-
ible with large-scale clinical applications where numerous
patients or sample numbers are present.
Isotope-coded affinity tags (ICAT), first described by
Gygi and Aberscold in 1999, is a chemical label consisting
 Free ebooks ==> www.Ebook777.com
of three functional elements: a specific chemical reactivity,
an isotopically coded linker, and an affinity tag (biotin).40
Labelling of reduced protein is achieved as the thiol-reactive
group is selective for the sulfhydryl group in the side chain
of reduced cysteine, which are then subsequently isolated
via the biotin-avidin affinity enrichment.47 The label is com-
pleted with an isotopically labelled linker available in two
forms, heavy and light. The heavy label, which originally
consisted of deuteriums, has subsequently been replaced
with C13, resulting in a 9 Da mass shift between the heavy
and the light labelled peptides.48 ICAT has been applied
extensively in biological research with applications in whole
cell lysates,49 conditioned culture medium,50 and subcellu-
lar fractions such as mitochondria.51 However, the limited
availability of labels does restrict its use; moreover, as label-
ling is achieved via cysteine binding and subsequent enrich-
ment of labelled peptides, there is a resulting loss of peptide
devoid of the presence of cysteine.
Stable isotope labelling by amino acids in culture
(SILAC), first established a decade ago, consists of the
incorporation of specific amino acids into (mammalian)
proteins by culturing cells in media depleted of an essen-
tial amino acid and replacing them with an isotopically
labelled form of the amino acid.39 In this process, two
cell populations are grown, one cultured in medium with
heavy-labelled amino acids containing 2H instead of H,
13
C instead of 12C, or 15N instead of 14N, and the other in
medium with the light amino acid (unlabeled). The result-
ing known mass shift in the heavy-labelled peptides enables
differentiation and quantification via the parent ion peak
height/area.52 Although SILAC was initially established
for the analysis of cell culture, this technique has now been
extended to animal models, thereby increasing its applica-
bility to the study of human disease.53–55
Label-Free
While labelling strategies are generally accepted as being
more accurate, the associated cost of label reagents and
dedicated software, combined with limited availability of
labels, has driven the development of label-free strategies.
Label-free quantification can be undertaken by two distinct
processes: spectral counting and peptide ion peak intensity/
area signal. Spectral counting is based on the observation
that the number of MS/MS spectra generated from a sin-
gle peptide is proportional to the abundance of that pep-
tide in the sample: that is, the more abundant, the greater
chance it will be selected for MS/MS fragmentation. While
this method has demonstrated linearity over two orders
of magnitude, bias towards highly abundant proteins is a
H u m an P rot e o m ic s   •   3 1
www.Ebook777.com
problem, along with its dependency on the generation of
high-quality MS/MS data.38 The quantification of peptides
via the ion peak intensity/area signal is based on the obser-
vation that, as a peptide is eluted from the LC system, the
resulting increase of observed ion signal in the mass spec-
trometer is proportional to the concentration of the peptide
in question. By mapping this peptide elution profile, a peak
area is generated, with the resulting area under the curve
used to determine the peptide peak abundance. Subsequent
retention time alignment of peptide peaks across multiple
samples thereby enables the relative quantification of pep-
tides, with no restriction on the number of samples that can
be analyzed in any one experiment.37 However, this data
processing is not a trivial task, with time alignment, peak
quantification, peak matching, peak identification, and
subsequent statistical analysis requiring sophisticated bio-
informatic tools.56 Fortunately, this bioinformatic support
is readily available, with numerous free online applications
available and commercial platforms with dedicated support
for persons of limited bioinformatic knowledge, further
opening the door of proteomics to the field of biology.57,58
TA R G ET E D P R OT E O M I C S
As global proteomics continued to produce increasingly
complex data and identified even greater numbers of pro-
teins deemed to be of biological interest, the need to vali-
date was becoming increasingly evident. This bottleneck in
development still poses a significant problem in all areas
of proteomic research, but significantly so for biomarker
research, where validation of multiple targets across large
numbers is required. Traditionally, validation of proteins
was conducted by Western Blotting and/or enzyme-linked
immunosorbent assay (ELISA) techniques that require
a lot of protein; however, although these techniques still
demonstrate superior sensitivity over MS, their ability to
multiplex is limited. Moreover, the development of anti-
bodies for every target identified can be a time-consuming
and expensive process and ultimately a death sentence for
the validation of biomarker candidates. An alternative
technique that could be considered the ELISA of the MS
world is multiple reaction monitoring (MRM), which has
been used for testing chemical analytes for decades. The
importance of targeted MS was highlighted in 2012 as
the Method of the Year by Nature Methods.59 Unlike the
shotgun approach, MRM has the ability to specifically tar-
get individual “proteotypic” peptides (peptides unique to
one protein) with the measured signal representative of its
constitutive protein. This process, traditionally undertaken
in a triple quadrupole mass spectrometer, consists of three
steps:  1)  proteotypic peptides (parent ions) are selected
in quadrupole (Q)1; 2)  the selected parent ion is then
fragmented in the second quadrupole (Q2) by allowing
it to collide with the molecules of an inert gas smashing;
3)  the resulting multiple fragments or “transitions” are
then selected to allow them to pass through Q3 and reach
a detector. The intensity of the transitions (fragments of
the parent ion) is a measure of the amount of the parent
ion (peptide). This serial selection process provides a fast,
sensitive, and cost-effective solution to validating multiple
protein targets in large numbers of patient samples.60–62 In
a move to increase sensitivity, Fortin et  al. demonstrated
the use of a new development, MRM(3).63 This technique
is similar to multiple reaction monitoring (MRM), with the
exception of the linear ion trap used in Q3. At this stage,
the transitions generated in Q2 are subjected to a second
round of fragmentation in Q3, with the resulting fragmen-
tation used for quantification. This process has been shown
to increase sensitivity of the lower limits of detection three-
to five-fold in plasma-based studies when compared to
MRM.64 However, the sensitivity of MRM when used for
measuring peptides in complex biological fluids continues
to be a challenge.65 In particular, the plasma proteome in
which proteins span at least 10 orders of magnitude in con-
centration range continues to challenge mass spectrome-
trists with its analytical complexity.66,67 As the sensitivity
of MRM and MRM(3) continues to improve, and given
their current ability to multiplex protein measurements, it
is expected that these techniques will play a leading role
in studies to validate of biomarkers. Moreover, with the
inclusion of isotopically labelled peptide standards, this
technique can provide absolute quantification of peptides/
proteins, providing a real alternative to ELISA and other
antibody-based quantification platforms.
B I O I N F O R M AT I C S O F M A SS
S P E C T R O MET RY
As the data output from mass spectrometers became
increasingly complex, the need for supportive bioinformat-
ics to decipher this information was critical. While tradi-
tionally bioinformatics, and to a lesser extent statistics,
were seen as a tax on proteomic scientists, the generation of
increasingly complex data ensured that bioinformatic anal-
ysis now required as much careful consideration as choos-
ing the instrument of analysis. In a report from HUPO,
where a standard containing 17 proteins was sent to pro-
teomic research centers across the globe to determine their
3 2   •   P rincip l e s o f G e no m ic M e dicin e
performance in analyzing the sample, in what was a worrying
outcome, many of the centers failed to identify all protein
species in the standard, and in many cases they identified
proteins that were not present.68 Moreover as the popularity
of proteomics continues to grow, the number of scientists
active in the field with limited bioinformatic capabilities
is on the increase, thus requiring user-friendly supportive
software. Fortunately such software has been developed in
abundance in recent years to supply users with a range of
abilities. The analysis of labelled or label-free shotgun data
has been provided with numerous open-source packages
such as SuperHirn and Maxquant, and commercial pack-
ages such as Progenesis LC-MS.37,69,70 The choice of software
can be difficult and is dependent on one’s bioinformatic
skill, the cost, and the software’s performance. Moreover,
the performance of data analysis can vary greatly from plat-
form to platform, as was recently highlighted by Mancuso
et  al. when comparing nine different platforms using a
large Orbitrap data set.71 The development and analysis of
MRM assays has also been supported with high-quality
and in many cases intuitive bioinformatic software. While
all mass spectrometer vendors provide machine-specific
software (e.g., Agilent-Mass Hunter, ABSciex-MultiQuant)
for the post-analysis of MRM data,32 the development of
such assays can be a more complex process. Fortunately,
numerous free online platforms (Skyline72, MRMer73) for
the development and analysis of MRM have provided users
with a single analysis tool for MRM assays.74 The developers
of such programs have significantly accelerated the analy-
sis of such data, preventing what would have been another
major bottleneck in the validation of many new candi-
date biomarkers identified in global proteomic studies. In
addition to the analysis of shotgun proteomic and target
proteomic data, other software tools that provide file con-
version capabilities, de novo sequencing, post-translational
modification and structure analysis, and more can be
sourced at www.ms-utils.org.
P R OT E I N M I C R OA R R AYS
Following the establishment of DNA microarray tech-
nologies in the 1990s, protein microarrays soon followed
suit, with a plethora of array-based technologies now
available with applications in drug discovery, biomarker
research, and pathway analysis of disease networks.75,76
These microarrays primarily fall into two distinct catego-
ries: 1) functional microarrays that identify protein inter-
actions, and 2)  abundance microarrays used in protein
quantification. Functional protein microarrays comprise
 Free ebooks ==> www.Ebook777.com
purified proteins, protein domains, or peptides immobi-
lized on glass support slides.77 Such microarrays have been
used to identify protein interactions with other proteins,
nucleic acids, lipids, small molecules, and biomolecules.
These microarrays have provided in-depth network signal-
ing information for proteins implicated in human disor-
ders,78 protein–pathogen interactions,79 and the human
interactome.80 Abundance-based microarrays, used in
quantitative protein-expression analysis, have been exten-
sively used in the areas of clinical trials and personalized
medicine.81,82 These arrays typically consist of either cap-
ture molecules such as antibodies (antibody arrays), or the
sample of interest (tissue microarrays, lysate microarrays)
bound to the glass support, with subsequent detection
via direct sample-labelling or labelled detection antibod-
ies.77 Antibody arrays and tissue microarrays have been
widely applied in biomarker research for bladder cancer,83
breast cancer,84,85 and pancreatic cancer.86,87 Lysate/reverse
microarrays have proven to be a particularly useful tool in
the analysis of cell line models of disease88,89 or tissue lysate
such as breast tissue aspirates.90 However, one of the inher-
ent limitations of these abundance-based microarrays is
the reliance on antibodies. The ability to produce spe-
cific and reliable antibodies that work consistently under
experimental conditions can significantly affect reliabil-
ity.91 In addition, protein microarrays, as with MS-based
proteomics, rely on genomic sequencing to code for pro-
tein expression, which can prove problematic where a par-
ticular species is not sequenced.
P O S T-T R A N SL AT I O N A L
M O D I F I C AT I O N S
Even with advancements in proteomic technologies such
as MS and microarray analysis, the complexity of the pro-
teome cannot be ignored due to gene splicing forming
different protein isoforms and various post-translational
modifications (more than 200 available). Protein
post-translational modifications (PTMs) play a role in the
regulation of both the structure and the function of cel-
lular proteins,92 and include, but are not limited to, phos-
phorylation, acetylation, ubiquitinylation, methylation,
and glycosylation. With advances in MS technologies
allowing more sensitive profiling of complete proteomes,
similar advances have allowed the characterization of
multiple PTMs. Phosphorylation, due to its role in a
wide variety of cellular processes, is the most widely stud-
ied PTM, with a number of experimental approaches
used:  2-DE gel-based PTM analysis, antibody-based
H u m an P rot e o m ic s   •   3 3
www.Ebook777.com
enrichment, affinity-based enrichment, and MS-based
PTM analysis.93–96 Affinity-based techniques such as
immobilized metal affinity chromatography (IMAC),97
and titanium dioxide (TiO2)98 have allowed global char-
acterization of phosphorylated proteins in complex
samples. Other PTMs such as ubiquitylation94 and glyco-
sylation99 are now receiving attention due to the techni-
cal advances in mass spectrometry. Standard PTM studies
focus either on PTMs of a single protein of interest, or
on PTMs of a protein population. Singular protein PTM
studies focus on antibody-specific precipitations and
affinity chromatography in order to analyze the protein
of interest and identify its PTMs. In contrast, PTM map-
ping of a protein population is a formidable task, as there
is a need to systematically assess modifications on a large
number of proteins. Both “bottom up” and “top down”
proteomics followed by MS have been used to identify
different PTM combinations and interactions in order to
understand the interplay between different cellular stim-
ulation and physiological responses.100
M A L D I-I M AG I N G
Matrix-assisted laser desorption ionization (MALDI)
imaging mass spectrometry (MALDI-MSI) is a pow-
erful and diverse technology for analyzing the spatial
distribution of endogenous and exogenous compounds
directly from a tissue section.101 MALDI imaging records
the spectra from thin tissue sections in order to produce
molecular-weight encoded images of the distribution
of constituent biomolecules, and it was first applied to
the visualization of proteins in 1997 by Capriloli and
co-workers.102 The technique requires the sectioning of
thin slices of tissue, which are then coated with an appro-
priate matrix solution. The laser cuts across the tissue
section, causing the solvent to evaporate. Upon solvent
evaporation, the extracted molecules are co-crystallized
with the matrix. The matrix functions to absorb the laser
energy and facilitate desorption/ionization of the ana-
lyte molecules. Mass spectra are then acquired across the
tissue at defined coordinates, resulting in a dataset that
contains hundreds to thousands of individual spectra,
consisting of all ions detected at each location of acqui-
sition. Like 2-DE, MALDI imaging suffers from lack of
reproducibility whereby sample preparation and han-
dling are crucial to obtaining good images despite incon-
sistencies in matrix crystal formation resulting in errors in
analyte abundances and distribution making the need for
normalization imperative for accurate results.
 C H EM I C A L P R OT E O M I C S
Where LC-MS/MS, protein microarrays, and 2-DE have
increased our understanding of protein expression within
a cell, the focus has now turned to functional proteomics
whereby chemical proteomics gives researchers the abil-
ity to assess the function of a protein within a cell or tissue.
Activity-based probe-profiling (ABPP) looks at the enzymatic
activity of a particular protein family.103 ABPP experiments
are designed to covalently label the active site of an enzyme
or enzymes of interest. Each probe contains a reactive group
and reporter tag.104 Activity-based probes can be used in both
a direct and indirect fashion. Direct probes can be developed
to target distinct enzyme families; for example, biotinylated/
fluorophore-tagged fluorophosphonates (FPs) target the ser-
ine hydrolase superfamily.104 Indirect strategies can be used
in the profiling of proteins lacking affinity labels. Libraries of
candidate probes synthesized and screened against complex
proteomes can identify “specific” protein-labelling events.104
Chemical proteomics can be used for targeted discovery
experiments and inhibitory experiments.
I N T E R AC TO M I C S
As well as expression, structural, and functional analysis
of an individual protein or group of proteins within a cell,
it is also important to note that proteins function as part of
multiple dynamic sets of proteins with the ability to interact
with each other and form complexes. An integral part of pro-
teomic analysis is to identify these interacting protein play-
ers of the cell, through “interactomics.” Throughout the past
15 years, high-throughput methods have been developed to
map protein–protein interactions (PPI) on a large scale. The
interactome can be accessed using three approaches:  firstly,
by applying computational methods for systematic iden-
tification of protein interactions whereby algorithms are
developed to predict PPI105; secondly, by using the yeast
two-hybrid (Y2H) system for binary interactions106; and
thirdly, co-immuno or co-affinity purification followed by
MS to identify multi-protein complexes.107 These method-
ologies each have their advantages and disadvantages, yet the
problem still remains in assessing biophysical interactions,
true biological interactions, and false positives, as currently
there is a low degree of overlap between PPI studies.
SYS T EMS B I O L O GY
Systems biology is an inter-discipline that makes use of the
“-omics” datasets, including genomics, transcriptomics,
3 4   •   P rincip l e s o f G e no m ic M e dicin e
proteomics, and metabolomics, which together provide
a better understanding of biological systems, creating a
framework for investigating the complexity of biological
systems by defining the components of the system (see also
Chapter  6). With the development of new technologies
such as high-throughput sequencing and MS, we can gain a
global profile of health and disease, allowing the combined
genomic, transcriptomic, and proteomic data to develop
new approaches in personalized medicine.108 Systems biol-
ogy has led to the development of systems medicine or “P4”
medicine: predictive, preventive, personalized, and partici-
patory.109 This type of medicine will provide deep insights
into disease complexity, allowing the development of treat-
ment strategies for disease due to networks perturbed either
by genetic or by environmental cues. Although this type of
medicine is attractive, it does have downsides. The cost of
an individual patient molecular fingerprint would be great;
also, this type of medicine would also depend heavily on
technology.108
C O N C LU S I O N S
It is safe to say that mass spectrometry has been critical
to the advances made in proteomics. Where genomics has
evolved to be able to map entire genomes, proteomics has
lagged behind, in part due to the enormous molecular
complexity and dynamic nature of the proteome, which
poses larger analytical challenges than genomics or tran-
scriptomics. Here we have aimed to highlight how pro-
teomic techniques have evolved and advanced in order
to meet the enormous task of deciphering the proteome.
Despite the hurdles outlined, proteomic technologies
have a significant role to play in biological research and the
life sciences. As MS capabilities continue to improve and
newer approaches are developed, the full characterization
of the complete human proteome and more routine analy-
sis of proteomes should come closer to fruition. There
remain, however, significant demand and opportunity for
the development of innovative and ground-breaking new
technologies—technologies that might operate with an
ability to amplify proteins or the analytical signals derived
from them and might be able to analyze proteins in a mul-
tiplexed manner with appropriate temporal and spatial
resolution.
R E C O MME N D E D R E A D I N G
1. Görg A, Drews O, Lück C, Weiland F, Weiss W. 2-DE with IPGs.
Electrophoresis. 2009 Jun; 30 Suppl 1:S122–S132.
 2. Nagaraj N, Wisniewski JR, Geiger T, Cox J, et al. Deep proteome
and transcriptome mapping of a human cancer cell line. Mol Syst
BiolMol Syst Biol. 2011;7:548.
3. Bantscheff M, Lemeer S, Savitski MM, Kuster B. Quantitative mass
spectrometry in proteomics: critical review update from 2007 to the
present. Anal Bioanal Chem. 2012;404(4):939–965.
4. Elliott MH, Smith DS, Parker CE, Borchers C. Current trends in
quantitative proteomics. J Mass Spectrom. 2009;44:1637–1660.
5. America AH, Cordewener JH. Comparative LC-MS: a landscape of
peaks and valleys. Proteomics. 2008;8:731–749.
6. Lemoine J, Fortin T, Salvador A, et al. The current status of clinical
proteomics and the use of MRM and MRM(3) for biomarker vali-
dation. Expert Rev Mol Diagn. 2012;12:333–342.
7. Gillette MA, Carr SA. Quantitative analysis of peptides and
proteins in biomedicine by targeted mass spectrometry. Nature
Methods. 2013;10:28–34.
8. Picotti P, Abersold R. Selected reaction monitoring-based pro-
teomics: workflows, potential, pitfalls and future directions. Nature
Methods. 2012;(9)6:555–566.
9. Mueller LN, Brusniak MY, Mani DR, Aebersold R. An assessment
of software solutions for the analysis of mass spectrometry based
quantitative proteomics data. J Proteome Res. 2008;7, 51–61.
10. Cham Mead JA, Bianco L, Bessant C. Free computational resources
for designing selected reaction monitoring transitions. Proteomics.
2010;10:1106–1126.
11. Lee JR, Magee DM, Gaster RS, LaBaer J, Wang SX. Emerging
protein array technologies for proteomics. Expert Rev Proteomics.
2013;10(1):66–75.
12. Kamath KS, Vasavada MS, Srivastava S. Proteomic databases and
tools to decipher post-translational modifications. J Proteomics
2011;75(1):127–144.
13. Vickerman J. Molecular imaging and depth profiling by mass
spectrometry—SIMS, MADI or DESI? Analyst. 2011;136(11):
2199–2217.
14. Miao Q, Zhang CC, Kast J. Chemical proteomics and its
impact on the drug discovery process. Expert Rev Proteomics.
2012;(9)3:281–291.
15. Rix U, Superti-Furga G. Target profiling of small molecules by
chemical proteomics. Nature Chem Biol. 2009;5(9):616–624.
16. Ngounou Wetie AG, Sokolowska I, Woods AG, Roy U, Loo JA, Darie
CC. Investigation of stable and transient protein-protein interac-
tions: past, present, and future. Proteomics 2013;13(3–4):538–557.
17. Cox J, Mann M. Quantitative, high-resolution proteomics for
data-driven systems biology. Annu Rev Biochem. 2011;80:273–299.
R EFE R E N C ES
1. Legrain P, Aebersold R, Archakov A, et al. The human proteome
project: current state and future direction. Mol Cell Proteomics:
MCP. 2011 Jul;10(7):M111.009993.
1a. Wilhelm M, Schlegl J, Hahne H, et al. Nature. 2014
May;29;509(7502):582–587. doi: 10.1038/nature13319.
1b. Kim MS, Pinto SM, Getnet D, et al. Nature. 2014
May;29;509(7502):575–581. doi: 10.1038/nature13302.
2. Nilsson T, Mann M, Aebersold R, Yates JR, Bairoch A, Bergeron
JJM. Mass spectrometry in high-throughput proteomics:  ready
for the big time. Nature Methods (Internet). Nature Publishing
Group; 2010 Sep [cited Mar 1,  2012];7(9):681–685. Available
from: http://www.ncbi.nlm.nih.gov/pubmed/20805795
3. Kropinski AM, Waddell T, Meng J, et  al. The host-range,
genomics and proteomics of Escherichia coli O157:H7 bacte-
riophage rV5. Virology J (Internet). Jan 2013 [cited Apr 19,
2013];10(1):76. Available from:  http://www.pubmedcentral.
nih.gov/articlerender.fcgi?artid=3606486&tool=pmcentrez&r
endertype=abstract
H u m an P rot e o m ic s   •   3 5
4. Smejkal GB. Genomics and proteomics: of hares, tortoises and the
complexity of tortoises. Expert Rev Proteomics (Internet). 2012
Oct;9(5):469–472. Available from: http://www.ncbi.nlm.nih.gov/
pubmed/23194260
5. Vangala RK, Ravindran V, Ghatge M, et  al. Integrative bioinfor-
matics analysis of genomic and proteomic approaches to under-
stand the transcriptional regulatory program in coronary artery
disease pathways. PloS One (Internet). 2013 Jan [cited Mar
8,  2013];8(2):e57193. Available from:  http://www.ncbi.nlm.nih.
gov/pubmed/23468932
6. Altelaar AFM, Munoz J, Heck AJR. Next-generation pro-
teomics: towards an integrative view of proteome dynamics. Nature
Rev Genet (Internet). 2013 Jan [cited Feb 28, 2013];14(1):35–48.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/23207911
7. Venter JC, Adams MD, Myers EW, et al. The sequence of the human
genome. Science. 2001 Feb;291(5507):1304–1351.
8. Lander ES, Linton LM, Birren B, et al. Initial sequencing and analy-
sis of the human genome. Nature. 2001 Feb;409(6822):860–921.
9. Collins F, Green E, Guttmacher A. A vision for the future of
genomics research. Nature (Internet). 2003 Apr [cited Jun
21,  2012];422(6934):835–847. Available from:  http://www.
nature.com/ng/journal/v36/n11s/full/doifinder/10.1038%2Fnat
ure01626
10. Klose J. From 2-D electrophoresis to proteomics. Electrophoresis
(Internet). 2009 Jun [cited Mar 17,  2013];30 Suppl 1:S142–149.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/19517494
11. Klose J. Protein mapping by combined isoelectric focusing and elec-
trophoresis of mouse tissues. Hum Genet. 1975;26(3):231–243.
12. Moreira JM, Cabezón T, Gromova I, et  al. Tissue proteomics of
the human mammary gland: towards an abridged definition of the
molecular phenotypes underlying epithelial normalcy. Mol Oncol
(Internet). 2010 Dec [cited Apr 26, 2013];4(6):539–561. Available
from: http://www.ncbi.nlm.nih.gov/pubmed/21036680
13. Domon B, Aebersold R. Mass spectrometry and protein analysis.
Science. 2006;312(5771):212–217.
14. Cox J, Mann M. Is proteomics the new genomics? Cell.
2007;130(3):395–398.
15. Pennington SR, Wilkins MR, Hochstrasser DF, Dunn MJ.
Proteome analysis: from protein characterization to biological func-
tion. Trends Cell Biol (Internet). 1997 Apr;7(4):168–173. Available
from: http://www.ncbi.nlm.nih.gov/pubmed/17708936
16. Herbert AP, Riesen M, Bloxam L, et al. NMR structure of Hsp12,
a protein induced by and required for dietary restriction-induced
lifespan extension in yeast. PloS One (Internet). 2012 Jan [cited Apr
9, 2013];7(7):e41975. Available from: http://www.pubmedcentral.
nih.gov/articlerender.fcgi?artid=3407059&tool=pmcentrez&rend
ertype=abstract
17. Fenn J, Mann M, Meng C, Wong S. Electrospray ionization for mass
spectrometry of large biomolecules. Science (Internet). 1989 [cited
Jun 21,  2012];246(4926):64–71. Available from:  http://www.sci-
encemag.org/content/246/4926/64.short
18. Karas M, Bachmann D, Bahr U, Hillenkamp F. Matrix-assisted
ultraviolet laser desorption of non-volatile compounds. Int J Mass
Spectrom. 1987;78:53–68.
19. Evans VC, Barker G, Heesom KJ, Fan J, Bessant C, Matthews DA.
De novo derivation of proteomes from transcriptomes for transcript
and protein identification. Nature Methods. 2012;9(12):1207–1211.
20. Hoopmann MR, Moritz RL. Current algorithmic solutions
for peptide-based proteomics data generation and identifica-
tion. Curr Opin Biotechnol (Internet). 2013 Feb [cited Mar
3,  2013];24(1):31–38. Available from:  http://www.ncbi.nlm.nih.
gov/pubmed/23142544
21. O’Farrell PH. High resolution two-dimensional electrophoresis of
proteins. J Biol Chem. 1975;250(10):4007–4021.
22. Görg A, Wilhelm P, Siegfried G. The current state of two-
dimensional electrophoresis with immobilized pH gradients.
Electrophoresis. 1988;9:531–564.
 23. Görg A, Wildgruber R. Review: The current state of two-dimensional
electrophoresis with immobilized pH gradients:  Proteomics and
2-DE. Electrophoresis. 2000;21:1037–1053.
24. Unlü M, Morgan ME, Minden JS. Difference gel electrophore-
sis:  a single gel method for detecting changes in protein extracts.
Electrophoresis. 1997;18(11):2071–2077.
25. Meyfour A, Tavirani MR, Sadeghi MR. Common proteomic tech-
nologies, applications, and their limitations. Journal of Paramedical
Sciences (JPS). 2013;4:115–125.
26. Görg A, Drews O, Lück C, Weiland F, Weiss W. 2-DE with IPGs.
Electrophoresis. 2009 Jun;30 Suppl 1:S122–S132.
27. Rogowska-Wrzesinska A, Le Bihan M-C, Thaysen-Andersen M,
Roepstorff P. 2D gels still have a niche in proteomics. J Proteomics
(Internet). 2013 Jan 24 [cited Mar 5,  2013];1–10. Available
from: http://www.ncbi.nlm.nih.gov/pubmed/23353020
28. Shevchenko A, Tomas H, Havlis J, Olsen JV, Mann M. In-gel
digestion for mass spectrometric characterization of proteins
and proteomes. Nature Protocols (Internet). 2006 Jan [cited Mar
5,  2012];1(6):2856–2860. Available from:  http://www.ncbi.nlm.
nih.gov/pubmed/17406544
29. Lim H, Eng J, Yates JR, et al. Identification of 2D-gel proteins: a com-
parison of MALDI/TOF peptide mass mapping to μ LC-ESI tandem
mass spectrometry. J Am Soc Mass Spectrom. 2003 Sep;14(9):957–970.
30. Bantscheff M, Lemeer S, Savitski MM, Kuster B. Quantitative mass
spectrometry in proteomics: critical review update from 2007 to the
present. Anal Bioanal Chem. 2012 Sep;404(4):939–965.
31. Yates JR, Ruse CI, Nakorchevsky A. Proteomics by mass spectrom-
etry: approaches, advances, and applications. Annu Rev Biomed Eng
(Internet). 2009 Jan [cited Mar 9,  2012];11(c):49–79. Available
from: http://www.ncbi.nlm.nih.gov/pubmed/19400705
32. Brusniak M-YK, Chu CS, Kusebauch U, Sartain MJ, Watts JD,
Moritz RL. An assessment of current bioinformatic solutions for
analyzing LC-MS data acquired by selected reaction monitoring
technology. Proteomics. 2012 Apr;12(8):1176–1184.
33. Reker D, Malmström L. Bioinformatic challenges in targeted pro-
teomics. J Proteome Res. 2012 Sep;11(9):4393–4402.
34. Gonzalez-Galarza FF, Lawless C, Hubbard SJ, et  al. A critical
appraisal of techniques, software packages, and standards for quan-
titative proteomic analysis. OMICS. 2012 Sep;16(9):431–442.
35. Michalski A, Damoc E, Hauschild J-P, et al. Mass spectrometry-based
proteomics using Q Exactive, a high-performance benchtop quad-
rupole Orbitrap mass spectrometer. Mol Cell Proteomics:  MCP
(Internet). 2011 Sep;10(9):M111.011015. Available from: http://
www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3284220&t
ool=pmcentrez&rendertype=abstract
36. Nagaraj N, Wisniewski JR, Geiger T, et  al. Deep proteome and
transcriptome mapping of a human cancer cell line. Mol Syst Biol.
(Internet). 2011 Jan [cited Feb. 29,  2012];7(548):548. Available
from:  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid
=3261714&tool=pmcentrez&rendertype=abstract
37. Neilson KA, Ali NA, Muralidharan S, et  al. Less label, more
free:  approaches in label-free quantitative mass spectrometry.
Proteomics. 2011 Jan;11(4):535–533.
38. Elliott MH, Smith DS, Parker CE, Borchers C. Current trends in
quantitative proteomics. J Mass Spectrom. 2009;44(12):1637–1660.
39. Ong S-E. Stable isotope labeling by amino acids in cell culture,
SILAC, as a simple and accurate approach to expression proteomics.
Mol Cell Proteomics. 2002 May;1(5):376–386.
40. Gygi SP, Rist B, Gerber SA, Turecek F, Gelb MH, Aebersold
R. Quantitative analysis of complex protein mixtures using
isotope-coded affinity tags. Nature Biotechnology. 1999 Oct;17(10):
994–999.
41. Ross PL, Huang YN, Marchese JN, et al. Multiplexed protein quan-
titation in Saccharomyces cerevisiae using amine-reactive isobaric
tagging reagents. Mol Cell Proteomics:  MCP. 2004 Dec;3(12):
1154–1169.
3 6   •   P rincip l e s o f G e no m ic M e dicin e
42. Yao X, Freas A, Ramirez J, Demirev PA, Fenselau C. Proteolytic
18O labeling for comparative proteomics: model studies with two
serotypes of adenovirus. Analytical Chemistry. 2001 Jul;73(13):
2836–2842.
43. Zhang Z, Li M, Zhang G, et al. Identification of human gastric car-
cinoma biomarkers by differential protein expression analysis using
18O labeling and nanoLC-MS/MS coupled with laser capture
microdissection. Medical Oncology (Northwood, London, England).
2010 Jun;27(2):296–303.
44. Qian W-J, Monroe ME, Liu T, et al. Quantitative proteome analysis
of human plasma following in vivo lipopolysaccharide administra-
tion using 16O/18O labeling and the accurate mass and time tag
approach. Mol Cell Proteomics: MCP. 2005 May;4(5):700–709.
45. Ye X, Luke B, Andresson T, Blonder J. 18O stable isotope label-
ing in MS-based proteomics. Briefings in Functional Genomics &
Proteomics. 2009 Mar;8(2):136–144.
46. Aggarwal K, Choe LH, Lee KH. Shotgun proteomics using the
iTRAQ isobaric tags. Briefings in Functional Genomics & Proteomics.
2006 Jun;5(2):112–120.
47. Smolka MB, Zhou H, Purkayastha S, Aebersold R. Optimization
of the isotope-coded affinity tag-labeling procedure for quantitative
proteome analysis. Analytical Biochemistry. 2001 Oct;297(1):25–31.
48. Shiio Y, Aebersold R. Quantitative proteome analysis using
isotope-coded affinity tags and mass spectrometry. Nature Protocols.
2006 Jan;1(1):139–145.
49. Shiio Y, Donohoe S, Yi EC, Goodlett DR, Aebersold R, Eisenman
RN. Quantitative proteomic analysis of Myc oncoprotein function.
The EMBO Journal. 2002 Oct;21(19):5088–5096.
50. Tam EM, Morrison CJ, Wu YI, Stack MS, Overall CM. Membrane
protease proteomics: isotope-coded affinity tag MS identification of
undescribed MT1-matrix metalloproteinase substrates. Proceedings
of the National Academy of Sciences of the United States of America.
2004 May;101(18):6917–6922.
51. Ramus C, Gonzalez de Peredo A, Dahout C, Gallagher M, Garin J.
An optimized strategy for ICAT quantification of membrane pro-
teins. Mol Cell Proteomics: MCP. 2006 Jan;5(1):68–78.
52. Mann M. Functional and quantitative proteomics using
SILAC. Nature Reviews. Molecular Cell Biology (Internet). 2006
Dec;7(12):952–958. Available from:  http://www.ncbi.nlm.nih.
gov/pubmed/17139335
53. Everley P, Krijgsveld J, Zetter B. Quantitative cancer proteomics: sta-
ble isotope labeling with amino acids in cell culture (SILAC) as
a tool for prostate cancer research. Cellular Proteomics (Internet).
2004 Jul [cited Jun 21, 2012];3(7):729–35. Available from: http://
www.mcponline.org/content/3/7/729.short
54. Moser M, Ussar S, Thievessen I. SILAC mouse for quantitative pro-
teomics uncovers kindlin-3 as an essential factor for red blood cell
function. Cell (Internet). 2008 Jul [cited Jun 21, 2012];134(2):353–
364. Available from:  http://www.sciencedirect.com/science/
article/pii/S0092867408006958
55. Westman-Brinkmalm A, Abramsson A, Pannee J, et  al. SILAC
zebrafish for quantitative analysis of protein turnover and tissue
regeneration. J Proteomics. 2011 Dec;75(2):425–434.
56. America AH, Cordewener JH. Comparative LC-MS: a landscape of
peaks and valleys. Proteomics. 2008;8(4):731–749.
57. Mueller LN, Brusniak MY, Mani DR, Aebersold R. An assessment
of software solutions for the analysis of mass spectrometry based
quantitative proteomics. Data Reviews. 2008;51–61.
58. Hoekman B, Breitling R, Suits F, Bischoff R, Horvatovich P.
msCompare:  a framework for quantitative analysis of label-free
LC-MS data for comparative biomarker studies. Mol Cell Proteomics.
2012;M111.015974.
59. Gillette MA, Carr SA. Quantitative analysis of peptides and pro-
teins in biomedicine by targeted mass spectrometry. Nature Methods
(Internet). 2013 Jan [cited Mar 1,  2013];10(1):28–34. Available
from: http://www.ncbi.nlm.nih.gov/pubmed/23269374
 60. Domanski D, Percy AJ, Yang J, et  al. MRM-based multiplexed
quantitation of 67 putative cardiovascular disease biomarkers
in human plasma. Proteomics (Internet). 2012 Apr [cited May
14,  2012];12(8):1222–1243. Available from:  http://www.ncbi.
nlm.nih.gov/pubmed/22577024
61. Da Costa LA, García-Bailo B, Borchers CH, Badawi A, El-Sohemy
A. Association between the plasma proteome and serum ascor-
bic acid concentrations in humans. The Journal of Nutritional
Biochemistry. 2012 24(5):842–847.
62. Garcı B, Brenner DR, Nielsen D, et  al. Dietary patterns and eth-
nicity are associated with distinct plasma proteomic groups 1–3.
2012;(3):352–361.
63. Fortin T, Salvador A, Charrier JP, et  al. Multiple reaction moni-
toring cubed for protein quantification at the low nanogram/mil-
liliter level in nondepleted human serum. Analytical Chemistry.
2009;81(22):9343–9352.
64. Lemoine J, Fortin T, Salvador A, Jaffuel A, Charrier J-P,
Choquet-Kastylevsky G. The current status of clinical proteomics
and the use of MRM and MRM(3) for biomarker validation. Expert
Rev Mol Diagn. 2012 May;12(4):333–342.
65. Shi T, Su D, Liu T, et al. Advancing the sensitivity of selected reac-
tion monitoring-based targeted quantitative proteomics. Proteomics.
2012 Apr;12(8):1074–1092.
66. Anderson NL, Anderson NG. The human plasma proteome:  his-
tory, character, and diagnostic prospects. Mol Cell Proteomics.
2002;1(11):845–867.
67. Anderson NL, Anderson NG, Pearson TW, et al. A human proteome
detection and quantitation project. Mol Cell Proteomics:  MCP.
2009 May;8(5):883–886.
68. Bell AW, Deutsch EW, Au CE, et al. A HUPO test sample study
reveals common problems in mass spectrometry-based proteomics.
Nature Methods. 2009;6(6):423–430.
69. Cox J, Mann M. MaxQuant enables high peptide identification rates,
individualized p.p.b.-range mass accuracies and proteome-wide pro-
tein quantification. Nature Biotech (Internet). 2008 Dec [cited Feb
28,  2013];26(12):1367–1372. Available from:  http://www.ncbi.
nlm.nih.gov/pubmed/19029910
70. Mueller LN, Brusniak MY, Mani DR, Aebersold R. An assessment
of software solutions for the analysis of mass spectrometry based
quantitative proteomics data. J Proteome Res. 2008;7(01):51–61.
71. Mancuso F, Bunkenborg J, Wierer M, Molina H. Data extraction from
proteomics raw data: An evaluation of nine tandem MS tools using
a large Orbitrap data set. J Proteomics. 2012;75(17):5293–5303.
72. Cham Mead JA, Bianco L, Bessant C. Free computational resources
for designing selected reaction monitoring transitions. Proteomics.
2010 Mar;10(6):1106–1126.
73. Martin DB, Holzman T, May D, et al. MRMer, an interactive open
source and cross-platform system for data extraction and visual-
ization of multiple reaction monitoring experiments. Mol Cell
Proteomics: MCP. 2008 Nov;7(11):2270–2278.
74. Maclean B, Tomazela DM, Shulman N, et  al. Skyline:  an open
source document editor for creating and analyzing targeted pro-
teomics experiments. Bioinformatics. 2010;26(7):966–968.
75. Berrade L, Garcia AE, Camarero JA. Protein microarrays:  novel
developments and applications. Pharmaceut Res. 2011
Jul;28(7):1480–1499.
76. Nand A, Gautam A, Pérez JB, Merino A, Zhu J. Emerging technol-
ogy of in situ cell free expression protein microarrays. Protein & Cell.
2012 Feb;3(2):84–88.
77. Wolf-Yadlin A, Sevecka M, MacBeath G. Dissecting protein func-
tion and signaling using protein microarrays. Curr Opin Chem Biol.
2009 Oct;13(4):398–405.
78. Hesselberth JR, Miller JP, Golob A, Stajich JE, Michaud GA, Fields
S. Comparative analysis of Saccharomyces cerevisiae WW domains
and their interacting proteins. Genome Biol. 2006 Jan;7(4):R30.
79. Lim CS, Seet BT, Ingham RJ, Gish G, Matskova L. The K15 protein
of Kaposi’s sarcoma-associated herpes virus recruits the endocytic
H u m an P rot e o m ic s   •   3 7
regulator intersectin 2 through a selective SH3 domain interaction †.
2007;3:9874–9885.
80. Rual J-F, Venkatesan K, Hao T, et al. Towards a proteome-scale map
of the human protein–protein interaction network. Nature. 2005
Oct;437(7062):1173–1178.
81. Mueller C, Liotta LA, Espina V. Reverse phase protein microarrays
advance to use in clinical trials. Mol Oncol. 2010 Dec;4(6):461–481.
82. Yu X, Schneiderhan-Marra N, Joos TO. Protein microarrays for per-
sonalized medicine. Clin Chem. 2010 Mar;56(3):376–387.
83. Sanchez-Carbayo M, Socci ND, Lozano JJ, Haab BB, Cordon-Cardo
C. Profiling bladder cancer using targeted antibody arrays. Am J
Pathol. 2006 Jan;168(1):93–103.
84. Carlsson A, Wingren C, Ingvarsson J, et al. Serum proteome profiling
of metastatic breast cancer using recombinant antibody microarrays.
Eur J Cancer (Oxford, England: 1990). 2008 Feb;44(3):472–480.
85. Abd El-Rehim DM, Ball G, Pinder SE, et  al. High-throughput
protein expression analysis using tissue microarray technology
of a large well-characterised series identifies biologically dis-
tinct classes of breast cancer confirming recent cDNA expression
analyses. Int J Cancer (Journal International du Cancer). 2005
Oct;116(3):340–350.
86. Orchekowski R, Hamelinck D, Li L, et  al. Antibody microarray
profiling reveals individual and combined serum proteins associated
with pancreatic cancer. Cancer Res. 2005 Dec;65(23):11193–11202.
87. Chung GG, Yoon HH, Zerkowski MP, et al. Vascular endothelial
growth factor, FLT-1, and FLK-1 analysis in a pancreatic cancer tis-
sue microarray. Cancer. 2006 Apr;106(8):1677–1684.
88. Pirnia F, Pawlak M, Thallinger GG, et al. Novel functional profiling
approach combining reverse phase protein microarrays and human
3-D ex vivo tissue cultures: expression of apoptosis-related proteins
in human colon cancer. Proteomics. 2009 Jul;9(13):3535–3548.
89. Spurrier B, Ramalingam S, Nishizuka S. Reverse-phase protein
lysate microarrays for cell signaling analysis. Nature Protocols. 2008
Jan;3(11):1796–1808.
90. Rapkiewicz A, Espina V, Zujewski JA, et  al. The needle in the
haystack:  application of breast fine-needle aspirate samples
to quantitative protein microarray technology. Cancer. 2007
Jun;111(3):173–184.
91. Saviranta P, Okon R, Brinker A, Warashina M, Eppinger J,
Geierstanger BH. Evaluating sandwich immunoassays in microar-
ray format in terms of the ambient analyte regime. Clin Chem. 2004
Oct;50(10):1907–1920.
92. Choudhary C, Mann M. Decoding signalling networks by
mass spectrometry-based proteomics. Nature Rev. Mol Cell Biol
(Internet). 2010 Jun [cited Mar 1, 2012];11(6):427–439. Available
from: http://www.ncbi.nlm.nih.gov/pubmed/20461098
93. Engholm-Keller K, Larsen MR. Technologies and challenges in
large-scale phosphoproteomics. Proteomics (Internet). 2013 Feb 13
[cited Mar 2,  2013];910–931. Available from:  http://www.ncbi.
nlm.nih.gov/pubmed/23404676
94. Sylvestersen KB, Young C, Nielsen ML. Advances in characterizing
ubiquitylation sites by mass spectrometry. Curr Opin Chem Biol
(Internet). 2013 Feb [cited Mar 18, 2013];17(1):49–58. Available
from: http://www.ncbi.nlm.nih.gov/pubmed/23298953
95. Guo M, Huang BX. Integration of phosphoproteomic, chemical,
and biological strategies for the functional analysis of targeted pro-
tein phosphorylation. Proteomics (Internet). 2013 Feb [cited Mar
7, 2013];13(3–4):424–437. Available from: http://www.ncbi.nlm.
nih.gov/pubmed/23125184
96. Mann M, Jensen ON. Proteomic analysis of post-translational mod-
ifications. Nature Biotech (Internet). 2003 Mar;21(3):255–261.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/12610572
97. Fedjaev M, Parmar A, Xu Y, et al. Global analysis of protein phos-
phorylation networks in insulin signaling by sequential enrichment
of phosphoproteins and phosphopeptides. Mol Biosyst (Internet).
2012 Apr [cited Apr 26,  2013];8(5):1461–1471. Available
from: http://www.ncbi.nlm.nih.gov/pubmed/22362066
 98. Zhou H, Ye M, Dong J, et al. Robust phosphoproteome enrichment
using monodisperse microsphere-based immobilized titanium (IV)
ion affinity chromatography. Nature Protocols (Internet). 2013 Feb
7 [cited Mar 1, 2013];8(3):461–480. Available from: http://www.
ncbi.nlm.nih.gov/pubmed/23391890
99. Hitchen PG, Twigger K, Valiente E, Langdon RH, Wren BW,
Dell A. Glycoproteomics:  a powerful tool for characterizing
the diverse glycoforms of bacterial pilins and flagellins. Biochem
Soc Transact (Internet). 2010 Oct [cited Apr 4,  2013];38(5):
1307–1313. Available from:  http://www.ncbi.nlm.nih.gov/
pubmed/20863304
100. 
Lanucara F, Eyers CE. Top-down mass spectrometry for the
analysis of combinatorial post-translational modifications. Mass
Spectrom Rev. 2013 Jan–Feb;32(1):27–42. Available from http://
www.ncbi.nlm.nih.gov/pubmed/22718314
101. Castellino S, Groseclose MR, Wagner D. MALDI imaging
mass spectrometry:  bridging biology and chemistry in drug
­development. Bioanalysis (Internet). 2011 Nov;3(21):2427–
2441. Available from:  http://www.ncbi.nlm.nih.gov/pubmed/
22074284
102. Caprioli RM, Farmer TB, Gile J. Molecular imaging of bio-
logical samples:  localization of peptides and proteins using
MALDI-TOF MS. Anal Chem (Internet). 1997 Dec 1;69(23):
4751–4760. Available from:  http://www.ncbi.nlm.nih.gov/
pubmed/9406525
103. Cravatt BF, Wright AT, Kozarich JW. Activity-based protein pro-
filing: from enzyme chemistry to proteomic chemistry. Annu Rev
Biochem (Internet). 2008 Jan [cited Mar 9,  2012];77:383–414.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/18366325
3 8   •   P rincip l e s o f G e no m ic M e dicin e
104. Speers AE, Cravatt BF. Chemical strategies for activity-based
proteomics. Chembiochem (Internet). 2004 Jan 3 [cited Mar 29,
2012];5(1):41–47. Available from:  http://www.ncbi.nlm.nih.
gov/pubmed/14695510
105. Zanivan S, Cascone I, Peyron C, et  al. A new computational
approach to analyze human protein complexes and predict novel
protein interactions. Genome Biol (Internet). 2007 Jan [cited Aug
10,  2012];8(12):R256. Available from:  http://www.pubmedcen-
tral.nih.gov/articlerender.fcgi?artid=2246258&tool=pmcentrez
&rendertype=abstract
106. Rajagopala SV, Sikorski P, Harry Caufield J, Tovchigrechko A,
Uetz P, Caufield JH. Studying protein complexes by the yeast
two-hybrid system. Methods (San Diego, Calif.) (Internet). 2012
Jul 24 [cited Aug 2, 2012]; Available from: http://www.ncbi.nlm.
nih.gov/pubmed/22841565
107. Voshol H, Ehrat M, Traenkle J, Bertrand E, Van Oostrum J.
Antibody-based proteomics: analysis of signaling networks using
reverse protein arrays. FEBS J (Internet). 2009 Dec [cited Aug
31, 2012];276(23):6871–6879. Available from: http://www.ncbi.
nlm.nih.gov/pubmed/19860827
108. Chen R, Snyder M. Systems biology:  personalized medicine for
the future? Curr Opin Pharmacol (Internet). 2012 Jul 31 [cited
Oct 5, 2012]; Available from:  http://www.ncbi.nlm.nih.gov/
pubmed/22858243
109. Hood L, Flores M. A personal view on systems medicine and the
emergence of proactive P4 medicine:  predictive, preventive, per-
sonalized and participatory. New Biotech (Internet). 2012 Oct
15 [cited Aug 31, 2012];29(6):613–624. Available from: http://
www.ncbi.nlm.nih.gov/pubmed/22450380
 4.
EPIGENETICS, EPIGENOMICS, AND HUMAN DISEASE
Aravind Ramesh, Cihangir Yandim, Theona Natisvili, Marta Mauri, Piu Pik Law,
Jackson P. K. Chan, Santiago Uribe Lewis, and Richard Festenstein*
INTRODUCTION
The different cellular phenotypes that compose multicel-
lular organisms are generated by the expression of house-
keeping and cell-type specific genes and repression of
inappropriate ones. The pattern of gene expression that
defines a cell type is termed the epigenotype, which is estab-
lished and maintained by “epigenetic” mechanisms able to
govern gene expression regardless of the underlying genetic
code.1 Genomic imprinting, where genes are expressed from
only one of the inherited parental alleles, represents a clas-
sical example of epigenetic gene regulation; memory of the
expression state, presumably established during gametic
meiosis, is thus transmitted to the zygote, maintained
throughout embryonic and post-embryonic develop-
ment, and reestablished during gametogenesis of the newly
formed organism in a sex-specific manner.2,3 It follows that
epigenetic “plasticity” would enable pluripotent stem cells
to give rise to a variety of epigenotypes. Conversely, induced
pluripotent stem cell (iPS) technology allows one to artifi-
cially stimulate epigenetic reprogramming of differentiated
adult cells and hence lead to their pluripotency.4–6
Cells can acquire an epigenotype by modulating the
availability of trans-acting factors to regulatory cis-acting
genetic elements that specify gene activity or inactivity.
Such availability can be controlled by the manner in which
DNA is packaged as chromatin inside the nucleus. Silent
genes may thus be packaged in “condensed” chromatin,
such as heterochromatin. Conversely, active genes may be
packaged in “open” chromatin, termed euchromatin.
Chromatin is formed when ~147 base pairs of DNA
associate with the histone octamer (two histone H3-H4
dimers plus two histone H2A-H2B dimers) to form the
nucleosome.7 Linker histones of the H1 class associate with
the DNA between single nucleosomes, facilitating a higher
level of organization, the so-called solenoid helical fibers
* Corresponding author
39
(30nm fibers).8,9 Although a detailed image of in vivo chro-
matin fibers has not been fully elucidated yet, it is gener-
ally accepted that higher order folding (condensation) and
unfolding (decondensation) are conceived as functionally
relevant chromatin states.10
Centromeric/pericentromeric heterochromatic regions
are gathered together in the mammalian interphase nucleus
as distinct domains, known as centromeric clusters.11 These
clusters have characteristic biochemical features that help
define heterochromatin:  abundant repetitive DNA (satel-
lite DNA),12 replication in mid- to late S phase,13 histone
hypoacetylation,14 and specific methylation at lysine 9 of
histone H3, the latter placed primarily by the histone meth-
yltransferase Suv39h.15 These modifications, together with
an RNA component16,17 and possibly the RNA interference
(RNAi) machinery,18–20 maintain the presence of hetero-
chromatin protein 1 (HP1), a component of constitutive
heterochromatin.21
Previously regarded as a static, condensed structure, het-
erochromatin is now known to be a highly dynamic struc-
ture. Fluorescence recovery after photobleaching (FRAP)
experiments using green fluorescent protein-tagged HP1
(HP1-GFP) revealed that HP1 is highly mobile at both het-
erochromatin and euchromatin.22–24 The dynamic exchange
of HP1 indicates the presence of “windows of opportunity”
for the binding of additional factors, and suggests that gene
regulation in heterochromatin results from a dynamic com-
petition between regulatory factors.25
Processes exist to allow changes in accessibility at both
the chromatin fiber and the core DNA. Covalent modifi-
cations in DNA (cytosine methylation or hydroxymeth-
ylation) and histones (lysine acetylation, methylation or
ubiquitination, serine and threonine phosphorylation,
and arginine methylation) modify the interaction between
histones, DNA, and chromatin binding factors,26 whereas
nucleosome remodeling factors modify core histone and
DNA accessibility.27 It is very likely that these biochemical
processes are regulated by an “epigenetic code” written as
modifications in DNA and histones and “read” by factors
that specifically recognize single or combined modifica-
tions.28–30 For example, histone acetylation, recognized by
proteins with bromodomains,31 as well as histone H3 lysine
4 methylation, recognized by WDR5 containing H3 lysine
4 methyltransferase complexes,32 are generally associated
with “open” chromatin and gene expression.14,33,34 On the
other hand, histone H3 lysine 9 or 27 methylation, recog-
nized by the chromodomains of HP1 and polycomb pro-
teins, respectively,35,36 and DNA methylation, recognized by
methyl binding domain (MBD) proteins,37 associate with
“condensed” chromatin and gene repression.38–40 Recently,
it has also been uncovered that DNA can be hydroxy-
methylated preferentially at cytosines of CpG residues
via ten eleven translocation (TET) enzymes.41 The exact
function of this mark has not been resolved yet; however,
it is thought to be an intermediary step between unmeth-
ylated CpGs and methylated ones.42 In line with this,
5-hydroxymethylcytosine residues were associated mainly
with transcriptionally active euchromatin and polycomb
silencing of formerly euchromatic promoters, where it pos-
sibly allows a rapid release of the repressed state in poised
genes.42,43
Methylated DNA, through MBDs, may target histone
deacetylases (HDACs)44,45 and H3 lysine 9 methyltransfer-
ases (e.g., Suv39h) to specific loci.46 Conversely, histone H3
lysine 9 methylation may bring about DNA methylation
(by association with DNA de novo methyltransferases)47 as
well as HP1.48,49 Therefore, both DNA and histone modi-
fication machineries may cross-talk to generate condensed
chromatin (see Figure  4.1a and 4.1b). DNA methylation
and H3 lysine 9 methylation may, however, function inde-
pendently49,50 and thus confer different degrees of epigen-
etic plasticity. DNA methylation, thought to be more stable
throughout meiosis and mitosis, may provide long-term
(repressive) epigenetic memory, whereas histone modifica-
tions are more labile epigenetic marks.
Many aspects of chromatin explained so far have been
evaluated using the chromatin immunoprecipitation
(ChIP) assay, which involves the immunoprecipitation of
protein-bound DNA using specific antibodies, followed
by PCR.51,52 Until recently, results were only obtained in
limited contexts. Importantly, though, development of
modern genomics techniques allowed high-throughput
sequencing of the genomic ChIP-DNA and thereby global
examination of chromatin.53 In line with this, the human
ENCODE project aimed to reveal a detailed picture (e.g.,
histone modifications, transcription factor binding) of how
4 0   •   P rincip l e s o f G e no m ic M e dicin e
a human genome functions.54–59 In addition, genome-wide
analyses on identical twins have furthered our understand-
ing of epigenetic differences.60,61
Where genes reside in the nucleus also provides a mech-
anism for epigenetic gene regulation. Genes may dynami-
cally relocate to associate with nuclear structures rich in
transcription factors required for expression,62 or be devel-
opmentally inactivated by their relocation to heterochro-
matic pericentromeric clusters.63 In the phenomenon of
position effect variegation (PEV), genes are repressed in a
proportion of the cells when abnormally juxtaposed to het-
erochromatin.64–67 That gene expression is affected by place-
ment of genes near heterochromatin has focused attention
on the associations between genes and their native cis-acting
regulatory elements. DNA sequences, such as locus control
regions68 or boundary elements/insulators,69 are thought to
regulate gene expression by establishing “permissive” chro-
matin domains70 or by facilitating nuclear relocation and/
or regulating interactions between gene promoters and
enhancer or silencer elements.69 Imprint control regions,
typically found as DNA differentially methylated regions
(DMRs) in imprinted gene clusters, are conceived as sites
from which chromatin structures are propagated bidirec-
tionally to control the expression of genes within imprinted
domains.71 Notably, recent advances now allow scientists to
study the three-dimensional organization of the genome
via chromosome-conformation capture (3C) approaches,
which are based on the restriction enzyme digestion of
crosslinked DNA and its subsequent ligation followed by
PCR.72 Within the human ENCODE project, a global
long-range interaction map of gene promoters was revealed
thanks to the fusion of chromosome confirmation-capture
technique with high-throughput sequencing.73
Disease states with an epigenetic basis are classified into
those where changes in chromatin structure at the deregu-
lated gene(s) result from mutations in DNA sequences in
the same chromosome; i.e., in cis, and those where genetic
mutations affect the genes that encode for factors that estab-
lish or maintain chromatin structures; i.e., in trans. Table 4.1
illustrates disease states with a confirmed or possible epi-
genetic basis, some of which we describe in greater detail
in the text. Importantly, epigenetic mechanisms appear to
play a key role in the development of numerous different
types of cancer, and given their potential reversible nature,
they offer exciting therapeutic targets. It has been also sug-
gested that epigenetic mechanisms might differ between
genders based on sex chromosome complements and the
male-determining gene “SRY”.74,75 These may explain dif-
ferent predisposition rates towards disease among different
genders.
 Free ebooks ==> www.Ebook777.com

(A) (B)
Suv39h Suv39h Suv39h

HP1 HP1 HP1 HP1 HP1 HP1

Suv39h CH3 CH3 CH3
CH3 CH3
HDAC
MBD
CH3

(C) Full frataxin expression (D) Silencing of frataxin expression

loose chromatin. Wild-type. condensed chromatin. FRDA.
(GAA)67–1700
Antisense
Hairpin

RNAPII
Triplex
MSH2/MSH3
(GAA)10–66

RNAPII

DNA methylation
Histone deacetylation

RNAPII
H3K4 demethylation
H3K9me2/me3
H3K27me3

(E) Domain 2 Domain 1

KNCQ1OT1
KCNQ1DN
SLC22A1L

CDKN1C
PHLDA2
NAP1L4

PHEMX
KCNQ1

ASCL2
TSSC4

CD81

IGF2

H19

L23
INS
TH

CTCF Enhancers

Cen. Tel.

ICR2 ICR1
(DMR-LIT1/KvDMR1) (H19DMR)

Figure 4.1 
Chromatin and disease. (A) Schematic representation of chromatin condensation. DNA (thick black line) wraps around histone octamers
(gray oval) to form nucleosomes. Methylated (CH3) DNA is recognized by DNA MBD proteins that complex with histone deacetylases (HDAC)
and histone H3 methyltransferases (e.g., Suv39h) to deacetylate histones, and specifically methylate lysine 9 of histone H3. A methylated lysine 9 of
H3 is specifically recognized by heterochromatin protein 1 (HP1), which interacts with Suv39h and normally exists as a dimer. (B) HP1 dimers may
link nucleosomes methylated at lysine 9 of H3 by Suv39h and facilitate chromatin compaction. (C) Potential mechanism underlying Friedreich’s
ataxia. In the presence of a normal GAA repeat in intron 1 of the frataxin gene, the region around the repeat and the nearby promoter may lie
in loosely packaged chromatin. Therefore, the gene is actively expressed. (D) In the presence of an expanded GAA repeat within intron1ofthe
frataxin gene,densely packagedheterochromatin mayform. This may induce similar changes at the nearby promoter (arrow). Variegated expression
of frataxin may then occur. The proportion of frataxin-positive cells may depend on GAA repeat length and the degree of heterochromatinization.
In transgenic mice, an untranslated GAA repeat expansion linked to a hCD2 reporter gene was associated with variegated expression and
heterochromatin formation close to the repeat and at the more distant promoter (see text). (E) Schematic representation of the imprinted domains
1 and 2 in chromosome 11p15.5 associated with the Beckwith–Wiedemann syndrome. Domain 1 contains insulin (INS), insulin-like growth factor
2 (IGF2), and H19. A differentially methylated region (DMR) upstream of H19 (H19 DMR) is the imprint control region 1 (ICR1) that, when
bound by CTCF on the unmethylated maternal chromosome (open lollipop), acts as a chromatin boundary which prevents an IGF2-downstream
enhancers (diamonds) interaction. Domain 2 contains the KCNQ1OT1 gene expressed from the paternal allele, and the PHLDA2, SLC22A1L,
CDKN1C, KCNQ1DN, and the KCNQ1 genes expressed from the maternal allele. A DNA DMR at the 50 end of KCNQ1OT1 (DMR-LIT1 or
KvDMR1; methylated on the maternal allele 002D filled lollipop) is the imprint control region for domain 2 required for repression of genes in the
paternal chromosome and expression of those in the maternal chromosome. Arrows indicate transcriptional orientation. Genes with two arrows
represent non-imprinted genes (modified from Robertson, 2005).

www.Ebook777.com
Table 4.1  GENETIC MUTATIONS GENERATING “EPIGENETIC” DISEASE IN CIS OR TRANS

MUTATION DISEASE OMIM

In cis Triplet repeat expansion Friedreich ataxia 229300
Myotonic dystrophy 160900
Fragile X syndrome 158900
Repeat contraction Facioscapulohumeral dystrophy 309550
Locus control region β-Thalassemia 141900
Imprint control region and/or chromatin boundary Beckwith–Wiedemann syndrome 130650
­genetic or epigenetic mutation
Imprint control region genetic or epigenetic mutation Prader–Willi/Angelman syndrome 176270/105830
Imprint control region genetic or epigenetic mutation TNMD 601410
Imprinted gene genetic and/or epigenetic mutation AHO/PHP-Ia and PHP-Ib 103580 and 603233
In trans DNMT3B, DNA methyltransferase ICF syndrome 242860
MECP2, methyl DNA binding protein Rett syndrome 312750
ATRX, chromatin remodeller ATR-X syndrome 301040
NSD1, histone methyl transferase Sotos syndrome 117550
RSK2, histone H3 kinase Coffin–Lowry syndrome 303600
SMARCAL1, chromatin remodeller? SIOD 242900
CBP, CREB binding protein. Rubinstein–Taybi syndrome 180849
Hairless Atrichia 209500
Emerin EDMD 310300
Lamin B receptor Pelger–Huet anomaly 169400
Abbreviations: OMIM, online Mendelian inheritance in man (www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=OMIM); TNMD, transient neonatal diabetes mel-
litus; AHO/PHP-Ia/PHPIb, Albright hereditary osteodystrophy/pseudo-hypoparathyroidism type Ia/Ib; ICF, immunodeficiency, centromere instability, and facial
anomalies; SIOD, Schimke immuno-osseous dysplasia; EDMD is X-linked Emery–Dreifuss muscular dystrophy (additional examples can be found in Hendrich and
Bickmore, 2001; Bickmore and van der Maarel, 2003; Jiang, Bressler, et al., 2004; Robertson, 2005).

G E N ET I C MU TAT I O N S 25. Progressive ataxia, cardiomyopathy, and associated

A FFE C T I N G   E P I G E N ET I C
R E GU L AT I O N   I N C I S
Several human diseases are associated with the expansion
of untranslated trinucleotide repeats, and their molecu-
lar pathogenesis may be mediated by effects on chromatin
packaging of nearby genes. These include Friedreich’s ataxia,
myotonic dystrophy, and fragile X syndrome. Deletion of
repetitive DNA may deregulate nearby genes via an epi-
genetic effect in facioscapulohumeral dystrophy. Nearby
genes may also be deregulated by genetic mutations in
imprint centers that control the expression of genes within
imprinted gene domains.
FR I E D R E I C H ’S ATAX I A
Friedreich’s ataxia (FRDA) was described by Nicholaus
Friedreich in 1863. It is the most common of the heredi-
tary ataxias, with a prevalence of 1 in 50,000. FRDA is an
autosomal recessive disease with age of onset usually before
diabetes are the core features.76 The most common genetic
abnormality is a homozygous expanded GAA trinucleo-
tide repeat in the first intron of the frataxin gene, located
on chromosome 9q.77 How this expansion is exactly trig-
gered remains elusive; however, mechanisms in relation to
DNA repair, replication, or transcription were suggested
to be important in this phenomenon.78 The frataxin gene
encodes the mitochondrial protein frataxin, which is
involved in the biogenesis of iron-sulfur clusters and is
therefore vital for iron homeostasis.79 The correlation of
the severity of certain clinical features and age of onset
with the shorter of the two expanded repeats80–84 may be
explained by frataxin expression’s being inversely propor-
tional to the length of the expanded GAA repeat, which
is particularly true for smaller expansions.85 Therefore,
residual expression from the allele with the shorter expan-
sion may be important in modulating disease severity.
Interestingly, some atypical patients have GAA repeats of
similar length to those in patients with more classical fea-
tures.80,86,87 Factors such as environment, modifier genes,

4 2   •   P rincip l e s o f G e no m ic M e dicin e
and somatic mosaicism may play a role in such phenotypi-
cal variation.
FRATA X IN G E N E R E P R E S S I O N
I N FR DA
Abnormally expanded GAA repeats within the frataxin
gene impair frataxin expression. This impairment could be
a direct result of physical blockage effects caused by unusual
conformations of DNA adopted by GAA triplets on the
elongation of transcription (Figure  4.1c and d). The most
common such structure is hairpin DNA resulting from
unusual hydrogen bonding between G•G, G•A and A•A,
and RRY (purine:purine:pyrimidine) triplexes, which is
also known as “sticky” DNA.88 Consistent with the idea of
transcriptional blockage effects, in vitro transcription assays
based on RNAse protection and northern blots revealed a
transcriptional elongation defect in the presence of expanded
GAA repeats.89–95 Moreover, ChIP assays performed on
lymphoblastoid cell lines using antibodies against initiating
and elongating forms of RNAPII, as well as histone marks
associated with transcriptional elongation (i.e., H3K36
and H3K79 methylation) also underlined a deficit in tran-
scriptional elongation on the frataxin gene.96–98 Whether
expanded GAA triplets affect the initiation of RNAPII at
the frataxin gene promoter, however, remains elusive.
Importantly, expanded GAA repeats are also associ-
ated with the heterochromatinization of the frataxin gene.
The first in vivo experiments to suggest that GAA-repeat
expansions could trigger heterochromatin formation were
performed in transgenic mice. Here, a (GAA)200 repeat
expansion was linked to a human CD2 (hCD2) reporter.
The direct inhibitory transcriptional effect of GAA repeats
on DNA transcription was excluded as the GAA repeat was
linked to the 3′ untranscribed region of the hCD2 trans-
gene. In this transgenic mouse model, the hCD2 reporter
gene alone is sensitive to juxtaposition to constitutively
tightly packaged DNA (heterochromatin), e.g., centro-
meres, and results in variegated hCD2 expression on T cells,
or PEV.66 In PEV, rather than gene expression being silenced
in all cells, a proportion of cells become silenced with the
remaining continuing to express. Linking a GAA repeat
expansion to the hCD2 transgene also resulted in PEV99 but
importantly this occurred even when the transgene was sit-
uated in regions of the chromosome that are usually loosely
packaged in euchromatin, suggesting that the presence of
GAA repeats induced chromatin condensation and hetero-
chromatin formation. In T cells where hCD2 was silenced,
DNase I hypersensitive site analysis showed condensation
of chromatin packaging at the promoter of the gene. This
Epig e n etic s , Epig e no m ic s , and H u m an D i s e a s e   •   4 3
silencing was also modified by altering the dosage of HP1β,
an important component of heterochromatin.99
In recent years, it has been shown that a similar het-
erochromatinization also takes place at the pathologically
silenced frataxin gene (Figure 4.1c and d). DNA methylation
levels in lymphoblastoid cells, peripheral blood mononuclear
cells (PBMC),100–102 as well as nerve tissues in FRDA mouse
models103 were found to be higher on the frataxin gene in
the presence of expanded GAA repeats. This result was also
accompanied by an increase in heterochromatic histone marks
(i.e., H3K9 di- and tri-methylation and H3K27 trimethyl-
ation) and an overall decrease in histone acetylation in the
flanking regions of expanded GAA repeats in patient-derived
lymphoblastoid cells/PBMCs,96–98,102,104 fibroblasts,105 and
FRDA mouse models.103 Indeed the pattern of heterochro-
matic marks on pathologically silenced alleles supports the
hypothesis that heterochromatin is spreading bidirectionally
from the GAA repeats causing the silencing of the gene.88
How exactly heterochromatin is triggered in FRDA
is an important question that remains to be answered.
Interestingly, unusual DNA conformations adopted by
expanded GAA repeats were shown to be recognized by the
cell’s mismatch repair mechanism. MSH2/MSH3 dimers
were shown to be attracted by expanded GAA repeats
(Figure 4.1d) in various FRDA models, including iPS cells
derived from patients.106–110 Some studies suggest hetero-
chromatinization as a protective response against faulty
transcription, which may be caused by DNA damage.111–116
Moreover, De Biase et  al.105 reported increased antisense
transcription in pathological frataxin alleles and suggest
this as a potential trigger for the heterochromatinization.
However, it is still unclear whether antisense transcription
may lead to heterochromatic silencing in mammals.117 De
Biase et al.105 also described a CTCF binding site near the
promoter of frataxin. CTCF is a chromatin insulator pro-
tein associated with enhancer blocking or chromatin insu-
lation activities.118,119 The effect of CTCF on frataxin has
not been clearly characterized yet; however, the study of De
Biase et al. reports a depletion of CTCF binding on silenced
alleles. Interestingly, knockdown of CTCF in healthy and
patient fibroblasts resulted in increased antisense tran-
scription in the promoter of the gene. This may imply that
CTCF has an inhibitory function against the spreading of
heterochromatin nucleated by GAA repeats.
One of the challenges in the FRDA field is limited access
to primary nerve tissue, which is predominantly affected in
this disease. Most of the studies so far have presented results
obtained from Epstein Barr virus–transformed lymphoblas-
toid cells derived from patients. Other sources of research were
fibroblasts and primary peripheral blood mononuclear cells,
 Free ebooks ==> www.Ebook777.com
as well as mouse models, which only exhibit a mild disease
phenotype.120 Notably, the arrival of iPS technology in the
recent years has allowed scientists to differentiate neuron-like
cells using fibroblasts obtained from FRDA patients.110,121
A lot has been resolved about the pathological silenc-
ing of the frataxin gene in the last decade. Importantly,
uncovering different aspects of heterochromatin brought
up the possibility of treating FRDA with HDAC inhibi-
tors, which could reduce histone deacetylation and thereby
a subsequent methylation. Indeed, synthetically derived
HDAC inhibitor BML-210 and its derivatives were shown
to upregulate frataxin expression significantly in FRDA
cells.104,122–126 Moreover, a recent study from our group
showed that an HDAC class III (Sirtuin) inhibitor nicotin-
amide also upregulates frataxin in vitro, ex vivo and in vivo
(mouse model) (Chan et al., 2013).126a Clinical trials for
BML-210 derivatives are currently ongoing. Importantly, a
recent phase IIa clinical trial with oral-dosing of high-dose
nicotinamide revealed significant up-regulation of the FXN
gene bringing its expression level towards that found in
asymptomatic carriers (Libri and Yandim et al. in press).126b
Undoubtedly, identification of specific epigenetic modifiers
responsible for the silencing of frataxin will help scientists
to further develop such radical therapeutic strategies, which
specifically address the primary cause of this currently incur-
able disease rather than its symptoms.
MYOTO N I C DYS T RO P H Y
Myotonic dystrophy is an autosomal dominant disease that
is the most common adult form of muscular dystrophy.
It is a multisystem neuromuscular disorder characterized
by clinical manifestations, including myotonia (skeletal
muscle hyperexcitability), progressive muscular dystrophy,
cataracts, cardiac conduction defects, cognitive deficits, and
endocrine anomalies.
Two forms of the disease with similar features are caused
by different microsatellite expansions in two different gene
loci. Myotonic dystrophy type 1 (DM1) is caused by a CTG
repeat expansion located in the 3′ untranslated region
(UTR) of the DMPK gene on chromosome 19q127–129
while myotonic dystrophy type 2 (DM2) is caused by
expansion of a CCTG repeat in the intron 1 of the Zinc
finger 9 (ZNF9) gene on chromosome 3q21.130 Expanded
CTG or CCTG repeats are highly unstable in both germ-
line and somatic tissues,130,131 and the length of the repeats
is correlated with the severity of symptoms and the earlier
disease onset.132,133
Although DM1 is now thought to be largely mediated
by an “RNA gain of function mechanism” in which CUG
4 4   •   P rincip l e s o f G e no m ic M e dicin e
www.Ebook777.com
repeat containing RNA transcripts cause sequestration
of muscle blind protein (MBNL) and increase CUGBP/
Elav-like family member 1 (CELF1) protein activi-
ties,134–137 creating a spliceopathy, the CTG expansion has
also been found to affect chromatin packaging of DNA.
For example, in fibroblasts from myotonic dystrophy
patients, the presence of a CTG expansion in the DMPK
gene is associated with condensation of chromatin, as indi-
cated by nuclease resistance at the six5 enhancer present in
the 3′ region of the DMPK gene, rendering it inaccessible
to transcription factors and causing downregulation of
six5 expression.138,139 CTG repeats also efficiently recruit
nucleosomes, the basic structural element of chroma-
tin.140,141 CTG expansions also behaved in a similar fashion
to pericentromeric heterochromatin, causing gene silenc-
ing and chromatin condensation in the hCD2 transgenic
mouse model.99 Some features of myotonic dystrophy
(particularly cataract formation as seen in six5 knock-out
mice142) may be secondary to deregulation of the six5 gene
located near the CTG repeat142–144; this deregulation may
be secondary to CTG repeat-induced chromatin conden-
sation. Evidence supporting this notion has emerged as
loss of CTCF binding to the regions flanking the CTG
repeat expansion has been shown at the DM1 locus, which
might be associated with DNA hypermethylation of these
regions.145,146 Spreading of heterochromatin at the CTG
expanded allele was indicated by the enrichment of his-
tone H3 lysine 9 (H3-K9) methylation and heterochro-
matin protein 1γ (HP1γ) recruitment where antisense
transcription of DMPK was activated. This might lead to
a wider dysregulation of the mRNA and protein amount
in DM1-affected cells.145 The exact mechanisms of CTG
repeat-mediated chromatin remodeling are still uncertain;
however, the repeat itself strongly stimulates nucleosome
formation.141 In addition, it has been suggested that DNA
binding proteins that recognize DNA triplet repeats147–149
contribute to the epigenetic changes seen at the DMPK
locus.
Other diseases where DNA repeats are found in
untranslated regions, like spinocerebellar ataxia type 10150
and myotonic dystrophy type 2,151 may share similar “epi-
genetic” molecular pathogenic mechanisms.
FR AG I L E X SY N D RO M E
Fragile X syndrome (FXS) is the most common inher-
ited form of mental retardation152 and is one of the
best-characterized forms of autism spectrum disorder
(ASD).153 It is an X-linked dominant disorder characterized
by variable penetrance. The name “fragile X” derives from
the observation that the cytogenetic band Xq27.3, where
the causative fragile X mental retardation 1 (FMR1) gene
resides, is a fragile site in individuals carrying the full muta-
tion (FM).154
As FMR1 is located on the X chromosome, the degree
of cognitive disability is more severe in males, who possess
only one X chromosome. Females (with two X chromo-
somes) manifest a less severe phenotype than males, which
is correlated to the extent of X inactivation on the abnormal
chromosome.155 FXS has an estimated frequency of 1/5,000
in males and 1/10,000 in females.156
The main disease manifestations are moderate to severe
intellectual disability, autistic features, seizures, hypersen-
sitivity to sensory stimuli, attention deficit, hyperactivity,
motor incoordination, growth abnormalities, sleep distur-
bances, connective tissue dysplasia, craniofacial abnormali-
ties, and macroorchidism.157
The syndrome is a trinucleotide repeat disorder caused
by the expansion of the triplet CGG in the 5′ untranslated
region (5′UTR) of the FMR1 gene.158 A CGG expansion
greater than 200 units results in hypermethylation at CpG
sites at the FMR1 promoter region; this is responsible
for the silencing of FMR1 and the subsequent loss of the
protein it codes for, the fragile X mental retardation pro-
tein (FMRP).154 FMRP is an RNA binding protein that is
able to bind to several mRNAs, including its own,159 and is
believed to be involved in the transportation of these target
mRNAs throughout neuronal dendrites and in the inhibi-
tion of their translation upon stimulation of the metabo-
tropic glutamate receptor 5 (mGluR5) at the synapse.160 It is
ubiquitously expressed until day 14 of embryonic develop-
ment, after which its expression becomes restricted to the
brain (specifically in neurons) and the gonads.161 Loss of
FMRP is thought to be critical due to the important role it
plays in neuronal function. Abnormal dendritic spines were
observed in both FXS patients and in FMR1 knockout
mice, supporting FMRP involvement in synaptic matura-
tion.162 More generally, the absence of FMRP appears to lead
to a global increase in brain protein synthesis,163 and several
studies in human patients are currently testing whether this
observation could be a biomarker of the disease.
If cells from FXS patients are treated with a DNA
methylation inhibitor (5-aza-2-deoxycitidine), the level
of CpG methylation decreases and FMR1 expression is
reactivated,164–168 suggesting that DNA methylation is the
major factor causing FMR1 silencing, rather than the trip-
let expansion itself. DNA methylation has been to shown
to also cause local histone deacetylation, creating another
mechanism of transcriptional silencing.169 Treatment of
the same lymphoblastoid fragile X cells with trichostatin
Epig e n etic s , Epig e no m ic s , and H u m an D i s e a s e   •   4 5
A  (TSA, an HDAC inhibitor) also resulted in moder-
ate transcriptional activation of FMR1, suggesting that
in fragile X, histones are deacetylated at the inactive
promoter. Furthermore, treatment of fragile X cells with
HDAC inhibitors and 5-azadC synergistically activated
transcription.170
To date, the exact mechanisms behind CGG expansion
and the consequent alteration in FMR1 transcription and
translation are still not fully understood. The generation of
stable cell lines harboring the FMR1 5′-UTR with varying
CGG repeat lengths targeted to the correct gene locus have
proven to be a useful model for studying FXS. The promoter
with variable (CGG)n length has been fused to the coding
sequence of a reporter gene. This construct has shown that
a full-mutation CGG repeat length inhibits reporter gene
expression, whilst a premutation CGG repeat does not
affect reporter gene expression. Therefore this model could
be a better tool to elucidate the molecular mechanisms of
FMR1 deregulation in FXS.171
Transcriptional silencing of the FMR1 gene due to
hypermethylation of CpG islands and the consequent loss
of FMRP expression is still considered to be the major
cause of the disease.154 However, a mouse model of FXS
with mice carrying long CGG repeats of nearly 230 units
showed high levels of the FMR1 mRNA, although low
levels of the FMRP protein, arguing against a purely tran-
scriptional deficit. Promoters in these mice do not show
the abnormal methylation described above, which suggests
that modeling FXS in mice requires more genetic manipu-
lation, and that perhaps the threshold number of repeats
in mice might be higher than the level observed in human
subjects.172
In addition, within humans there are cases of males
expressing FMR1 to some extent despite carrying the full
mutation. Mosaicism of the gene promoter methylation
pattern is thought to underlie this, allowing some transcrip-
tion to occur. The presence of three types of mosaicism was
tested in cells derived from male expressing patients, and
the data suggests that inter-cell mosaicism in DNA methyl-
ation patterns might explain the presence of FMR1 mRNA
in some FXS=affected individuals.173
The CGG repeat region is unstable, and repeat length
can vary in unaffected individuals from 6–55 repeats. The
instability of the region can result in an expansion of this
region upon maternal transmission to the next generation.
A  (CGG)n range between 55–200 is referred to as the
premutation (PM).174 It was previously thought that carri-
ers of the PM simply had a higher risk of developing FXS
upon transmission of the PM allele to the next generation.
However, it has become evident that certain carriers show
autistic features or anxiety disorders.175–177 Nearly 20%
of female PM carriers manifest premature ovarian failure
(POF:  cessation of menstruation at or before 40  years of
age).178 PM carriers are also at risk of developing a progres-
sive neurodegenerative disorder called fragile X-associated
tremor/ataxia syndrome (FXTAS).179 The PM expansion
results in a 2- to 8-fold upregulation of FMR1 mRNA,
but the FMRP translation is inhibited.180–182 It is currently
believed that a toxic RNA gain of function of the CGG
expanded FMR1 mRNA is responsible for FXTAS.152,183,184
Recently, the use of human embryonic stem cells
derived from genetically diagnosed preimplantation
embryos has allowed scientists to look at the temporal
order of the events that lead to FMR1 silencing. This sys-
tem has shown that transcriptional downregulation of
FMR1 and some other chromatin modifications occur
before DNA methylation and that they might be responsi-
ble for the initiation of FMR1 inactivation. In fact, undif-
ferentiated human embryonic stem cells express FMR1,
and their DNA is unmethylated despite the presence of an
FM. But DNA methylation and other histones’ modifica-
tions occur upon differentiation of these cells, establishing
epigenetic silencing of FMR1 and leading to intellectual
impairment.172
More recently, human-induced pluripotent stem cells
from FXS patients and from healthy controls were used
as an FXS model. This model looked at the FMRP pro-
tein in early neurodevelopment prior to synaptogenesis,
and from it, a direct link has been identified between the
epigenetic modifications of the FMR1 gene, the conse-
quent loss of FMRP expression, and aberrant neuronal
differentiation.185
In addition to methylation, hydroxymethylation is
increasingly thought to play a role in FXS pathophysi-
ology. 5-hydroxymethylcytosine (5-hmC) is a modified
form of cytosine thought to play an important role in
neurodevelopment. For example, genome-wide DNA
hydroxymethylation studies have revealed a positive cor-
relation between 5-hmC levels and cerebellum devel-
opment.186 This modified cytosine is enriched in exons
and untranslated regions of protein-coding genes, but
it is absent on introns and intergenic regions. Different
regions of DNA also show different hydroxymethylation
patterns, depending on the stage of development. Several
mRNA targets of FMRP are differentially enriched in
hydroxymethylated regions during development, and
these hydroxymethylation patterns are disrupted in
several autism genes. Consequently, the disruption of
5-hmC-mediated epigenetic regulation might contribute
to the pathogenesis of FXS.186
4 6   •   P rincip l e s o f G e no m ic M e dicin e
FAC I O S C A P U L O HUM E R A L
DYS T RO P H Y
Facioscapulohumeral dystrophy (FSHD) is the third
-most-common inherited muscular dystrophy, inherited
in an autosomal dominant fashion and affecting approxi-
mately one in 20,000 individuals worldwide.187 The con-
dition usually manifests clinically by the second decade,
beginning with progressive weakness of the facial, scapular,
and humeral muscles, and later involving the abdominal
musculature and lower limbs.187 It is primarily a disease of
skeletal muscle, but retinal telangiectasia, sensorineural
deafness, cardiac arrhythmias, and mental retardation also
occur.188,189 Its molecular genetic pathogenesis is thought to
be mediated by epigenetic mechanisms.
Contraction of a tandem array of 3.3 kilobase (kb)
D4Z4 repeats lying in the subtelomeric region of chromo-
some 4q is associated with FSHD.190,191 Ninety-five percent
of FSHD cases are due to this deletion, and contraction to
a threshold of less than 11 repeats results in disease.191,192
There are problems in correlating genotype and phenotype
in FSHD, due to the large variability in symptom severity,
even in patients from the same affected family193; however,
generally, the shorter the remaining D4Z4 tandem array,
the earlier the onset and more severe the disease.192,194 Each
D4Z4 repeat unit of the array contains several GC-rich
sequences and also an open reading frame (ORF) of a double
homeobox transcription factor designated DUX4.195 Several
sequences similar to this D4Z4 array are found throughout
the human genome, with the array at 10q26 almost com-
pletely identical to the one at 4q35.196 Large contractions of
this other D4Z4 array, found in 10% of the normal popula-
tion, are not associated with FSHD, confirming 4q35 as the
causative genome region.197 Importantly, haploinsufficiency
of the entire chromosome 4q subtelomeric region, includ-
ing the D4Z4 tandem repeats and nearby genes, does not
cause FSHD,198 suggesting a gain of function underlying the
pathogenesis of this disease.
The genetic basis of FSHD is further complicated by
the fact that additional 4q variants have been described; for
example, 4qA, 4qB, and 4qC. Of these, FSHD appears to
be only associated with the 4qA variant.199–201 A particular
simple sequence length polymorphic site proximal to the 4q
array is found in three haplotypes, which, if a large D4Z4
repeat contraction is also present, seems to be specifically
associated with the FSHD phenotype.202,203 It therefore
appears that the D4Z4 repeat contraction will only cause
disease when found on a certain genetic background.
The way this repeat contraction causes disease is
thought to be epigenetically mediated. It was hypothesized
initially that, in healthy patients, D4Z4 repeat tracts are
heterochromatinized, and that variable spreading of het-
erochromatin silences nearby genes. Loss of these repeats
may produce a more open chromatin configuration, result-
ing in inappropriate de-repression of these nearby genes.204
However, early ChIP analysis looking at H4 acetylation
of the FSHD locus in human lymphoid cells as well as
human-rodent somatic cell hybrids suggested that regions
close to the D4Z4 repeats and of two nearby genes (FRG1
and ANT1) showed histones acetylated more in the pat-
tern of euchromatin than heterochromatin.205 Subsequent
ChIP analysis has looked at other histone modifica-
tions:  both repressive, such as trimethylation at lysine 9
and 27 on histone 3 (H3K9me3 and H3K27me3), as well
as markers of transcriptional activation such as dimeth-
ylation at lysine 4 on histone 3 (H3K4me2). This has in
fact shown that, in healthy controls, D4Z4 arrays display
both heterochromatic and euchromatic regions; while
in FSHD patients, there are significantly reduced levels
of H3K9me3, with unaltered levels of H3K27me3 and
H3K4me2,206 indicating a relative reduction in repressive
modifications, a more open chromatin configuration as a
consequence, and perhaps gene upregulation.
Along with histone modifications, DNA meth-
ylation patterns also reflect chromatin configuration.
Heterochromatic regions are usually hypermethylated, and
normal D4Z4 arrays reflect this. Contracted D4Z4 arrays,
however, display hypomethylation,207 and, rather akin to
phenotype, the level of this hypomethylation correlates
with repeat length: the shorter the array, the lower the level
of methylation.208 However, hypomethylation is also seen in
some unusual asymptomatic individuals who carry both the
contracted D4Z4 array as well as a permissive haplotype,
perhaps suggesting that hypomethylation is not sufficient
for disease onset.207 In addition, more profound levels of
hypomethylation are seen at D4Z4 repeats in the immu-
nodeficiency, centromere instability, and facial anomalies
(ICF) syndrome,207 which is phenotypically entirely differ-
ent from FSHD. In the same syndrome, H3K9me3 enrich-
ment is normal in at D4Z4,206 suggesting that it is the loss of
H3K9me3 that is crucial in FSHD, rather than any change
in DNA methylation.
The above indicates that transcriptional upregulation
is key to FSHD pathogenesis, and given the clearly dem-
onstrated association with the chromosome 4q35 region,
a number of candidate genes found in that area have been
investigated as potentially causative. These include FRG1,
FRG2, ANT1, and, more recently, DUX4c and DUX4.
FRG1 in particular seemed a promising candidate, as func-
tional studies suggested a role in muscle development.209
Epig e n etic s , Epig e no m ic s , and H u m an D i s e a s e   •   4 7
Early work suggested that FRG1, FRG2, and ANT1 seemed
to be upregulated in muscle cells from FSHD patients in a
manner specific to FSHD (not found in other hereditary
myopathies) and also to muscle (not replicated in patient
lymphocytes).210 Subsequent studies failed to fully corrobo-
rate these findings, being unable to replicate any increase
in FRG1 or ANT1 expression205,211–213 and showing that
FRG2 was only upregulated in FSHD myoblasts and not
in mature myocytes.214 In fact, studies in certain FSHD
families have largely ruled out these genes as causative
for the disease. In one family, the disease-associated allele
showed a large deletion in the D4Z4 repeat array, includ-
ing both the DUX4c and FRG2 genes, arguing against a role
for them in causing disease.215 Similarly, in another family,
the pathological allele was actually found on chromosome
10q26, where the short repeat array had been extended by a
D4Z4 fragment translocated from 4q35. This translocated
fragment included part of the distal D4Z4 fragment, allow-
ing stable DUX4 mRNA expression, but did not include
FRG1, ANT1 or DUX4c, again arguing against a role for
these genes,202 and suggesting that DUX4 is the key gene
involved in FSHD.
Early work looking at DUX4 was greatly hampered by
the difficulty of detecting specific mRNA transcripts. This,
coupled with a lack of an identifiable polyadenylation site,
led to the theory that DUX4 was actually a non-functional
pseudogene.216 The major breakthrough arose with the
identification of a mature mRNA transcript containing the
DUX4 ORF, with RT-PCR.217 This transcript was shown
to originate from the distal D4Z4 unit and extend to an
adjacent region conferring the polyadenylated tail, termed
pLAM1.217 There is evidence that DUX4 expression at high
levels in muscle has numerous effects, including inhibi-
tion of differentiation,218 and induction of genes involved
in muscle atrophy, apoptosis and cell death, implying that
DUX4 overexpression could be responsible for FSHD.219,220
However, DUX4 mRNA abundance is very low; it was
not always detectable in FSHD muscle biopsies, and only
around one per 1,000 FSHD myoblasts expressed DUX4
in culture.221 Snider et al. (2010) have shown that the low
abundance in muscle actually reflects a small number of
nuclei expressing abundant amounts of DUX4, and that
DUX4 is highly expressed in human testis and germline
cells.221 This has led to a developmental model being pro-
posed for FSHD, whereby in normal individuals, DUX4 is
expressed in early development and is heterochromatically
silenced in mature tissues. In FSHD, there is a defect in this
silencing mechanism, which leads to occasional escape from
repression in muscle cells, with consequent DUX4 expres-
sion and cell death.221,222
 There have been great advances in our understanding
of the epigenetic mechanisms underlying FSHD recently.
Outstanding questions concern the exact function of
DUX4 and how expression in so few cells in adult tissue can
lead to a progressive myopathy.
G E N ET I C MU TAT I O N S A FFE C T I N G
E P I G E N ET I C R E GU L AT I O N I N T R A N S
I C F SY N D RO M E : T H E I M P O RTA N C E
O F D NA M ET H Y L AT I O N
ICF syndrome is an extremely rare autosomal reces-
sive disorder in which patients display immunodefi-
ciency, centromere instability (association, breakage, and
stretching of the pericentromeric heterochromatin of
chromosomes 1, 9, and 16), and facial anomalies. Mental
retardation and developmental delay are also observed.223
The disease maps to chromosome 20, and the gene respon-
sible is the DNA de novo methyltransferase DNMT3B.224
Mutations in both the catalytic domain responsible for
methyltransferase activity and the aminoterminal end of
the protein, which is likely to be responsible for its target-
ing to pericentromeric sequences,225,226 are present in ICF
patients.227,228 Even though deletion of the Dnmt3b cata-
lytic domain in mice results in perinatal lethality,229 mis-
sense mutations in this domain in ICF patients probably
impair rather than completely abolish enzymatic activ-
ity.227 Indeed, the majority of ICF patients have missense
mutations in the catalytic domain of DNMT3B, and no
patients have been found to be homozygous for nonsense
alleles.230 This has been further confirmed in a mouse
model of the disease where mutant alleles carrying mis-
sense mutations as in ICF were constructed. Unlike total
knockout, which is embryonically lethal, these mice are
alive at birth and display higher methylation levels sug-
gestive of a residual level of DNMT3B activity as hypoth-
esized in ICF patients.231
Around 40% of ICF patients do not have mutations in
DNMT3B yet have exactly the same clinical phenotype;
this group is denoted ICF type 2.232,233 Early hypotheses
for this subgroup of patients suggested alternative cata-
lytically inactive splice variants of DNMT3B might be
overexpressed, a phenomenon noticed in human hepato-
carcinogenesis with overexpression of the splice variant
DNMT3B4.234 However, the DNMT3B3 isoform remains
the most abundant isoform in ICF type 2 patients, arguing
against this alternative splicing mechanism.232 Recent work
has attempted to find another candidate gene that might
4 8   •   P rincip l e s o f G e no m ic M e dicin e
be responsible for ICF type 2. Homozygosity mapping and
whole-exome sequencing of patients with ICF syndrome
type 2 has identified a mutation in exon 3 of ZBTB24,235
and another mutation in the same gene has been found in
another pedigree.236 This gene is thought to be involved in
B cell differentiation,235 and it is mutations in this gene that
are thought to be responsible for ICF type 2. This gene also
appears to have a hypermethylated promoter in ICF Type
1 patients, thus ZBTB24 dysfunction might be a common
mechanism to account for the similar phenotypes seen in
both classes of patient.237
Reduced DNA methylation levels in ICF cells are not
global but primarily observed in specific DNA sequences
at specific loci. As expected from the sites of chromo-
somal instability, there is hypomethylation of satellite 2
and 3 of pericentromeric heterochromatin but only on
chromosomes 1, 16, and occasionally chromosome 9. At
these sites, a number of aberrant chromatin rearrange-
ments are seen,238 and these cytological abnormalities are
seen primarily in lymphocytes, although they have also
been noted in other cell types in ICF patients.239 Such
ICF-specific cytogenetic abnormalities are also seen in
similar frequency in normal lymphoblastoid cell lines
and lymphocytes that are treated with inhibitors of
DNA methylation,240,241 again confirming a direct link
between DNA methylation abnormalities and chromo-
somal instability as seen in ICF. Undermethylation of
sequences is seen in other locations such as at Yqh in males
and within the inactive X chromosome in females,239,242
but these changes are not likely to be biologically signifi-
cant, as sex-specific differences in disease severity are not
observed.243 A whole genome scan to identify sequences
that are consistently hypomethylated in lymphoblasts
from ICF patients compared to controls demonstrated
a methylation deficit on only a small proportion of the
genome; in particular, two types of repeats, one of which
was the D4Z4 repeats implicated in facioscapulohumeral
muscular dystrophy (FSHD; Table 4.1, and see above).244
Further work has raised the possibility that, in addition
to the hypomethylation seen in pericentromeric hetero-
chromatin mentioned above, there may be hypermeth-
ylation in other regions within genes and promoters,
suggesting that it may be the overall methylation pattern
in ICF that is perturbed, rather than hypomethylation
specifically.237
Why should particular sequences have altered meth-
ylation states in ICF cells? One possibility is that these
sequences are present at the chromosomal sites where the
DNMT3B enzyme is normally localized. The sub-cellular
localization of endogenous human DNMT3B has not yet
been reported, but exogenously tagged murine Dnmt3b
co-localizes with pericentric heterochromatin in some cell
types.225 The domain necessary for targeting DNMT3B to
heterochromatin has not been determined, but it is likely to
be at the aminoterminus where there are two domains com-
monly found in other chromatin-associated proteins. One
of these is a PHD finger similar to that found in ATRX,
a chromatin remodeling protein that is concentrated at
sites of heterochromatin and repetitive DNA sequences.245
There is also hypomethylation of specific repetitive DNA
sequences in ATR-X patients (but hypermethylation of
other sequences). However, there is no chromosome insta-
bility reported at the sites of hypomethylation in ATR-X
cells, and little phenotypical overlap between ATR-X and
ICF syndromes is observed (see ATR-X below). The other
domain that may be involved in targeting DNMT3B is a
PWWP (conserved proline and tryptophan) domain,
which binds DNA246 and, based on similarity to tudor and
chromodomains,247 may also recognize methylated proteins.
A homozygous missense mutation in the PWWP domain
of DNMT3B has been identified in ICF sibs, resulting in
a serine to proline change that may have a profound con-
sequence for the mutant protein’s structure.228 In addition,
the PWWP domain of murine Dnmt3b has been shown to
bind DNA non-specifically and to be required for target-
ing DNA methyltransferase activity to murine pericentric
sequences.226
Presumably, DNA hypomethylation in ICF syndrome
leads to deregulation of genes that perturb craniofacial,
cerebral, and immunological development. Microarray
analysis has been used to identify genes with significantly
altered mRNA levels in ICF lymphoblastoid cell lines
compared with controls.248 Many of the genes identified
have known roles in immune function in both B and T
cell lineages that could account for the immunodeficiency
consistently manifested in ICF syndrome. No alteration of
DNA methylation was detected at the promoters of any of
the deregulated genes tested,248 consistent with the find-
ings of the whole-genome scan,244 arguing against a direct
cis effect of a methylation abnormality. Furthermore, none
of these genes is located on chromosomes 1, 9, or 16. This
raises the question of how hypomethylation of specific
repetitive DNA sequences in ICF patients can lead to
altered expression of genes located at distant genomic sites.
One possibility is that the hypomethylation of satellite
DNA alters their heterochromatin properties, and that it
is the physical association of deregulated genes with these
domains in the nucleus that is aberrant in ICF cells.249
Silenced genes in human B and T lymphocytes have been
shown to co-localize with domains of pericentromeric
Epig e n etic s , Epig e no m ic s , and H u m an D i s e a s e   •   4 9
heterochromatin and to relocate away from these domains
upon gene activation.250 Centromere distribution within
the nucleus of lymphoid cells has been shown to vary with
different stages of differentiation, suggesting that the distri-
bution of heterochromatin can influence gene expression
in trans.251,252 The 3D organization of intranuclear pericen-
tric heterochromatin has been shown to be abnormal in
ICF patient lymphoblastoid B cells, at least for chromo-
some 1, and treatment with a demethylating agent partially
induces this heterochromatic remodeling.253,254 Further
work is needed to see if particular genes also show altered
spatial distributions in ICF cells in parallel to heterochro-
matic changes.
Alternatively, the loss of DNA methylation at large
arrays of satellite repeats may release or recruit protein
complexes and affect the balance of regulatory complexes
throughout the genome. Interestingly, ICF lymphoblastoid
cell lines showed altered binding patterns of HP1, with the
formation of large foci containing HP1 and components of
promyelocytic (PML) nuclear bodies that co-localize with
chromosome 1qh and 16qh DNA sequences.255 The altered
pattern was, however, only observed during the G2 phase
of the cell cycle and not in fibroblasts. These results suggest
that binding of HP1 is not dependent on DNA methyla-
tion (since the abnormal HP1 distribution only occurs at
one stage in the cell cycle) and also indicate that cell type–
specific defects in the timing of heterochromatin packaging
may be major determinants of chromosomal abnormalities
and gene deregulation.255 The aggregation of such chroma-
tin proteins in ICF may simply be a result of underconden-
sation of heterochromatin at those specific loci, but it also
may have an effect in trans on gene expression elsewhere in
the genome.243
Finally, in addition to hypomethylation and any
changes in transcription, a recent study has shown that
DNA replication itself is altered in ICF, either as a result
of altered transcription of genes involved in replication, or
perhaps due to alteration of chromatin structure affecting
the access of the replication machinery.256 Given that one
of the key features of ICF is the chromosomal instabil-
ity at certain locations, a parallel is seen here with other
conditions in which DNA replication defects result in
chromosomal instability, such as certain cancers or cancer
syndromes.256
Since the mutation responsible for ICF was first
described, it has been clear that hypomethylation plays a
key role in this disorder. However, many questions remain
unanswered. It is not clear why only certain regions of the
genome are hypomethylated in ICF, and while it is plausi-
ble that reduced methylation could lead to de-repression of
normally silenced genes, specific target genes are yet to be
confirmed.
R ET T SY N D RO M E : FA I LU R E TO R E A D A N D/
O R M A I N TA I N M ET H Y L AT E D D NA
Rett syndrome is a severe mental retardation syndrome char-
acterized by loss of intellectual functioning, motor skills,
and communicative abilities; microcephaly; development
of stereotypical hand movements; respiratory abnormali-
ties; seizures; scoliosis; growth defects; and hypotonia. The
syndrome is characterized by normal early infant develop-
ment, followed by developmental regression from around
6–18  months of age, and then stabilization of phenotype,
which may persist for the lifetime of the patient. Patients nor-
mally survive into adulthood but require intensive support.257
The vast majority of cases (90–95%) of Rett syndrome
result from sporadic de novo258 mutations in the MECP2 gene,
which is located on the X chromosome (Xq28) and encodes
the nuclear methyl-CpG binding protein 2.259 To date, more
than 600 pathogenic MECP2 mutations, including non-
sense, missense, and frameshift, have been described.260
Males carrying comparable MECP2 mutations present
with infantile encephalopathy and rarely survive beyond
two years,261 so Rett syndrome patients are usually hetero-
zygous females. Following random X inactivation, around
half of the cells express the wild-type allele and the other
half a mutated MECP2. Cases of symptom-free female car-
riers of MECP2 mutations are very rare, where skewing
of X chromosome inactivation prevents expression of the
mutant allele.261
MeCP2 is a member of the family of methyl-CpG
binding (MBD) proteins, a group of transcriptional repres-
sors.262,263 The functional domains of MeCP2 consist of an
aminoterminal methyl-cytosine binding domain (MBD), a
transcriptional repressor domain (TRD), and a C-terminal
domain (CTD). The MBD enables binding to methyl-
ated CpG dinucleotides, preferentially adjacent to A/T
sequences.264 The TRD interacts with co-repressor com-
plexes, such as Sin3a, to recruit histone deacetylases and
heterochromatic proteins, resulting in chromatin conden-
sation and transcriptional repression.44 The CTD has been
shown to interact with splicing factors.265 MeCP2 has been
shown to suppress transcription of repetitive elements, thus
controlling transcriptional “noise” resulting from transcrip-
tion of these sequences within the cell.266
As well as playing a role in transcriptional repression,
MeCP2 has been identified in recent studies as involved in
transcriptional activation via interaction with the cAMP
response element binding (CREB) protein,267 chromatin
5 0   •   P rincip l e s o f G e no m ic M e dicin e
compaction and nucleosome clustering,268,269 as well as
RNA processing.270
The MeCP2 gene is expressed in various tissues, with
the highest level in postnatal neurons and the lowest level
in glial cells.271,272 Two major splice variants, MeCP2_e2
(MeCP2-β or MeCP2A form) and MeCP2_e1 (MeCP2-α
or MeCP2B form), have been characterized, the latter more
abundantly expressed in somatic tissues, including the
brain.273,274
Much recent work in Rett syndrome has focused on
MeCP2 knockdown mice. Deletion of MeCP2 in mice
results in a remarkably similar phenotype to that observed
in Rett syndrome.275,276 Male MeCP2-null mice develop
normally until around 6 weeks of age, following which there
is a period of rapid regression resulting in reduced sponta-
neous movement, clumsy gait, irregular breathing, hind
limb clasping, and tremors, leading to death by 20 weeks. In
contrast, heterozygous female mice show delayed onset of
symptoms (4–12 months) with stabilization of the pheno-
type later on. Neurobiological examination of MeCP2-null
mice has revealed reduced synaptic plasticity, reduced syn-
aptic connectivity, altered network excitability, and over-
all reduced neuron size. Interestingly, the activation of the
MeCP2 gene in MeCP2-null mice after the onset of Rett
syndrome symptoms was discovered to rescue disease phe-
notype in both males and females.277–279 This reversibility
suggests an essential role for MeCP2 in the maintenance of
normal neurological functions in the developed brain, and
questions its involvement in early stage brain development.
In addition, MeCP2 inactivation at different time points
in postnatal mouse brain cells results in the appearance of
Rett-like phenotypes, further supporting this hypothesis.280
Another interesting finding is that MeCP2 has different
functions based on different post-translational modifica-
tions.281 To date, two phosphorylation modifications have
been studied, linking Ser 421 phosphorylation exclusively
to neuronal activity.282
To fully understand the role of the MeCP2 in brain
function, it is essential to identify downstream target genes
of MeCP2. Analysis of the genomic distribution of MeCP2
has revealed affinity for both methylated as well as unmeth-
ylated DNA regions.283 An interesting target for MeCP2
is the promoter site of the brain-derived neurotrophic fac-
tor (BDNF) in resting neurons. Recent evidence suggests
an in vivo interaction of MeCP2 and BDNF and a role for
BDNF in the Rett syndrome phenotype. Overexpression
of BDNF in mice has been shown to reverse the Rett-like
phenotype, in contrast with knockout, which worsens
symptoms.284 In a different experiment, ChIP analysis has
identified MeCP2 binding within the mouse chromosome
 Free ebooks ==> www.Ebook777.com
6 imprinted domain where Dlx5 and its non-imprinted
neighbor Dlx6 are present. Dlx5 and Dlx6 encode proteins
that regulate neurotransmitter production,285 and were
shown be upregulated two-fold in MeCP2-null mice brains
with accompanying changes in histone modifications and
lack of chromatin loop formation.286
Around 5–10% of individuals clinically diagnosed with
Rett syndrome do not appear to have mutations in the
MeCP2 gene. In addition, MeCP2 mutations have been
found in other disorders, including neonatal-onset encepha-
lopathy, autism, patients exhibiting Angelman phenotype,
nonsyndromic X-linked mental retardation, and psychosis,
pyramidal signs, and macroorchidism (PPM-X) syndrome.
Recently, mutations in cyclin-dependent kinase 5 (CDKL5),
which are shown to directly interact with MeCP2, and
Netrin G1, have been found in patients with a very similar
phenotype to that of Rett syndrome.287–289
Recent work has confirmed the importance of epigene-
tic mechanisms in Rett syndrome pathophysiology, but spe-
cific downstream targets are still to be determined. Perhaps
most significantly, the accepted hypothesis that Rett syn-
drome is purely a disorder of neurodevelopment has been
recently called into question, with reversal of phenotype in
adult mouse models of disease.
AT R-X SY N D RO M E : A C O N N EC T I O N
TO C H RO M AT I N R E MO D E L I N G
The X-linked α-thalassemia mental retardation syndrome is
another example of genetic mutation in a factor involved
in chromatin organization affecting disease loci in trans.
The ATRX gene at chromosome Xq13.3 encodes for a
SNF2-like chromatin remodeling helicase. Functional
domains of ATRX include a PHD zinc finger–like motif
at its aminoterminus (homologous to the PHD motifs in
DNMT3A and DNMT3B), an adjacent coil-coil motif
termed ATRX-DNMT3-DNMT3L (ADD), and a heli-
case domain at its carboxyterminus.290 Mutations in ATRX
are clustered in these domains and are thought to impair its
nuclear localization, protein–protein interactions, or chro-
matin remodeling functions.245,291–293
Affected individuals have low levels of α-globin subunits
that favor the formation of unstable β-globin tetramers that
precipitate within erythrocytes, causing varying degrees of
hemolysis and splenomegaly. Affected males have relatively
severe mental retardation together with facial and skeletal
abnormalities, urogenital abnormalities, and microcephaly,
whereas heterozygous females are usually asymptomatic due
to a skewed pattern of X-chromosome inactivation prefer-
entially silencing the mutated allele.294–296
Epig e n etic s , Epig e no m ic s , and H u m an D i s e a s e   •   5 1
www.Ebook777.com
ATRX may be involved in gene activation, suggested
by the reduced expression of the α-globin locus in ATR-X
syndrome. This, however, does not explain the additional
phenotypical traits observed, and presumably deregulation
of many other loci gives rise to the complex phenotype.
The observation that diverse DNA methylation defects
(hypermethylation at DYZ2 Y-chromosome repeats and
hypomethylation of ribosomal DNA) are present in dis-
ease297 indicates ATRX is able to regulate chromatin struc-
ture at several distinct loci. Direct (stimulatory) effects of
ATRX upon loci are also indicated by its association with
the transcriptional regulatory death-associated protein 6
(DAXX). The DAXX-ATRX-containing complex, levels
of which are reduced in ATRX patient cell lines, is able to
remodel nucleosomes in vitro,298 and patient mutations in
ATRX cause a reduction in this function.293 Furthermore,
ATR-X associates with PML bodies, which are thought to
function as regulatory (activator) factor reservoirs in the
nucleus.298,299
On the other hand, chromatin remodeling by ATR-X
may facilitate chromatin condensation and gene silenc-
ing. The ADD domain of ATR-X has been shown to bind
directly to histone H3 trimethylated at lysine 9 (H3K9me3),
a hallmark of pericentric chromatin, and disease-causing
mutations impair this association.300 ATR-X was also found
to associate with the histone methyltransferase enhancer of
zeste homologue 2 (EZH2).301 EZH2 is part of the poly-
comb group repressor complexes (PRC) that methylate
histone H3 lysine 27 (PRC2) and histone H1 lysine 26
(PRC3).302 H3 lysine 27 methylation is a mark recognized
by the chromodomain of the Polycomb protein (contained
within the PRC1 complex) implicated in chromatin con-
densation and developmentally regulated gene silenc-
ing.303–305 Methylation at H1 lysine 26 may also be involved
in chromatin condensation, as HP1 has been shown to
specifically interact with this mark.306 These interactions
reveal potential new pathways of gene regulation where the
HP1 and the polycomb group–silencing pathways may be
synergistic. For example, a novel polycomb group complex,
PRC4, was recently shown to be upregulated upon onco-
genic transformation.307 PRC4 preferentially methylates
histone H1b lysine 26, particularly in the presence of native
complex subunit isoforms (i.e. Eed2), and histone deacety-
lases (i.e. Sirt1). Excess of PRC4 might lead to increased
lysine 26 methylation and subsequent recognition by HP1.
Dependent on the genomic context and associated com-
plexes, HP1 may either silence or de-repress the affected
loci.308 Furthermore, ATR-X and HP1 have been shown to
co-localize by immunofluorescence245 and to directly inter-
act.309–311 These interactions of ATR-X with other molecular
chaperones may generate regulatory complexes that directly
affect chromatin condensation at pericentric sites.
Mutations in ATRX are frequent at the PHD-zinc
finger motif, which is likely to be responsible for the tar-
geting of ATR-X to pericentromeric heterochromatin.245
Inadequate targeting of catalytically active ATR-X may
therefore result in ectopic binding at loci deregulated in dis-
ease. Additionally, due to the predominant localization of
ATR-X to pericentromeric heterochromatin, ATR-X may
be indirectly regulating additional loci by modulating the
nature of gene association with pericentromeric heterochro-
matin. Interestingly, different patients with identical muta-
tions in the ATRX gene exhibit significant phenotypical
variation.290,312 Recent ChIP sequencing data have suggested
that ATR-X is found associated with variable number tan-
dem repeats (VNTRs), and given that the size of these
VNTRs varies between individuals, this may explain some
of this variation.313
Acquired forms of alpha-thalassemia myelodysplastic
syndrome (ATMDS) with mutations in ATRX exhibit a
much more severe form of alpha-thalassemia compared to
inherited forms even when the mutation itself is identical,
again implicating epigenetic mechanisms in causing sever-
ity.314 The above observations plainly demonstrate that
ATRX is clearly a chromatin modifier acting in trans, par-
ticularly at pericentric heterochromatin, with the potential
for either gene activation or silencing. However, the down-
stream targets and the exact nature of modifying complexes
formed are yet to be elucidated.
E P I G E N ET I C ME C H A N I SMS
IN CANCER
Cancer is a complex multifactorial disease.315 Genetic muta-
tions may promote tumorigenesis by hampering tumor
suppressor activity or by hyper-activation of oncogenic
pathways. The balance between these activities promotes
gene expression programs that fuel neoplastic metabolic
states. How the genome responds to these signals in terms
of transcriptional output is thus central to the generation
and/or maintenance of the cancer phenotype.
Access to genetic information and subsequent transcrip-
tional readout are modulated by epigenetic factors—DNA
methylation, chromatin histone modifications, non-coding
RNA, and higher-order chromatin structures. The epig-
enome is thought to constitute a “buffering” system that
regulates gene expression “noise”.This buffering system is
grossly disrupted in cancer, leading to increased noise and
heterogeneity of gene expression.316
5 2   •   P rincip l e s o f G e no m ic M e dicin e
The cancer methylome is characterized by a global
reduction in DNA methylation that enhances genomic
instability, and by focal gains of methylation, in particu-
lar at promoters enriched in CpG dinucleotides (CpG
islands). DNA hypermethylation of promoter CpG islands
frequently is associated with repression of the associated
genes—more often than not of tumors’ suppressor loci.317
Nevertheless, the function of DNA methylation is now
understood to be dependent on where exactly it is located;
it has been shown to also associate with active genes.318
It is becoming increasingly clear that an interplay exists
between DNA methylation and histone modification.319 In
colonic tumors, increased DNA methylation is frequent at
promoters of tumour suppressor genes, which in common
with embryonic stem cells, are marked by both histone H3
lysine 4 and lysine 27 trimethylation.320 This bivalency for
histone H3 modification is therefore part of an instructive
process that guides the DNA methylation machinery to
the affected loci. Indeed, polycomb group complexes that
methylate H3 lysine 27 are known to interact with the de
novo DNA methyltransferases.321
Long non-coding RNA (>200bp) and small non-coding
RNA (~22bp) such as microRNAs are novel functional ele-
ments capable of regulating gene expression, and they seem
to be involved in tumorigenesis.322,323 For example, a long
non-coding RNA that is antisense to the CDKN2B (p15)
tumor suppressor locus can regulate the chromatin and
DNA methylation of the p15 locus,324 and a similar effect in
cis was seen at the CDKN1A (p21) locus.325 Effects in trans,
where the non-coding RNA has an effect on a chromosome
other than the one where it originates, have been described
where large intergenic non-coding RNAs appear to target
chromatin-modifying machineries to affect the expression
of distant loci.326 MicroRNAs are very powerful regulatory
elements327 grossly de-regulated in cancers323 and thus pro-
vide potentially attractive targets for cancer therapy.328
Effects at a distance are also brought about by chroma-
tin looping, where higher-order chromatin architectures
enable the interaction of physically distant regulatory ele-
ments, enhancer–promoter interaction for example, or
place loci within nuclear-chromosomal territories that
facilitate or inhibit gene activity. Higher-order architec-
tures such as those mediated by CTCF binding sites and
cohesin are thought to be part of networks of long-range
interactions that control developmentally regulated tran-
scriptional programs.329 Since CTCF binding is sensitive
to DNA methylation, further refinement of the role that
long-distance interactions play in tumorigenesis is needed.
Unlike genetic mutation, epigenetic mutation/modi-
fication is potentially reversible and thus a very attractive
 Free ebooks ==> www.Ebook777.com
target for therapeutic intervention. Inhibitors of DNA
methylation and histone deacetylation are already in use for
some types of tumours,319,328 and extensive research is under-
way to identify further epigenetic targets and to understand
how the epigenetic modifications interrelate such that com-
binatorial therapies can be designed. Together with the
technical advances that make genome-wide profiling of epi-
genetic modifications in individuals possible, the future of
epigenetic therapy in cancer is promising.
C O N C LU S I O N
It is clear that previously mysterious aspects of gene regu-
lation that can be grouped under the terms “epigenetic”
or “epigenomic” are finally yielding to molecular biology
approaches and have revealed a new level of genome orga-
nization and regulation. Already, the rapid increase in our
understanding of the control of gene expression patterns has
revealed potentially powerful new therapeutic avenues for an
ever-increasing number of human diseases, and a great vari-
ety of human cancers, many of which are currently incurable.
R EFE R E N C ES
1. Holliday R. DNA methylation and epigenotypes. Biochemistry
(Mosc). May 2005;70:500.
2. Reik W, Walter J. Genomic imprinting:  parental influence on the
genome. Nat Rev Genet. Jan 2001;2:21.
3. Ferguson-Smith AC. Genomic imprinting: the emergence of an epi-
genetic paradigm. Nat Rev Genet. Aug 2011;12:565.
4. Hochedlinger K, Plath K. Epigenetic reprogramming and induced
pluripotency. Development. Feb 2009;136:509.
5. Takahashi K, et al. Induction of pluripotent stem cells from adult
human fibroblasts by defined factors. Cell. Nov 30, 2007;131:861.
6. Takahashi K, Yamanaka S. Induction of pluripotent stem cells from
mouse embryonic and adult fibroblast cultures by defined factors.
Cell. Aug 25, 2006;126:663.
7. Kornberg RD, Lorch Y. Twenty-five years of the nucleosome,
fundamental particle of the eukaryote chromosome. Cell. Aug 6,
1999;98:285.
8. Thoma F, Koller T, Klug A. Involvement of histone H1 in the orga-
nization of the nucleosome and of the salt-dependent superstruc-
tures of chromatin. J Cell Biol. Nov 1979;83:403.
9. Rattner JB, Hamkalo BA. Nucleosome packing in interphase chro-
matin. J Cell Biol. May 1979;81:453.
10. Tremethick DJ. Higher-order structures of chromatin:  the elusive
30 nm fiber. Cell. Feb 23, 2007;128:651.
11. Hsu TC, Cooper JE, Mace ML Jr, Brinkley BR. Arrangement of
centromeres in mouse cells. Chromosoma. 1971;34:73.
12. Joseph A, Mitchell AR, Miller OJ. The organization of the mouse
satellite DNA at centromeres. Exp Cell Res. Aug 1989;183:494.
13. Guenatri M, Bailly D, Maison C, Almouzni G. Mouse centric and
pericentric satellite repeats form distinct functional heterochroma-
tin. J Cell Biol. Aug 16, 2004;166:493.
14. Jeppesen P, Mitchell A, Turner B, Perry P. Antibodies to defined
histone epitopes reveal variations in chromatin conformation and
Epig e n etic s , Epig e no m ic s , and H u m an D i s e a s e   •   5 3
www.Ebook777.com
underacetylation of centric heterochromatin in human metaphase
chromosomes. Chromosoma. Mar 1992;101:322.
15. Peters AH, et  al. Loss of the Suv39h histone methyltransferases
impairs mammalian heterochromatin and genome stability. Cell.
Nov 2, 2001;107:323.
16. Maison C, et al. Higher-order structure in pericentric heterochro-
matin involves a distinct pattern of histone modification and an
RNA component. Nat Genet. Mar 2002;30:329.
17. Maison C, et  al. SUMOylation promotes de novo targeting
of HP1alpha to pericentric heterochromatin. Nat Genet. Mar
2011;43:220.
18. Schramke V, et al. RNA-interference-directed chromatin modifica-
tion coupled to RNA polymerase II transcription. Nature. Jun 30,
2005;435:1275.
19. Fukagawa T, et al. Dicer is essential for formation of the heterochro-
matin structure in vertebrate cells. Nat Cell Biol. Aug 2004;6:784.
20. Pal-Bhadra M, et al. Heterochromatic silencing and HP1 localiza-
tion in Drosophila are dependent on the RNAi machinery. Science.
Jan 30, 2004;303:669.
21. Eissenberg JC, Elgin SC. The HP1 protein family: getting a grip on
chromatin. Curr Opin Genet Dev. Apr 2000;10:204.
22. Festenstein R, et  al. Modulation of heterochromatin protein 1
dynamics in primary mammalian cells. Science. Jan 31, 2003;299:719.
23. Cheutin T, et al. Maintenance of stable heterochromatin domains
by dynamic HP1 binding. Science. Jan 31, 2003;299:721.
24. Schmiedeberg L, Weisshart K, Diekmann S, Meyer Zu Hoerste G,
Hemmerich P. High- and low-mobility populations of HP1 in het-
erochromatin of mammalian cells. Mol Biol Cell. Jun 2004;15:2819.
25. Dillon N, Festenstein R. Unravelling heterochromatin: competition
between positive and negative factors regulates accessibility. Trends
Genet. May 2002;18:252.
26. Bannister AJ, Kouzarides T. Regulation of chromatin by histone
modifications. Cell Res. Mar 2011;21:381.
27. Hargreaves DC, Crabtree GR. ATP-dependent chromatin remodel-
ing: genetics, genomics and mechanisms. Cell Res. Mar 2011;21:396.
28. Turner BM. Histone acetylation and an epigenetic code. Bioessays.
Sep 2000;22:836.
29. Strahl BD, Allis CD. The language of covalent histone modifica-
tions. Nature. Jan 6, 2000;403:41.
30. Narlikar GJ, Fan HY, Kingston RE. Cooperation between com-
plexes that regulate chromatin structure and transcription. Cell. Feb
22, 2002;108:475.
31. de la Cruz X, Lois S, Sanchez-Molina S, Martinez-Balbas MA. Do
protein motifs read the histone code? Bioessays. Feb 2005;27:164.
32. Wysocka J, et al. WDR5 associates with histone H3 methylated at
K4 and is essential for H3 K4 methylation and vertebrate develop-
ment. Cell. Jun 17, 2005;121:859.
33. Bernstein BE, et al. Methylation of histone H3 Lys 4 in coding regions
of active genes. Proc Natl Acad Sci U S A. Jun 25, 2002;99:8695.
34. Schubeler D, et al. The histone modification pattern of active genes
revealed through genome-wide chromatin analysis of a higher
eukaryote. Genes Dev. Jun 1, 2004;18:1263.
35. Lachner M, O’Carroll D, Rea S, Mechtler K, Jenuwein T.
Methylation of histone H3 lysine 9 creates a binding site for HP1
proteins. Nature. Mar 1, 2001;410:116.
36. Fischle W, et al. Molecular basis for the discrimination of repressive
methyl-lysine marks in histone H3 by polycomb and HP1 chromo-
domains. Genes Dev. Aug 1, 2003;17:1870.
37. Hermann A, Gowher H, Jeltsch A. Biochemistry and biology
of mammalian DNA methyltransferases. Cell Mol Life Sci. Oct
2004;61:2571.
38. Bird AP, Wolffe AP. Methylation-induced repression—belts, braces,
and chromatin. Cell. Nov 24, 1999;99:451.
39. Lachner M, Jenuwein T. The many faces of histone lysine methyla-
tion. Curr Opin Cell Biol. Jun 2002;14:286.
40. Dellino GI, et al. Polycomb silencing blocks transcription initiation.
Mol Cell. Mar 26, 2004;13:887.
41. Tahiliani M, et  al. Conversion of 5-methylcytosine to
67. Henikoff S. Dosage-dependent modification of position-effect var-
5-hydroxymethylcytosine in mammalian DNA by MLL partner
TET1. Science. May 15, 2009;324:930.
42. Branco MR, Ficz G, Reik W. Uncovering the role of
5-hydroxymethylcytosine in the epigenome. Nat Rev Genet. Jan
2012;13:7.
43. Williams K, et al. TET1 and hydroxymethylcytosine in transcrip-
tion and DNA methylation fidelity. Nature. May 19, 2011;473:343.
44. Nan X, et al. Transcriptional repression by the methyl-CpG-binding
protein MeCP2 involves a histone deacetylase complex. Nature.
May 28, 1998;393:386.
45. Ng HH, et al. MBD2 is a transcriptional repressor belonging to the
MeCP1 histone deacetylase complex. Nat Genet. Sep 1999;23:58.
46. Fuks F, et  al. The methyl-CpG-binding protein MeCP2 links
DNA methylation to histone methylation. J Biol Chem. Feb 7,
2003;278:4035.
47. Fuks F, Hurd PJ, Deplus R, Kouzarides T. The DNA methyltransfer-
ases associate with HP1 and the SUV39H1 histone methyltransfer-
ase. Nucleic Acids Res. May 1, 2003;31:2305.
48. Aagaard L, et  al. Functional mammalian homologues of the
Drosophila PEV-modifier Su(var)3–9 encode centromere-associated
proteins which complex with the heterochromatin component
M31. Embo J. Apr 1, 1999;18:1923.
49. Lehnertz B, et al. Suv39h-mediated histone H3 lysine 9 methylation
directs DNA methylation to major satellite repeats at pericentric
heterochromatin. Curr Biol. Jul 15, 2003;13:1192.
50. Lewis A, et al. Imprinting on distal chromosome 7 in the placenta
involves repressive histone methylation independent of DNA meth-
ylation. Nat Genet. Dec 2004;36:1291.
51. Ebralidse KK, Hebbes TR, Clayton AL, Thorne AW,
Crane-Robinson C. Nucleosomal structure at hyperacetylated
loci probed in nuclei by DNA-histone crosslinking. Nucleic Acids
Research. Oct 11, 1993;21:4734.
52. Orlando V, Strutt H, Paro R. Analysis of chromatin structure by
in vivo formaldehyde cross-linking. Methods. Feb 1997;11:205.
53. Shendure J, Ji H. Next-generation DNA sequencing. Nature Biotech.
Oct 2008;26:1135.
54. Arvey A, Agius P, Noble WS, Leslie C. Sequence and chromatin
determinants of cell-type-specific transcription factor binding.
Genome Res. Sep 2012;22:1723.
55. EP Consortium, et al. An integrated encyclopedia of DNA elements
in the human genome. Nature. Sep 6, 2012;489:57.
56. Djebali S, et al. Landscape of transcription in human cells. Nature.
Sep 6, 2012;489:101.
57. Harrow J, et al. GENCODE: the reference human genome annota-
tion for The ENCODE Project. Genome Res. Sep 2012;22:1760.
58. Thurman RE, et  al. The accessible chromatin landscape of the
human genome. Nature. Sep 6, 2012;489:75.
59. Wang J, et  al. Sequence features and chromatin structure around
the genomic regions bound by 119 human transcription factors.
Genome Res. Sep 2012;22:1798.
60. Wong CC, et  al. A longitudinal study of epigenetic variation in
twins. Epigenetics. Aug 16, 2010;5:516.
61. Bell JT, Spector TD. A twin approach to unraveling epigenetics.
Trends Genet. Mar 2011;27:116.
62. Osborne CS, et  al. Active genes dynamically colocalize to shared
sites of ongoing transcription. Nat Genet. Oct 2004;36:1065.
63. Brown K, et  al. Association of transcriptionally silent genes with
Ikaros complexes at centromeric heterochromatin. Cell. 1997;91:845.
64. Locke J, Kotarski MA, Tartof KD. Dosage-dependent modifiers of
position effect variegation in Drosophila and a mass action model
that explains their effect. Genetics. Sep 1988;120:181.
65. Cattanach BM. Position effect variegation in the mouse. Genet Res.
Jun 1974;23:291.
66. Festenstein R, et  al. Locus control region function and
heterochromatin-induced position effect variegation. Science. Feb
23, 1996;271:1123.
5 4   •   P rincip l e s o f G e no m ic M e dicin e
iegation in Drosophila. Bioessays. May 1996;18:401.
68. Grosveld F. Activation by locus control regions? Curr Opin Genet
Dev. Apr 1999;9:152.
69. West AG, Fraser P. Remote control of gene transcription. Hum Mol
Genet. Apr 15, 2005;14:Spec No 1, R101.
70. Festenstein R, Kioussis D. Locus control regions and epigenetic
chromatin modifiers. Curr Opin Genet Dev. Apr 2000;10:199.
71. Soejima H, Wagstaff J. Imprinting centers, chromatin structure, and
disease. J Cell Biochem. May 15, 2005;95:226.
72. Gavrilov A, et  al. Chromosome conformation capture (from
3C to 5C) and its ChIP-based modification. Methods Mol B.
2009;567:171.
73. Sanyal A, Lajoie BR, Jain G, Dekker J. The long-range interaction
landscape of gene promoters. Nature. Sep 6, 2012;489:109.
74. Wijchers PJ, et  al. Sexual dimorphism in mammalian autosomal
gene regulation is determined not only by Sry but by sex chromo-
some complement as well. Dev Cell. Sep 2010;19:477.
75. Wijchers PJ, Festenstein RJ. Epigenetic regulation of autosomal gene
expression by sex chromosomes. Trends Genet. Apr 2011;27:132.
76. Harding AE. Friedreich’s ataxia: a clinical and genetic study of 90
families with an analysis of early diagnostic criteria and intrafamilial
clustering of clinical features. Brain. 1981;104:589.
77. Campuzano V, et  al. Friedreich’s ataxia:  autosomal recessive dis-
ease caused by an intronic GAA triplet repeat expansion. Science.
1996;271:1423.
78. Mirkin SM. Expandable DNA repeats and human disease. Nature.
Jun 21, 2007;447:932.
79. Pandolfo M, Pastore A. The pathogenesis of Friedreich ataxia and
the structure and function of frataxin. J Neurol. Mar 2009;256
Suppl. 1:9.
80. Montermini L, et al. Phenotypic variability in Friedreich ataxia: role
of the associated GAA triplet repeat expansion. Ann Neurol.
1997;41:675.
81. Monros E, et  al. Phenotype correlation and intergenerational
dynamics of the Friedreich ataxia GAA trinucleotide repeat. Am J
Hum Genet. Jul 1997;61:101.
82. Filla A, et al. The relationship between trinucleotide (GAA) repeat
length and clinical features in Friedreich ataxia. Am J Hum Genet.
Sep 1996;59:554.
83. Durr A, et al. Clinical and genetic abnormalities in patients with
Friedreich’s ataxia. N Engl J Med. 1996;335:1169.
84. Lamont PJ, Davis MB, Wood NW. Identification and sizing of the
GAA trinucleotide repeat expansion of Friedreich’s ataxia in 56
patients. Clinical and genetic correlates. Brain. 1997;120:673.
85. Campuzano V, et al. Frataxin is reduced in Friedreich ataxia patients
and is associated with mitochondrial membranes. Hum Mol Genet.
1997;6:1771.
86. Geschwind DH, et  al. Friedreich’s ataxia GAA repeat expan-
sion in patients with recessive or sporadic ataxia. Neurology.
1997;49:1004.
87. Gellera C, et al. Very late onset Friedreich’s ataxia without cardiomy-
opathy is associated with limited GAA expansion in the X25 gene.
Neurology. 1997;49:1153.
88. Kumari D, Usdin K. Is Friedreich ataxia an epigenetic disorder?
Clin Epigenet. 2012;4:2.
89. Bidichandani SI, Ashizawa T, Patel PI. The GAA triplet-repeat
expansion in Friedreich ataxia interferes with transcription and may
be associated with an unusual DNA structure. Am J Hum Genet.
Jan 1998;62:111.
90. Grabczyk E, Usdin K. Alleviating transcript insufficiency caused
by Friedreich’s ataxia triplet repeats. Nucleic Acids Res. Dec 15,
2000;28:4930.
91. Grabczyk E, Usdin K. The GAA*TTC triplet repeat expanded in
Friedreich’s ataxia impedes transcription elongation by T7 RNA
polymerase in a length and supercoil dependent manner. Nucleic
Acids Res. Jul 15, 2000;28:2815.
 92. Krasilnikova MM, et  al. Effects of Friedreich’s ataxia (GAA)
n*(TTC)n repeats on RNA synthesis and stability. Nucleic Acids
Res. 2007;35:1075.
93. Sakamoto N, et al. Sticky DNA: self-association properties of long
GAA.TTC repeats in R.R.Y triplex structures from Friedreich’s
ataxia. Mol Cell. Apr 1999;3:465.
94. Ohshima K, Montermini L, Wells RD, Pandolfo M. Inhibitory
effects of expanded GAA.TTC triplet repeats from intron I  of
the Friedreich ataxia gene on transcription and replication in vivo.
J Biol Chem. Jun 5, 1998;273:14588.
95. Sakamoto N, Ohshima K, Montermini L, Pandolfo M, Wells RD.
Sticky DNA, a self-associated complex formed at long GAA*TTC
repeats in intron 1 of the frataxin gene, inhibits transcription. J Biol
Chem. Jul 20, 2001;276:27171.
96. Kim E, Napierala M, Dent SY. Hyperexpansion of GAA repeats
affects post-initiation steps of FXN transcription in Friedreich’s
ataxia. Nucleic Acids Research. Oct 1, 2011;39:8366.
97. Kumari D, Biacsi RE, Usdin K. Repeat expansion affects both tran-
scription initiation and elongation in friedreich ataxia cells. J Biol
Chem. Feb 11, 2011;286:4209.
98. Punga T, Buhler M. Long intronic GAA repeats causing Friedreich
ataxia impede transcription elongation. EMBO Mol Med. Apr
2010;2:120.
99. Saveliev A, Everett C, Sharpe T, Webster Z, Festenstein R. DNA
triplet repeats mediate heterochromatin-protein-1-sensitive varie-
gated gene silencing. Nature. Apr 24, 2003;422:909.
100. Castaldo I, et al. DNA methylation in intron 1 of the frataxin gene
is related to GAA repeat length and age of onset in Friedreich
ataxia patients. J Med Genet. Dec 2008;45:808.
101. Evans-Galea MV, et  al. FXN methylation predicts expression
and clinical outcome in Friedreich ataxia. Ann Neurol. Apr
2012;71:487.
102. Greene E, Mahishi L, Entezam A, Kumari D, Usdin K. Repeat-induced
epigenetic changes in intron 1 of the frataxin gene and its conse-
quences in Friedreich ataxia. Nucleic Acids Res. 2007;35:3383.
103. Al-Mahdawi S, et al. The Friedreich ataxia GAA repeat expansion
mutation induces comparable epigenetic changes in human and
transgenic mouse brain and heart tissues. Hum Mol Genet. Mar 1,
2008;17:735.
104. Herman D, et  al. Histone deacetylase inhibitors reverse gene
silencing in Friedreich’s ataxia. Nat Chem Biol. Oct 2006;2:551.
105. De Biase I, Chutake YK, Rindler PM, Bidichandani SI. Epigenetic
silencing in Friedreich ataxia is associated with depletion of CTCF
(CCCTC-binding factor) and antisense transcription. PloS One.
2009;4(11):e7914.
106. Bourn RL, Rindler PM, Pollard LM, Bidichandani SI. E. coli mis-
match repair acts downstream of replication fork stalling to stabi-
lize the expanded (GAA.TTC)(n) sequence. Mutat Res. Feb 10,
2009;661:71.
107. Du J, et al. Role of mismatch repair enzymes in GAA{middle dot}
TTC triplet-repeat expansion in Friedreich ataxia induced plurip-
otent stem cells. J Biol Chem. Aug 24, 2012;287:29861.
108. Ezzatizadeh V, et al. The mismatch repair system protects against
intergenerational GAA repeat instability in a Friedreich ataxia
mouse model. Neurobiol Dis. Apr 2012;46:165.
109. Kim HM, et  al. Chromosome fragility at GAA tracts in yeast
depends on repeat orientation and requires mismatch repair.
EMBO J. Nov 5, 2008;27:2896.
110. Ku S, et al. Friedreich’s ataxia induced pluripotent stem cells model
intergenerational GAATTC triplet repeat instability. Cell Stem
Cell. Nov 5, 2010;7:631.
111. Ayoub N, Jeyasekharan AD, Bernal JA, Venkitaraman AR.
HP1-beta mobilization promotes chromatin changes that initiate
the DNA damage response. Nature. May 29, 2008;453:682.
112. Goodarzi AA, et  al. ATM signaling facilitates repair of DNA
double-strand breaks associated with heterochromatin. Mol Cell.
Jul 25, 2008;31:167.
Epig e n etic s , Epig e no m ic s , and H u m an D i s e a s e   •   5 5
113. Luijsterburg MS, et al. Heterochromatin protein 1 is recruited to
various types of DNA damage. J Cell Biol. May 18, 2009;185:577.
114. Peng JC, Karpen GH. Heterochromatic genome stability requires
regulators of histone H3 K9 methylation. PLoS Genet, Mar,
2009;5:e1000435.
115. Sun Y, et al. Histone H3 methylation links DNA damage detection
to activation of the tumour suppressor Tip60. Nat Cell Biol. Nov
2009;11:1376.
116. Zarebski M, Wiernasz E, Dobrucki JW. Recruitment of heterochro-
matin protein 1 to DNA repair sites. Cytometry A. Jul 2009;75:619.
117. Djupedal I, Ekwall K. Epigenetics: heterochromatin meets RNAi.
Cell Res. Mar 2009;19:282.
118. Barkess G, West AG. Chromatin insulator elements:  establish-
ing barriers to set heterochromatin boundaries. Epigenomics. Feb
2012;4:67.
119. Bushey AM, Dorman ER, Corces VG. Chromatin insulators: reg-
ulatory mechanisms and epigenetic inheritance. Mol Cell. Oct 10,
2008;32:1.
120. Puccio H. Multicellular models of Friedreich ataxia. J Neurol. Mar
2009;256 Suppl 1:18.
121. Liu J, et al. Generation of induced pluripotent stem cell lines from
Friedreich ataxia patients. Stem Cell Rev. Sep 2011;7:703.
122. Chou CJ, Herman D, Gottesfeld JM. Pimelic diphenylamide 106
is a slow, tight-binding inhibitor of class  I  histone deacetylases.
J Biol Chem. Dec 19, 2008;283:35402.
123. Rai M, et al. Two new pimelic diphenylamide HDAC inhibitors
induce sustained frataxin upregulation in cells from Friedreich’s
ataxia patients and in a mouse model. PLoS One. 2010:5: e8825.
124. Rai M, et  al. HDAC inhibitors correct frataxin deficiency in a
Friedreich ataxia mouse model. PLoS One. 2008;3:e1958.
125. Sandi C, et al. Prolonged treatment with pimelic o-aminobenzamide
HDAC inhibitors ameliorates the disease phenotype of a
Friedreich ataxia mouse model. Neurobiol Dis. Jun 2011;42:496.
126. Xu C, et al. Chemical probes identify a role for histone deacety-
lase 3 in Friedreich’s ataxia gene silencing. Chem Biol. Sep 25,
2009;16:980.
126a. Chan PK, Torres R, Yandim C, Law PP, Khadayate S, Mauri M,
Grosan C, Chapman-Rothe N, Giunti P, Pook M, Festenstein R.
Heterochromatinization induced by GAA-repeat hyperexpansion
in Friedreich's ataxia can be reduced upon HDAC inhibition by
vitamin B3. Human Molecular Genetics. 2013;22:2662–2675.
126b. Libri V*, Yandim C*, Athanasopoulos S, Loyse N, Natisvili T, Law
PP, Chan PK, Mohammad T, Mauri M, Tam KT, Leiper J, Piper
S, Ramesh A, Parkinson MH, Huson L, Giunti P, Festenstein R.
Epigenetic and neurological effects and safety of high-dose nicotin-
amide in patients with Friedreich’s ataxia: an exploratory, open-label,
dose-escalation study. The Lancet (in press-1st May 2014).
127. Fu YH, et al. An unstable triplet repeat in a gene related to myo-
tonic muscular dystrophy. Science. Mar 6, 1992;255:1256.
128. Mahadevan M, et  al. Myotonic dystrophy mutation:  an unstable
CTG repeat in the 3′ untranslated region of the gene. Science.
Mar 6, 1992;255:1253.
129. Brook JD, et al. Molecular basis of myotonic dystrophy: expansion
of a trinucleotide (CTG) repeat at the 3′ end of a transcript encod-
ing a protein kinase family member. Cell. Apr 17, 1992;69:385.
130. Liquori CL, et al. Myotonic dystrophy type 2 caused by a CCTG
expansion in intron 1 of ZNF9. Science. Aug 3, 2001;293:864.
131. Gomes-Pereira M, Monckton DG. Chemical modifiers of unsta-
ble expanded simple sequence repeats: what goes up, could come
down. Mutat Res. Jun 25, 2006;598:15.
132. Ashizawa T, Sarkar PS. Myotonic dystrophy types 1 and 2. Handb
Clin Neurol. 2011;101:193.
133. Lee JE, Cooper TA. Pathogenic mechanisms of myotonic dystro-
phy. Biochem Soc Trans. Dec 2009;37:1281.
134. Mankodi A, Lin X, Blaxall BC, Swanson MS, Thornton CA.
Nuclear RNA foci in the heart in myotonic dystrophy. Circ Res.
Nov 25, 2005;97:1152.
 135. Jiang H, Mankodi A, Swanson MS, Moxley RT, Thornton CA.
Myotonic dystrophy type 1 is associated with nuclear foci of mutant
RNA, sequestration of muscleblind proteins and deregulated alter-
native splicing in neurons. Hum Mol Genet. Dec 15, 2004;13:3079.
136. Lin X, et al. Failure of MBNL1-dependent post-natal splicing transi-
tions in myotonic dystrophy. Hum Mol Genet. Jul 1, 2006;15:2087.
137. Kuyumcu-Martinez NM, Wang GS, Cooper TA. Increased
steady-state levels of CUGBP1 in myotonic dystrophy 1 are due
to PKC-mediated hyperphosphorylation. Mol Cell. Oct 12,
2007;28:68.
138. Otten AD, Tapscott SJ. Triplet repeat expansion in myotonic dys-
trophy alters the adjacent chromatin structure. Proc Natl Acad Sci
U S A. 1995;92:5465.
139. Klesert TR, Otten AD, Bird TD, Tapscott SJ. Trinucleotide repeat
expansion at the myotonic dystrophy locus reduces expression of
DMAHP. Nat Genet. 1997;16:402.
140. Wang YH, Griffith J. Expanded CTG triplet blocks from the myo-
tonic dystrophy gene create the strongest known natural nucleo-
some positioning elements. Genomics. Jan 20, 1995;25:570.
141. Wang YH, Amirhaeri S, Kang S, Wells RD, Griffith JD. Preferential
nucleosome assembly at DNA triplet repeats from the myotonic
dystrophy gene. Science. Jul 1994;265:669.
142. Klesert TR, et al. Mice deficient in Six5 develop cataracts: implica-
tions for myotonic dystrophy. Nat Genet. 2000;25:105.
143. Inukai A, et  al. Reduced expression of DMAHP/SIX5 gene in
myotonic dystrophy muscle. Muscle Nerve. 2000;23:1421.
144. Sarkar PS, et al. Heterozygous loss of Six5 in mice is sufficient to
cause ocular cataracts. Nat Genet. 2000;25:110.
145. Cho DH, et al. Antisense transcription and heterochromatin at the
DM1 CTG repeats are constrained by CTCF. Mol Cell. Nov 11,
2005;20:483.
146. A. Lopez Castel, et al. Expanded CTG repeat demarcates a bound-
ary for abnormal CpG methylation in myotonic dystrophy patient
tissues. Hum Mol Genet. Jan 1, 2011;20:1.
147. Maueler W, Bassili G, Hardt C, Keyl HG, Epplen JT. A complex
containing at least one zinc dependent HeLa nuclear protein binds
to the intronic (GAA)(n) block of the frataxin gene. Gene. May 30,
2001;270:131.
148. Naumann F, Remus R, Schmitz B, Doerfler W. Gene structure and
expression of the 5′-(CGG)(n)-3′-binding protein (CGGBP1).
Genomics. Jan 2004;83:106.
149. Timchenko LT, Timchenko NA, Caskey CT, Roberts R. Novel
proteins with binding specificity for DNA CTG repeats and RNA
CUG repeats:  implications for myotonic dystrophy. Hum Mol
Genet. Jan 1996;5:115.
150. Matsuura T, et  al. Large expansion of the ATTCT pentanu-
cleotide repeat in spinocerebellar ataxia type 10. Nat Genet.
2000;26:191.
151. Liquori CL, et al. Myotonic dystrophy type 2 caused by a CCTG
expansion in intron 1 of ZNF9. Science. 2001;293:864.
152. Bardoni B, Davidovic L, Bensaid M, Khandjian EW. The fragile X
syndrome: exploring its molecular basis and seeking a treatment.
Expert Rev Mol Med. 2006;8:1.
153. Penagarikano O, Mulle JG, Warren ST. The pathophysiology of
fragile X syndrome. Annu Rev Genomics Hum Genet. 2007;8:109.
154. Verkerk AJ, et  al. Identification of a gene (FMR-1) contain-
ing a CGG repeat coincident with a breakpoint cluster region
exhibiting length variation in fragile X syndrome. Cell. May
1991;65:905.
155. Merenstein SA, et al. Molecular-clinical correlations in males with
an expanded FMR1 mutation. Am J Med Genet. Aug 1996;64:388.
156. Bourgeois JA, et  al. A review of fragile X premutation disor-
ders: expanding the psychiatric perspective. J Clin Psychiatry. Jun
2009;70:852.
157. Bhakar AL, Dölen G, Bear MF. The pathophysiology of fragile
X (and what it teaches us about synapses). Annu Rev Neurosci.
2012;35:417.
5 6   •   P rincip l e s o f G e no m ic M e dicin e
158. Oberle I, et  al. Instability of a 550-base pair DNA segment and
abnormal methylation in fragile X syndrome. Science. May
1991;252:1097.
159. Ashley CT, Wilkinson KD, Reines D, Warren ST. FMR1 pro-
tein: conserved RNP family domains and selective RNA binding.
Science. Oct 1993;262:563.
160. Gantois I, et al. Chronic administration of AFQ056/Mavoglurant
restores social behaviour in Fmr1 knockout mice. Behav Brain Res.
Feb 2013;239:72.
161. Bakker CE, et  al. Immunocytochemical and biochemical charac-
terization of FMRP, FXR1P, and FXR2P in the mouse. Exp Cell
Res. Jul 2000;258:162.
162. Zhao X, Pak C, Smrt RD, Jin P. Epigenetics and neural develop-
mental disorders:. Epigenetics. 2007;2:126.
163. Qin M, Kang J, Burlin TV, Jiang C, Smith CB. Post-adolescent
changes in regional cerebral protein synthesis: an in vivo study in
the FMR1 null mouse. J Neurosci. May 2005;25:5087.
164. Chiurazzi P, Pomponi MG, Willemsen R, Oostra BA, Neri G.
In vitro reactivation of the FMR1 gene involved in fragile X syn-
drome. Hum Mol Genet. Jan 1998;7:109.
165. Chiurazzi P, et al. Synergistic effect of histone hyperacetylation and
DNA demethylation in the reactivation of the FMR1 gene. Hum
Mol Genet. Nov 1999;8:2317.
166. Coffee B, Zhang F, Warren ST, Reines D. Acetylated histones are
associated with FMR1 in normal but not fragile X-syndrome cells.
Nat Genet. May 1999;22:98.
167. Coffee B, Zhang F, Ceman S, Warren ST, Reines D. Histone modi-
fications depict an aberrantly heterochromatinized FMR1 gene in
fragile x syndrome. Am J Hum Genet. Oct 2002;71:923.
168. Tabolacci E, et  al. Differential epigenetic modifications in the
FMR1 gene of the fragile X syndrome after reactivating pharmaco-
logical treatments. Eur J Hum Genet. May 2005;13:641.
169. Eden S, Hashimshony T, Keshet I, Cedar H, Thorne AW.
DNA methylation models histone acetylation. Nature. Aug
1998;394:842.
170. Neri G, Chiurazzi P. X-linked mental retardation. Adv Genet.
1999;41:55.
171. Sølvsten C, Nielsen AL. FMR1 CGG repeat lengths mediate
different regulation of reporter gene expression in compara-
tive transient and locus specific integration assays. Gene. Oct
2011;486:15.
172. Ben-Yosef D, Malcov M, Eiges R. PGD-derived human embryonic
stem cell lines as a powerful tool for the study of human genetic
disorders. Mol Cell Endocrinol. Jan 2008;282:153.
173. Stöger R, et al. Testing the FMR1 promoter for mosaicism in DNA
methylation among CpG sites, strands, and cells in FMR1-expressing
males with fragile X syndrome. PLoS One. 2011;6, e23648.
174. Maddalena A, et  al. Technical standards and guidelines for frag-
ile X:  the first of a series of disease-specific supplements to the
Standards and Guidelines for Clinical Genetics Laboratories of
the American College of Medical Genetics. Quality Assurance
Subcommittee of the Laboratory Practice Committee. Genet Med.
2001;3:200.
175. Hagerman RJ, Hagerman PJ. The fragile X premutation: into the
phenotypic fold. Curr Opin Genet Dev. Jun 2002;12:278.
176. Hagerman PJ, Hagerman RJ. The fragile-X premutation: a matur-
ing perspective. Am J Hum Genet. May 2004;74:805.
177. Bacalman S, et  al. Psychiatric phenotype of the fragile
X-associated tremor/ataxia syndrome (FXTAS) in males: newly
described fronto-subcortical dementia. J Clin Psychiatry. Jan
2006;67:87.
178. Sherman SL. Premature ovarian failure among fragile X premu-
tation carriers:  parent-of-origin effect? Am J Hum Genet. Jul
2000;67:11.
179. Jacquemont S, et  al. Fragile X premutation tremor/ataxia syn-
drome:  molecular, clinical, and neuroimaging correlates. Am
J Hum Genet. Apr 2003;72:869.
 Free ebooks ==> www.Ebook777.com
180. Tassone F, et al. Elevated levels of FMR1 mRNA in carrier males: a
new mechanism of involvement in the fragile-X syndrome. Am J
Hum Genet. Jan 2000;66:6.
181. Tassone F, et al. Elevated FMR1 mRNA in premutation carriers is
due to increased transcription. RNA. Apr 2007;13:555.
182. Kenneson A, Zhang F, Hagedorn CH, Warren ST. Reduced
FMRP and increased FMR1 transcription is proportionally associ-
ated with CGG repeat number in intermediate-length and premu-
tation carriers. Hum Mol Genet. Jul 2001;10:1449.
183. Hagerman RJ, Hagerman PJ. Fragile X syndrome:  a model of
gene-brain-behavior relationships. Mol Genet Metab. 74:89.
184. Hoem G, et  al. CGG-repeat length threshold for FMR1 RNA
pathogenesis in a cellular model for FXTAS. Hum Mol Genet. Jun
2011;20:2161.
185. Sheridan SD, et  al. Epigenetic characterization of the FMR1
gene and aberrant neurodevelopment in human induced pluripo-
tent stem cell models of fragile X syndrome. PLoS One. 2011;6,
e26203.
186. Wang T, et  al. Genome-wide DNA hydroxymethylation changes
are associated with neurodevelopmental genes in the developing
human cerebellum. Hum Mol Genet. Dec 2012;21:5500.
187. Padberg G. In: Upadhyaya M, Cooper DN, eds. Facioscapulohumeral
Muscular Dystrophy:  Clinical Medicine and Molecular Cell
Biology. New York: Garland Science/BI OS Scientific, Abingdon;
2004: 41–54.
188. Tawil R, Figlewicz DA, Griggs RC, Weiffenbach B. Facioscapulo­
humeral dystrophy:  a distinct regional myopathy with a novel
molecular pathogenesis. FSH Consortium. Ann Neurol. Mar
1998;43:279.
189. Tawil R. Facioscapulohumeral muscular dystrophy. Curr Neurol
Neurosci Rep. Jan 2004;4:51.
190. van Deutekom JC, et al. FSHD associated DNA rearrangements
are due to deletions of integral copies of a 3.2 kb tandemly repeated
unit. Hum Mol Genet. Dec 1993;2:2037.
191. Wijmenga C, et al. Chromosome 4q DNA rearrangements associ-
ated with facioscapulohumeral muscular dystrophy. Nat Genet. Sep
1992;2:26.
192. Lunt PW. 44th ENMC International Workshop: Facioscapulo­
humeral Muscular Dystrophy:  Molecular Studies. July 19–21,
1996, Naarden, The Netherlands. Neuromuscul Disord. Apr
1998;8:126.
193. Richards M, Coppée F, Thomas N, Belayew A, Upadhyaya M.
Facioscapulohumeral muscular dystrophy (FSHD):  an enigma
unravelled? Hum Genet. Mar 2012;131:325.
194. Tawil R, et al. Evidence for anticipation and association of deletion
size with severity in facioscapulohumeral muscular dystrophy. The
FSH-DY Group. Ann Neurol. Jun 1996;39:744.
195. Gabriels J, et al. Nucleotide sequence of the partially deleted D4Z4
locus in a patient with FSHD identifies a putative gene within each
3.3 kb element. Gene. Aug 5, 1999;236:25.
196. Lyle R, Wright TJ, Clark LN, Hewitt JE. The FSHD-associated
repeat, D4Z4, is a member of a dispersed family of
homeobox-containing repeats, subsets of which are clustered on
the short arms of the acrocentric chromosomes. Genomics. Aug
1995;28:389.
197. Wijmenga C, et al. Chromosome 4q DNA rearrangements asso-
ciated with facioscapulohumeral muscular dystrophy. Nat Genet.
Sep 1992;2:26.
198. Tupler R, et al. Monosomy of distal 4q does not cause facioscapu-
lohumeral muscular dystrophy. J Med Genet. May 1996;33:366.
199. Lemmers RJ, et  al. Facioscapulohumeral muscular dystrophy is
uniquely associated with one of the two variants of the 4q subtelo-
mere. Nat Genet. Oct 2002;32:235.
200. Lemmers RJ, et al. Worldwide population analysis of the 4q and
10q subtelomeres identifies only four discrete interchromosomal
sequence transfers in human evolution. Am J Hum Genet. Mar
2010;86:364.
Epig e n etic s , Epig e no m ic s , and H u m an D i s e a s e   •   5 7
www.Ebook777.com
201. Thomas NS, et al. A large patient study confirming that facioscapu-
lohumeral muscular dystrophy (FSHD) disease expression is
almost exclusively associated with an FSHD locus located on a
4qA-defined 4qter subtelomere. J Med Genet. Mar 2007;44:215.
202. Lemmers RJ, et  al. A unifying genetic model for facioscapulo-
humeral muscular dystrophy. Science. Sep 2010;329:1650.
203. Spurlock G, Jim HP, Upadhyaya M. Confirmation that the specific
SSLP microsatellite allele 4qA161 segregates with fascioscapulo-
humeral muscular dystrophy (FSHD) in a cohort of multiplex and
simplex FSHD families. Muscle Nerve. Nov 2010;42:820.
204. Fisher J, Upadhyaya M. Molecular genetics of facioscapulohumeral
muscular dystrophy (FSHD). Neuromuscul Disord. Jan 1997;7:55.
205. Jiang G, et  al. Testing the position-effect variegation hypothesis
for facioscapulohumeral muscular dystrophy by analysis of histone
modification and gene expression in subtelomeric 4q. Hum Mol
Genet. Nov 15, 2003;12:2909.
206. Zeng W, et al. Specific loss of histone H3 lysine 9 trimethylation
and HP1gamma/cohesin binding at D4Z4 repeats is associated
with facioscapulohumeral dystrophy (FSHD). PLoS Genet. Jul
2009;5, e1000559.
207. van Overveld PG, et al. Hypomethylation of D4Z4 in 4q-linked
and non-4q-linked facioscapulohumeral muscular dystrophy. Nat
Genet. Dec 2003;35:315.
208. van Overveld PG, et  al. Variable hypomethylation of D4Z4
in facioscapulohumeral muscular dystrophy. Ann Neurol. Oct
2005;58:569.
209. Gabellini D, et  al. Facioscapulohumeral muscular dystrophy in
mice overexpressing FRG1. Nature. Feb 2006;439:973.
210. Gabellini D, Green MR, Tupler R. Inappropriate gene activation
in FSHD: a repressor complex binds a chromosomal repeat deleted
in dystrophic muscle. Cell. Aug 2002;110:339.
211. Klooster R, et  al. Comprehensive expression analysis of FSHD
candidate genes at the mRNA and protein level. Eur J Hum Genet.
Dec 2009;17:1615.
212. Arashiro P, et  al. Transcriptional regulation differs in affected
facioscapulohumeral muscular dystrophy patients compared to asymp-
tomatic related carriers. Proc Natl Acad Sci U S A. Apr 2009;106:6220.
213. Masny PS, et  al. Analysis of allele-specific RNA transcription in
FSHD by RNA-DNA FISH in single myonuclei. Eur J Hum
Genet. Apr 2010;18:448.
214. Rijkers T, et al. FRG2, an FSHD candidate gene, is transcription-
ally upregulated in differentiating primary myoblast cultures of
FSHD patients. J Med Genet. Nov 2004;41:826.
215. Deak KL, et  al. Genotype-phenotype study in an FSHD fam-
ily with a proximal deletion encompassing p13E-11 and D4Z4.
Neurology. Feb 2007;68:578.
216. Yip DJ, Picketts DJ. Increasing D4Z4 repeat copy number com-
promises C2C12 myoblast differentiation. FEBS Lett. Feb
2003;537:133.
217. Dixit M, et  al. DUX4, a candidate gene of facioscapulohumeral
muscular dystrophy, encodes a transcriptional activator of PITX1.
Proc Natl Acad Sci U S A. Nov 2007;104:18157.
218. Geng LN, et al. DUX4 activates germline genes, retroelements, and
immune mediators: implications for facioscapulohumeral dystro-
phy. Dev Cell. Jan 2012;22:38.
219. Kowaljow V, et al. The DUX4 gene at the FSHD1A locus encodes
a pro-apoptotic protein. Neuromuscul Disord. Aug 2007;17:611.
220. Vanderplanck C, et al. The FSHD atrophic myotube phenotype is
caused by DUX4 expression. PLoS One. 2011;6, e26820.
221. Snider L, et  al. Facioscapulohumeral dystrophy:  incomplete sup-
pression of a retrotransposed gene. PLoS Genet 6, e1001181 (Oct,
2010).
222. SM van der Maarel, Tawil R, Tapscott SJ. Facioscapulohumeral
muscular dystrophy and DUX4: breaking the silence. Trends Mol
Med. May 2011;17:252.
223. Smeets DF, et al. ICF syndrome: a new case and review of the lit-
erature. Hum Genet. Sep 1994;94:240.
 224. Hansen RS, et al. The DNMT3B DNA methyltransferase gene is
mutated in the ICF immunodeficiency syndrome. Proc Natl Acad
Sci U S A. Dec 1999;96:14412.
225. Bachman KE, Rountree MR, Baylin SB. Dnmt3a and Dnmt3b are
transcriptional repressors that exhibit unique localization proper-
ties to heterochromatin. J Biol Chem. Aug 24, 2001;276:32282.
226. Chen T, Tsujimoto N, Li E. The PWWP domain of Dnmt3a and
Dnmt3b is required for directing DNA methylation to the major
satellite repeats at pericentric heterochromatin. Mol Cell Biol. Oct
2004;24:9048.
227. Wijmenga C, et al. Genetic variation in ICF syndrome: evidence
for genetic heterogeneity. Hum Mutat. Dec 2000;16:509.
228. Shirohzu H, et al. Three novel DNMT3B mutations in Japanese
patients with ICF syndrome. Am J Med Genet. Sep 15, 2002;112:31.
229. Okano M, Bell DW, Haber DA, Li E. DNA methyltransferases
Dnmt3a and Dnmt3b are essential for de novo methylation and
mammalian development. Cell. Oct 29, 1999;99:247.
230. Matarazzo MR, De Bonis ML, Vacca M, Della Ragione F,
D’Esposito M. Lessons from two human chromatin diseases,
ICF syndrome and Rett syndrome. Int J Biochem Cell Biol. Jan
2009;41:117.
231. Ueda Y, et  al. Roles for Dnmt3b in mammalian develop-
ment:  a mouse model for the ICF syndrome. Development. Mar
2006;133:1183.
232. Jiang YL, et  al. DNMT3B mutations and DNA methylation
defect define two types of ICF syndrome. Hum Mutat. Jan
2005;25:56.
233. Hagleitner MM, et  al. Clinical spectrum of immunodeficiency,
centromeric instability and facial dysmorphism (ICF syndrome).
J Med Genet. Feb 2008;45:93.
234. Saito Y, et  al. Overexpression of a splice variant of DNA meth-
yltransferase 3b, DNMT3b4, associated with DNA hypo-
methylation on pericentromeric satellite regions during
human hepatocarcinogenesis. Proc Natl Acad Sci U S A. Jul 23,
2002;99:10060.
235. de Greef JC, et  al. Mutations in ZBTB24 are associated with
immunodeficiency, centromeric instability, and facial anomalies
syndrome type 2. Am J Hum Genet. Jun 2011;88:796.
236. Chouery E, et al. A novel deletion in ZBTB24 in a Lebanese family
with immunodeficiency, centromeric instability, and facial anoma-
lies syndrome type 2. Clin Genet. Nov 2012;82:489.
237. Heyn H, et  al. Whole-genome bisulfite DNA sequencing of a
DNMT3B mutant patient. Epigenetics. Jun 2012;7:542.
238. Gisselsson D, et  al. Interphase chromosomal abnormalities and
mitotic missegregation of hypomethylated sequences in ICF syn-
drome cells. Chromosoma. Jul 2005;114:118.
239. Jeanpierre M, et al. An embryonic-like methylation pattern of clas-
sical satellite DNA is observed in ICF syndrome. Hum Mol Genet.
Jun 1993;2:731.
240. Ji W, et  al. DNA demethylation and pericentromeric rearrange-
ments of chromosome 1. Mutat Res. Sep 1997;379:33.
241. Prada D, et al. Satellite 2 demethylation induced by 5-azacytidine
is associated with missegregation of chromosomes 1 and 16 in
human somatic cells. Mutat Res. Jan 2012;729:100.
242. Miniou P, et  al. Abnormal methylation pattern in constitutive
and facultative (X inactive chromosome) heterochromatin of ICF
patients. Hum Mol Genet. Dec 1994;3:2093.
243. Ehrlich M, et  al. ICF, an immunodeficiency syndrome:  DNA
methyltransferase 3B involvement, chromosome anomalies, and
gene dysregulation. Autoimmunity. May 2008;41:253.
244. Kondo T, et  al. Whole-genome methylation scan in ICF syn-
drome: hypomethylation of non-satellite DNA repeats D4Z4 and
NBL2. Hum Mol Genet. Mar 1, 2000;9:597.
245. McDowell TL, et al. Localization of a putative transcriptional reg-
ulator (ATRX) at pericentromeric heterochromatin and the short
arms of acrocentric chromosomes. Proc Natl Acad Sci U S A. Nov
23, 1999;96:13983.
5 8   •   P rincip l e s o f G e no m ic M e dicin e
246. Qiu C, Sawada K, Zhang X, Cheng X. The PWWP domain of
mammalian DNA methyltransferase Dnmt3b defines a new family
of DNA-binding folds. Nat Struct Biol. Mar 2002;9:217.
247. Maurer-Stroh S, et al. The Tudor domain “Royal Family”: Tudor,
plant Agenet, Chromo, PWWP and MBT domains. Trends
Biochem Sci. Feb 2003;28:69.
248. Ehrlich M, et al. DNA methyltransferase 3B mutations linked to
the ICF syndrome cause dysregulation of lymphogenesis genes.
Hum Mol Genet. Dec 1, 2001;10:2917.
249. Bickmore WA, van der Maarel SM. Perturbations of chromatin
structure in human genetic disease:  recent advances. Hum Mol
Genet. 2003;12(2): R207.
250. Brown KE, Baxter J, Graf D, Merkenschlager M, Fisher AG.
Dynamic repositioning of genes in the nucleus of lymphocytes pre-
paring for cell division. Mol Cell. Feb 1999;3:207.
251. Alcobia I, Quina AS, Neves H, Clode N, Parreira L. The spatial
organization of centromeric heterochromatin during normal
human lymphopoiesis: evidence for ontogenically determined spa-
tial patterns. Exp Cell Res. Nov 2003;290:358.
252. Gasser SM. Positions of potential: nuclear organization and gene
expression. Cell. Mar 2001;104:639.
253. Jefferson A, et  al. Altered intra-nuclear organisation of het-
erochromatin and genes in ICF syndrome. PLoS One. 2010;5,
e11364.
254. Dupont C, et al. 3D position of pericentromeric heterochromatin
within the nucleus of a patient with ICF syndrome. Clin Genet.
Aug 2012;82:187.
255. Luciani JJ, et al. Subcellular distribution of HP1 proteins is altered
in ICF syndrome. Eur J Hum Genet. Jan 2005;13:41.
256. Lana E, et  al. DNA replication is altered in Immunodeficiency
Centromeric instability Facial anomalies (ICF) cells carrying
DNMT3B mutations. Eur J Hum Genet. Oct 2012;20:1044.
257. Neul JL, et  al. Rett syndrome:  revised diagnostic criteria and
nomenclature. Ann Neurol. Dec 2010;68:944.
258. Girard M, et al. Parental origin of de novo MECP2 mutations in
Rett syndrome. Eur J Hum Genet. Mar 2001;9:231.
259. Amir RE, et al. Rett syndrome is caused by mutations in X-linked
MECP2, encoding methyl-CpG-binding protein 2. Nat Genet.
Oct 1999;23:185.
260. Christodoulou J, Grimm A, Maher T, Bennetts B. RettBASE: The
IRSA MECP2 variation database-a new mutation database in evo-
lution. Human Mutation. May 2003;21(5):466.
261. Villard L, et al. Two affected boys in a Rett syndrome family: clini-
cal and molecular findings. Neurology. Oct 2000;55:1188.
262. Klose RJ, Bird AP. Genomic DNA methylation: the mark and its
mediators. Trends Biochem Sci. Feb 2006;31:89.
263. Meehan RR, Lewis JD, Bird AP. Characterization of MeCP2,
a vertebrate DNA binding protein with affinity for methylated
DNA. Nucleic Acids Res. Oct 11, 1992;20:5085.
264. Klose RJ, et  al. DNA binding selectivity of MeCP2 due to a
requirement for A/T sequences adjacent to methyl-CpG. Mol Cell.
Sep 2, 2005;19:667.
265. Buschdorf JP, Stratling WH. A WW domain binding region in
methyl-CpG-binding protein MeCP2: impact on Rett syndrome.
J Mol Med (Berl). Feb 2004;82:135.
266. Muotri AR, et al. L1 retrotransposition in neurons is modulated by
MeCP2. Nature. Nov 18, 2010;468:443.
267. Chahrour M, et  al. MeCP2, a key contributor to neurological
disease, activates and represses transcription. Science. May 30,
2008;320:1224.
268. Nikitina T, et al. Multiple modes of interaction between the meth-
ylated DNA binding protein MeCP2 and chromatin. Mol Cell
Biol. Feb 2007;27:864.
269. Georgel PT, et  al. Chromatin compaction by human
MeCP2. Assembly of novel secondary chromatin structures
in the absence of DNA methylation. J Biol Chem. Aug 22,
2003;278:32181.
 270. Long SW, Ooi JY, Yau PM, Jones PL. A brain-derived MeCP2
complex supports a role for MeCP2 in RNA processing. Biosci Rep.
Oct 1, 2011;31:333.
271. Shahbazian MD, Antalffy B, Armstrong DL, Zoghbi HY. Insight
into Rett syndrome: MeCP2 levels display tissue- and cell-specific
differences and correlate with neuronal maturation. Hum Mol
Genet. Jan 15, 2002;11:115.
272. Ballas N, Lioy DT, Grunseich C, Mandel G. Non-cell autonomous
influence of MeCP2-deficient glia on neuronal dendritic morphol-
ogy. Nat Neurosci. Mar 2009;12:311.
273. Kriaucionis S, Bird A. The major form of MeCP2 has a novel
N-terminus generated by alternative splicing. Nucleic Acids Res.
2004;32:1818.
274. Mnatzakanian GN, et al. A previously unidentified MeCP2 open
reading frame defines a new protein isoform relevant to Rett syn-
drome. Nat Genet. Apr 2004;36:339.
275. Chen RZ, Akbarian S, Tudor M, Jaenisch R. Deficiency of
methyl-CpG binding protein-2 in CNS neurons results in a
Rett-like phenotype in mice. Nat Genet. Mar 2001;27:327.
276. Guy J, Hendrich B, Holmes M, Martin JE, Bird A. A mouse
MeCP2-null mutation causes neurological symptoms that mimic
Rett syndrome. Nat Genet. Mar 2001;27:322.
277. Guy J, Gan J, Selfridge J, Cobb S, Bird A. Reversal of neuro-
logical defects in a mouse model of Rett syndrome. Science. Feb
2007;315:1143.
278. Giacometti E, Luikenhuis S, Beard C, Jaenisch R. Partial rescue of
MeCP2 deficiency by postnatal activation of MeCP2. Proc Natl
Acad Sci U S A. Feb 2007;104:1931.
279. Robinson L, et  al. Morphological and functional reversal of
phenotypes in a mouse model of Rett syndrome. Brain. Sep
2012;135:2699.
280. Cheval H, et  al. Postnatal inactivation reveals enhanced require-
ment for MeCP2 at distinct age windows. Hum Mol Genet. Sep
2012;21:3806.
281. Tao J, et al. Phosphorylation of MeCP2 at Serine 80 regulates its
chromatin association and neurological function. Proc Natl Acad
Sci U S A. Mar 2009;106:4882.
282. Zhou Z, et al. Brain-specific phosphorylation of MeCP2 regulates
activity-dependent BDNF transcription, dendritic growth, and
spine maturation. Neuron. Oct 2006;52:255.
283. Yasui DH, et  al. Integrated epigenomic analyses of neuronal
MeCP2 reveal a role for long-range interaction with active genes.
Proc Natl Acad Sci U S A. Dec 2007;104:19416.
284. Chang Q, Khare G, Dani V, Nelson S, Jaenisch R. The disease pro-
gression of MeCP2 mutant mice is affected by the level of BDNF
expression. Neuron. Feb 2006;49:341.
285. Stühmer T, Puelles L, Ekker M, Rubenstein JL. Expression from
a Dlx gene enhancer marks adult mouse cortical GABAergic neu-
rons. Cereb Cortex. Jan 2002;12:75.
286. Horike S, Cai S, Miyano M, Cheng JF, Kohwi-Shigematsu T. Loss
of silent-chromatin looping and impaired imprinting of DLX5 in
Rett syndrome. Nat Genet. Jan 2005;37:31.
287. Borg I, et al. Disruption of Netrin G1 by a balanced chromosome
translocation in a girl with Rett syndrome. Eur J Hum Genet. Aug
2005;13:921.
288. Evans JC, et al. Early onset seizures and Rett-like features asso-
ciated with mutations in CDKL5. Eur J Hum Genet. Oct
2005;13:1113.
289. Tao J, et al. Mutations in the X-linked cyclin-dependent kinase-like
5 (CDKL5/STK9) gene are associated with severe neurodevelop-
mental retardation. Am J Hum Genet. Dec 2004;75:1149.
290. De la Fuente R, Baumann C, Viveiros MM. Role of ATRX in chro-
matin structure and function: implications for chromosome insta-
bility and human disease. Reproduction. Aug 2011;142:221.
291. Gibbons RJ, et  al. Mutations in transcriptional regulator ATRX
establish the functional significance of a PHD-like domain. Nat
Genet. Oct 1997;17:146.
Epig e n etic s , Epig e no m ic s , and H u m an D i s e a s e   •   5 9
292. Cardoso C, et al. ATR-X mutations cause impaired nuclear loca-
tion and altered DNA binding properties of the XNP/ATR-X
protein. J Med Genet. Oct 2000;37:746.
293. Tang J, et al. A novel transcription regulatory complex containing
death domain-associated protein and the ATR-X syndrome pro-
tein. J Biol Chem. May 7, 2004;279:20369.
294. Hendrich B, Bickmore W. Human diseases with underlying defects
in chromatin structure and modification. Hum Mol Genet. Oct 1,
2001;10:2233.
295. Ausio J, Levin DB, De Amorim GV, Bakker S, Macleod PM.
Syndromes of disordered chromatin remodeling. Clin Genet. Aug
2003;64:83.
296. Gibbons RJ, Higgs DR. Molecular-clinical spectrum of the ATR-X
syndrome. Am J Med Genet. 2000;97:204.
297. Gibbons RJ, et al. Mutations in ATRX, encoding a SWI/SNF-like
protein, cause diverse changes in the pattern of DNA methylation.
Nat Genet. Apr 2000;24:368.
298. Xue Y, et  al. The ATRX syndrome protein forms a
chromatin-remodeling complex with Daxx and localizes in promy-
elocytic leukemia nuclear bodies. Proc Natl Acad Sci U S A. Sep 16,
2003;100:10635.
299. Wang J, et  al. Promyelocytic leukemia nuclear bodies associate
with transcriptionally active genomic regions. J Cell Biol. Feb 16,
2004;164:515.
300. Dhayalan A, et  al. The ATRX-ADD domain binds to H3 tail
peptides and reads the combined methylation state of K4 and K9.
Hum Mol Genet. Jun 2011;20:2195.
301. Cardoso C, et al. Specific interaction between the XNP/ATR-X
gene product and the SET domain of the human EZH2 protein.
Hum Mol Genet. Apr 1998;7:679.
302. Kuzmichev A, Jenuwein T, Tempst P, Reinberg D. Different
EZH2-containing complexes target methylation of histone H1 or
nucleosomal histone H3. Mol Cell. Apr 23, 2004;14:183.
303. Kuzmichev A, Nishioka K, Erdjument-Bromage H, Tempst P,
Reinberg D. Histone methyltransferase activity associated with a
human multiprotein complex containing the Enhancer of Zeste
protein. Genes Dev. Nov 15, 2002;16:2893.
304. Cao R, et  al. Role of histone H3 lysine 27 methylation in
Polycomb-group silencing. Science. Nov 1, 2002;298:1039.
305. Czermin B, et  al. Drosophila enhancer of Zeste/ESC complexes
have a histone H3 methyltransferase activity that marks chromo-
somal Polycomb sites. Cell. Oct 18, 2002;111:185.
306. Daujat S, Zeissler U, Waldmann T, Happel N, Schneider R. HP1
binds specifically to Lys26-methylated histone H1.4, whereas
simultaneous Ser27 phosphorylation blocks HP1 binding. J Biol
Chem. Nov 11, 2005;280:38090.
307. Kuzmichev A, et al. Composition and histone substrates of poly-
comb repressive group complexes change during cellular differen-
tiation. Proc Natl Acad Sci U S A. Feb 8, 2005;102:1859.
308. Hiragami K, Festenstein R. Heterochromatin protein 1: a pervasive
controlling influence. Cell Mol Life Sci. Nov 2, 2005;62(23):2711.
309. Le Douarin B, et  al. A possible involvement of TIF1 alpha and
TIF1 beta in the epigenetic control of transcription by nuclear
receptors. Embo J. Dec 2, 1996;15:6701.
310. Lechner MS, Schultz DC, Negorev D, Maul GG, Rauscher FJ
III. The mammalian heterochromatin protein 1 binds diverse
nuclear proteins through a common motif that targets the
chromoshadow domain. Biochem Biophys Res Commun. Jun 17,
2005;331:929.
311. Kourmouli N, Sun YM, van der Sar S, Singh PB, Brown JP.
Epigenetic regulation of mammalian pericentric heterochromatin in
vivo by HP1. Biochem Biophys Res Commun. 2005;337(3):901–907.
312. Gibbons RJ, et al. Mutations in the chromatin-associated protein
ATRX. Hum Mutat. Jun 2008;29:796.
313. Law MJ, et  al. ATR-X syndrome protein targets tandem repeats
and influences allele-specific expression in a size-dependent man-
ner. Cell. Oct 2010;143:367.
 314. Steensma DP, Gibbons RJ, Higgs DR. Acquired alpha-thalassemia
in association with myelodysplastic syndrome and other hemato-
logic malignancies. Blood. Jan 2005;105:443.
315. Hanahan D, Weinberg RA. Hallmarks of cancer: the next genera-
tion. Cell. Mar 4, 2011;144:646.
316. Pujadas E, Feinberg AP. Regulated noise in the epigenetic land-
scape of development and disease. Cell. Mar 16, 2012;148:1123.
317. Kulis M, Esteller M. DNA methylation and cancer. Adv Genet.
2010;70:27.
318. Jones PA. Functions of DNA methylation: islands, start sites, gene
bodies and beyond. Nat Rev Genet. Jul 2012;13:484.
319. Baylin SB, Jones PA. A decade of exploring the cancer epig-
enome—biological and translational implications. Nat Rev Cancer.
Oct 2011;11:726.
320. Ohm JE, et al. A stem cell-like chromatin pattern may predispose
tumor suppressor genes to DNA hypermethylation and heritable
silencing. Nat Genet. Feb 2007;39:237.
321. Vire E, et al. The Polycomb group protein EZH2 directly controls
DNA methylation. Nature. Feb 16, 2006;439:871.
6 0   •   P rincip l e s o f G e no m ic M e dicin e
322. Huarte M, Rinn JL. Large non-coding RNAs: missing links in can-
cer? Hum Mol Genet. Oct 15, 2010;19:R152.
323. Lovat F, Valeri N, Croce CM. MicroRNAs in the pathogenesis of
cancer. Semin Oncol. Dec 2011;38:724.
324. Yu W, et al. Epigenetic silencing of tumour suppressor gene p15 by
its antisense RNA. Nature. Jan 10, 2008;451:202.
325. Morris KV, Santoso S, Turner AM, Pastori C, Hawkins PG.
Bidirectional transcription directs both transcriptional gene
activation and suppression in human cells. PLoS Genet. Nov
2008;4,:e1000258.
326. Khalil AM, et al. Many human large intergenic noncoding RNAs
associate with chromatin-modifying complexes and affect gene
expression. Proc Natl Acad Sci U S A. Jul 14, 2009;106:11667.
327. Baek D, et  al. The impact of microRNAs on protein output.
Nature. Sep 4, 2008;455:64.
328. Heyn H, Esteller M. DNA methylation profiling in the clinic: appli-
cations and challenges. Nat Rev Genet. Oct 2012;13:679.
329. Merkenschlager M, Odom DT. CTCF and cohesin: linking gene
regulatory elements with their targets. Cell. Mar 14, 2013;152:1285.
 Free ebooks ==> www.Ebook777.com
5.
GENES, GENOME, AND DEVELOPMENTAL
MALFORMATIONS
Dhavendra Kumar
D
isorders that affect tissue differentiation, organo-
genesis, and morphogenesis constitute a signifi-
cant proportion of human genetic disease. These
disorders may result from any of the known genetic mecha-
nisms and may occur singly, in combination with other
malformations, or as part of the multisystem complex phe-
notype. Clinical and laboratory analyses of these disorders
have led to the emergence of clinical dysmorphology as a
distinct discipline, now an integral part of medical genetics
and clinical genomics. All practicing pediatricians, clinical
geneticists, and genetic counselors are required to have the
basic skills in order to deal with patients who present with
a developmental malformation that may occur either singly
or as part of a recognizable dysmorphic syndrome ( Jones
and Smith, 2006).
During the past two decades, clinical dysmorphologists
(specialist clinicians dealing with clinical management of
malformations) have delineated a large number of mal-
formation syndromes that comprise multiple malforma-
tions affecting unrelated organs and tissues. Researchers
and clinicians continue to generate more information,
and the number of related syndromes continues to
increase. Dedicated dysmorphology databases (London
Medical Databases, www.lmdatabases.com [London
Dysmorphology, LDDB and London Ophthalmic
Genetics, GeneEye], Pictures of Standardized Syndromes
and Undiagnosed Malformations [www.POSSUM.net.
au]) are now available to help clinical geneticists and
other clinicians for carrying out comprehensive database
searches. Although considerable progress has been made in
understanding the developmental pathology and possible
causes of some of these malformation syndromes, there
remains a huge gap in our understanding of the abnormal
human development and clinical challenges.
Recent advances in genetics and genomics have made
it possible to understand the molecular basis of develop-
mental malformations. These fall into two broad groups.
61
www.Ebook777.com
The first group, which includes advances in gene map-
ping, cloning, and identification of genes, has received
a tremendous boost from the Human Genome Project
(HGP) (see Chapter  1). Advances in genome sequenc-
ing, particularly exome genome sequencing, following
the successes of HGP and other projects, have helped
in identifying mutations and polymorphisms in spe-
cific human genes associated with particular syndromes.
The HGP has also permitted comparisons between the
human genome and the genomes of other organisms such
as mouse and Drosophila, which has helped in the iden-
tification of new genes, deciphering evolutionary con-
served genes and genomic regions, and the subsequent
study of their role in abnormal development. The second
group includes a number of highly interactive molecular
pathways that govern the developmental process. These
molecular pathways include several specific genes and
polymorphisms belonging to a particular molecular fam-
ily; for example, RAS/MAPK, NOTCH signaling, and
many more. These advances have made it possible to
develop a better understanding of development and mor-
phogenesis and to work out a plausible genotype–pheno-
type correlation.
This chapter reviews some of the basic concepts in
developmental biology, which are intricately related to
genes and genomes. This chapter is not intended to pro-
vide the reader with a comprehensive account of normal
and abnormal human development, as this field is enor-
mous and beyond the scope of this review. The interested
reader may refer to excellent texts available on this subject
(see “Further Reading”). The main aim of this chapter is to
introduce the subject from the perspective of genomics. It
is anticipated that the reader may find the contents of this
chapter helpful in understanding aspects of advances in
genomic-related science and technology that are relevant
to understanding developmental malformations, which
may in turn influence their clinical practice.
 G ENOME P RO J EC T S
Since the 1990s, the efforts to sequence a number of eukary-
otic genomes have been successful, resulting in the avail-
ability of complete genome sequences. In addition, several
prokaryotic organisms have been sequenced (http://www.
tigr.org/). Apart from completion of the human genome
sequence (Lander, Linton, et al., 2001; Istrail, Sutton et al.,
2004), genomes that have been sequenced in their entirety
include mouse, Drosophila melanogaster (Goldstein and
Gunawardena, 2000), worm (Caenorhabditis elegans) (Li,
Stoeckert, et al., 2003), and budding yeast (Saccharomyces
cerevisiae) (Krogan, Cagney, et al., 2006). Despite limited
information on the precise gene content and gene order
in the human genome, most of the approximately 23,000
human genes (see Chapter 2) are comparable to the corre-
sponding mouse genome. It is widely agreed that the avail-
ability of genome sequences is of paramount importance
to our understanding of fundamental structural and func-
tional units across a broad range of eukaryotic organisms,
from the fruit fly to higher primates like Homo sapiens.
Such understanding has already led to the discovery of sev-
eral gene families and gene-sequence polymorphisms that
regulate embryonic development and differentiation, and
are involved in the pathogenesis of several human malfor-
mation disorders.
Among the recent initiatives, new relevant genome proj-
ects include the Human Variome/Phenome Project (www.
hvp.org), the GenPhen (www.Gen2Phen.org) and Sanger’s
the Decipher Developmental Delay (www.DDDUK.org).
The common theme of these projects is to decipher devel-
opmental phenotypes in the most comprehensive genomic
context. The positive outcomes of these projects could revo-
lutionize the clinical and preventive approach to a number
of human developmental disorders.
G E N O M E S EQ U E N C E
P O LY MO R P H I S M S
Human beings do not all possess the same genomic
sequence. Every person, except probably among the mono-
zygotic twins, has a unique sequence that is different from
any other person’s by at least one change in every 500–1000
base pairs (bp). This variation or polymorphism presum-
ably does not carry any significance in terms of structural
or functional phenotype (however, it could carry some
evolutionary biological significance). Nevertheless, the
process of genome sequencing needs to take into account
the existence of this variation between individuals, popu-
lations, and ethnic/racial groups and its possible influence
6 2   •  P r i n c i p l e s o f G e n o m i c M e d i c i n e
on the phenotype. Although their role is yet unclear, these
polymorphisms could affect the level or activity of certain
classes of proteins through direct or epigenetic influence on
a range of genes or alleles. In humans, some of these dif-
ferences may be disease-causing or result in developmental
malformations. Genomic polymorphisms could be bio-
logically important by virtue of being physically within the
coding region (exons) of genes, outside of or close to the
promoter region of genes, altering or modifying the func-
tion of non-coding RNAs, or influencing the structure
or function of the imprinting control regions with major
epigenomic outcomes. In addition, these sequence poly-
morphisms could be extremely important in the identifica-
tion of both single loci and sets of loci that produce disease.
The most common type of polymorphic genomic variation
includes single nucleotide polymorphisms (SNPs) and
copy number variations (CNVs). Several research insti-
tutions and organizations are engaged in collecting data
on SNPs (http://snp.cshl.org/; http://www.humgen.nl/
SNP_databases.html, and many more) and CNVs (http://
cnv.gene-quantification.info/ and many more) that are
correlated with disease states, including human malfor-
mations. The analysis of such data must take into account
the population structure, since the expression of a particu-
lar allele may depend on the presence or absence of other
genetic variability in the genome. This is a mammoth and
challenging task, but it is likely to identify which of these
polymorphisms are associated with abnormal human devel-
opment and disease.
C O M PA R AT I VE G E N O M I C S A N D
HUM A N D EVE L O PM E N T
The sequencing of genomes from different organisms now
provides a splendid opportunity to compare the functional
significance of known or predicted genes in one species
with those in another species. A systematic comparison of
closely related organisms, such as mouse and human, in
which most genes are evolutionarily conserved, has revealed
the location of homologous genes. This information is made
available and regularly updated on the Internet (www.infor-
matics.jax.org/; www.geneontology.org).
Similarly, by comparisons between the genomes of dis-
tantly related species, such as humans and fruit flies, that
share common developmental processes, a vast amount of
information about the genetic basis of development has
been accumulated. These comparisons have helped in iden-
tifying genes belonging to gene/molecular families that
are derived from an ancestral gene in a common branch
of the evolutionary tree. The genes in such a family are
 Free ebooks ==> www.Ebook777.com

called paralogous, and they originate from repeated gene
duplication events followed by evolutionary divergence of
the duplicated genes (Figure  5.1). The proteins encoded
by such genes share common sequence characteristics and
are referred to as belonging to a gene family. The activi-
ties of such proteins may differ slightly, or the proteins
may perform a different function as a result of mutation
or natural selection, and they are also likely to have been
produced at different stages of development. The function
of the proteins may also vary in different tissues due to
alternative splicing and mRNA editing, causing modifica-
tion of the mRNA sequence and the subsequent protein
sequence. It is known that the same gene family identified
by the same function can occur in two or more different
species. Such genes are called orthologous. Using a cluster
analysis of such proteins, it is possible to determine which
of the synthesized proteins are homologous and therefore
likely to have similar functions. This type of analysis has
helped in understanding the evolution of gene/ molecular
families that influence development (Figure 5.2) (Mount,
2004).
P H Y LO G ENOMICS AND H U MAN
DEVELO PMEN T
Developmental biologists remain intrigued by the com-
plexities of developmental mechanisms. Among many

Anterior Posterior

lab pb bcd, Dfd Scr ftz Antp Ubx abd-A Abd-B

zen
Drosophila HOM-C

Hox1 Hox2 Hox3 Hox4 Hox5 Hox6 (central) Hox7 (posterior)

Ancestral urbilaterian
HOM-C

A1 A2 A3 A4 A5 A6 A7 A9 A10 A11 A13

HOxA

B1 B2 B3 B4 B5 B6 B7 B8 B9 B13
HOxB
Human

C4 C5 C6 C8 C9 C10 C11 C12 C13

HOxC

D1 D2 D4 D8 D9 D10 D11 D12 D13

HOxD
1 2 3 4 5 6 7 8 9 10 11 12 13
Homology group
3' 5'
Transcription

Anterior Posterior

Evolutionary conservations of genomic organization and expression patterns of Drosophila fruit fly and mammal Hox genes. The human
Figure 5.1 
embryo shows anterior-posterior cluster of four Hox genes based on mouse expression studies. The fruit fly shows Drosophila Hox genes aligned with
their mammalian orthologues and corresponding expression patterns mapped onto the body plan. (Robert, 2008)

G e n e s , G e n o m e , a n d D e ve l o pm e n ta l M a l f o r m at i o n s   •   6 3

www.Ebook777.com
 Gene
Chromosome
One function A

Gene duplication

A1 A2
Speciation

Two functions A1 A2 A1 A2

Species A Species B

(same function)
Orthologs

Paralogs
(related by gene duplication)

Repeated duplications

Species A Species B

All-by-all sequence analysis

followed by clustering of the
most alike sequences
reveals orthologs ands paralogs

Orthologous pairs
(rest are members of a paralogous family)

Figure 5.2  Orthologous and paralogous origin of gene families through successive gene duplication events. (Reproduced from Mount, 2004, with permission from Oxford
University Press, New York.)

challenging avenues, by far the most complex is molecu-
lar evolutionary biology. In the genomic context it is also
referred to as phylogenomics. The most challenging ques-
tion remains, how can species share similar developmental
genetic or genomic toolkits but still generate diverse life
forms, ranging from an invertebrate worm to a human,
the most advanced vertebrate? Conversely, how can simi-
lar forms develop from different toolkits? To large extent,
advances and unraveling of genomics have offered many
answers. Genomics bridges the gap between evolutionary
and developmental biology and thus helps answer several
questions on the evolution and development (evo-devo)
philosophy of the developmental biology (Cañestro,
2012). There numerous ways by which phylogenomics has
emerged as a leading sub-field in developmental biology:
1) It sets out the direction of evolutionary diversification
by clarifying historical relationships;
2) it provides the necessary genetic and genomic toolkits
for studying the genome expansion and contraction;
3) it offers logical clarifications to underlying mechanisms
for evolution and developmental functions; and
finally;
4) it helps in the identification of conserved non-coding
elements and their relationship to genome architecture
and development.
In order to identify genes that share sequences with sim-
ilar biological function developmental, biologists search the
genomes of different organisms. This search is commonly

6 4   •  P r i n c i p l e s o f G e n o m i c M e d i c i n e
referred to as comparative genomics. It is important because
developmental genetic or genomic toolkits in the two dif-
ferent organisms are likely to have originated from a com-
mon ancestral gene in the evolutionary past. It is likely that
if mutations in one of the genes result in a developmental
malformation, then the other gene presumably has a simi-
lar role in the organism’s development. This inference holds
true even for strikingly dissimilar organisms, such as fruit
flies and humans.
Biologists agree that evolutionarily divergent organ-
isms use similar fundamental gene systems and associ-
ated encoded proteins during developmental processes.
Developmental variation is dependent, not solely on the
structural variations in these genes, but also on gene function
and regulation during evolutionary cycles. This is referred to
as the evo-devo concept of development (David, 2004) and
focuses on identifying basic sets of genes in organisms and
studying how they are regulated in developing cells and tis-
sues. One approach to identifying such gene sets is a detailed
functional identification of specific transcription factors and
cis-acting binding sites for these factors (Davidson, McClay,
et al., 2003). Another approach is the search for common bio-
logical functions in distantly related organisms, which may
help in tracing the evolutionary and developmental origins
of genes. For example, the genes regulating light perception
in a microbe are similar to those that regulate the function
of chloroplasts, and, eventually, the evolution of complex
functions of the eye (Gehring, 2002). Other evo-devo studies
support the premise that variations in the genes themselves
are important for evolution and development.
Comparison of genomes from different organisms has
revealed the fascinating phenomenon of genome expansion
and contraction. Comparative genomics provides a power-
ful tool to discover trait-specific genes on the basis of the
assumption that most genes that were expressed exclusively
in a trait are lost if the trait was secondarily lost (Cañestro,
Yokoi, et  al., 2007). This hypothesis is supported by the
observation in Drosophila melanogaster and C.  elegans
that genes for cilia or flagella in organisms that have them
are absent in other organisms that lack these organelles
(Figure 5.3).
The power of comparative genomics is best exemplified
by studying genes for cilia and flagella, microtubule-based
organelles that are important for development of left–right
asymmetries, heart formation, vertebrate photoreceptors,
and invertebrate mechano- and chemoreceptors (Pazour
and Witman, 2003). Comparative genomics of organisms
with cilia (such as flies, roundworms, green algae, proto-
zoans, and humans), and organisms that lack cilia (such as
plants, yeasts, and slime molds; Avidor-Reiss, Maer, et al.,
G e n e s , G e n o m e , a n d D e ve l o pm e n ta l M a l f o r m at i o n s   •   6 5
2004), and comparison of organisms with flagella (such
as green algae, flies, roundworms, sea squirts, mice, and
humans) and organisms that lack flagella (such as plants; Li,
Gerdes, et al., 2004), identified several hundred candidate
genes related to cilia or flagella. An exhaustive search led to
the detection of more than 80% of ancestral genes that are
known to be involved in ciliary function. The proteomic
analysis identified a novel family of proteins (OSEG) that
are essential for the development of cilia in Drosophila
melanogaster (Holland, 2007). Studies in silico, in vitro, and
in vivo in Caenorhabditis elegans validated flagella-related
genes, and identified a novel human gene (BBS5) as defec-
tive in Bardet-Biedl syndrome (Li, Gerdes, et al., 2004). It
is acknowledged that further applications of this genomic
strategy will facilitate the identification of candidate genes
that are important for the development and evolution of a
variety of developmental traits.
G ENOME ARC H I T EC T U RE AND
DEVELO PMEN T
Developmental biologists have demonstrated that, by incor-
porating comparative genomics, a clear picture of the devel-
opmentally relevant part of genomes can be built. These
so-called genomic developmental toolkits include genome
contraction, genome expansion, epigenetic segments of the
genome, and evolutionarily conserved non-coding elements
(Figure 5.4). In simple terms, these elements constitute the
genome “architecture” relevant to the organism’s develop-
ment. The net contribution of the genome architecture is
evidenced by different expression patterns in different spe-
cies (orthologues) compared to different expression pat-
terns within species (paralogues). Comparative genomics
has offered an understanding of mechanisms that govern
difference in expression patterns and the evolutionary sta-
bility of the genome architecture (Cañestro, Yokoi, et  al.,
2007).
Another strategy for finding evolutionarily relevant
genes is studying the effects of environmental factors, such
as a dietary supplement. For example, a particular strain
of inbred mice can have either a normal or a highly defec-
tive skeleton due to the presence of immature cartilage,
simply by altering the diet (Wu, Li, et  al., 2008). This is
probably related to the suppression of mutant Hox gene
expression by a dietary substance, supporting the argument
that expression of accumulated mutations in an important
developmental gene may not occur until accompanied by
an environmental change (Wu, Li, et  al., 2008). It is pos-
sible that some of these evolutionary events may be neutral
 Free ebooks ==> www.Ebook777.com

(A)
Trait present (+)
or absent (–) – + + + – + + – +

Trait loss

Trait loss
Trait loss
+

Trait present in last

common ancestor

(B)
Genomes lacking the trait – – –

Genomes with the trait

+ + + + + +

Intersection enriched
in trait-related genes

Core of shared candidate Candidate genes for

trait-specific genes trait subcomponents
shared among some
organisms

Figure 5.3 
Comparison of genomes with lost traits: (A) identifies candidate trait-related genes that are present in organisms that have the trait but are
absent in organisms that lack the trait; (B) Comparison of different levels of evolutionary distance for trait-specific genes that are considered lost
without the trait (Cañestro, Yokoi, et al., 2007).
Subphylum Cephalochordates Urochordates Vertebrates

Class Ascidians Larvaceans

1 2 3 4 5 6 7 8 9 10 111213
1 2 3 4 5 6 7 8 9 10111213 14 1 2–4 10 65 12 13 1 2 4 9 9’ 10111213 A
B
Hox clusters C
D

Loss of classic RA machinery

Loss of Dnmt1 and Dnmt3
CNS reorganization (absence of midbrain)
Axial patterning becomes independent of RA
Hox cluster rupture and loss of temporal collinearity

Genome diminution
Determinative development, rapid embryogenesis and life cycle

Figure 5.4 
Genome contraction and morphology—note reduced the size of urochordates’ genomes, resulting in the loss of temporal collinearity of
Hox-gene expression by breaking up their Hox-gene cluster; loss of the need to use retinoic acid (RA) for anteroposterior axial patterning associated
with the reorganization of their CNS. Larvaceans lack the classic genetic machinery to synthesize, degrade, and detect RA, and they also lack a
complete genetic system for DNA methylation (carried out by DNA methyltransferases; Dnmts). This illustration demonstrates that the genome
contraction builds a complete chordate body plan that is retained throughout life (Cañestro, Yokoi, et al., 2007).

6 6   •  P r i n c i p l e s o f G e n o m i c M e d i c i n e

www.Ebook777.com
 Eyespot
Sensory
appendages?
Mouth
A/P Axis
Hox genes
Conserved developmental patterning systems in a hypothetical ancestral creature; Anterior/posterior axis segmental identity by Hox genes;
Figure 5.5 
A/P axis patterning by hedgehog genes and through suppression of BMP signaling; dorsal/ventral division by Notch signaling and promotion of
appendage outgrowth by Distalless; and formation of light-sensitive organs by Eyeless/Pax6. (Reproduced from Bier and McGinnis, In ‘Inborn Errors of Development’, Eds.
Epstein, Erickson et al., 2004, with permission from Oxford University Press.)
or deleterious. However, as with evolution in general, some
of these events may be beneficial, leading to a new develop-
mental scheme under the influence of natural selection. For
example, the great diversity in butterfly wings is linked to
genetic variation, and the resulting phenotypes are attrib-
uted to the cumulative action of natural selection on devel-
opmental processes (Beldade and Brakefield, 2002).
In brief, there are many developmental pathways com-
mon to invertebrates and vertebrates, including humans.
These have originated from a sophisticated, bilaterally
symmetrical ancestral creature regulated by many architec-
tural and organ-specific genetic systems (Figure 5.5). These
genetic systems have been conserved through successive
evolutionary cycles.
RE GU LAT ION OF
DEVELO PMEN TAL G ENES
Previous sections have elucidated some basic aspects of
genetics and genomics that are intricately connected with
biological development across a wide range of organisms.
Taken together, this is referred to as developmental biol-
ogy, which is a science concerned with the genetic basis of
anatomy. During the past few years, we have developed a
far better understanding of the basic mechanisms of devel-
opmental biology, one that is beyond the scope of this
chapter (see Table 5.1).
It is important to appreciate that, for each of these fun-
damental processes, many genes are involved that relate to
each other through developmental pathways. These genes
function in a coordinated manner and encode proteins that
share common biological properties. For this reason, these
G e n e s , G e n o m e , a n d D e ve l o pm e n ta l M a l f o r m at i o n s   •   6 7
Head
Abdomen Tail
Anus/
Genitals
Gills?
D/V Axis
Dpp/BMP4 Non-neural Ectoderm
Photosensitive Protrusions or
Sog/Cod Neural Ectoderm
organs appendages
Table 5.1 MECHANISMS IN DEVELOPMENTAL
BIOLOGY
1. Mechanisms of differential gene expression
2. Role of enhancers and promoters
3. Signal-transduction pathways linking cell membrane and nucleus
4. Mechanisms of origin of syndromes
5. Mechanisms producing dominant and recessive traits
6. Molecular mechanisms for morphogenetic interactions
7. Role of genome variation in morphogenesis
SOURCE: Adopted from ‘Inborn Errors of Development’, Eds. Epstein, Erickson
et al., 2004, with permission from Oxford University Press, NY.
are also referred to as being in a gene family. Several such
gene families include genes that encode for a special class
of proteins that bind to enhancer or promoter regions and
interact to activate or repress the transcription of a particu-
lar gene. These are called transcription factors (trans). Most
transcription factors can bind to specific DNA sequences.
These proteins can be grouped together into protein families,
based on structural similarities (Table 5.2). DNA-binding
sites in a particular transcription factor family are similar.
Any alteration in the amino acids at the binding sites can
alter the DNA sequence to which the factor binds.
T R A N S C R I P T I O N FAC TO R S
Transcription factors have three major domains:  firstly,
a DNA-binding domain, which recognizes a particular
DNA sequence; secondly, a trans-activating domain, which
activates or suppresses the transcription of the gene whose
enhancer or promoter it has found; and finally, there may
be a protein–protein interaction domain, which allows the
Table 5.2  TRANSCRIPTION FACTOR FAMILIES AND FUNCTIONS
FAMILY REPRESENTATIVE TRANSCRIPTION FACTORS
Homeodomain
  HOX HOXA-1, HOXB-2, HOXC, HOXD-1 etc.
 POU PIT1, UNC-86, Oct-2
 LIM LIM-1, Forkhead
 PAX PAX-1, -2, -3, etc.
Basic helix-loop-helix MYOD, achaete
Basic leucine zipper C/EBP, AP1
Zinc finger
 Standard WT1, Krüppel
 Sry–Sox Sry, SOXD, Sox2
SOURCE: Adopted from ‘Inborn Errors of Development’, Eds. Epstein, Erickson et al., 2004, with permission from Oxford University Press, NY.
transcription factor’s activity to be modulated by transcrip-
tion binding protein (TBP), transcription-associated fac-
tors (TAFs), or other transcription factors.
There are numerous diseases that result from a defi-
ciency of transcription factors, often termed “transcription
factoropathy” (Gilbert, 2004). The first such human dis-
ease was androgen insensitivity syndrome (AIS), in which,
despite normal testosterone production, the affected male
externally develops as a phenotypical female and fails to
develop secondary sexual characteristics. In AIS, the testos-
terone receptor is either absent or deficient, and its DNA
fails to bind the DNA of male-specific genes (Migeon,
Brown, et  al., 1981). Conversely, the binding of DNA
and consequent activation of the receptor site can lead
to Waardenburg syndrome, type II. In this disorder, the
affected heterozygotes have a white forelock, are deaf, and
have multicolored irides. They possess a wild-type allele
of the microphthalmia (MITF) gene. Activation of this
transcription factor through the protein tyrosine kinase
cascade enables it to dimerize and bind to the regulatory
regions of particular genes that open a region of DNA for
transcription.
Transcription factors work in conjunction with other
transcription factors to activate particular genes. However,
the binding of a specific transcription factor to the enhancer
or promoter of a gene does not always cause transcription of
the gene. Some of these transcription factors, called “mas-
ter regulatory genes,” are important and take the lead in the
transcription process. For example, PAX6 (eye) and MYOD
(muscles) work in concert to initiate cellular differentia-
tion. The use of PAX6 by different organs illustrates the
modular nature of transcriptional regulatory units. PAX6 is
needed for mammalian eye, nervous system, and pancreatic
6 8   •  P r i n c i p l e s o f G e n o m i c M e d i c i n e
KEY FUNCTIONS
Axis formation, patterning
Pituitary development, neural fate
Head development
Neural specification, eye development
Muscle and nerve specification
Liver differentiation, fat cell specification
Kidney & gonad development, hormone receptors,
­estrogen receptor, secondary sex determination
Bone, primary sex determination
development. In addition to ocular abnormalities, muta-
tions in the PAX6 gene can cause severe nervous system and
pancreatic optic abnormalities (Cvekl and Tamm, 2004).
PAX6-binding sequences have been found in the enhanc-
ers of vertebrate lens crystalline genes and in the genes for
insulin, glucagon, and somatostatin, which are expressed in
the endothelial cells of the pancreas.
There are some basic principles that govern the role
played by transcription factors. Firstly, transcription fac-
tors function in combination with other transcription fac-
tors. Secondly, transcription factors find their way to the
nucleus either through cell lineage or induction. Thirdly,
transcription factors can continue to be synthesized after
the original signal has ceased. Finally, post-translational
modification is often necessary to ensure adequate func-
tioning of the transcriptional factors. This mechanism
is probably of major importance in differentiation and
morphogenesis.
OT H E R M EC H A N I S M S
Differential transcription of DNA is not solely dependent
on transcriptional factors. There are other mechanisms that
influence developmental gene regulation. Even though
a particular RNA transcript may be synthesized, it is not
always possible to generate a functional protein. In order
for the mRNA to become an active protein, a number of
other steps are important, including processing of mRNA
by removal of introns, translocation from the nucleus to the
cytoplasm, translation by the protein-synthesizing appara-
tus, and post-translational modification to make an active
protein molecule. Regulation can occur at any of these steps
during development.
 RE GU LAT ION OF EM B RYO G ENESIS
AND OR G ANO G ENESIS
Soon after fertilization, the rapid cell division in the result-
ing embryo paves the way for the formation of ectoderm,
mesoderm, and endoderm, the three primordial germ lay-
ers. Early embryogenesis is regulated by a complex system
of proteins that regulate complex processes of differentia-
tion and morphogenesis (Table 5.1; Figure 5.6). Toward
the end of embryonic development and differentiation,
specific organ formation is initiated. Organogenesis is
highly complex, as different organs are composed of tissues
derived from different primordial layers. Disruption in one
of these tissues will lead to either a structural malformation
or a functional abnormality in the organ. For example, in
the eye, a precise arrangement of tissues forming the trans-
parent cornea, lens, vitreous, choroid, and neural retina is
necessary for normal shape and function. Thus, construc-
tion of organs is accomplished by a group of cells chang-
ing the behavior of an adjacent set of cells, causing them to
change their shape, mitotic rate, and eventual fate. This is
called interaction. The interaction between closely located
cells or tissues is referred to as proximate interaction. This
process is continued throughout organogenesis. Proximate
interaction consists of two components—inducers, the tis-
sue producing the signal, and responders, the tissue being
induced. The ability of the tissue to respond to the induc-
tion signal is called competence. It is an active process,
as the responding tissue undergoes several changes and
interaction with other factors to ensure formation of the
intended organ. For example, in the formation of the lens
Differentiation and morphogenesis
(A)
Ligand
Ligand-binding
domain
Extracellular
Cytoplasm Inacive P P
responding
protein
Dormant ATP
tyrosine
kinase
ADP
domain
Acive
responding
protein
Figure 5.6  Major molecular steps in embryonic differentiation and morphogenesis.
G e n e s , G e n o m e , a n d D e ve l o pm e n ta l M a l f o r m at i o n s   •   6 9
in the eye, Pax6 acts as the inducer for the ectoderm in the
optic vesicle, which in turn acts with fibroblast growth fac-
tor-8 (FGF8) and other factors (Sox2, Sox3, and L-Maf )
to ensure the production of the lens. Further induction
is called instructive, when the responding tissue depends
on a specific tissue to begin the process. In general, there
are three broad defining principles of instructive induction
(Souza, Kuliszewski, et al., 1995): (1) tissue A is necessary
for tissue B to respond in a desired manner; (2) tissue B
does not respond in the desired manner in the absence of
tissue A; (3) tissue B may not respond in the desired man-
ner in the presence of another tissue, but in the absence
of tissue A. An example of this form of instruction is in
the optic vesicle, which, when placed in another part of
the developing head ectoderm, can form an ectopic lens.
Induction also depends on environmental factors, which is
called permissive induction.
R EG I O NA L S P EC I FI C IT Y O F I N D U C T I O N
Induction is a dynamic process that governs the early cell and
tissue differentiation leading to specific organ formation. It
particularly involves an interaction between various tissues,
particularly those that lie adjacent to each other (Gilbert,
2004). For example, the interaction between epithelial cells
and mesenchymal cells is probably the most important in the
development of several organs (Table 5.3); the best example
of which is the skin. Skin comprises epidermis (epithelial)
and dermis (mesodermal), developed from the interactions
of sheets of epithelial cells and mesenchymal cells derived
from the mesoderm (see also Chapters 39.1 and 39.2). This
(B) Steel (stem.cell factor)
Receptor (Kit)
Extracellular
Cytoplasm P P
Adapter
protein
Ras
Acive
tyrosine GDP ERK Mitf Melanoblast-
kinase
specific gene
MEK
Ras GTP
p300/CBP
P
P P
RAF
Transcription
 Table 5.3 ORGANS DERIVED FROM EPITHELIAL–MESENCHYMAL INTERACTIONS (GILBERT, 2004)
ORGAN EPITHELIAL COMPONENT
Skin appendages Epidermis (ectoderm)
Tooth Jaw epithelium (ectoderm)
Gut organs Endodermal epithelium
Respiratory organs Endodermal epithelium
Kidney Ureteric bud epithelium (mesoderm)
phenomenon is regionally specific. For example, the devel-
oping epidermis signals the underlying dermis, probably
through Sonic hedgehog and transforming growth factor
β (TGF-β) proteins, and the condensed dermal mesen-
chyme responds by secreting factors that cause the epider-
mis to form regionally specific cutaneous structures. These
structures could be any of the skin appendages, such as hair,
nails, or sweat glands (see Chapter 45). Several other organs
develop from such interactions where the mesenchymal
cells take the lead in instructing different sets of genes in the
responding epithelial cells.
PA R AC R I N E FAC TO R S
Transmission of signals from the inducer to the responder
is a complex process and depends on several factors. The
interaction is juxtacrine when cell membrane proteins of
the responding cell are physically in close contact with the
cell membrane proteins of the inducing cell. In contrast,
paracrine interaction depends on the diffusion of proteins
synthesized by the inducing cell over to the cell membrane
of the responding cell. This process involves several special
kinds of protein families, collectively referred to as para-
crine factors. Essentially, these are growth and differentia-
tion factors (GDFs). Paracrine factors differ from endocrine
factors (hormones), as they do not travel through the
blood but are secreted into spaces surrounding the tar-
get cells. These proteins are inducers and are biologically
similar throughout the animal kingdom, from the fruit fly
Drosophila to humans. There are four major classes of pro-
tein families that comprise the majority of the paracrine
factors.
FI B RO B L A S T G ROW T H FAC TO R S
Several FGF genes are important for mammalian develop-
ment. The FGFs code for specific proteins, of which there
are a number of isoforms produced by alternate RNA splic-
ing or varying initiation codons in different tissues. These
7 0   •  P r i n c i p l e s o f G e n o m i c M e d i c i n e
MESODERMAL COMPONENT
Dermis (mesoderm)
Neural crest (ectodermal) mesenchyme
Mesodermal mesenchyme
Mesodermal mesenchyme
Metanephrogenic (mesodermal) mesenchyme
factors activate a set of receptor tyrosine kinases, namely the
FGF receptors (FGFRs). Mutations in some of the FGFRs
result in certain skeletal disorders. For example, mutations
in FGFR3 result in sporadic lethal thanotophoric dysplasia
and autosomal dominant achondroplasia. Receptor tyro-
sine kinases are proteins that extend from the cell surface to
the nucleus. The extracellular part binds with FGFs, and the
intracellular component activates dormant tyrosine kinase.
FGFs are associated with a number of developmental func-
tions, including angiogenesis (blood vessel formation),
mesoderm formation, and axon extension. FGF2 is particu-
larly important for angiogenesis, and FGF8 is important for
the development of the mid-brain, eyes, and limbs (Gilbert,
2004).
H E D G E H O G P ROT E I NS
The hedgehog proteins are a family of important paracrine
factors that induce particular cell types and create bound-
aries between tissues. There are at least three homologues
known for the Drosophila hedgehog gene—sonic hedgehog
(shh), desert hedgehog (dhh), and Indian hedgehog (ihh).
Desert hedgehog is expressed in the Sertoli cells of the tes-
tes, and mice homozygous for a null allele of dhh exhibit
abnormal spermatogenesis. Indian hedgehog is expressed in
the gut and cartilage, and is important for postnatal skeletal
growth.
Sonic hedgehog is perhaps the most widely used hedge-
hog protein. It is expressed in the developing notochord,
and is responsible for the patterning of the neural tube in
such a manner that the ventral neurons develop motor neu-
rons and the sensory neurons are formed from the dorsal
neurons (Yamada, Pfaff, et al., 1993). It is also responsible
for patterning the somites. Sonic hedgehog is crucial for the
formation of the left–right axis in many vertebrates. It initi-
ates the anterior–posterior axis in limbs, induces regional
specificity in the gut, and induces hair formation (Gilbert,
2004). Sonic hedgehog works in conjunction with other
paracrine factors, for example, Wnt (wingless integrated)
and FGF proteins.
 Free ebooks ==> www.Ebook777.com
WN T FA M I LY P ROT E I NS
The Wnt family proteins are cysteine-rich glycoproteins, and
comprise at least 15 members that are important for skeletal
and muscle development in vertebrates. Wnt1 appears to be
active in inducing the dorsal cells of the somites to become
muscles, while sonic hedgehog proteins are important in
patterning of the ventral portion of somites (Stern, Brown,
et al., 1995). In addition, Wnt proteins are also important in
establishing the polarity of insect and vertebrate limbs, and
are used in several steps of urogenital system development.
TG F-Β S U P E R FA M I LY P ROT E I NS
The TGF-β superfamily comprises 30 structurally related
proteins that regulate some of the most important inter-
actions in development. The important members are the
TGF-β family, the activin family, the bone morphogenic pro-
teins (BMPs), the Vg1 family, glia-derived neurotrophic fac-
tor (necessary for kidney and enteric neuron differentiation),
and Müllerian inhibitory factor (important for mammalian
sex differentiation). Some of the TGF-β proteins (TGF-β-
1, -2, -3, and -5) are important in regulating the formation
of the extracellular matrix between cells and for regulating
cell division. The members of the BMP family are named
because of their ability to induce bone formation. However,
bone formation is just one of their many functions, including
regulation of cell division, programmed cell death (apopto-
sis), cell migration, and differentiation (Hogan, 1996).
OT H E R PA R AC R I N E FAC TO R S
Apart from the above family of proteins, there are several
other paracrine factors that are important for vertebrate
development. These include epidermal growth factor, hepa-
tocyte growth factor, neurotrophins, and stem cell factor.
Several factors are exclusively associated with the develop-
ment and maturation of erythrocytes, and include erythro-
poietin, the cytokines, and the interleukins.
S I G NA L-T R A NS D U C T I O N PAT H WAY S
As described earlier, the paracrine factors include several
protein families that act as inducers. The next major class of
molecules is involved in the cellular response. These act on
the cell membrane by binding to different receptor mole-
cules and trigger a cascade of interacting proteins that trans-
mit a signal through a pathway from the bound receptor
to the nucleus. These pathways between the cell membrane
and the genome are called signal-transduction pathways.
Each receptor spans the cell membrane and consists of an
G e n e s , G e n o m e , a n d D e ve l o pm e n ta l M a l f o r m at i o n s   •   7 1
www.Ebook777.com
extracellular region, a transmembrane region, and a cyto-
plasmic region. When one of the paracrine factor protein
molecules binds to the extracellular region, it induces a
conformational change in the receptor. This is transmit-
ted across the cell membrane and changes the shape of
the cytoplasmic domains. This physical change induces an
enzymatic activity, usually one of the kinases that rely on
ATP for phosphorylation. This further triggers phosphory-
lation of other kinases and activates a dormant transcrip-
tion factor, which activates or represses a particular set of
genes. There are several signal-transduction and apoptosis
pathways (Figure  5.7, A–F). A  detailed discussion on the
molecular biology of these pathways is beyond the scope of
this chapter. The interested reader is advised to refer to one
of the major texts on developmental biology (see “Further
Reading”). However, a brief outline is summarized in
Table 5.4.
T H E RO L E O F T H E E X T R AC E L LU L A R
M AT R I X I N D EVE L O PM E N T
Cells secrete several large molecules into their immedi-
ate vicinity. These form the extracellular matrix, which is
largely noncellular material and resides in the interstices
between the cells. The extracellular matrix has an impor-
tant role in vertebrate development. A  number of cellu-
lar functions, including cell migration, cell adhesion, and
the formation of epithelial sheets and tubes, depend on
the ability of cells to form attachments with extracellular
matrices. Depending upon the requirements of specific
cells and tissues, these attachments are of variable strength
and rely largely upon the properties of the extracellular
matrix.
The extracellular matrix consists of collagen, proteo-
glycans, and several specialized glycoproteins, including
fibronectin and laminin. These large glycoproteins are
responsible for organizing the matrix and cells into an
ordered structure. Fibronectin is important in cell migra-
tion, especially the mesenchymal cells. These large mol-
ecules form a kind of “road” on which specific types of
mesenchymal cells travel to the desired organ; for example,
germ cells reach the gonads, and heart cells move to the
midline of the embryo, which is the mesodermal region for
the development of the heart.
In contrast to fibronectin, the laminin helps to keep cells
firmly in place. Laminin has a greater affinity for epithelial
cells, and fibronectin for mesenchymal cells. Laminin, along
with type IV collagen, is an important component of the
basal lamina. The basal lamina ensures adhesion of epithe-
lial cells in sheets. Like fibronectin, the laminin also plays a
 (A) (B) (C) (D) (E) (F)

Ligand TGF-β Ligand Ligand Wnt Hedgehog Bcl 2 Bik

RTK Intracellular
membrane
Receptor II Receptor Frizzled Patched Bax
GNRP
Receptor I JAK Disheveled Smoothened
RAS
Smad Stat GSK-3 Ci protein
RAF activation made activator Apaf-1
Stat β-catenin
MEK Smad dimerization Transcription
dimerization Caspase-9
Transcription
Transcription
ERK
New
transcription Apoptosis
Transcription
factor
Smad JAK-STAT WNT Hedgehog
Transcription

RTK-MAPK

Figure 5.7 
Five of the major signal-transduction pathways through which signals from the cell surface are sent into the nucleus. (A) The receptor
tyrosine kinase-mitogen-activated protein kinase (RTK-MAPK pathway); (B) the Smad pathway used by transforming growth factor β (TGF-β)
superfamily proteins; (C) the JAK-STAT pathway; (D) the Wnt pathway; (E) the Hedgehog pathway; and (F) one of the apoptosis pathways used
by mammalian neurons. Abbreviations: ERK, extracellular signal-regulated kinase; GNRP, guanine nucleotide-releasing protein; GSK, glycogen synthase kinase; JAK, Janus kinase; MAPK, mitogen-activated
protein kinase pathway; MEK, MAPK/ERK-kinases; STAT, signal transduction and activator of transcription.

role in the assembly of the extracellular matrix, promoting
cell adhesion and growth, changing cell shape, and permit-
ting cell migration.
In addition to the above-mentioned mechanical roles,
the extracellular matrix also plays a role in regulating gene
function. It helps in inducing specific gene expression in
developing tissues; for instance, liver, testis, and mammary
glands, through binding the cell substrate that is important
for specific transcription factors. The extracellular matrix
also has a role in inhibiting apoptosis through integrin,
which is the cell membrane receptor for fibronectin and
other extracellular matrix molecules. Another major cel-
lular function of the extracellular matrix is the ability to
regulate differentiation of the chondrocytes to produce the
cartilage for developing vertebrae and limbs, which is also
achieved by binding to the integrin. In the absence of inte-
grin or experimentally blocking the binding of integrin, the
developing chondrocytes fail to differentiate into cartilage
and bone (Schwab, Kasper, et al., 2000). Lastly, it has also
been shown that branching of some parenchymal organs,
such as kidney and lungs, depends on the extracellular
matrix (Kumar, 2008).

Table 5.4  THE SIGNAL-TRANSDUCTION PATHWAYS IN MAMMALIAN DEVELOPMENT

PATHWAY LIGANDS INTERMEDIARY PROTEINS

The receptor tyrosine kinase-protein FGFs, EGFs, PDGFs, STF GTPase-activating (GAP)
mitogen-activated protein kinase pathway
(RTK-MAPK)
The Smad pathway TGF-β superfamily Sma and Mad proteins (1–5)
The JAK-STAT pathway FGFRs JAK-STAT tyrosine kinases
The Wnt-β-Catenin pathway Frizzled proteins Disheveled, GSK-3, β-catenin
The Hedgehog pathway Patched proteins Smoothened, Ci protein
The Notch pathway Delta, Jagged, Serrate CBF-1, Suppressor of Hairless, Lag-1
Abbreviations: Ci, Cubitus interruptus; EGFs, epidermal growth factors; FGFs, fibroblast growth factors; JAK, Janus-activating kinase; PDGFs,
platelet-derived growth factors; STAT, signal transducers and activators of transcription; STF, stem cell trigger factor.

7 2   •  P r i n c i p l e s o f G e n o m i c M e d i c i n e
 A P O P TO S I S A N D D EVE L O PM E N T It is now confirmed that CED-3 and CED-4 proteins
act at the center of the apoptosis pathway. These regulate
Apoptosis, or programmed cell death, is a normal part of
initiation of other genes in the pathway, such as BMP-4.
development. Cells in all animals are programmed to die
Homologues for these proteins are important for auto-cell
every day, and approximately equal numbers of cells are
digestion. The CED-4 protein homologue is called Apaf-1
replaced. For example, adult humans lose as many as 1011
(apoptotic protease activating factor-1), which participates
cells every day, and these are regularly replaced by other
in the cytochrome-dependent activation of the mamma-
cells. It is estimated that the total weight lost every year
lian CED-3 homologues caspase-9 and caspase-3 ( Joza,
through programmed cell death could be equal to the adult
Susin, et al., 2001). Activation of these caspases causes cell
body weight. Apoptosis begins immediately after birth. It
auto-digestion, leading to cell death. Mice homozygous
is estimated that the total number of neurons accumulated
for Apaf-1 deletions have severe craniofacial abnormalities,
throughout the gestation period of nine months is approxi-
brain overgrowth, and syndactyly (webbing between toes).
mately three times that in an adult of average intelligence.
It is important to appreciate that apoptosis can follow
Programmed cell death is a continuous process. It is
more than one pathway. For instance, the “death domain,”
essential for creating proper spaces within an organ, as
containing receptors of the tumor necrosis factor (TNF)
well as between organs or body parts. Examples include
family, can induce apoptosis in several cell systems that can
the middle-ear space, the separation of digits to create the
also be triggered by other apoptosis-inducing factors. This is
proper shape and size of fingers and toes, and the lower
accomplished by blocking the anti-apoptosis signals sent by
vaginal space and opening (Newton and Strasser, 1998).
other factors. One of the developmentally important TNF
Clearly, through apoptosis, redundant tissues and struc-
receptors with a death domain is Edar, a protein required
tures are pruned away. Different tissues use varying signals
for the development of hair, teeth, and other cutaneous
for apoptosis. Among vertebrates, the BMP4-mediated
appendages. Mutations of this gene or its ligand, Eda, cause
signals are important. For example, the connective tissues
X-linked hypohidrotic ectodermal hypoplasia (Online
respond to BMP4 to differentiate into bone. Similarly, the
Mendelian Inheritance in Man [OMIM] 305100), a syn-
surface ectoderm responds to this by differentiating into
drome characterized by lack of sweat glands, sparse hairs,
skin. Another good example is the development of tooth
and poorly formed teeth. An identical syndrome results
enamel. After the tooth cusp has grown, the enamel knot
from deficiency of the adapter protein that binds the death
synthesizes BMP4, which, through apoptosis, stops further
domain of this receptor (Headon, Emmal, et  al., 2001).
enamel differentiation (Vaahtokari, Aberg, et al., 1996). As
Instead of resulting in cell death, the activation of the recep-
previously described, the erythropoietin-induced red blood
tor enables continued development of skin appendages.
cell population is programmed for apoptosis. In its absence,
the red cells will undergo apoptosis. This works through
I N FLU E N C E O F E N VI RO N M E N TA L
the JAK-STAT signal-transduction pathway (see previous
A N D O P P O RT U N I S T I C FAC TO R S O N
section).
D EVE L O PM E N T
Apoptosis works through several pathways. One of the
pathways is regulated by genes that were discovered from Developmental biologists agree that the environment plays
studies on C. elegans, appropriately designated as ced-3 and a significant part in producing a phenotype. Nutritional fac-
ced-4 genes. The gene product of these two genes initiates tors undoubtedly result in a number of disease phenotypes,
apoptosis. However, the protein product of another gene such as marasmus, kwashiorkor, rickets, diabetes mellitus,
(ced-9) is shown to inhibit the programmed cell death. and coronary heart disease. There are several genetic factors
Mutations in this gene will accentuate apoptosis by with- that are equally important for creating a pathophysiological
drawing the control. This has been confirmed experimen- state, which predispose to morbid effects of dietary factors
tally when inactivated CED-9 protein led to the death of (see Chapter 12). Dietary supplementation and modifica-
an entire embryo. On the contrary, gain-of-function ced-9 tion are known to significantly alter the phenotype. For
mutations can help cells survive that would have other- example, a normal daily dietary intake of vitamin C pre-
wise died. In other words, wild ced-9 acts as a binary switch vents the development of the clinical effects of vitamin C
between life and death at the cellular level. In mammals, deficiency, as human beings lack naturally occurring vita-
members of the Bcl-2 gene family are the CED-9 protein min C (hypoascorbemia, OMIM 240400) due to deficient
homologues. This gene family is important for red blood gulonolactone oxidase as a result of a mutation in the gulo-
cell development and differentiation. nolactone oxidase gene on the short arm of chromosome

G e n e s , G e n o m e , a n d D e ve l o pm e n ta l M a l f o r m at i o n s   •   7 3
8.  This can result in severe childhood connective tissue
disease leading to death. Gulonic acid oxidase enzyme is
the final enzyme leading to the synthesis of ascorbic acid
(vitamin C). In contrast to humans, several other mammals
have normal gulonolactone oxidase enzyme activity offer-
ing natural protection from the clinical effects of vitamin
C deficiency.
Periconceptional folic acid supplementation is now
commonplace for the prevention of recurrent neural tube
defects, and probably even for the primary prevention
of some other congenital anomalies. Fetuses with muta-
tions in genes associated with folate metabolism are at an
increased risk for neural tube defects (De Marco, Calevo,
et al., 2003). One such gene is methylene tetrahydrofolate
reductase (MTHFR), which incorporates folic acid in the
methylation of homocysteine to methionine. Mutations or
polymorphisms in this gene result in increased homocyste-
ine levels, resulting in peripheral vascular disease, and are
associated with myocardial infarction. However, the mech-
anisms by which folate deficiency or lack of bio-availability
result in neural tube defect are not fully understood.
Several developmental malformation syndromes are
associated with defects in cholesterol biosynthesis. One
of the enzymes in this pathway is 7-dehydoxycholesterol
reductase. Mutations in the gene for this enzyme (7DHCR)
result in lack of downregulation of Sonic hedgehog (SHh),
resulting in a number of abnormal phenotypes such as
Smith-Lemli-Opitz syndrome (Figure 5.8). It is possible to
ameliorate some of the deleterious effects of this downregu-
lation by dietary cholesterol supplementation. Similarly,
dietary restriction of the excess metabolite in a number of
inherited metabolic conditions can alter the phenotype. An
excellent example is that of phenylketonuria, in which the
behavioral and cognitive effects of excessive accumulation
Smith–Lemli–Opitz (DHCR7)
107
152
7-DHCR
76
Pallister–Hall (GLI3)
SHH
DHCR7 CR
Figure 5.8  The gene pathway of 7-dehydroxy cholesterol reductase, Sonic Hedgehog, and GLI3 complex (Brunner and van Driel, 2004).
7 4   •  P r i n c i p l e s o f G e n o m i c M e d i c i n e
of phenylalanine can be significantly reduced by dietary
restriction of phenylalanine.
The genomes of primate mammals determine their
final physical shape, which is also under the direct influ-
ence of the environment. For example, the facial phenotype
depends on firm and regular chewing, which stimulates the
facial muscle and bone (maxilla and mandible) develop-
ment (Corruccini, 1990). The increased prevalence of orth-
odontic problems in young children and adults in modern
times is attributed to a soft or mid-textured diet. This has
been shown in experimental primates who were fed a soft
diet and developed lower jaw malocclusion similar to that
in children requiring orthodontic treatment (Corruccini,
Whitley, et al., 1985).
In brief, the production of a phenotype such as a devel-
opmental malformation depends on the genotype, which is
regulated at numerous levels. The cellular phenotype is the
direct consequence of the genome within the cell and the
fate of the community of cells in which it resides. It is also
argued that probably even the environment can alter gene
expression (Gilbert and Epel, 2009)!
MO D E L O RG A N I S M S F O R
U N D E R S TA N D I N G D EVE L O PM E N T
For several years, geneticists have used a number of model
organisms to dissect how genes control metabolism, repro-
duction, and development. Similarly, using the same model
organisms, biochemists and molecular biologists have dis-
covered mechanisms through which the protein products
of these genes regulate biological processes. These model
organisms include different microbes, the budding yeast
(S.  cerevisiae), the worm C.  elegans (nematode), the plant
Arabidopsis thaliana, and the fruit fly D. melanogaster. For
Cytoplasm
GLI complex
Nucleus
PTCH GLI acivator
SMO
humans, the mouse model, Mus musculus, is ideal, since
the two species are closely related through evolution. With
the rapid progress in DNA sequencing, sequences of the
best-understood genes were deposited into sequence data-
bases such as GeneBank, maintained at the National Center
of Biotechnology Information (NCBI; http://www.ncbi.
nlm.nih.gov). This information is used to predict an mRNA
sequence, which can be readily used to predict the amino
acid sequence of the corresponding protein. These pro-
tein sequences are known as the sequence signature for that
particular protein in other organisms. Thus, homologous
genes in other organisms can be found through searching
for nucleic acid sequences that, when translated, produce a
similar protein sequence. This newly identified gene is then
predicted to have similar biological function in the second
organism.
Complete genome sequences of model organisms and
the human genome sequence are available on GeneBank,
which is now used in searching for new genes and the cor-
responding proteins. A  newly identified gene sequence
can be compared to the existing database of sequences
by aligning the predicted protein sequences. If a gene of
unknown function is similar to another of a known func-
tion, the newly discovered gene can be predicted to have
similar function. However, not all gene functions can be
predicted in this manner. In some situations, new genes
of unknown function are found in two organisms. This
problem can be alleviated by building a protein domain
knowledge base that can be searched for amino acid sig-
natures indicative of structure or biochemical activity.
If present, these signatures provide a clue to the gene
function. If two predicted protein sequences from two
genomes are similar, then their amino acid alignments
would be the same. In such a situation, the genes can be
predicted to have similar functions.
Biological makeup of the model organism determines
the extent of information applicable to human develop-
ment. Unicellular organisms are an excellent model for
studying eukaryotic cell function. For example, genetic
studies in the yeast (S.  cerevisiae) and slime molds
(Dictyostelium discoideum) can be carried out and repeated
several times, as a billion progeny can be produced in a
relatively short time. These studies provide vital informa-
tion about intragenic primary and secondary suppressor
loci. Therefore, unicellular organisms are immensely use-
ful in establishing the networks of gene action involved in
basic cell biological processes. However, such studies have
limited applications in studying complex cellular func-
tions such as those of the nervous system, which depend
on the interaction between cells. Nevertheless, studies on
G e n e s , G e n o m e , a n d D e ve l o pm e n ta l M a l f o r m at i o n s   •   7 5
unicellular organisms have helped us in understanding
enzymes that are involved in complex energy metabolism
defects resulting in neuromuscular disorders (Berardo,
DiMauro, et al., 2010).
Studies on invertebrate model organisms, for example,
the Drosophila fruit fly and the nematode (C.  elegans),
have contributed to our understanding of several basic
biological mechanisms, such as the organization of genes
into independently segregating linear chromosomes, the
creation of the first chromosome “maps,” the one gene–
one protein hypothesis, radiation-induced mutagenesis,
the principles of pattern formation, and the identification
of genetic pathways implicated in human disease. As both
fruit flies and nematodes have closely related gene coun-
terparts to many human disease genes, the identification
of new genes in these invertebrates will help define new
candidate disease genes that are likely to be involved in
the same disease processes. A useful example is the Notch
signaling pathway that has been shown to be important
in invertebrate development. Mutations in the compo-
nent genes of the Notch signaling pathway have been
shown to result in notching of the wing margin in fruit
flies and defects in vulval development in worms (Gupta,
Wang, et  al., 2003). Notched wings were also observed
in mutations in the ligand Delta, the Notch receptor, or
the signal transducer suppressor (Hairless). Vertebrates
have several common paralogues of the Notch signaling
pathway components. Reduced function of the related
ligand, Delta3, or the Notch homologue (Notch1) itself
results in axial skeletal malformations. A good example is
spondylocostal dysostosis, an autosomal recessive heter-
ogenous condition (OMIM 277300, 608681, 609813).
The characteristic feature of this disease is the spinal seg-
mental anomaly, which is also associated with anal and
urogenital anomalies (OMIM 271520).
One of the major contributions of studies in inverte-
brate model organisms has been the discovery of homeo-
tic selector genes, now referred to as Hox genes. The term
homeostasis was coined by William Bateson (1894) for
the phenomenon in which one segment of an organism is
“transformed in whole or part to another” (Reid, 2004).
The genetic basis of these transformations is explained by
mutations in Hox genes. Systematic analyses of Hox gene
mutations in the Drosophila fruit fly revealed an extra
pair of wings due to mutations in the Ultrabithorax (Ubx)
gene and an extra thoracic leg attached to the head result-
ing from dominant mutations in the Antennapedia (Antp)
gene (Gehring, Kloter, et al., 2009). Molecular analysis of
the genomes has revealed that humans and other bilateral
animals have multiple Hox genes (Figure 5.1), which carry
a common DNA sequence motif called the homeobox. The
homeobox motif encodes a similar 60-amino acid motif in
Hox proteins, termed the homeodomain. The Hox proteins
belong to the transcription factor family (Table 5.2). These
exert their function through activation and repression of
multiple target genes. Arrangement of these genes is strik-
ingly similar in the fruit fly and humans. These genes are
arranged in clusters. There is evidence that this clustered
arrangement of Hox genes has been maintained for more
than 500 million years, because different genes in the clus-
ters are controlled by the same cis-acting DNA regulatory
regions. In general, there are four clusters—HOXA, HOXB,
HOXC, and HOXD. Each gene has a role in the ante-
rior–posterior axis patterning of various organs and body
parts. This is evident as specific human malformation syn-
dromes are now recognized to be associated with HOXA
(OMIM 609296; 601536), HOXB (OMIM 249000,
Meckel syndrome, MKS1), and HOXD (OMIM 606708,
split hand-foot-absent uterus syndrome; OMIM 127300,
Leri-Well dyschondrosteosis; OMIM 113200, brachydac-
tyly type D; OMIM 112500, brachydactyly type A1). In
addition to these malformations, the HOX genes interact
with several other patterning genes with a crucial role in
development.
MO L ECU L A R BA S I S O F
M A L F O R M AT I O N SY N D RO M E S
Developmental biologists and dysmorphologists have been
intrigued for a considerable time about the molecular basis
of phenotypical variability in malformation syndromes
(Katsanis, Lupski, et al., 2001; Liao, Kochilas, et al., 2004).
Among their questions have been:
(1) Why are mutations at a single genetic locus associated
with a multitude of phenotypical features (pleiotropy)?
(2) What is the mechanism for the same phenotypes
being caused by mutations in several different genes
(genetic heterogeneity)?
(3) What is the mechanism of a dominant phenotype?
P L E I OT RO P Y
Observations of the expression patterns of transcription
factors and paracrine factors have revealed mechanisms
that lead to different malformations caused by mutations
at a single locus. This is called pleiotropy. There are several
examples of this phenomenon. For instance, it is known
that heterozygosity for MITF (in humans and mice) causes
7 6   •  P r i n c i p l e s o f G e n o m i c M e d i c i n e
multiple malformations that include iris defects, pigmenta-
tion abnormalities, deafness, and the inability to produce a
normal number of mast cells. Moreover, these abnormali-
ties are not related to each other and can occur indepen-
dently. This occurs because all body parts can use the MITF
protein as a transcription factor. This type of pleiotropy is
called mosaic pleiotropy, as the relevant organ or body part
is separately affected by the mutant gene. In contrast, some
malformations in the related part do not result directly
from the abnormal gene function, as the mutant protein
is not expressed. For example, the failure of MITF expres-
sion results in the pigmented retina’s not being fully differ-
entiated. This in turn causes a malformation involving the
choroidal tissue, which results in drainage of the vitreous
humor fluid. This further leads to failure of ocular develop-
ment, causing microphthalmia (small eye). This phenom-
enon, in which several developing tissues or organs might
be sequentially affected even though they do not express
the mutant gene, is called relational pleiotropy. This concept
is important in dealing with complex clinical genetic situ-
ations, particularly in prenatal diagnosis where the predic-
tion of phenotype is important in making informed choices.
G E N ET I C H ET E RO G E N E IT Y
An important aspect of dysmorphology is the recognition
of the phenotype and the search for a syndrome of which
that particular phenotype is known to be a major compo-
nent. However, the search is made difficult by the fact that
many syndromes can feature the same phenotype. At the
molecular level, this refers to mutations in different genes
being responsible for these syndromes. This is possible if the
genes are part of the same signal-transduction pathway. This
phenomenon is referred to as genetic heterogeneity. There
are several examples. Genetic heterogeneity could be either
within the gene (allelic heterogeneity) or due to genes at dif-
ferent loci (locus heterogeneity). A good example is cyclopia,
which is caused by both mutations in the sonic hedgehog
gene or the genes regulating cholesterol biosynthesis. Since
they are in the same pathway, mutations in one gene gener-
ate a phenotype similar or identical to those generated by
mutations in other genes. Another good example is hypo-
hidrotic ectodermal dysplasia, which can result from muta-
tions in EDA (receptor), EDAR (ligand), or EDARADD
(adapter protein) (Chassaing, Bourthoumieu, et al., 2006).
M EC H A N I S M S O F D O M I NA N C E
Molecular analysis can help in delineating whether a partic-
ular syndrome is dominant or recessive. The true dominant
phenotype only occurs in the heterozygous state. In other
words, the homozygous state never exists, probably because
it is lethal to the embryo. There are several possibilities that
can result in a dominant phenotype (Wilkie, 1997).
Haploinsufficiency
Haploinsufficiency refers to the situation when one copy of
the gene (haploid) is not enough to produce the normal
amount of wild protein required for normal development.
In other words, an abnormal phenotype can result if one
of the two copies of the gene is either absent or nonfunc-
tional. For example, in type 2 Waardenburg’s syndrome, the
variable phenotype is associated with only about half the
amount of wild-type MITF protein being present. This is
not enough for full pigment cell proliferation, mast cell dif-
ferentiation, or inner ear development, which manifested
clinically in the variable phenotype of this autosomal domi-
nant condition. The phenotype in several micro-deletion
multiple malformation syndromes is also due to haploinsuf-
ficiency when one or more genes are lost that would have
been within the deleted chromosomal band. Thus it is likely
that mutations in some of these genes could behave in a
dominant manner.
Gain-of-Function Mutations
An abnormal phenotype can also result from mutations in a
gene causing a gain in additional functions, or acquiring new
function. A good example is the FGFR gene, wherein differ-
ent mutations can result in a constitutively active gene prod-
uct, which can result in the potentially lethal thanatophoric
dysplasia as well as the milder form of related skeletal dys-
plasia (achondroplasia). Other examples of gain-in-function
type dominant phenotypes include late-onset Huntington’s
disease and other neurodegenerative disorders associated
with expanded triplet-repeats (see Chapter 6).
Dominant-Negative Allele
Another mechanism of dominance is a dominant-negative
allele. This can occur when the active protein product is a
multitimer. All constituent proteins of the multitimer have
to be wild-type in order for the multitimer to be active.
Mutations in the gene for any of the member proteins can
render that unit inactive, resulting in a nonfunctional or rel-
atively ineffective multitimer. A good example is autosomal
dominant Marfan syndrome, whose variable phenotype
is due to mutations in the fibrillin-1 gene. The wild-type
product of this gene is a glycoprotein that forms multitimer
G e n e s , G e n o m e , a n d D e ve l o pm e n ta l M a l f o r m at i o n s   •   7 7
microfibrils in elastic connective tissue. The presence of
even minute amounts of mutant fibrillin prohibits the asso-
ciation of wild-type fibrillin into microfibrils (Watt and
Chung, 2009).
Allelic Interactions
Another mechanism of dominance is related to qualitative
differences in the product made from the interactions of
different alleles. Such interactions can result in a superior
protein dimer made from two alleles compared to an infe-
rior or less active product made by one allele alone (Trehan
and Gill, 2002).
F U NC T IONAL G ENOMICS IN
SY NDROME -FAMILIES
Delineation of a dysmorphic syndrome is complex, and
it involves continuous refining of the phenotype by the
identification of new cases and revisiting those previously
described. There are several syndromes that are apparently
identical or nearly identical to the original description, and
this might create a false impression of homogeneity for that
syndrome (Cohen Jr, 1989). However, this is often not true,
as close inspection of the phenotype can reveal distinctive
features. Reports on patients with partly overlapping phe-
notypes lead to frequent debates between “lumpers and
splitters” (McKusick, 1969). Apart from a few quantitative
analyses, syndrome definition is largely a matter of compar-
ing the phenotype of a suspected patient with the “best-fit”
case (Verloes, 1995). Identification of pathogenic muta-
tions in a gene might help in defining the core phenotype
and its variants. However, resolving the molecular genetics
does not always solve the problem of understanding the
phenotypical variability, because allelic heterogeneity and
the action of modifier genes can influence the phenotype
(Nadeau, 2001). This is further complicated as mutations in
different genes can manifest with the same or related pheno-
types. Thus, a strict molecular classification of syndromes is
clinically not relevant, as it would obscure the relationship
between molecular pathology and the phenotype (Adams,
Smith, et al., 2008).
It has been recognized for some time that several syn-
dromes share overlapping dysmorphic features. It is argued
that these syndromes might have a common biological
relationship, and may be referred to as “syndrome com-
munity,” “phenotype communities,” and “syndrome fam-
ily.” The concept of a “syndrome family” is based on the
observations that several dysmorphic syndromes have
common phenotypical features (Oti and Brunner, 2007).
The syndrome-family approach was first systematically
applied to skeletal dysplasias; for example, the family of
chondrodysplasias includes several distinct skeletal dyspla-
sias such as achondrodysplasia, hypochondrodysplasia, and
so on (Spranger, Winterpacht, et al., 1994). The advances in
molecular genetics have vindicated this concept. For exam-
ple, mutations in FGFR3 result in three distinct members
of the chondrodysplasia family. The converse is true for the
phenotype of the Stickler-Kniest family, which is related
to mutations in three different collagen genes (COL2A1,
COL11A2, and COL11A1) (Reginato and Olsen, 2002).
Thus, rapid advances in molecular genetics and genomics
could see merging of syndromes, more splitting of syn-
drome families, and even the complete disappearance of
syndromes. However, the description and definition of
syndromes and syndrome families is important. Several
genetic databases (McKusick’s OMIM, the LDDB, and
POSSUM) continue to record these syndromes as inde-
pendent entities.
There are approximately 200 Mendelian syndromes in
man, which appear in OMIM and other databases. Each
syndrome is recognized by a specific phenotypical feature
or pattern. Some syndromes differ only by a few features.
It is acceptable logic that there could be some biological
relationship in syndrome families that share the same phe-
notype. It is successfully argued that a systematic analysis
of phenotypical relationships could be applied in the iden-
tification of new genes, providing clues to gene interactions,
molecular pathways, and functions (Brunner and van Driel,
2004). This has been applied in large-scale mutagenesis pro-
grams that aim to define the function of genes in a genome
in relation to mutant phenotypes. This has been completed
in yeast, and work is underway for other model systems
such as C. elegans, the mouse, and the zebrafish. Although
this strategy has been successful, it is not clear how many
different mutants will be required for such screens to be
comprehensive. It is accepted that creating a single knock-
out mouse model might not be sufficient to probe a specific
gene function in development and homeostasis. A compre-
hensive analysis of the mutant phenotype would require
studying the functional effects of several mutations. Starting
from interesting phenotypical differences and then compar-
ing the underlying mutations might be more productive
(Brunner and van Driel, 2004). Spontaneous mutations are
frequent, and studying the phenotypical effect can contrib-
ute to our understanding of gene function. Thus, it is impor-
tant to adopt a “phenotype-driven” approach that saturates
the genome with mutations, either experimentally shown in
animal models or observed in human disease states.
7 8   •  P r i n c i p l e s o f G e n o m i c M e d i c i n e
Phenotypical similarities in different syndromes
belonging to a so-called syndrome family could be a reli-
able indicator of shared biological mechanisms. Apparently,
single-gene human genetic disease could in fact be associated
with mutations in different genes, probably contributing to
a common molecular pathway. This was first demonstrated
in familial elliptocytosis, because linkage to the Rhesus (Rh)
blood group was not seen in all families (Morton, 1956).
It is now clear that nonallelic heterogeneity is extensive in
human disease. Although genetic heterogeneity could be
problematic in conducting familial genetic studies, this can
be viewed positively, as it might reflect interactions at the
protein level: for example, ligand–receptor interactions, dif-
ferent subunits of a multiprotein complex, or proteins that
function at different steps of a metabolic pathway. Using
this strategy, the other genes could be found once the first
gene is discovered. It is also possible that unrelated genes
that result in the same phenotype could also be found and
will ultimately be shown to have a functional relationship.
Clinical classification is paramount, followed by molecular
verification. In other words, defining the phenotype and
syndrome identification could become a functional genom-
ics tool (Brunner and van Driel, 2004).
T R A N S L AT I O NA L R E S E A RC H
I N DYS MO R P H O L O GY
Clinical studies on human multiple malformation syn-
dromes are not only helpful in medical management but
are also a useful resource for research in understanding the
basic mechanism of human development, and, eventually,
mammalian development in general. This approach allows
researchers to gain insight into basic mechanisms of devel-
opment and how genes program organisms to achieve per-
manent or adult morphological shapes (Martínez‐Frías,
2004). A  number of malformation syndromes have over-
lapping manifestations, despite being phenotypically
and genetically dissimilar. This fact can be used in basic
research; for example, on developing animal model systems,
such as fruit flies, mice, worms, and other simple organisms.
The data thus generated can be applied in the clinical set-
ting to understand both the problems patients suffer and
the mechanisms of development. This is referred to as trans-
lational research. However, there are limitations to using
human subjects in translational research. In humans, it is
obviously not possible or ethically appropriate to experi-
mentally manipulate genes to test hypotheses. In addition,
there is inevitable difficulty in obtaining tissues or organs
for study due to the lack of consent and/or availability. The
final limitation or disadvantage in carrying out studies on
humans is the enormous cost and length of time it can take
to complete the study. Moreover, as a genetic model system,
the human has a long generation time; hence, individuals
and families must be followed over long periods of time.
The typical example of human translational research is
to creatively apply a basic laboratory discovery to the clini-
cal management of a patient with the clinical phenotype of
a particular dysmorphic syndrome. This approach would
allow testing of the hypothesis as well as providing an
opportunity to validate the basic laboratory finding. Any
further improvement or information can be fed back to the
patient, and the system thus works bidirectionally. Such
a system would allow extracting the maximum amount
of basic information from the patient. This could then be
applied back in the clinical setting to explain the cause
and the likely outcome of the diagnosis. Thus, researchers
can develop novel hypotheses about the mechanisms of
development, modifying variables, or other disorders with
overlapping manifestations. The aim is to keep the cycle in
motion, continually collecting and examining new infor-
mation, and constantly applying it back in improved clini-
cal care.
The above concept has been applied in a number of
multiple malformation syndromes with overlapping phe-
notypical manifestations, or families of syndromes (see
previous section). One example is the Pallister-Hall syn-
drome (PHS), a typical syndrome family disorder in which
affected individuals can present with one or more of a range
of malformations that include hypothalamic hamartoma,
imperforate anus, laryngeal anomalies, and central polydac-
tyly, with shortened terminal digits (Biesecker and Graham,
1996). The disorder is inherited in an autosomal dominant
manner with significant inter- and intra-familial variabil-
ity. Clinical and genetic studies in families affected with
PHS revealed mutations in the zinc finger transcription
factor gene, GLI3 (Kang, Graham, et al., 1997). This was
an interesting finding, as previously this gene was inciden-
tally found to be causally related to the Greig cephalopoly-
syndactyly syndrome (GCPS) (Vortkamp, Gessler, et  al.,
1991). Patients with GCPS had balanced chromosomal
rearrangements that had apparently disrupted the GLI3
gene. It is now accepted that GCPS is a distinct develop-
mental syndrome caused by haploinsufficiency of GLI3.
This was a challenging observation, as etiologically
significant mutations in the same gene were found in two
distinct genetic syndromes. The research group led by
Dr.  Leslie Biesecker at the National Institutes of Health
(NIH) discovered that mutations in Cubitus interruptus
or ci, the Drosophila homologue of GLI3, were linked to a
wide range of phenotypes in fruit flies, including abnormal
G e n e s , G e n o m e , a n d D e ve l o pm e n ta l M a l f o r m at i o n s   •   7 9
wing-vein patterning and lethal malformations manifesting
with larval death (Kinzler, Ruppert, et al., 1988). It trans-
pired that ci belonged to the genetic pathway downstream
of the hedgehog signaling molecule (Hh). It was shown that
the ci protein negatively controlled downstream genes, and
the cleaved ci protein turned off the expression of down-
stream genes.
When the same information was applied to humans,
it soon became apparent that the truncated GLI3 protein
had a different clinical outcome (PHS) compared to that
resulting from haploinsufficiency of the gene (GCPS). In
other words, PHS phenotype was due to the qualitative
change, and the quantitative change led to the GCPS phe-
notype. It is now shown that GLI3 protein, like ci protein,
is proteolytically processed (Liu, Wang, et al., 2005). This
example illustrates that molecular studies in rare devel-
opmental disorders like PHS and GCPS can illustrate
basic mechanisms of mammalian development. The phe-
notype related to GLI3 mutations also includes polydac-
tyly, imperforate anus, and vertebral anomalies (OMIM
174100).
The genetic pathway involving GLI3 also includes
the DCHR7 gene (Figure  5.6), mutations in which cause
Smith-Lemli-Opitz syndrome (OMIM 270400), one
of the malformation syndromes related to disruption in
cholesterol biosynthesis. The DCHR7 gene codes for an
enzyme called 7-dehydroxycholesterol reductase, which reg-
ulates sonic hedgehog (SHH) gene function. One of the
downstream effectors of SHH is GLI3. Other SHH protein
homologues include patched protein homologue (PTCH)
and smoothed homologue precursor (SMO), both of
which play key roles in regulating GLI3 gene function. It
is likely that the phenotypical similarity in other malforma-
tion syndromes like Optiz G (OMIM 145410) and Mohr
syndrome (OMIM 252100) could be due to mutations in
other genes in this gene family (Brunner and van Driel,
2004).
Another illustrative example is McKusick-Kauffman
syndrome (MKS), which was first described among the Old
Order Amish of Lancaster County, Pennsylvania (Biesecker,
2002). This disorder is inherited in an autosomal recessive
manner, and the phenotype includes polydactyly (central
and post-axial), congenital heart disease, and hydrome-
trocolpos due to congenital uterine outflow obstruction.
This disorder is rare, with fewer than 100 cases described
in the literature. It is likely that the disorder, or a pheno-
typically similar disorder, probably occurs in other inbred
population groups. Increased incidence of autosomal reces-
sive disorders, including multiple malformation syndromes,
is recognized among the Amish and other highly inbred
ethnic population groups. These populations groups are
ideal for conducting homozygosity mapping studies.
Using the whole-genome-wide scan with 385 markers,
the gene for MKS was mapped to chromosome 20. Further
molecular studies identified two substitution mutations in
a single mutant chromosome in one of the candidate genes.
However, this was not associated with a known function.
Since the Amish are closely related, it was not possible to
prove that the sequence variants were pathogenic. A search
for non-Amish cases was then made. This proved difficult,
as the reported cases were either deceased or no longer avail-
able. In some cases the diagnosis was changed. Eventually,
a newborn girl with features of MKS was recruited to the
study, and was found to have a 2-bp deletion on the same
allele in the same gene that was identified in the Amish
(Stone, Slavotinek, et al., 2000). This confirmed that muta-
tions in the novel Amish gene (MKKS) caused MKS.
Since that time, many patients who were originally diag-
nosed with MKS have developed other clinical features and
have been diagnosed with Bardet-Biedl syndrome (BBS)
(Slavotinek, Stone, et al., 2000). Affected individuals with
BBS have post-axial polydactyly, mental retardation, pro-
gressive pigmentary retinopathy, and obesity complicated
with diabetes mellitus. It soon became obvious that both
MKS and BBS share common phenotypical features. BBS
is caused by mutations in different genes. It is an illustrative
example of a digenic or multigenic disorder, wherein a com-
bination of mutations in different genes results in the BBS
phenotype. Some patients who were not known to have
mutations in one of the BBS genes have now been reported
to have mutations in the MKKS gene (Slavotinek, Searby,
et al., 2002; Katsanis, 2004). These mutations are different
from that found in the Amish. These findings show that it
is important to review all patients with a clinical diagno-
sis of BBS or MKS and apply the molecular techniques to
establish diagnosis. This is important in clinical care and
for offering accurate genetic counseling to the family. This
is another example where the basic scientific data could be
taken back to the clinical setting for the benefit of patients,
parents, and other family members.
CONCLU SION
In conclusion, understanding normal human develop-
ment is a prerequisite for analyzing the pathophysiology of
developmental malformations occurring either singly or as
part of a multiple-malformation syndrome. Development
in humans is strikingly similar to that in small creatures,
as well as in several other mammals, including our fellow
8 0   •  P r i n c i p l e s o f G e n o m i c M e d i c i n e
primates. The mechanisms are similar by which the individ-
ual genome specifies the physical and functional state of the
human body. This chapter, in brief, summarizes a number
of gene families and numerous related protein families that
regulate the complex process of development. The availabil-
ity of human and mouse genomes and that of several other
small organisms and animals have provided developmental
biologists with powerful new tools for identifying genes,
and their mutations, that control development. The inter-
ested reader is urged to explore other literature in this com-
plex and stimulating field.
F U RT H ER READIN G
Epstein CJ, Erickson RP, Wynshaw-Boris A (eds.) (2004). Inborn Errors
of Development. New York: Oxford University Press.
Jones K (2006). Smith’s Recognizable Patterns of Human Malformations
(5th ed.). New York: Elsevier.
Stevenson R, Hall JH (eds.) (2006). Human Malformations and Related
Anomalies (2nd ed.). New York: Oxford University Press.
REFERENCES
Adams M, Smith UM, et al. (2008). Recent advances in the molecular
pathology, cell biology and genetics of ciliopathies. Journal of medi-
cal genetics. 45(5):257–267.
Avidor-Reiss T, Maer AM, et al. (2004). Decoding cilia function: defin-
ing specialized genes required for compartmentalized cilia biogen-
esis. Cell. 117(4):527–539.
Beldade P, Brakefield PM (2002). The genetics and evo-devo of butterfly
wing patterns. Nature reviews genetics. 3(6):442–452.
Berardo A, DiMauro S, et al. (2010). A diagnostic algorithm for meta-
bolic myopathies. Current neurology and neuroscience reports.
10(2):118–126.
Biesecker LG (2002). Coupling genomics and human genetics to
delineate basic mechanisms of development. Genetics in Medicine.
4:39S–42S.
Biesecker LG, Graham J (1996). Pallister-Hall syndrome. Journal of med-
ical genetics. 33(7):585–589.
Brunner HG, van Driel MA (2004). From syndrome families to func-
tional genomics. Nature reviews genetics. 5(7):545–551.
Cañestro C (2012). Two rounds of whole-genome duplication: evidence
and impact on the evolution of vertebrate innovations. In Polyploidy
and Genome Evolution. Springer: 309–339.
Cañestro C, Yokoi H, et al. (2007). Evolutionary developmental biology
and genomics. Nature reviews genetics. 8(12):932–942.
Chassaing N, Bourthoumieu S, et  al. (2006). Mutations in EDAR
account for one-quarter of non-ED1-related hypohidrotic ectoder-
mal dysplasia. Human Mutation. 27(3):255–259.
Cohen Jr M (1989). Syndromology:  an updated conceptual overview.
III. Syndrome delineation. International journal of oral and maxil-
lofacial surgery. 18(5):281–285.
Corruccini RS (1990). Australian aboriginal tooth succession, interprox-
imal attrition, and Begg’s theory. American Journal of Orthodontics
and Dentofacial Orthopedics. 97(4):349–357.
Corruccini, RS, Whitley LD, et al. (1985). Facial height and breadth
relative to dietary consistency and oral breathing in two populations
(North India and US). Human Biology. 57(2):151–161.
Cvekl A, Tamm ER (2004). Anterior eye development and ocular mes-
enchyme:  new insights from mouse models and human diseases.
Bioessays. 26(4):374–386.
David, WM (2004). 4′Consequences of the Genome Project for under-
standing development. The Molecular Basis of Clinical Disorders of
Morphogenesis. In ‘Inborn Errors of Development;’ Eds. Epstein,
Erickson et al. Oxford University Press, NY, 2004.
Davidson, EH, McClay DR, et  al. (2003). Regulatory gene networks
and the properties of the developmental process. Proceedings of the
National Academy of Sciences. 100(4):1475–1480.
De Marco P, Calevo MG, et  al. (2003). Reduced folate carrier poly-
morphism (80A→ G) and neural tube defects. European journal of
human genetics. 11(3):245–252.
Epstein, CJ, Erickson RP, et al. (2004). Inborn Errors of Development: The
Molecular Basis of Clinical Disorders of Morphogenesis.
New York: Oxford University Press.
Gehring WJ (2002). The genetic control of eye development and its
implications for the evolution of the various eye-types. International
Journal of Developmental Biology. 46(1):65–74.
Gehring, WJ, Kloter U, et al. (2009). Evolution of the Hox gene complex
from an evolutionary ground state. Current topics in developmental
biology. 88:35–61.
Gilbert SF (2004). General Principles of Differentiation and Morphogenesis.
New York: Oxford University Press. 10–24.
Gilbert SF, Epel D (2009). Ecological Developmental Biology: Integrating
Epigenetics, Medicine, and Evolution. Sinauer Associates Sunderland,
Mass.
Goldstein LS, Gunawardena S (2000). Flying through the Drosophila
cytoskeletal genome. Journal of cell biology. 150(2):F63–F68.
Gupta, BP, Wang M, et al. (2003). The C. elegans LIM homeobox gene
lin-11 specifies multiple cell fates during vulval development.
Development. 130(12):2589–2601.
Headon, DJ, Emmal SA, et al. (2001). Gene defect in ectodermal dys-
plasia implicates a death domain adapter in development. Nature.
414(6866):913–916.
Hogan B (1996). Bone morphogenetic proteins:  multifunctional
regulators of vertebrate development. Genes and development.
10(13):1580–1594.
Holland LZ (2007). Developmental biology:  a chordate with a differ-
ence. Nature. 447(7141):153–155.
Istrail S, Sutton GG, et  al. (2004). Whole-genome shotgun assem-
bly and comparison of human genome assemblies. Proceedings of
the National Academy of Sciences of the United States of America.
101(7):1916–1921.
Jones KL, Smith DM (2006). Recognizable Patterns of Human
Malformation. New York: Elsevier.
Joza N, Susin SA, et  al. (2001). Essential role of the mitochondrial
apoptosis-inducing factor in programmed cell death. Nature.
410(6828):549–554.
Kang S, Graham JM, et  al. (1997). GLI3 frameshift mutations cause
autosomal dominant Pallister-Hall syndrome. Nature genetics.
15(3):266–268.
Katsanis N (2004). The oligogenic properties of Bardet-Biedl syndrome.
Human molecular genetics. 13(suppl 1):R65–R71.
Katsanis N, Lupski JR, et  al. (2001). Exploring the molecular
basis of Bardet-Biedl syndrome. Human molecular genetics.
10(20):2293–2299.
Kinzler, KW, Ruppert JM, et al. (1988). The GLI gene is a member of the
Kruppel family of zinc finger proteins. Nature. 332(6162):371–374.
Krogan, NJ, Cagney G, et  al. (2006). Global landscape of pro-
tein complexes in the yeast Saccharomyces cerevisiae. Nature.
440(7084):637–643.
Kumar D (2008). Genomic perspectives of human development. In
Genomics and Clinical Medicine' Eds. Kumar, D & Weatherall
DW, Oxford University Press, NY.
Lander ES, Linton LM, et al. (2001). Initial sequencing and analysis of
the human genome. Nature. 409(6822):860–921.
G e n e s , G e n o m e , a n d D e ve l o pm e n ta l M a l f o r m at i o n s   •   8 1
Li, JB, Gerdes JM, et al. (2004). Comparative genomics identifies a fla-
gellar and basal body proteome that includes the BBS5 human dis-
ease gene. Cell. 117(4):541–552.
Li L, Stoeckert CJ, et al. (2003). OrthoMCL: identification of ortholog
groups for eukaryotic genomes. Genome research. 13(9):2178–2189.
Liao J, Kochilas L, et  al. (2004). Full spectrum of malformations
in velo-cardio-facial syndrome/DiGeorge syndrome mouse
models by altering Tbx1 dosage. Human molecular genetics.
13(15):1577–1585.
Liu A, Wang B, et al. (2005). Mouse intraflagellar transport proteins reg-
ulate both the activator and repressor functions of Gli transcription
factors. Development. 132(13):3103–3111.
Martínez-Frías M (2004). Segmentation anomalies of the vertebras and
ribs: one expression of the primary developmental field. American
Journal of Medical Genetics, Part A. 128(2):127–131.
McKusick VA (1969). On lumpers and splitters, or the nosology of
genetic disease. Birth Defects. 5(1):23–32.
Migeon, BR, Brown TR, et al. (1981). Studies of the locus for androgen
receptor: localization on the human X chromosome and evidence
for homology with the Tfm locus in the mouse. Proceedings of the
National Academy of Sciences. 78(10):6339–6343.
Morton NE (1956). The detection and estimation of linkage between
the genes for elliptocytosis and the Rh blood type. American journal
of human genetics. 8(2):80.
Mount DW (2004). Bioinformatics: Sequence and genome analysis. Cold
Spring Harbour, NY: Cold Spring Harbour Laboratory Press.
Nadeau JH (2001). Modifier genes in mice and humans. Nature reviews
genetics. 2(3):165–174.
Newton K, Strasser A (1998). The Bcl-2 family and cell death regulation.
Current opinion in genetics & development. 8(1):68–75.
Oti M, Brunner H (2007). The modular nature of genetic diseases.
Clinical Genetics. 71(1):1–11.
Pazour GJ, Witman GB (2003). The vertebrate primary cilium is a sen-
sory organelle. Current opinion in cell biology. 15(1):105–110.
Reginato AM, Olsen BR (2002). The role of structural genes in
the pathogenesis of osteoarthritic disorders. Arthritis research.
4(6):337–345.
Reid R (2004). Epigenetics and Environment:  The Historical Matrix of
Matsuda’s Pan-Environmentalism. Cambridge, MA: The MIT Press.
Robert JS (2008). Taking old ideas seriously:  evolution, development,
and human behavior. New Ideas in Psychology. 26(3):387–404.
Schwab W, Kasper M, et al. (2000). Characterization of caveolins from
human knee joint cartilage: expression of caveolin-1, -2, and -3 in
chondrocytes and association with integrin β1. Histochemistry and
cell biology. 113(3):221–225.
Slavotinek A, Searby C, et al. (2002). Mutation analysis of the MKKS
gene in McKusick-Kaufman syndrome and selected Bardet-Biedl
syndrome patients. Human genetics. 110(6):561–567.
Slavotinek AM, Stone EM, et  al. (2000). Mutations in MKKS cause
Bardet-Biedl syndrome. Nature genetics. 26(1):15–16.
Souza P, Kuliszewski M, et al. (1995). PDGF-AA and its receptor influ-
ence early lung branching via an epithelial-mesenchymal interaction.
Development. 121(8):2559–2567.
Spranger J, Winterpacht A, et al. (1994). The type II collagenopathies: a
spectrum of chondrodysplasias. European journal of pediatrics.
153(2):56–65.
Stern HM, Brown A, et  al. (1995). Myogenesis in paraxial meso-
derm:  preferential induction by dorsal neural tube and by cells
expressing Wnt-1. Development. 121(11):3675–3686.
Stone DL, Slavotinek A, et  al. (2000). Mutation of a gene encoding a
putative chaperonin causes McKusick-Kaufman syndrome. Nature
genetics. 25(1):79–82.
Trehan KS, Gill KS (2002). Epigenetics of dominance for enzyme activ-
ity. Journal of Biosciences. 27(2):127–134.
Vaahtokari A, Aberg T, et  al. (1996). Apoptosis in the developing
tooth: association with an embryonic signaling center and suppres-
sion by EGF and FGF-4. Development. 122(1):121–129.
Verloes A (1995). Numerical syndromology: a mathematical approach to
the nosology of complex phenotypes. American journal of medical
genetics. 55(4):433–443.
Vortkamp A, Gessler M, et  al. (1991). GLI3 zinc-finger gene inter-
rupted by translocations in Greig syndrome families. Nature.
352(6335):539–540.
Watt AJ, Chung KC (2009). Generalized skeletal abnormalities. Hand
Clinics. 25(2):265–276.
Wilkie AO (1997). Craniosynostosis:  genes and mechanisms. Human
molecular genetics. 6(10):1647–1656.
Wu M, Li J, et al. (2008). Persistent expression of Pax3 in the neural crest
causes cleft palate and defective osteogenesis in mice. Journal of clini-
cal investigation. 118(6):2076.
Yamada T, Pfaff SL, et al. (1993). Control of cell pattern in the neural
tube: motor neuron induction by diffusible factors from notochord
and floor plate. Cell. 73(4):673–686.

8 2   •  P r i n c i p l e s o f G e n o m i c M e d i c i n e
 6.
BIOINFORMATICS, SYSTEMS BIOLOGY,
AND SYSTEMS MEDICINE
Binay Panda and Neeraja M. Krishnan
INTRODUCTION
Bioinformatics is an interdisciplinary approach that broadly
aims at applying computational power and analytical skills
to solve biological problems.1 In the last two decades, due to
the advent of high-throughput technology, biology has wit-
nessed a data deluge that has demanded that scientists and
policymakers alike rethink sustainable and long-term solu-
tions towards data analyses, storage, archiving, and shar-
ing.2–4 Whole-genome and exome-sequencing was made
possible by the introduction of newer generations of DNA-
sequencing instruments, the availability of lower-cost con-
sumables, automated and easier experimental workflows,
and better computational pipelines. These have resulted in
better understanding of complex diseases like cancer. This,
along with the need to understand and apply genome sci-
ence to medicine, is making computational biology indis-
pensable to scientists involved in genome medicine and
systems biology research. A new breed of scientists, savvy
in dealing with high-throughput data and systems biology,
is emerging to be at the forefront of research in genome sci-
ence and medicine.
The rate at which genome-scale data is being produced
has already surpassed Moore’s law in terms of the amount of
total data produced year after year and the cost involved to
sequence a megabase (mb) of DNA. The field is already at a
stage where the cost associated with data storage and analy-
ses dwarfs the cost of producing the primary sequencing
data. Hence, in those instances where biological material
can be easily procured, it might be cheaper and quicker to
produce the data from scratch, if need be, rather than stor-
ing the raw data. However, in medical sequencing, where
samples are limited and precious, it is imperative to come up
with better solutions for data analyses, storage, and manage-
ment in order to arrive at clinically meaningful conclusions.
Bioinformatics plays a vital role in sifting through
unstructured data and storing valid aspects of the data in
83
organized databases, thus allowing biologist to access and
curate preexisting information, and add further data in a
structured manner.5 The next important level of bioinfor-
matics aims at developing tools and algorithms to aid in the
analysis of the data. Computational tools and mathematical
modeling have paved the way towards a systems approach
for understanding the biological functions and behavior of
complex systems6,7 and have gone through iterations of con-
ceptual development with the advent of high-throughput
data. Systems biology attempts to map out the complexity
of interactions among individual components, rather than
reduce its complexity by elimination of parts (as in classi-
cal genetics8) and integrates concepts of theoretical biology
with experimental validation.9 Systems biology and medi-
cine is becoming an integral part of large biology studies
that largely rest on the premise that the whole is greater
than the sum of its parts; therefore, the system is looked at
in its entirety.10–13 Needless to say, both the integrated and
reductionist approaches are needed to further our under-
standing of biological systems.
Systems medicine extends the principles of systems
thinking to the design and development of holistic and
affordable healthcare with an improved understanding of
disease states and disease progression. Systems medicine
will begin to provide mechanistic insights into human
diseases, and facilitate the development of better biomark-
ers for prognosis and diagnosis of diseases such as cancer.
Plugging the pathway connectivity of disease-susceptible
loci into drug design will help scientists find new ways to
develop drugs. Ultimately, one can envision a modern,
open-minded, and affordable healthcare system developed
through a systems approach to medicine, catering to all sec-
tors of human life.
As most systems biology and medicine is heavily depen-
dent on high-throughput data and bioinformatics analyses,
this chapter will explain in detail high-throughput data
generation, downstream analyses, and interpretation, with
a focus on data generated by the newer generations of DNA
sequencers. Lastly, biological structure and function along
with its utility in clinical effectiveness will also be discussed.
D N A SE Q U E N C I N G I N T H E
P O S T–H U M A N G E N O ME
P R O J E C T P H A SE
Post–human genome sequencing (which was achieved pri-
marily by using the automated Sanger capillary sequencing
instruments), the second generation of DNA sequencers have
contributed the most to the data deluge in biology. The first
instrument in the second generation of DNA sequencers was
commercially introduced by a Connecticut-based American
company called 454 Life Sciences in 2005, which was later
bought by Roche Diagnostics. Although many other com-
panies introduced their technology in the market post-2005,
the one from Solexa (which was later acquired by Illumina)
employing a massive parallel sequencing-by-synthesis (SBS)
method using reversible terminator chemistry14 remains the
dominant player, both in terms of the total amount of data
produced so far globally, and of the market penetration in
large genome centers and individual laboratories. Hence, this
chapter shall cover data analysis mainly related to the sequenc-
ing reads produced by sequencing-by-synthesis technology
from Illumina, as it is the technology most widely used by
researchers. Additionally, due to accessibility constraints, this
review will focus on open-source tools or tools available as
free downloads from commercial providers.
The second or new generation of DNA sequencers
(popularly called next-generation sequencers or NGS) has
brought about a major transformation in biology, medicine,
and agriculture in the last few years by aiding the discov-
ery of genetic, transcriptomic, and epigenetic markers in a
genome-wide scale. With sequencing becoming ubiqui-
tous, transcriptome sequencing is increasingly preferred to
the hybridization-based experiments using whole-genome
microarrays, since sequencing is annotation-independent,
digital, and quantitative. RNA and cDNA sequencing also
overcome disadvantages typically inherent in the microar-
ray experiments: like probe cross-hybridization, “noise,” and
limited dynamic range of detection.15 Chip-Seq, also a next-
generation technology, assays the entire genome/transcrip-
tome for protein-bound DNA/RNA regions, and therefore
has advantages over the array-based ChIp-chip technique,
which relies on probes16,17 and suffers from the same draw-
backs as array-based gene expression experiments.
As the focus of this chapter is to elaborate the bioin-
formatics and systems biology aspects of high-throughput
8 4   •   P rincip l e s o f G e no m ic M e dicin e
genome medicine, we first provide a short description of the
standard analyses of workflow and post raw data retrieval
from the sequencing instruments. Various sequencing tech-
nology, the chemistry of producing sequencing libraries,
and study design considerations are reviewed extensively
elsewhere.18–24
A N A LY T I C A L WO R K FL OW
S EQ U E N C E R E A D S A N D P ROTO C O L S
The Illumina platform offers two kinds of protocols for
sequencing, the short-insert and the long-insert protocols.
The former protocol allows users to sequence libraries with
a shorter insert size of up to 600bp, whereas the latter can
sequence long insert libraries ranging from 1.5kb to 20kb,
making it useful for understanding structural variations in the
genome. In both protocols, users have an option to produce
either single- or paired-end reads. The basic format for stor-
ing the sequenced short reads is the fastq format. The fastq
format comprises both the sequence and the ASCII-coded
quality scores for each nucleotide sequenced (Figure 6.1A).
Q UA L I T Y C H E C K S A N D R E A D
P R E -P R O C ESS I N G
Before the short sequenced reads are used further in any
analytical workflow, it is important to check the individ-
ual read quality and quality of each nucleotide in a read.
Different sequencing platforms estimate quality differently,
and there are several read quality control (QC) tools that
access the quality. FastQC25 is a standard modular tool used
to perform quality checks on high-throughput sequence
data derived from Illumina instruments by running a bat-
tery of tests on the raw data, and it sets a flag when certain
quality conditions are not met. Trimmomatic26 is a read
QC and trimming tool, which trims the reads based on
known Illumina-like adapter sequences or a sliding window
of trailing base qualities. Once the standard QC checks are
done, researchers studying archival or metagenomic sam-
ples often run another set of tools, like DeconSeq27 that are
used as pre-processing filters to eliminate contamination
from genomic or metagenomic sequenced reads. Cancer
researchers often use a tool called ContEst28 for estimat-
ing the amount of cross-sample contamination in sequenc-
ing data. ContEst uses a Bayesian framework to estimate
contamination levels from array-based genotypes and
­
sequencing reads.
(A)

(B)

Figure 6.1  Sequenced reads in Fastq (A) and SAM (B) formats. Various attributes of the file are shown.
 A L I G N I N G R AW R E A D S TO T H E
R EFE R E N C E SE Q U E N C E
Once checked for quality and pre-processed, reads are
aligned to a reference genome using a mapper or read aligner.
There are numerous open-source or freely downloadable
aligners available, each tailored to perform optimally under
specific criteria that fall in two major classes. The first class
of aligners is built on hash table–based approach of the
reference genome that includes Bfast,29 Ssaha,30 Smalt,31
Stampy,32 and Novoalign.33 The second class of aligners
relies on creating an efficient index of the reference genome,
and in this category fall BWA34 and Bowtie.35,36 BWA,34 an
aligner based on the Burrows Wheeler transform, is the most
popular short-read aligner used by the research community.
In a comparison study37 involving five such short-read align-
ers, we found Novoalign to perform best in read alignment
in terms of sensitivity, perhaps owing to its post-alignment
base quality recalibration functionality. Following the read
alignment to the reference genome, reads are usually stored
in Sequence Alignment/Map (SAM) format, which is
evolving as a consensus, flexible, and generic format among
the research community. Figure 6.1B provides a snippet of
the SAM format.
C A LL I N G VA R I A N T S
After alignment of reads, single nucleotide polymorphisms
or variants (SNPs or SNVs), insertion and deletion vari-
ants (indels), and copy number alterations or variations
(CNAs or CNVs) are identified in the sequenced genomes
and exomes. There are several tools designed to call SNVs
using aligned sequence data. The most popular one is
the Genome Analyses Took Kit (GATK)38 developed by
the Broad Institute in Cambridge, Massachusetts. In a
comparative study among the other variant calling algo-
rithms,37 GATK performed best in calling SNVs, most
likely due to the base call recalibration and local realign-
ment step that it incorporates for variant calling. In addi-
tion to GATK, tools like Samtools,39 Bambino,40 and
Freebayes41 are some of the other open-source variant call-
ers that are widely used. For indels, Samtools,39 Pindel,42
and Dindel43 are widely used to detect indels in the short,
medium, and long range, in addition to tandem duplica-
tions and inversions. Copy number alterations (CNAs)
are an important category of structural aberrations in
human diseases. CNAs range from one kilobase (kb) to
several megabases (mb) and have been implicated in sev-
eral human diseases, including cancer, and in the normal
8 6   •   P rincip l e s o f G e no m ic M e dicin e
population.44–55 SVDetect,56 Breakway,57 Breakdancer,58
and CREST59 are a few open-source tools that can detect
structural variations and genome breakpoints. Copy num-
ber detection tools either measure absolute copy number
changes in a single sample or somatic copy number altera-
tions (SCNA) in cancer and other disease samples where
both control and diseased sample data are used. Both
these sets of tools can use whole genome or exome or both
types of data. RDXplorer,60 CNVNator,61 and Freec62 are
CNV detection tools meant for whole genomes of single
sample analyses, and COPS,63 SVDetect,56 and CNV-Seq64
are whole-genome SCNA detection tools that use disease
and matched normal paired samples. In a rigorous bench-
marking of these tools, we found COPS to perform well,
in terms of both sensitivity and specificity, in detecting a
wide size range of CNAs, using libraries with varying read
lengths and sequencing coverage using the whole-genome
dataset.63 The fragmented spread of exonic regions makes it
difficult to determine CNA boundaries, and hence, break-
points accurately. Control-Freec,65 ExomeCNV,66 and
ExomeDepth67 are paired CNA callers, specially designed
for exome dataset.
G E N E A N D VA R I A N T
A N N OTAT I O N
Once the variants are called, several tools are used for gene
and variant annotation. GPAT68 is a rapid gene-annotation
tool based on the genomic positions. It operates currently
on various genome versions of popular organisms such as
human (H. sapiens), mouse (M. musculus), Arabidopsis
(A. thaliana), fruit fly (D. melanogaster), and zebrafish
(D. rerio). The purpose of GPAT is to provide an easy-to-
use and convenient tool for rapidly annotating reasonably
large sets of genomic positions. In addition to gene anno-
tation, GPAT allows the retrieval of some expression data
and chromosome position data. Several in silico tools have
been developed to predict the effect of single nucleotide
variants on a protein. Some of the more popular ones are
SIFT,69 Polyphen,70 VEP,71,72 and PROVEAN.73 Most of
these tools use predictions based on the degree of con-
servation of amino acid residues in sequence alignments
derived from closely related sequences, collected through
PSI-BLAST.74,75 PROVEAN73 (Protein Variation Effect
Analyzer) predicts the impact of SNP and indel on protein
function, and can thus identify non-synonymous SNPs and
functionally relevant indels. It compares well in its speed
and performance to SIFT69 and PolyPhen-2, VEP,71,72 an
Ensembl tool, provides additional annotation of variants
by determining their effects on genes, transcripts, protein
sequences, and even regulatory regions.
VI S UA L I Z I N G G E N O ME DATA
Several visualization platforms have been developed to
aid genome and variant visualization. The common and
most popular ones are the UCSC76 and Ensembl77 genome
browsers. Both genome browsers can import user data in
different formats for visualization alongside the publicly
available data and data from multiple types of experiments
(Figure 6.2). Although genome browsers, like that from
UCSC, are immensely popular, easy to browse, can be cus-
tomized to include external data, and provide data visual-
ization for most community and consortium projects (like
HapMap and Encode), they lack the ability to visualize data
from two or more faraway regions of the genome at the same
resolution in a single window. Recently, a nonlinear repre-
sentation of data that are megabases away, where multiple
regions of the genome can be visualized together in a single
window, termed the Elastic Genome Browser, has been pro-
posed.78 For easy and simultaneous visualization of all types
of variants in individual genomes, Circos plots79 provide us
with the best option (Figure 6.3). As represented in Figure
6.3, the chromosomes, positions of multiple types of vari-
ants, allelic imbalances, genes bearing the SNVs, indels, and
CNVs can be visualized simultaneously using Circos plots,
aiding visualization at the whole-genome level.
PAT H WAY A N A LYSES
At the end of next-generation sequencing data analyses, biol-
ogists often face the problem of finding meaning in the list
of genes discovered in a particular disease study. This entails
extensive functional analyses and validation that require
skills and substantial time to complete. The pathway-based
approach reduces complexity, increases the explanatory
power of high-throughput analyses, and simplifies the task
of looking at many variants throughout the genome. There
are various databases for performing knowledge-based path-
way analyses, like Kegg,80–83 BioCarta,84 Reactome,85–89 and
Panther.90–93 DAVID94 offers an application interface that
links all the above knowledge databases. Besides these, sev-
eral commercial tools can perform pathway analyses, such
as AvadisNGS,95 which uses a Natural Language Processing
(NLP)–based tool to look up biological interactions from
literature and Ingenuity Pathway Analysis (IPA).96 IPA is
the most popular pathway analysis tool among biologists,
B ioin f or m atic s , Sys t e m s B io l ogy, and Sys t e m s M e dicin e   •   8 7
as it relies on a expert-curated database storing biological
interactions and functional annotations from published
literature. Performing pathway analyses after discovery of
variants has many implications for personalized therapy,
especially for pathway-based drug repositioning, where
existing drugs can be tried for new therapeutic uses based
on pathway information.
SE Q U E N C E A SSEM B LY
In the absence of a valid reference genome, de novo sequence
assembly is performed, involving self-alignment of the reads
and merging of the longest stretch of self-aligned reads
to reconstruct the longest contiguous sequence (contig).
Contigs are then assembled into scaffolds using gaps between
paired sequence reads. Generation of short sequence reads
of any genome, followed by de novo assembly, can be com-
pared to shredding a paper into several pieces and trying to
assemble them back to recreate the original shape and form.
This process is a complex one and is often difficult and error-
prone.97 There is a possibility of misplacing a piece of paper
from its original location. Particularly, identical parts of the
paper may get exchanged, and in this manner, get misas-
sembled. The best methods available today towards produc-
ing complete assemblies of large genomes are methods that
take sequence reads from multiple technologies (Sanger,
second- and third-generation), read lengths (using both
long- and short-insert libraries) and chemistry (sequencing-
by-synthesis, sequencing-by-ligation and single molecule
long reads). Several de novo genome assemblers have been
developed in the past.98–111 These tools use short sequence
reads, many from both short- and long-insert libraries, to
assemble them into different contigs and further assemble
the contigs into “scaffolds,” using appropriate gaps spanning
those contigs. Long insert mate-pair libraries are typically
used in the process of scaffolding, and provide the informa-
tion on the distance between two contiguous stretches of
DNA elements. This reduces fragmentation of scaffolds and
pieces together the multitude of contigs to constitute fewer
scaffolds while creating several gaps; e.g., stretches of Ns
or non-ATGC characters within the scaffolds. Assemblers
are also tailored to a particular sequencing platform.
Velvet,110 ABySS,108 and SOAPdenovo111 are commonly
used genome assemblers for Illumina and SOLiD reads.
AMOS112 can be used to assemble Sanger and 454 reads.
CLC Genomics Workbench,113 a proprietary tool, is used
to assemble Sanger, 454, Illumina, and reads. Newbler114 is
exclusively used to assemble 454 reads. MIRA115 is used for
assembling Sanger, 454, Ion Torrent and Illumina PacBio
 (A)

(B)

Figure 6.2 
Visualization using UCSC (A) and Ensembl (B) genome browsers. Genomic coordinate for the gene TP53 is shown with self-explanatory
attributes from the public databases.
 (A)

1
X
22
21
20

2
19

18

17 3

16

4
15

CTX

SV
14 CNS ELs
5
-INS IND

13
L SNV
-DE s
12
6 CNV

Het-A
-B
Het
11

7
10

8
6

1 22
(B)

2 21

3 20

4 19

5 18
Chromosome

6 17

7 16

8 15

9 14

10 13

11 12

Figure 6.3 
Variant visualization using circular representation of the genome using Circos (A) and linear representation of the genome using the
statistical program R (B). Various variants depicted are: SNVs—single nucleotide variants, CTX—chromosomal translocations; SV—large
insertions/deletions and inversions, detected using the program BreakDancer; CNV-INS and CNV-DEL—insertion and deletion copy number
variants, respectively; Het-A, Het-B—variants found using whole-genome microarrays in A and B allele, respectively; and INDEL—short
insertion/deletion variants found using the program Pindel.
reads. ALLPATHS-LG116 assembles Illumina and SOLiD
reads, and requires overlapping read libraries. While the
tools mentioned above can produce both contigs and scaf-
folds, there are tools that perform only scaffolding using
pre-assembled contigs, such as SSPACE,117 Bambus,118
and Opera.101 Reads from the Pacific BioSciences (PacBio)
instrument (which are typically of varying lengths but are
often long) can be used along with the short reads from
Illumina to perform hybrid assembly using the overlap-
layout-consensus (OLC) approach. PacBiotoCA119 is an
error-correction pipeline that corrects the long and error-
prone reads by mapping accurate shorter reads from other
sequencing instruments to those from PacBio. Another
algorithm developed by Koren and his co-workers, which
is implemented as part of the Celera assembler,120 improves
read accuracy for assembly using long reads from PacBio.121
The algorithm does this by calculating an accurate hybrid
consensus sequence by mapping higher-accuracy short
reads from Illumina to individual lower-accuracy long reads
from PacBio.
Even with the availability of multiple sequencing tech-
nology and chemistry, some parts of the genome are inher-
ently hard to sequence. This is primarily due to either the
sequence context (repetitive elements) or the failure of
a particular technology to read through a stretch of the
same nucleotide (homopolymer issues in pyrosequenc-
ing). In order to be practically useful, the completeness of
the sequenced genomes is a prerequisite and imperative for
our understanding the underlying biology. Several methods
have been proposed to fill the gaps in the assembly process
with the existing short reads.104,111,122,123 We are currently
working on a proposed method called FILAM (FILling the
gAp by re-Mapping) that takes into account the partially
mapped reads to the assembly gaps and recreates scaffolds
post gap filling. In the proposed method, once the partially
mapped reads are identified, the exact gap sizes in the scaf-
folds are determined followed by filling the assembly gaps
iteratively by extending the reads.124
Transcriptome and EST assembly are more complex
than genome assembly. This is due to the presence of
spliced site junctions, representation of low copy number
transcripts, genes with varying expression levels, and the
presence of homologues. Seed length for alignment125 and
k-mer size126,127 are important parameters for transcrip-
tome assembly. De novo transcriptome assembly can either
make use of the available genome sequence of the indi-
vidual organism (genome-assisted) or be performed inde-
pendently. The tools that do the latter use the seed length
parameter while aligning reads to a reference genome prior
to assembling transcriptome. De novo assembly tools either
9 0   •   P rincip l e s o f G e no m ic M e dicin e
involve construction of de Bruijn graphs using k-mers (for
short reads) or use an overlap-layout-consensus (OLC)
approach (for longer reads).128 Trinity,129 Trans-ABySS,127
Oases,130 and SOAPdenovo-Trans131 are commonly used de
novo transcriptome assemblers. Genome-guided Trinity,132
Scripture,133 and Tophat1-Cufflinks134 are genome-guided
transcriptome assemblers. Genome-guided Trinity,132
although it requires a genome, only uses the genome to
cluster reads locally and then proceeds to assemble each of
these clusters using the same approach as Trinity, which is
different from the others in the same group. The former
is used mostly for short reads, while the latter is used for
long reads from 454 or PacBio. A detailed account of
various transcriptome assembly tools has been reviewed
elsewhere.135 Both the de novo and the genome-guided tran-
scriptome assembly processes have their own advantages
and disadvantages. De novo transcriptome assembly tools
can recover transcript fragments from regions missing in
the genome assembly, while the genome-guided assembly
ones can filter out contamination and sequencing artifacts,
can recover low-abundance transcripts, and can fill gaps
using the genome sequence, resulting in full-length tran-
scripts. Recently, from a comparison study of the existing
de novo and genome-guided assembler tools, we proposed
an augmented method for the transcriptome assembly
using the de novo tool with a genome-guided approach has
been suggested.136
F U N C T I O N A L A N N OTAT I O N
The process of assembly is usually followed by gene predic-
tion and annotation. Gene prediction tools utilize gene
models and training sets, based on previously annotated
genes from closely related organisms. In addition, the
assembled genome or transcriptome can also be subjected to
similarity-based annotation by performing BLAST against
pre-existing databases. The AutoFACT137 pipeline system-
atically annotates assemblies based on performing BLAST
against the nr74 (nucleotide and protein), uniref90,138 uni-
ref100,138 kegg,80,82,139,140 and cog141,142 databases. KEGG
orthologies using the Kegg Automatic Annotation Server
(KAAS),143 annotate sequences similar to known ortholo-
gous genes from a selected range of organisms, across a wide
spectrum of pathways. However, BLAST-based analyses
can only recover known annotations. Un-annotated genes
specific to a given organism need to be validated experi-
mentally to confirm their functional role. A potential
trick to reduce complexity is to identify known domains
within the un-annotated genes using domain databases
such as PFam,144 and increase knowledge of these predic-
tions/transcripts. Furthermore, the assembled genome can
be characterized with respect to its repeat content. This is
typically done using known organism-specific repeat librar-
ies by tools like RepeatMasker145 and RepeatModeler,146 or
using de novo repeat predictions by tools like LTRFinder,147
MITE-hunter,148 and RepeatScout149 using certain consen-
sus repeat motifs.
SE Q U E N C E , S T RU C T U R E , A N D
F U N C T I O N O F N U C LE I C AC I D S
A N D P R OT E I N S
Studying the sequence, structure, and function of DNA,
RNA, and protein forms the foundation of all biological
questions. The ultimate aim of all high-throughput technol-
ogy, biochemical and analytical, which has led to increasing
amounts of sequences and structures of biological macro-
molecules in the databases, is to understand biological func-
tion. Below, a few key methods that apply to sequence and
structure analyses, both for nucleic acids and for protein, are
outlined.
MU LT I P LE SE Q U E N C E A L I G N ME N T
A multiple sequence alignment (MSA) of DNA, RNA, or
protein helps identify conserved regions of high similarity
across the sequences that might have arisen out of a com-
mon evolutionary origin to support different functions,
or alternatively, with some shared and some unique func-
tional domains of proteins. As (proposed by Francois Jacob)
Nature is a tinkerer and not an inventor,150 the existing
information is adapted and reused rather than new infor-
mation being created. Pfam151 is a database of all known
protein domains. ClustalW152 is commonly used in pro-
ducing a multiple sequence alignment. The MSA is used to
produce a phylogeny, a graphical, treelike representation of
the relatedness between sequences. There are various evo-
lutionary models153 to produce phylogenies. The simplest
non-model approach is the principle of parsimony, which
assumes the minimal change required to go from one node
to another. Methods that use Maximum Likelihood and
Bayesian approaches rely on an optimized substitution-
rate matrix to reconstruct ancestral states of the phylogeny,
and are known to perform better with minimal composi-
tional bias. There are various tools available for reconstruct-
ing phylogeny, such as MrBayes,154 PAUP*,155 RAxML,156
Phyml,157 MEGA,158 and PHYLIP.159
B ioin f or m atic s , Sys t e m s B io l ogy, and Sys t e m s M e dicin e   •   9 1
SE C O N DA RY S T RU C T U R E
Base pairing within regions of nucleic acids, in single-
stranded DNA or RNA resulting in a stem-loop second-
ary structure, plays an important role in biological func-
tion. Mfold provides secondary structure estimates for
DNA and RNA, within a certain percent of sub-optimality
of the most optimal (minimum free energy) structure.160
Sfold provides a more accurate representation of the ther-
modynamic spread of all secondary structures by sampling
from the Boltzmann-weighted structure ensemble.161 The
tRNAScan-SE server162 uses the clover-leaf secondary
structural fold of a tRNA for detection in an un-annotated
sequence format. Nucleic acid tertiary structures like the
double helix, G-quadruplexes, A-minor motifs, ribose
zipper, pseudoknot, riboswitches, and others have been
implicated to carry specific functions in vivo. Molecular
dynamics simulations have become crucial tools in areas
such as biomolecular engineering, synthetic biology, and
drug design.163 Tools can render, model, and compare
three-dimensional structures of nucleic acids, based on
molecular dynamics of their primary and secondary struc-
tures. NAST164 and SETTER165 are two such tools.
For proteins, alpha helices and beta-sheets are highly
local sub-structures of a protein, defined by hydrogen
bonding. The meta-server for protein sequence analysis
(MESSA) is one of the most widely used protein struc-
ture and function prediction servers,166 which uses mul-
tiple tools to predict protein properties. YASPIN167 uses
Hidden Neural Networks and PSI-BLAST75,168 to produce
Position-Specific Scoring Matrix (PSSM169), used for sec-
ondary structure prediction.
While majority of high-resolution protein structures
have been solved using X-ray crystallography, nuclear mag-
netic resonance (NMR) can provide a time-dependent pro-
tein conformational change despite compromising on the
resolution of the predicted structures. Protein misfolding
has been linked to human diseases170 such as Alzheimer’s,
Huntington’s, osteogenesis imperfecta, and even some can-
cers. Folding@home,171 a distributed computing project
housed in the Stanford University, is used to analyze the
tertiary and quaternary structures of proteins. Abalone172
and Ascalaph Designer173 are two software suites from
Agile Molecule that perform biomolecular simulations
using molecular dynamics and protein folding and molecu-
lar building and dynamics using DNA, proteins, hydrocar-
bons, and nanotubes respectively. The DisoveryStudio174 is
an entire suite dedicated to discovery of drugs, and it allows
modeling and simulation of small drug molecules docking
to macromolecules.
 HEAD AND NECK CANCER
COLORECTAL CANCER
PANCREATIC CANCER
LUNG CANCER
ESOPHAGEAL CANCER
RENAL CELL CARCINOMA
Tumor/matched PROSTRATE CANCER
normal pair
BLADDER CANCER
TUMOR
CERVICAL/UTERUS CANCER
NORMAL
OVARIAN CANCER
BREAST CANCER
Figure 6.4  A cartoon depicting an integrative multi-omics approach, from disease to pathways.
I N T E G R AT I VE B I O I N F O R M AT I C S
TO SYS T EMS B I O L O GY A N D
SYS T EMS ME D I C I N E
Whole-genome sequencing, whole-transcriptome, and
whole-genome methylation profiling of humans and other
animals, microbes, plants, and other life forms give us a
handle to understand tissue-specificity, diseased states,
treatment response, and environmental stress response.
Moreover, these techniques provide new biological insights.
Having said that, many genome-wide studies are stochastic
in nature; hence, they may not be linked with true func-
tionality. An example is the Encode project, which claims
that more than 80% of the human genome is functional.
In fact, there are reports that criticize such genome-wide
projects attempting to link function with a set of biologi-
cal measurements.175 Attempts should be made to elimi-
nate any stochastic “noise” before arriving at a conclusion
about the functional association of any molecular signa-
ture. Another case in point is cancer. In the last few years,
there have been many attempts by multiple international
initiatives to come up with a comprehensive cancer-derived
signature176,177 where multiple types of genetic, transcrip-
tomic, and epigenetic signatures have been linked with a
particular tumor type. This brings integrative analyses to
the forefront (Figure  6.4) where data from genetic, tran-
scriptomic, and epigenetic studies are combined to gain
functional insights. Although these integrative studies have
identified hundreds of genetic mutations in any particu-
lar tumor type, only a fraction of these mutations is really
responsible for the tumor initiation and/or progression
(driver mutation for those responsible for tumor progres-
sion). Most of these mutations are neutral or passenger
9 2   •   P rincip l e s o f G e no m ic M e dicin e
Integrated data
METHYLATION
analysis/filtering
WHOLE GENOME/
EXOME SEQUENCING
ARRAY GENOTYPING
METABOLOME PROFILING
Pathway
analysis
TRASCRIPTOME
SEQUENCING
GENE EXPRESSION
mutations.178 Hence, it is imperative to distinguish the
driver or functional mutations that can enable better diag-
nosis, prognosis, and treatment options to drive personal-
ized and preventative medicine forward. Systems medicine,
in the context of modern medicine integrating data from
multiple analyses and technologies, can thus be envisioned
to become predictive, preventative, personalized, and par-
ticipatory with a joint effort of academia, industry, and
health care professionals.
Lastly, efforts should be made to combine age-old tra-
ditional medicine practices developed empirically in old
cultures with Western medicine to treat diseases. This has
gained momentum recently with the publication of a few
scientific reports.179,180 Before the use of such age-old prac-
tices in modern clinic becomes ubiquitous, a systematic
review of the literature and rigorous analyses of data from
systems biology studies and high-throughput technologies,
suggesting their mode of action, will be required.
R EFE R E N C ES
1. Wooley JC, Lin H, National Research Council (U.S.), Committee
on Frontiers at the Interface of Computing and Biology. Catalyzing
inquiry at the interface of computing and biology. Washington,
DC: National Academies Press; 2005.
2. Pang T. Infections, genomics and global health. In: Kumar D, ed.
Genomics and health in the developing world. Oxford, UK: Oxford
University Press; 2012. pp. 152–156.
3. Reis-Filho JS. Next-generation sequencing. Breast Can Res: BCR.
2009;11(Suppl 3):S12.
4. Soon WW, Hariharan M, Snyder MP. High-throughput sequencing
for biology and medicine. Mol syst biol. 2013;9:640.
5. Luscombe NM, Greenbaum D, Gerstein M. What is bioinformat-
ics? A proposed definition and overview of the field. Methods inf
med. 2001;40(4):346–358.
 6. Lobo D, Malone TJ, Levin M. Towards a bioinformatics of pattern-
ing:  a computational approach to understanding regulative mor-
phogenesis. Biol open. 2013;2(2):156–169.
7. Schneider HC, Klabunde T. Understanding drugs and diseases by
systems biology? Bioorg med c l. 2013;23(5):1168–1176.
8. Laszlo A, Kripner S. Systems theories:  their origins, foundations
and development. In:  Systems Theories and A  Priori Aspects of
Perception. Amsterdam: Elsevier Science; 1998:47–74.
9. Kritikou E, Pulverer B, Heinrichs A. All systems go! Nat Rev Mol
Cell Biol. 2006;7:801.
10. Auffray C, Chen Z, Hood L. Systems medicine: the future of medi-
cal genomics and healthcare. Genome med. 2009;1(1):2.
11. Auffray C, Hood L. Editorial:  Systems biology and personalized
medicine—the future is now. Biotech j. 2012;7(8):938–939.
12. Auffray C, Imbeaud S, Roux-Rouquie M, Hood L. From functional
genomics to systems biology: concepts and practices. Comptes ren-
dus biologies. 2003;326(10–11):879–892.
13. Auffray C, Noble D. Origins of systems biology in William Harvey’s
masterpiece on the movement of the heart and the blood in animals.
Int J Mol Sci. 2009;10(4):1658–1669.
14. Bentley DR, Balasubramanian S, Swerdlow HP, et  al. Accurate
whole human genome sequencing using reversible terminator chem-
istry. Nature. 2008;456(7218):53–59.
15. Wang Z, Gerstein M, Snyder M. RNA-Seq: a revolutionary tool for
transcriptomics. Nature Rev Genet. 2009;10(1):57–63.
16. Barski A, Cuddapah S, Cui K, et  al. High-resolution pro-
filing of histone methylations in the human genome. Cell.
2007;129(4):823–837.
17. Johnson DS, Mortazavi A, Myers RM, Wold B. Genome-wide
mapping of in vivo protein-DNA interactions. Science.
2007;316(5830):1497–1502.
18. Mwenifumbo JC, Marra MA. Cancer genome-sequencing study
design. Nat Rev Genet. 2013;14(5):321–332.
19. Mardis ER. Next-generation DNA sequencing methods. Annu Rev
Genom Hum Genet. 2008;9:387–402.
20. Mardis ER. New strategies and emerging technologies for massively
parallel sequencing: applications in medical research. Genome Med.
2009;1(4):40.
21. Shendure J. Next-generation human genetics. Genome Biol.
2011;12(9):408.
22. Shendure J, Ji H. Next-generation DNA sequencing. Nat Biotech.
2008;26(10):1135–1145.
23. Metzker ML. Emerging technologies in DNA sequencing. Genome
Res. Dec 2005;15(12):1767–1776.
24. Metzker ML. Sequencing technologies—the next generation. Nat
Rev Genet. 2010;11(1):31–46.
25. Fast  QC.  http://www.bioinformatics.babraham.ac.uk/projects/
fastqc/.
26. Lohse M, Bolger AM, Nagel A, et  al. RobiNA:  a user-friendly,
integrated software solution for RNA-Seq-based transcriptomics.
Nucleic Acids Res. Jul 2012;40(Web Server issue):W622–W627.
27. Schmieder R, Edwards R. Fast identification and removal of
sequence contamination from genomic and metagenomic datasets.
PloS one. 2011;6(3):e17288.
28. Cibulskis K, McKenna A, Fennell T, Banks E, DePristo M,
Getz G. ContEst:  estimating cross-contamination of human
samples in next-generation sequencing data. Bioinformatics.
2011;27(18):2601–2602.
29. Homer N, Merriman B, Nelson SF. BFAST: an alignment tool for
large scale genome resequencing. PloS one. 2009;4(11):e7767.
30. Ning Z, Cox AJ, Mullikin JC. SSAHA:  a fast search method for
large DNA databases. Genome Res. 2001;11(10):1725–1729.
31. Ponstingl H. SMALT. 2011; http://www.sanger.ac.uk/resources/
software/smalt/.
32. Lunter G, Goodson M. Stampy:  a statistical algorithm for sensi-
tive and fast mapping of Illumina sequence reads. Genome Res.
2011;21(6):936–939.
B ioin f or m atic s , Sys t e m s B io l ogy, and Sys t e m s M e dicin e   •   9 3
3 3. Novocraft. Novoalign. 2011. Accessed 25 November 2011.
34. Li H, Durbin R. Fast and accurate short read alignment with
Burrows-Wheeler transform. Bioinformatics. 2009;25(14):
1754–1760.
35. Langmead B. Aligning short sequencing reads with Bowtie.
Current  protocols in bioinformatics / editoral board, Andreas D.
Baxevanis . . . [et al.]. Dec 2010;Chapter 11:Unit 11.7.
36. Langmead B, Salzberg SL. Fast gapped-read alignment with
Bowtie 2. Nat Methods. 2012;9(4):357–359.
37. Pattnaik S, Vaidyanathan S, Pooja DG, Deepak S, Panda B.
Customisation of the exome data analysis pipeline using a combina-
torial approach. PloS one. 2012;7(1):e30080.
38. McKenna A, Hanna M, Banks E, et  al. The Genome Analysis
Toolkit:  a MapReduce framework for analyzing next-generation
DNA sequencing data. Genome Res. 2010;20(9):1297–1303.
39. Li H, Handsaker B, Wysoker A, et al. The Sequence Alignment/Map
format and SAMtools. Bioinformatics. 2009;25(16):2078–2079.
40. Edmonson MN, Zhang J, Yan C, Finney RP, Meerzaman DM,
Buetow KH. Bambino:  a variant detector and alignment viewer
for next-generation sequencing data in the SAM/BAM format.
Bioinformatics. 2011;27(6):865–866.
41. Garrison E, Marth M. Haploytype-based variant detection from
short-read sequencing. arXiv:1207.3907v2 [q-bio.GN] 2012.
42. Ye K, Schulz MH, Long Q, Apweiler R, Ning Z. Pindel:  a pat-
tern growth approach to detect break points of large dele-
tions and medium sized insertions from paired-end short reads.
Bioinformatics. 2009;25(21):2865–2871.
43. Albers CA, Lunter G, MacArthur DG, McVean G, Ouwehand
WH, Durbin R. Dindel: accurate indel calls from short-read data.
Genome Res. 2011;21(6):961–973.
44. Seeger RC, Brodeur GM, Sather H, et al. Association of multiple
copies of the N-myc oncogene with rapid progression of neuroblas-
tomas. N Engl J Med. 1985;313(18):1111–1116.
45. Robbins CM, Tembe WA, Baker A, et  al. Copy number and tar-
geted mutational analysis reveals novel somatic events in metastatic
prostate tumors. Genome Res. 2011;21(1):47–55.
46. Prasad SE, Howley S, Murphy KC. Candidate genes and the behav-
ioral phenotype in 22q11.2 deletion syndrome. Dev Disabil Res
Rev. 2008;14(1):26–34.
47. Newkirk HL, Bittel DC, Butler MG. Analysis of the Prader-Willi
syndrome chromosome region using quantitative micro-
sphere hybridization (QMH) array. Am J Med Genet A.
2008;146A(18):2346–2354.
48. Elsea SH, Girirajan S. Smith-Magenis syndrome. Eur J Hum
Genet: EJHG. 2008;16(4):412–421.
49. Sebat J, Lakshmi B, Malhotra D, et  al. Strong association
of de novo copy number mutations with autism. Science.
2007;316(5823):445–449.
50. Sebat J, Lakshmi B, Troge J, et al. Large-scale copy number polymor-
phism in the human genome. Science. 2004;305(5683):525–528.
51. Kusenda M, Sebat J. The role of rare structural variants in the
genetics of autism spectrum disorders. Cytogenet Genome Res.
2008;123(1–4):36–43.
52. Lucito R, Suresh S, Walter K, et al. Copy-number variants in patients
with a strong family history of pancreatic cancer. Cancer Biol Ther.
2007;6(10):1592–1599.
53. Mills RE, Walter K, Stewart C, et  al. Mapping copy num-
ber variation by population-scale genome sequencing. Nature.
2011;470(7332):59–65.
54. Conrad DF, Pinto D, Redon R, et  al. Origins and functional
impact of copy number variation in the human genome. Nature.
2010;464(7289):704–712.
55. Redon R, Ishikawa S, Fitch KR, et al. Global variation in copy num-
ber in the human genome. Nature. 2006;444(7118):444–454.
56. Zeitouni B, Boeva V, Janoueix-Lerosey I, et al. SVDetect: a tool to
identify genomic structural variations from paired-end and mate-pair
sequencing data. Bioinformatics. 2010;26(15):1895–1896.
 57. Clark MJ, Homer N, O’Connor BD, et al. U87MG decoded: the
genomic sequence of a cytogenetically aberrant human cancer cell
line. PLoS genet. 2010;6(1):e1000832.
58. Chen K, Wallis JW, McLellan MD, et  al. BreakDancer:  an algo-
rithm for high-resolution mapping of genomic structural variation.
Nat Methods. 2009;6(9):677–681.
59. Wang J, Mullighan CG, Easton J, et al. CREST maps somatic struc-
tural variation in cancer genomes with base-pair resolution. Nat
Methods. 2011;8(8):652–654.
60. Yoon S, Xuan Z, Makarov V, Ye K, Sebat J. Sensitive and accurate
detection of copy number variants using read depth of coverage.
Genome Res. 2009;19(9):1586–1592.
61. Abyzov A, Urban AE, Snyder M, Gerstein M. CNVnator:  an
approach to discover, genotype, and characterize typical and atypi-
cal CNVs from family and population genome sequencing. Genome
Res. 2011;21(6):974–984.
62. Boeva V, Zinovyev A, Bleakley K, et al. Control-free calling of copy
number alterations in deep-sequencing data using GC-content nor-
malization. Bioinformatics. 2011;27(2):268–269.
63. Krishnan NM, Gaur P, Chaudhary R, Rao AA, Panda B. COPS: a
sensitive and accurate tool for detecting somatic copy number altera-
tions using short-read sequence data from paired samples. PloS one.
2012;7(10):e47812.
64. Xie C, Tammi MT. CNV-seq, a new method to detect copy number
variation using high-throughput sequencing. BMC bioinformatics.
2009;10:80.
65. Boeva V, Popova T, Bleakley K, et al. Control-FREEC: a tool for
assessing copy number and allelic content using next-generation
sequencing data. Bioinformatics. 2012;28(3):423–425.
66. Sathirapongsasuti JF, Lee H, Horst BA, et  al. Exome
sequencing-based copy-number variation and loss of heterozygosity
detection: ExomeCNV. Bioinformatics. 2011;27(19):2648–2654.
67. Plagnol V, Curtis J, Epstein M, et al. A robust model for read count
data in exome sequencing experiments and implications for copy
number variant calling. Bioinformatics. 2012;28(21):2747–2754.
68. Krebs A, Frontini M, Tora L. GPAT: retrieval of genomic annota-
tion from large genomic position datasets. BMC bioinformatics.
2008;9:533.
69. Kumar P, Henikoff S, Ng PC. Predicting the effects of coding
non-synonymous variants on protein function using the SIFT algo-
rithm. Nat Protoc. 2009;4(7):1073–1081.
70. Adzhubei I, Jordan DM, Sunyaev SR. Predicting functional effect
of human missense mutations using PolyPhen-2. Curr Protoc Hum
Genet. Jan 2013;Chapter 7:Unit 7.20.
71. McLaren W, Pritchard B, Rios D, Chen Y, Flicek P, Cunningham
F. Deriving the consequences of genomic variants with the
Ensembl API and SNP Effect Predictor. Bioinformatics.
2010;26(16):2069–2070.
72. Rios D, McLaren WM, Chen Y, et al. A database and API for varia-
tion, dense genotyping and resequencing data. BMC bioinformat-
ics. 2010;11:238.
73. Choi Y, Sims GE, Murphy S, Miller JR, Chan AP. Predicting the
functional effect of amino acid substitutions and indels. PloS one.
2012;7(10):e46688.
74. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local
alignment search tool. J Mol Biol. 1990;215(3):403–410.
75. Altschul SF, Koonin EV. Iterated profile searches with
PSI-BLAST—a tool for discovery in protein databases. Trend bio-
chem sci. 1998;23(11):444–447.
76. Kent WJ, Sugnet CW, Furey TS, et al. The human genome browser
at UCSC. Genome Res. 2002;12(6):996–1006.
77. Flicek P, Aken BL, Ballester B, et al. Ensembl’s 10th year. Nucleic
Acids Res. Jan 2010;38(Database issue):D557–D562.
78. Hariharan R. Elastic genome browser, Strand Life Sciences.
Bangalore, India, Personal communications.
79. Krzywinski M, Schein J, Birol I, et al. Circos: an information aesthetic
for comparative genomics. Genome Res. 2009;19(9):1639–1645.
9 4   •   P rincip l e s o f G e no m ic M e dicin e
80. Kanehisa M. The KEGG database. Novartis Foundation sym-
posium. 2002;247:91–101; discussion:  101–103, 119–128,
244–152.
81. Kanehisa M, Araki M, Goto S, et al. KEGG for linking genomes to
life and the environment. Nucleic Acids Res. Jan 2008;36(Database
issue):D480–D484.
82. Kanehisa M, Goto S. KEGG:  kyoto encyclopedia of genes and
genomes. Nucleic Acids Res. 2000;28(1):27–30.
83. Kanehisa M, Goto S, Furumichi M, Tanabe M, Hirakawa
M. KEGG for representation and analysis of molecular net-
works involving diseases and drugs. Nucleic Acids Res. Jan
2010;38(Database issue):D355–D360.
84. Nishmura D. Biocarta. Biotech Software Internet Report. Vol
22001:4.
85. Croft D, Mundo AF, Haw R, et al. The Reactome pathway knowl-
edgebase. Nucleic Acids Res. Nov 15 2013.
86. Croft D, O’Kelly G, Wu G, et  al. Reactome:  a database of reac-
tions, pathways and biological processes. Nucleic Acids Res. Jan
2011;39(Database issue):D691–D697.
87. Joshi-Tope G, Gillespie M, Vastrik I, et al. Reactome: a knowledge-
base of biological pathways. Nucleic Acids Res. 2005;33(Database
issue):D428–D432.
88. Matthews L, Gopinath G, Gillespie M, et al. Reactome knowledge-
base of human biological pathways and processes. Nucleic Acids
Res. Jan 2009;37(Database issue):D619–D622.
89. Vastrik I, D’Eustachio P, Schmidt E, et  al. Reactome:  a knowl-
edge base of biologic pathways and processes. Genome Biol.
2007;8(3):R39.
90. Mi H, Thomas P. PANTHER pathway: an ontology-based path-
way database coupled with data analysis tools. Methods Mol Biol.
2009;563:123–140.
91. Thomas PD, Kejariwal A, Campbell MJ, et  al. PANTHER:  a
browsable database of gene products organized by biological func-
tion, using curated protein family and subfamily classification.
Nucleic Acids Res. 2003;31(1):334–341.
92. Mi H, Dong Q, Muruganujan A, Gaudet P, Lewis S, Thomas PD.
PANTHER version 7: improved phylogenetic trees, orthologs and
collaboration with the Gene Ontology Consortium. Nucleic Acids
Res. Jan 2010;38(Database issue):D204–D210.
93. Mi H, Guo N, Kejariwal A, Thomas PD. PANTHER version
6:  protein sequence and function evolution data with expanded
representation of biological pathways. Nucleic Acids Res. Jan
2007;35(Database issue):D247–D252.
94. Huang da W, Sherman BT, Lempicki RA. Systematic and inte-
grative analysis of large gene lists using DAVID bioinformatics
resources. Nat Protoc. 2009;4(1):44–57.
95. Limited SLSP. Avadis NGS. 2012. Accessed Sept 20, 2012.
96. Systems I. Ingenuity Pathway Analysis (IPA). 2012; http://www.
ingenuity.com/products/ipa.
97. Alkan C, Sajjadian S, Eichler EE. Limitations of next-generation
genome sequence assembly. Nat Methods. 2011;8(1):61–65.
98. Ariyaratne PN, Sung WK. PE-Assembler: de novo assembler using
short paired-end reads. Bioinformatics. 2011;27(2):167–174.
99. Butler J, MacCallum I, Kleber M, et  al. ALLPATHS:  de novo
assembly of whole-genome shotgun microreads. Genome Res.
2008;18(5):810–820.
100. Dohm JC, Lottaz C, Borodina T, Himmelbauer H. SHARCGS, a
fast and highly accurate short-read assembly algorithm for de novo
genomic sequencing. Genome Res. 2007;17(11):1697–1706.
101. Gao S, Sung WK, Nagarajan N. Opera:  reconstructing optimal
genomic scaffolds with high-throughput paired-end sequences.
J Comput Biol. 2011;18(11):1681–1691.
102. Hossain MS, Azimi N, Skiena S. Crystallizing short-read assem-
blies around seeds. BMC bioinformatics. 2009;10(Suppl 1):S16.
103. Jeck WR, Reinhardt JA, Baltrus DA, et  al. Extending assem-
bly of short DNA sequences to handle error. Bioinformatics.
2007;23(21):2942–2944.
 104. Li R, Fan W, Tian G, et al. The sequence and de novo assembly of
the giant panda genome. Nature. 2010;463(7279):311–317.
105. Narzisi G, Mishra B. Comparing de novo genome assembly:  the
long and short of it. PloS one. 2011;6(4):e19175.
106. Narzisi G, Mishra B. Scoring-and-unfolding trimmed tree assem-
bler:  concepts, constructs and comparisons. Bioinformatics.
2011;27(2):153–160.
107. Schmidt B, Sinha R, Beresford-Smith B, Puglisi SJ. A fast
hybrid short read fragment assembly algorithm. Bioinformatics.
2009;25(17):2279–2280.
108. Simpson JT, Wong K, Jackman SD, Schein JE, Jones SJ, Birol I.
ABySS: a parallel assembler for short read sequence data. Genome
Res. 2009;19(6):1117–1123.
109. Warren RL, Sutton GG, Jones SJ, Holt RA. Assembling mil-
lions of short DNA sequences using SSAKE. Bioinformatics.
2007;23(4):500–501.
110. Zerbino DR, Birney E. Velvet: algorithms for de novo short read
assembly using de Bruijn graphs. Genome Res. 2008;18(5):821–829.
111. Luo R, Liu B, Xie Y, et al. SOAPdenovo2: an empirically improved
memory-efficient short-read de novo assembler. GigaScience.
2012;1(1):18.
112. Arbuckle JL. Amos 7.0 User’s Guide. 2007.
113. CLCBio. CLC Workbench. 2012; www.clcbio.com.
114. Margulies M, Egholm M, Altman WE, et  al. Genome sequenc-
ing in microfabricated high-density picolitre reactors. Nature.
2005;437(7057):376–380.
115. Chevreux B. Using the miraEST assembler for reliable and auto-
mated mRNA transcript assembly and SNP detection in sequenced
ESTs. Genome Res. 2004;14:13.
116. Gnerre S, Maccallum I, Przybylski D, et  al. High-quality draft
assemblies of mammalian genomes from massively parallel sequence
data. Proc Natl Acad Sci U S A. 2011;108(4):1513–1518.
117. Boetzer M, Henkel CV, Jansen HJ, Butler D, Pirovano W.
Scaffolding pre-assembled contigs using SSPACE. Bioinformatics.
2011;27(4):578–579.
118. Pop M. Hierarchical scaffolding with Bambus. Genome Res.
2004;14(1):11.
119. Au KF, Underwood JG, Lee L, Wong WH. Improving PacBio
long read accuracy by short read alignment. PloS one. 2012;7(10):
e46679.
120. Assembler C. Whole-genome Shotgun DNA Sequence Assembler.
2012; http://sourceforge.net/apps/mediawiki/wgs-assembler/
index.php?title=Main_Page.
121. Koren S, Schatz MC, Walenz BP, et  al. Hybrid error correction
and de novo assembly of single-molecule sequencing reads. Nat
Biotech. 2012;30(7):693–700.
122. Boetzer M, Pirovano W. Toward almost closed genomes with
GapFiller. Genome Biol. 2012;13(6):R56.
123. Tsai IJ, Otto TD, Berriman M. Improving draft assemblies by
iterative mapping and assembly of short reads to eliminate gaps.
Genome Biol. 2010;11(4):R41.
124. A gap filling and re-scaffolding model for improved genome assem-
bly. Ganit Labs, Bio-IT Centre, Institute of Bioinformatics and
Applied Biotechnology; Unpublished Data.
125. Schatz MC, Delcher AL, Salzberg SL. Assembly of large genomes
using second-generation sequencing. Genome Res. 2010;20(9):
1165–1173.
126. Gruenheit N, Deusch O, Esser C, Becker M, Voelckel C, Lockhart
P. Cutoffs and k-mers: implications from a transcriptome study in
allopolyploid plants. BMC genomics. 2012;13:92.
127. Robertson G, Schein J, Chiu R, et al. De novo assembly and analy-
sis of RNA-seq data. Nat Methods. 2010;7(11):909–912.
128. Nagarajan N, Pop M. Sequence assembly demystified. Nat Rev
Genet. 2013;14(3):157–167.
129. Grabherr MG, Haas BJ, Yassour M, et al. Full-length transcriptome
assembly from RNA-Seq data without a reference genome. Nat
Biotech. 2011;29(7):644–652.
B ioin f or m atic s , Sys t e m s B io l ogy, and Sys t e m s M e dicin e   •   9 5
130. Schulz MH, Zerbino DR, Vingron M, Birney E. Oases: robust de
novo RNA-seq assembly across the dynamic range of expression
levels. Bioinformatics. 2012;28(8):1086–1092.
131. Xie Y, Wu W, Tang J, et al. SOAPdenovo-Trans: De novo transcrip-
tome assembly with short RNA-Seq reads. arXiv:1305.6760v2
[q-bio.GN]. 2013.
132. TrinityTeam. Genome-guided Trinity. Vol 20132013.
133. Guttman M, Garber M, Levin JZ, et  al. Ab initio reconstruc-
tion of cell type-specific transcriptomes in mouse reveals the
conserved multi-exonic structure of lincRNAs. Nat Biotech.
2010;28(5):503–510.
134. Trapnell C, Roberts A, Goff L, et al. Differential gene and tran-
script expression analysis of RNA-seq experiments with TopHat
and Cufflinks. Nat Protoc. 2012;7(3):562–578.
135. Martin JA, Wang Z. Next-generation transcriptome assembly. Nat
Rev Genet. 2011;12(10):671–682.
136. Jain P, Krishnan NM, Panda B. Augmenting transriptome assem-
bly combinatorially. arXiv:1305.6259v22013.
137. Koski LB, Gray MW, Lang BF, Burger G. AutoFACT:  an auto-
matic functional annotation and classification tool. BMC bioin-
formatics. 2005;6:151.
138. Suzek BE, Huang H, McGarvey P, Mazumder R, Wu CH.
UniRef:  comprehensive and non-redundant UniProt reference
clusters. Bioinformatics. 15 2007;23(10):1282–1288.
139. Kanehisa M, Goto S, Kawashima S, Okuno Y, Hattori M. The
KEGG resource for deciphering the genome. Nucleic Acids Res.
Jan 1 2004;32(Database issue):D277–D280.
140. Kanehisa M, Goto S, Sato Y, Furumichi M, Tanabe M. KEGG for
integration and interpretation of large-scale molecular data sets.
Nucleic Acids Res. 2012;40(1):D109–D114.
141. Tatusov RL, Galperin MY, Natale DA, Koonin EV. The COG
database: a tool for genome-scale analysis of protein functions and
evolution. Nucleic Acids Res. 2000;28(1):33–36.
142. Tatusov RL, Fedorova ND, Jackson JD, et  al. The COG data-
base: an updated version includes eukaryotes. BMC bioinformat-
ics. 2003;4:41.
143. Moriya Y, Itoh M, Okuda S, Yoshizawa AC, Kanehisa M. KAAS: an
automatic genome annotation and pathway reconstruction server.
Nucleic Acids Res. 2007;35(Web Server issue):W182–W185.
144. Punta M, Coggill PC, Eberhardt RY, et  al. The Pfam protein
families database. Nucleic Acids Res. Jan 2012;40(Database
issue):D290–D301.
145. Smit AFA, Hubley R. Repeat Masker. 1989; http://www.repeat-
masker.org. Accessed October 12, 2011.
146. Smit AFA, Hubley R. Repeat Modeller. 1989; http://www.repeat-
masker.org. Accessed December 12, 2011.
147. Xu Z, Wang H. LTR_FINDER: an efficient tool for the predic-
tion of full-length LTR retrotransposons. Nucleic Acids Res. Jul
2007;35(Web Server issue):W265–W268.
148. Han Y, Wessler SR. MITE-Hunter:  a program for discovering
miniature inverted-repeat transposable elements from genomic
sequences. Nucleic Acids Res. 2010;38(22):e199.
149. Price AL, Jones NC, Pevzner PA. De novo identification of
repeat families in large genomes. Bioinformatics. 2005;21(Suppl
1):i351–i358.
150. Jacob F. Evolution and tinkering. Science. 1977;196(4295):
1161–1166.
151. Bateman A, Birney E, Cerruti L, et al. The Pfam protein families
database. Nucleic Acids Res. 2002;30(1):276–280.
152. Thompson JD, Gibson TJ, Higgins DG. Multiple sequence align-
ment using ClustalW and ClustalX. Current protocols in bioin-
formatics / editoral board, Andreas D. Baxevanis. . . [et al.]. Aug
2002;Chapter 2:Unit 2.3.
153. Lio P, Goldman N. Models of molecular evolution and phylogeny.
Genome Res. Dec 1998;8(12):1233–1244.
154. Huelsenbeck JP, Ronquist F. MRBAYES:  Bayesian inference of
phylogenetic trees. Bioinformatics. 2001;17(8):754–755.
 155. Wilgenbusch JC, Swofford D. Inferring evolutionary trees with
PAUP*. Current protocols in bioinformatics/editoral board,
Andreas D. Baxevanis . . . [et al.]. Feb 2003;Chapter 6:Unit 6.4.
156. Stamatakis A, Hoover P, Rougemont J. A rapid bootstrap algorithm
for the RAxML Web servers. Syst Biol. 2008;57(5):758–771.
157. Anisimova M, Gascuel O. Approximate likelihood-ratio test for
branches:  a fast, accurate, and powerful alternative. Syst Biol.
2006;55(4):539–552.
158. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S.
MEGA5: molecular evolutionary genetics analysis using maximum
likelihood, evolutionary distance, and maximum parsimony meth-
ods. Mol Biol Evol. 2011;28(10):2731–2739.
159. Felsenstein J. PHYLIP—Phylogeny Inference Package (Version
3.2). Cladistics. 1989;5:164–166.
160. Zuker M. Mfold web server for nucleic acid folding and hybrid-
ization prediction. Nucleic Acids Res. 2003;31(13):3406–3415.
161. Ding Y, Chan CY, Lawrence CE. Sfold web server for statistical
folding and rational design of nucleic acids. Nucleic Acids Res. Jul
1 2004;32(Web Server issue):W135–W141.
162. Lowe TM, Eddy SR. tRNAscan-SE:  a program for improved
detection of transfer RNA genes in genomic sequence. Nucleic
Acids Res. 1997;25(5):955–964.
163. Perez A, Luque FJ, Orozco M. Frontiers in molecular dynamics
simulations of DNA. Acc Chem Res. 2012;45(2):196–205.
164. Jonikas MA, Radmer RJ, Laederach A, et al. Coarse-grained mod-
eling of large RNA molecules with knowledge-based potentials
and structural filters. RNA. 2009;15(2):189–199.
165. Cech P, Svozil D, Hoksza D. SETTER: web server for RNA struc-
ture comparison. Nucleic Acids Res. Jul 2012;40(Web Server
issue):W42–W48.
166. Cong Q, Grishin NV. MESSA: MEta-Server for protein Sequence
Analysis. BMC biology. 2012;10:82.
167. Lin K, Simossis VA, Taylor WR, Heringa J. A simple and fast
secondary structure prediction method using hidden neural net-
works. Bioinformatics. 2005;21(2):152–159.
9 6   •   P rincip l e s o f G e no m ic M e dicin e
168. Altschul SF, Madden TL, Schaffer AA, et al. Gapped BLAST and
PSI-BLAST:  a new generation of protein database search pro-
grams. Nucleic Acids Res. 1997;25(17):3389–3402.
169. Beckstette M, Homann R, Giegerich R, Kurtz S. Fast index based
algorithms and software for matching position specific scoring
matrices. BMC bioinformatics. 2006;7:389.
170. Gregersen N, Bross P, Vang S, Christensen JH. Protein misfolding and
human disease. Annu Rev Genom Hum Genet. 2006;7:103–124.
171. Team FH. Folding@home. 2012; folding.stanford.edu/home.
172. Agile Molecule. Abalone. 2012; http://www.biomolecular-modeling.
com/Abalone/index.html.
173. Agile Molecule. Ascalaph Designer. 2012; http://www.
biomolecular-modeling.com/Ascalaph/Ascalaph_Designer.html.
174. Accelrys Software Inc. Discovery Studio Modeling Environment.
2012; Release 3.5, 2012.
175. Graur D, Zheng Y, Price N, Azevedo RB, Zufall RA, Elhaik E. On
the immortality of television sets: “function” in the human genome
according to the evolution-free gospel of ENCODE. Genome Biol
Evol. 2013;5(3):578–590.
176. Kandoth C, McLellan MD, Vandin F, et  al. Mutational land-
scape and significance across 12 major cancer types. Nature.
2013;502(7471):333–339.
177. Cline MS, Craft B, Swatloski T, et al. Exploring TCGA Pan-Cancer
data at the UCSC Cancer Genomics Browser. Sci Reports.
2013;3:2652.
178. Haber DA, Settleman J. Cancer:  drivers and passengers. Nature.
2007;446(7132):145–146.
179. Aggarwal BB, Ichikawa H, Garodia P, et  al. From traditional
ayurvedic medicine to modern medicine: identification of thera-
peutic targets for suppression of inflammation and cancer. Expert
Opin Ther Tar. 2006;10(1):87–118.
180. Garodia P, Ichikawa H, Malani N, Sethi G, Aggarwal BB. From
ancient medicine to modern medicine:  ayurvedic concepts of
health and their role in inflammation and cancer. J Soc Integr
Oncol. 2007;5(1):25–37.
 7.
PHARMACOGENOMICS—CRITICAL COMPONENT
OF GENOMIC MEDICINE
Wolfgang Sadee
INTRODUCTION:
P H A R M AC O G E N ET I C S
A N D   P H A R M AC O G E N O M I C S
The role of genetic factors that determine how we interact
with exogenous substances was realized early in the twen-
tieth century, but it took several decades before these insights
became relevant to drug therapy. Identifying inborn errors
of metabolism as a genetic basis of some diseases in 1909,
Archibald Garrod presciently concluded that “every active
drug is a poison, when taken in large enough doses, and in
some subjects a dose which is innocuous to the majority of
people has toxic effects, whereas others show exceptional
tolerance to the drug.” In 1932, Arthur L. Fox found that
phenylthiocarbamide—a component of some vegetables
such as broccoli—has a bitter taste for most people, but
some do not taste it at all. Ability to taste this compound
varied dramatically by ethnicity, presaging the importance
of the patient’s ethnic origin in drug response. Later studies
revealed mutations in the bitter-taste receptor TAS2R38 to
account for such differences, presumably a result of envir-
onmental selection pressures.1 Similarly, lactose intolerance
is widespread in Asian populations, but greatly reduced in
other populations, in parallel with increasing use of dairy
products—demonstrating strong evolutionary pressures
that contribute to genetic causes of variable responses to
ingested substances, either toxins or nutrients—a topic
that bears upon the general relevance of evolution in public
health.2 Inadvertently, such genetic variants can modu-
late drug effects—with large numbers of subjects exposed
to thousands of new chemicals over the past century. One
might view drug therapy as a worldwide experimental stage
that can be exploited to probe for genetic differences im-
portant to humans’ interactions with the environment.
The main concepts in pharmacogenetics—so termed by
Friedrich Vogel in 1959—emerged in the 1950s and 1960s,
anchored by discoveries of dramatic differences in drug
97
response as a result of genetic deficiencies. These include slow
acetylation of isoniazid because of low N-acetyltransferase
activity (NAT2) activity, primaquine hemolysis resulting
from deficient glucose-6-phosphate dehydrogenase (G6PH),
and succinylcholine apnea caused by plasmacholinesterase
deficiency. Somewhat later, in 1975, a quintessential pharma-
cogenetic deficiency was identified in cytochrome P4502D6
(CYP2D6), with strong adverse reactions to debrisoquine
and sparteine (for reviews, see 3,4). With approximately 7% of
Caucasians being poor CYP2D6 metabolizers, numerous ad-
verse drug reactions may have been caused by CYP2D6 null
mutations,5 so that today, drug discovery seeks novel chemical
entities that are not metabolized by CYP2D6. In the 1980s
and 1990s, many of the genes encoding drug-metabolizing
enzymes were cloned, enabling the search for mutations and
raising the possibility that genotyping could be performed
to avoid adverse drug reactions (ADRs) or select the proper
drug for the right patient.
These seminal discoveries have dominated the direction
of pharmacogenetics, with the expectation that single gen-
etic variants are a main cause of different drug response. And
possibly this expectation was justified, at least with respect
to adverse drug reactions. Although these are expected to be
infrequent events—otherwise the drug would not be clinic-
ally useful for most indications—the likelihood is neverthe-
less high that a single mutation causes or strongly enhances
the risk of an ADR. For example, a unique HLA allele can
dramatically enhance the risk of rare but serious drug-related
injuries, also referred to as “idiosyncratic drug reactions”
(IDRs). Much attention has focused on the numerous genes
encoding drug-metabolizing enzymes, some carrying fre-
quent null mutations. Indeed, ADRs have been identified
as a prominent cause of death and morbidity in the United
States. We have found that the main ADR-causing drugs
are metabolized by polymorphic enzymes,5 suggesting that
ADRs—defined as toxic drug effects that occur upon appro-
priate drug therapy but are deemed “unavoidable”—could in
fact be avoided or at least reduced if we were to better under-
stand genetic differences between individuals.
With the emergence of genomics, our view of genetic
factors in disease and therapy broadened to account for
biological complexity on an ever-increasing scale. Beyond
monogenic Mendelian diseases, common disorders such as
cardiovascular and central nervous system (CNS) disorders,
diabetes, and cancer are considered to have polygenic origins.
Moreover, gene–environment interactions are critical deter-
minants of well-being and disease, with epigenetic modula-
tion of cellular regulation playing a key role. In contrast to
a less complex genetic architecture of ADRs, we therefore
expect drug efficacy to depend on complex processes so that
variability in drug response is multifaceted—with genetic
factors contributing to a different extent in each specific
drug therapy. With functionally relevant genetic variants
largely uncertain,6 known genetic factors in drug response
often account for only a small portion of variability, reduc-
ing the value of predictive clinical biomarker tests. Since
drug efficacy in the treatment of complex disorders ranges
from 20–80%, a substantial portion of the patient popula-
tion fails to benefit while nevertheless being exposed to risk
of ADRs. Understanding the causes underlying the failure
of a patient to respond to a specific drug therapy would yield
a quantum leap in optimal care and prevention.
In view of the complex biology impinging on drug re-
sponse, it became compelling to embrace the emerging
genomics concepts and technologies:  most prominently,
large-scale sequencing, leading to a transformation of
pharmacogenetics into pharmacogenomics. Also, concepts
guiding the study of the genomic DNA sequence, its epigen-
etic modifications and chromatin regulation, and the tran-
scriptome, proteome, and metabolome, merged and indeed
became inseparably intertwined. Therefore, pharmacogen-
omics now is understood to encompass all these areas, inte-
grating complex patterns in both genotype and phenotype,
in an attempt to optimize and individualize drug therapy.
In the context of genomic medicine, pharmacogenomics has
assumed a critical role in implementing genomic principles
in clinical practice. In his article “Genomics, Health Care,
and Society,” Hudson states:  “One of the most promising
areas of genomic medicine is the ability to match an individ-
ual’s genetic profile to the likely effect of particular drugs.”7
R E S E A R C H A N D D E VE L O PM E N T
I N P H A R M AC O G E N O M I C S
Early pharmacogenetics studies focused on indi-
vidual variability in drug response and toxicity, while
9 8   •   P rincip l es o f G enomic M edicine
pharmacogenomics was initially employed in the pharma-
ceutical industry to accelerate drug target discovery. With
genome sequencing on the horizon, researchers expected
to find the genes and their protein products critical to the
disease process, and hence countless suitable drug targets.
However, the complexity of human diseases has foiled
rapid discovery of valid targets and drugs, leading to a
reemergence of multi-target and multi-drug therapies—
surreptitiously confounding attempts to understand drug
response at the genetic level. On the other hand, upon
growing adoption of genomics concepts, pharmacogen-
etics morphed into pharmacogenomics with the expressed
goal of optimizing personalized drug therapy, including
multi-component therapies. Therefore, pharmacogenomics
has become an integral part of the burgeoning “personal-
ized medicine,” or “personalized health care”—the latter
emphasizing the importance of early intervention and pre-
vention. A deeper understanding of the causes underlying
variable drug response has the potential to minimize ADRs
and maximize efficacy; therefore, the use of biomarker tests
with drug therapy, or for that matter any other intervention
or preventive measure, emerges as a main goal, predicting
an individual’s disease risk and response, and guiding op-
timal therapeutic strategies. Table 7.1 summarizes the main
goals of pharmacogenomics, leading to biomarkers based
on DNA sequence variants, but also other -omics panels,
such as transcriptomics, proteomics, and metabolomics
(Figure  7.1). It is noted that such -omics biomarkers rep-
resent phenotypes rather than genotypes but are neverthe-
less included with the term “pharmacogenomics”. Because
all these areas cannot be adequately addressed here, this
chapter will focus on genetic factors predicting drug re-
sponse and toxicity, arguably one central focus of current
research. Epigenetic factors and gene–environment interac-
tions, and also phenotypical markers derived from RNA,
protein, and metabolite profiles, will be addressed where
appropriate.
P H A R M AC O G E N O M I C S —BA S I C
R E S E A RC H
Distinct from medical genetics/genomics, pharmacogen-
omics largely employs drug-related phenotypes, including
pharmacokinetic parameters (PK; metabolism, transport,
clearance, half-life, protein binding, etc.) and pharmaco-
dynamics (PD; drug response and toxicity). On the other
hand, differential diagnosis of a disease profile can also
guide selection of the drug class, showing that medical
issues and pharmacogenomics cannot be neatly separated.
Moreover, recorded family histories typically cover disease
Table 7.1  GOALS AND CONCEPTS, METHODS, AND DELIVERABLES OF PHARMACOGENOMICS FOR EACH AP-
PLICATION, IT IS IMPERATIVE THAT MOLECULAR GENETIC STUDIES BE CARRIED OUT TO VALIDATE THE
PRESUMED FUNCTIONAL POLYMORPHISM/MUTATION/PHENOTYPICAL MARKERS AND UNDERSTAND THE
MOLECULAR AND CELLULAR CONTEXT (E.G.: PROMOTER/ENHANCER MUTATIONS DEPEND ON THE PRES-
ENCE OF TRANSCRIPTION FACTORS AND THE EPIGENETIC MACHINERY THAT IS PRESENT IN THE TARGET
TISSUE). ADRS MAY BE CAUSED BY EXCESSIVE DRUG EXPOSURE ACTING AT THE INTENDED TARGET,
OFTEN A RESULT OF POOR METABOLIZER STATUS, BUT OFF-TARGET EFFECTS ARE ALSO COMMON, SUCH
AS TORSADE DE POINTES (VENTRICULAR TACHYCARDIA) UPON DRUG BINDING TO CARDIAC CHANNELS
INVOLVED IN CARDIAC PACING (LONG QT SYNDROME, A MAIN PROBLEM LEADING TO FAILURE OF INVES-
TIGATIONAL DRUGS).

GOALS AND CONCEPTS APPROACH AND METHODS DELIVERABLES

Drug target identification All genomics techniques, systems biology, trans-
genic animal models, etc.
Predict adverse drug reactions Genotyping-sequencing and clinical association
(ADRs) studies to detect variants/mutations, often in
drug-metabolizing enzymes and transporters
Predict idiosyncratic drug reac- Large-scale clinical association studies to detect
tions (IDRs) highly penetrant rare variants
Predict drug response (efficacy) Clinical association studies with drug effect as
phenotype; often involving multiple genes (drug
targets and pathways); systems biology
Monitor drug response Measure phenotypical changes in target tissues;
detect residuals aberrant/transformed cells with
genotyping assays
Early treatment and prevention Determine genetic predisposition to disease; risk
factor analysis; factors predicting long-term re-
sponse to drug therapy and prophylaxis; system
biology
Druggable protein or other target critical to the disease
process
Genetic biomarker tests involving relatively frequent
mutations, predicting pharmacokinetic parameters or
drug [in]activation
Genetic biomarker tests involving single mutations in
key off-target genes (such as HLAs)
Multigenic biomarker tests and other genomics bio-
marker panels (e.g., mRNA, protein profiles), predict-
ing pharmacodynamic parameters
Phenotypical biomarker tests (RNA, protein, metab-
olite profiles); genotype test detecting residual cells
carrying mutations (oncogenic) as a measure of therapy
response
Multigenic biomarker tests predicting disease risk and
type of effective drug treatment; genomics biomarker
panels as possible indicators of long-term risk (e.g.,
mRNA, protein lipid profiles)

Pharmacogenomics biomarkers
phenotype versus genotype
almost exclusively, with drug-effect GWAS results emerg-
ing only recently.
Pharmacogenetics It is perhaps instructive to consider the nature of genetic
Transcriptomics
Sequence variations
mRNA profiles variability associated with disease risk and drug response. On
(genotype)
the basis of traditional genetic studies and GWAS results, a
common assumption is that highly penetrant disease-risk
variants are rare—being under negative evolutionary selec-
Proteomics Metabolomics tion pressure—while numerous gene variants exerting low
protein profiles metabolite profiles
penetrance can be frequent and associate with complex dis-
orders (Figure 7.2). The latter have been revealed by GWAS,
Figure 7.1 
Main types of pharmacogenomics biomarkers. Note that only pointing to the pathways involved in the disease process;
the genomics/genetics panel represents a genotype; the other panels are however, assuming additive effects of all known genetic vari-
phenotypes. Biomarkers derived from these areas vary in their clinical
applications. ants associated with a complex disorder, one accounts only
for a small portion of the estimated high degree of herit-
ability in a population. This gap in our current knowledge
incidence but not heritable response to drug therapy. The is called the “missing heritability,”11 also applicable to drug
strong heritability often associated with drug response response phenotypes—we often do not know as yet the
therefore is not readily obvious from family histories or degree of heritability of a drug response trait, and which
linkage studies, but an assessment has to rely on newly genetic factors contribute, even when environmental inter-
designed studies of drug therapies in target populations, or actions are factored in. We can look for two possible paths
involving twin studies.8–10 Also, genome-wide association to clarifying the “missing heritability.” First, using mathem-
studies (GWAS) initially targeted disease phenotypes atical modeling, Lander and colleagues12 have proposed that

P harmacogenomics — C ritica l C omponent o f G enomic M edicine   •   9 9

 Frequency and penetrance of genetic variants Types of functional polymorphisms
in complex traits
Polymorphisms Regulatory Structural RNA
in coding region polymorphyisms polymorphisms
high
Rare, cSNPs rSNPs srSNPs
high
impact
Effect size

??

Frequent,
low Altered protein Altered mRNA
Altered
impact sequence and processing and
transcription
function translation
low

low high
Allele frequency Figure 7.3 
The nature of genetic variation with phenotypical
consequences. “SNPs” is used here to represent all sequence variations,
Figure 7.2 
Frequency and penetrance of genetic variants in complex traits. including SNPs, insertion/deletions, copy number variants, and
We postulate the existence of frequent variants (mostly regulatory) that chromosomal rearrangements, such as inversions and translocations. For
have accumulated as a result of genetic drift or under positive selection more information, see Sadee et al.6
pressure, and therefore are not primary risk factors per se—only under
specific environmental challenges and with amplified epistatic effects.
Early genetic studies have focused on non-synonymous
SNPs (cSNPs) that alter a protein’s function directly.
epistatic gene–gene interactions have the potential to fill the
However, regulatory variants appear to be more frequent,
gap, thereby moving away from the common additive mod-
affecting gene expression (rSNPs) and RNA functions,
els adopted from GWAS results considering many candidate
such as elongation, splicing, turnover, editing, cellular traf-
single nucleotide polymorphisms [SNPs] each with low
ficking, and translation. We have termed these latter vari-
penetrance. The authors further predict that search for add-
ants “structural RNA SNPs” (srSNPs) (Figure  7.3).7,13 Of
itional variants will account only for a rather small fraction
course, sequence variations also include repeats, insertions/
of the missing heritability under additive model assumption.
deletions, inversions, and translocations, etc., and the term
Second, I propose the existence of frequent polymorphisms
“SNP” is used here for the sake of simplicity (unless speci-
with relatively high penetrance13; however, the affected phe-
fied otherwise). The effect of regulatory variants is subject
notypes may not be directly related to disease risk, except
to the cellular, tissue, whole body, and external environ-
when an individual is exposed to specific environmental
ment, opening many pathways for evolutionary pressures
conditions, including drug exposure, that reveal the genetic
to select optimal conditions for physiological target tissues
effect. Such variants can be frequent because of population
while avoiding deleterious effects in other tissues; hence
substructures, such as bottleneck populations with founder
we can expect some regulatory variants to be frequent
mutations, or they could have been under positive evolu-
without showing obvious disease-risk phenotypes, with
tionary selection because of some reproductive advantage;
effects that are restricted to target organs. I will discuss how
clearly, these genetic variants are unlikely to confer strong
cSNPs and regulatory variants affect drug response and
disease risk and therefore typically are missed by GWAS.
ADRs. Some mutations may be rare with high penetrance,
Multiple studies support the notion that countless regions
leading to idiosyncratic drug reactions; others are quite fre-
of the human genome bear evidence of positive selection,14
quent, affecting ADRs and PK/PD parameters (Table 7.1).
indicating selectable features with phenotypical pene-
Each of these drug phenotypes requires different experi-
trance. Yet, such genetic variants could act as a double-edged
mental approaches to understand the underlying molecular
sword: given certain environmental conditions, the genetic
and genetic mechanisms, and clinical implications (the
variants can turn deleterious, particularly when effects are
Pharmacogenomics Knowledge Base PharmGKB [https://
amplified epistatically (non-linear gene–gene interactions;
www.pharmgkb.org/index.jsp] provides detailed informa-
examples will be provided further below). Drug therapy is
tion on drug–gene interactions).
one such environmental condition that humans have only
recently been exposed to on a massive scale, and therefore,
can reveal hidden variants of strong effect. A case in point is
Discovery of Variants in Genes Relevant to Drug
the frequent mutations in drug-metabolizing enzymes that
Response and Toxicity
can dramatically increase drug exposure in poor metabo-
lizers, or affect their response to environmental toxins and We must tailor the approach to identifying variants with clin-
carcinogens. ical pharmacological relevance to the problem at hand. First,

1 0 0   •   P rincip l es o f G enomic M edicine

one estimates the frequency of a drug-related phenotype in
the treated population; if only 60% of patients respond favor-
ably and one suspects a SNP with 1% minor allele frequency
to affect drug response, only a small portion of the variability
can be accounted for. On the other hand, if severe adverse
drug reactions occur only in one of 1,000 subjects, a frequent
allele is unlikely to be the cause, although it could increase
the risk to a limited extent. Expected allele frequency deter-
mines the approach to be taken, and the population size and
structure needed to resolve the genetic factors. As discussed
earlier, drug exposure can reveal strong effects of frequent
mutations in genes that do not immediately affect disease
(whether acquired by random population drift or by posi-
tive selection pressure). Genes encoding drug-metabolizing
enzymes and transporters often carry such frequent variants,
which have emerged as important pharmacogenetic factors
in drug therapy. Relatively rare alleles in the HLA cluster of
genes, on the other hand, have been associated with severe
IDRs; for example, of carbamazepine.15
There is some debate over whether rare alleles in genes
already shown to harbor frequent strong variants could
contribute substantially to the variability of treatment out-
comes (see Figure 7.2), but few studies have addressed this
issue directly. One such study, on methotrexate clearance,
has explored the contribution of frequent and rare vari-
ants in the hepatic anionic drug transporter SLCO1B1
(OATP), which is known to carry at least two frequent
non-synonymous SNPs.16 Using extensive exome sequenc-
ing, several rare, non-synonymous SNPs were identified
and predicted to have reduced transport function. All
SLCO1B1 variants combined accounted for ~11% of clear-
ance variability, mostly attributable to the known frequent
variants, whereas rare damaging variants accounted for an-
other 2%. Rare variants would have a limited effect in a re-
cessive genetic model, where homozygosity occurs in a very
small number of subjects but is needed for robust impact
on drug transport. Nevertheless, the authors concluded that
the rare variants may be clinically relevant,16 in particular
because the likelihood of compound heterozygosity is high
when combined with a frequent variant. Yet, clinical imple-
mentation of this insight might turn out to be difficult, as
the phasing of two different functional variants is typically
uncertain; if both are on the same haplotype, the other al-
lele is presumed to have wild-type activity, and the overall
effect is limited (unless the variant is dominant negative).
Genome-Wide Association Studies
GWAS studies were initially intended to find frequent vari-
ants with relatively strong impact as risk factors in disease,
P harmacogenomics — C ritica l C omponent o f G enomic M edicine   •   1 0 1
but the vast majority of significant SNPs conveyed only a
small increase in risk, requiring ever larger cohorts to be ana-
lyzed, a trend facilitated by the technological advances of
massively parallel genotyping platforms and bioinformatics
software for analysis. Growing application of GWAS to
drug therapy outcomes has revealed a number of instances
where frequent variants have substantial effect sizes. One
such example is the high odds ratio for suffering from ser-
ious statin-induced myotoxicity (rabdomyelysis) attribut-
able to frequent variants in SLCO1B1, an effect particularly
pronounced with high doses of simvastatin.17 Another case
involves necrolysis upon treatment with carbamazepine, at-
tributable to the HLA-B1502,15 both now included with
drug labels issued by the Food and Drug Administration
(FDA; see below). With very large subject cohorts, even
rare ADRs can be identified; for example, in nearly all
Canadian children covered under a program supported by
Genome Canada.18 Looking ahead, full genome sequencing
is likely to replace other methodologies, with application to
family studies and in large populations, as the cost of whole
genome sequencing will continue to drop sharply. However,
it will take much longer to interpret such data, in particular
with respect to regulatory sequence variation (SNPs, indels,
CNVs, and chromosomal rearrangements). Sequence vari-
ants affecting transcription can be located at large distances
from the immediate gene locus, and therefore are difficult
to detect.6
It is argued that regulatory variants (rSNPs and
srSNPs, Figure  7.3) are probably more frequent than
non-synonymous SNPs7; therefore, several approaches
have been developed to discover them. Combining tran-
scriptome analysis in tissues (typically with hybridiza-
tion micro-arrays but more recently using RNAseq) with
GWA-derived genotype data has proven powerful in detect-
ing regulatory variants. Each mRNA level across tissues
from multiple subjects serves as a quantitative trait used to
identify any SNP locus in the genome significantly associ-
ated with the mRNA profile of that gene (yielding expres-
sion quantitative trait loci, or eQTLs).19 If the associated
SNP is considered to reside in the same gene locus and act
on it in cis, it is termed a cis-eQTL, and those in other gen-
omic regions are called trans-eQTLs. This general approach
has been extended to analyze drug responses by measuring
the transcriptome before and after drug treatment of the
tissue.20 Similar analyses can be performed with protein lev-
els (proteomics), or metabolites (metabolomics), and other
genomic parameters, such as epigenetic marks, chromatin
reactions, non-coding RNAs, etc. Integration of these ele-
ments characterizes the emergence of “functional genomics.”
For example, we have used the National Cancer Institute
 Regulatory variants affecting gene expression
eQTL mapping Compare cis-eQTL mapping
RNA
eQTL
expression
mappingarrays
RNA expression arrays AE measurements
Trans-acting variants cis-acting variants
Non-coding RNAs Protein coding mRNAs
Multiple transcription and polyadenylation sites,
alternative splicing, RNA editing, non-colinear transcripts,
antisense transcripts, RNA trafficking and sequestration,
mRNA at ribosomes and translation
Figure 7.4 
Methodologies for finding regulatory polymorphisms, focusing
on RNA transcriptome analyses (adapted from Sadee et al.6).
(NCI) drug discovery panel of over 50,000 compounds
tested against 60 cancer cell lines to determine associations
between cytotoxic drug potencies and transporter mRNA
expression, revealing many new drug–transporter interac-
tions.21 As regulatory variants are highly tissue-dependent,
performing these analyses in relevant target tissues is im-
portant; however, thus far most studies have adopted trans-
formed B-lymphocytes for which GWA data are already
available. A  variation of the eQTL analysis is to measure
allelic mRNA ratios, where one detects different amounts
of mRNA and mRNA isoforms generated in a target tissue
from one allele over the other.6,13 This is probably associ-
ated with a cis-acting regulatory variant (cis-eQTL), or with
allele-selective epigenetic effects (uneven X-inactivation for
X-linked genes in females; maternal or paternal imprinting,
chromatin remodeling facilitated by a DNA variant, etc.).
Figure  7.4 illustrates this approach, which has led to the
discovery of numerous cis-eQTL with a genotyping SNP
chip capable of determining the allelic mRNA ratios.19 We
are currently exploring the use of next-generation RNAseq
for this purpose, but have previously implemented a rapid
and robust analysis of allelic mRNA ratios in comparison
to allelic gDNA ratios, using a primer extension assay, to
examine select key drug candidate genes for regulatory vari-
ants (Figure  7.5).7,22 This approach has revealed frequent
regulatory variants in drug metabolizing enzymes such as
CYP3A423 and drug targets such as DRD2, DAT, TPH2,
ACE, and CETP, each with robust clinical effects.24–28
Epigenetic Factors in Drug Response
Epigenetic factors are similarly important in regulating drug
response but will not be discussed in detail here. Epigenetic
gene silencing or activation is particularly important in can-
cer and antineoplastic chemotherapy, as epigenetic processes
are germane to cellular transformation and progression.

mRNA

B mRNA (cDNA)

34 35 36 37 38 39 40

DNA

DNA

Figure 7.5 
Detection of allelic mRNA ratios, in comparison to allelic gDNA ratios. mRNA and DNA are extracted from a human tissue (e.g., liver),
PCR amplified, and the alleles labeled with fluorescent deoxyribose trinucleotides. The fluorescent amplicons are then analyzed by capillary
electrophoresis. In this example, the allelic gDNA and mRNA ratios differ twofold, revealing a robust allelic expression imbalance (AEI), indicative
of the presence of regulatory variants in this gene locus (cis-eQTL) in this tissue.

1 0 2   •   P rincip l es o f G enomic M edicine

Table 7.2  GENES (PROTEINS) AND GENE FAMILIES INVOLVED IN PHARMACOKINETICS (PK) AND
PHARMACODYNAMICS (PD)
SOME PK GENE FAMILIES ARE LARGE, SUCH AS THE SOLUTE CARRIERS (SLCS; ~350 GENES), ABC TRANS-
PORTERS (ABCS; 49 GENES), OR CYPS (65 GENES). PD GENES ARE EXTREMELY DIVERSE AND INCLUDE
­R ECEPTOR GENES’ FAMILIES, SUCH AS G PROTEIN COUPLED RECEPTORS (GPCRS, >500), VARIOUS KINASES
(>300), AND NUMEROUS ENZYME FAMILIES OF VARYING SIZE. IDIOSYNCRATIC DRUG REACTIONS (IDRS)
ARE LISTED SEPARATELY, AS THESE ARE OFTEN ATTRIBUTABLE TO SINGLE GENES WITH RARE, HIGHLY
PENETRANT VARIANTS. ONLY A FEW PROMINENT GENES/PROTEINS ARE LISTED.

PK GENES PD GENES IDRS

Drug-metabolizing enzymes Receptors Immune factors
CYPs: CYP1A2, 2A6, 2B6, 2C6, 2C9, 2C19, GPCRs: adrenergic receptors, peptide receptors, HLA gene cluster and alleles-haplotypes
3A4, 3A5 neurotransmitter receptors (glutamate, GABA,
Other redox enzymes dopamine, serotonin, etc.)
Conjugating enzymes catalyzing: Kinases: tyrosine kinases and tyrosine kinase
­glucuronidation, sulfation, GSH1 transfer, receptors, Ser/Thre kinases
acetylation Ion channels: sodium, calcium, potassium
Hydrolases channels
Nuclear receptors
Drug transporters Enzymes ROS6 metabolism
SLCs: SLCO1B1, amino acid and sugar Lipid metabolism Peroxidases, dismutases, etc.
­transporters, etc. Peptidases: ACE4 GSH biosynthesis and transferases
ABCs: ABCB1 (MDR12), ABCC2 (MRP2), Neurotransmitter metabolism: MAOA5 (GTS enzymes)
ABCG23, etc.
Binding proteins Various Inflammatory agents
Plasma proteins: albumin, Signaling molecules: many components of Cytokines and receptors,
alpha-acidglycoprotein, cellular proteins ­receptor signaling cascade many other factors
1
GSH: glutathione; 2 MDR1: multidrug resistance polypeptide 1; 3 BCRP, breast cancer resistance protein; 4 ACE: angiotensin converting enzyme;
5
MAO: ­monoamine oxidases; 6 ROS: reactive oxygen species

Similarly, non-coding RNAs such as microRNAs and lin-
cRNAs are now recognized to play critical roles in all types
of diseases and drug therapies but have yet to advance to the
stage of serving as routine biomarkers.
The following sections focus on genetic sequence varia-
tions, and implications for individualized drug therapy.
I  will survey genetic findings related to pharmacokinetics
and pharmacodynamics (Table 7.2), the latter divided into
drug efficacy and adverse effects (including IDRs). As many
reviews and monographs cover the genes relevant to these
topics, the discussion will focus on overarching principles,
with only a few select examples.
G E N ET I C FAC TO R S I N
P H A R M AC O K I N ET I C S
Drug absorption, distribution, metabolism, and excretion
(ADME) varies with health status, sex, age, body weight,
diurnal and seasonal rhythms, and environmental condi-
tion (for example, xenobiotics inducing drug-metabolizing
enzymes). In the introduction of this chapter, I  have dis-
cussed how genetic factors can have dramatic effects on drug
response; however, to what extent genetics (and epigen-
etics) contributes to overall drug response remains a topic of
debate. Most prominent genes encoding drug-metabolizing
enzymes and transporters have been intensely studied,
while regulatory variants still are insufficiently explored.
Currently, SNP genotyping panels are commercially avail-
able, containing hundreds to thousands of ADME vari-
ants in numerous genes, but one suspects that these panels
only account for a portion of the genetic variability; more-
over, for the vast majority of the genotyped SNPs, a clear
understanding of their function and clinical impact is still
pending. While clinical association studies require rigorous
replication, molecular genetics studies are often taken at
face value before the underlying mechanisms fully under-
stood. Moreover, the prevailing use of surrogate markers ra-
ther than validated causative variants in clinical biomarker
tests introduces additional error, lowering any predictive
value, even though we do have the research tools to fully
understand the underlying biology—an effort that should
be of high importance.
Drug Metabolizing Enzymes (DMEs)
Several metabolizing enzymes are critical for drug elimin-
ation from the body, as most drugs are lipophilic and there-
fore not readily excreted into the urine, because of tubular

P harmacogenomics — C ritica l C omponent o f G enomic M edicine   •   1 0 3

reabsorption. Metabolic oxidation followed by conjugation
to hydrophilic products is the typical pathway by which
drugs are excreted. However, oxidative metabolism can also
yield reactive intermediates that are toxic and contribute to
ADRs, including idiosyncratic ADRs, IDRs, and cancer
risk. Early pharmacogenetic studies have revealed abun-
dant mutations in DMEs, including null alleles, far above
the mean for genes under more stringent internal selection
pressures. Such variety could reflect the selective pressure
of varying environmental exposure that differs between
regions. Whether and to what extent a mutation in a DME
gene affects a drug’s clearance, and hence the exposure of
an individual to a given drug dose, depends on the com-
plete mode of elimination; drugs can be substrates for mul-
tiple enzymes or follow alternative excretion paths, such as
urinary excretion, thereby diminishing the impact of muta-
tions in a single gene. Drug–drug interactions must also be
considered, since patients often take multiple drugs that
could interfere with each other. For example, an antidepres-
sant metabolized equally by both CYP2D6 and 2C19 is not
expected to have drastically reduced clearance in CYP2D6
poor metabolizers (i.e., homozygous carriers of two null
alleles), but upon co-administration of a CYP2C19 in-
hibitor, such as omeprazole, a severe reduction in clear-
ance can lead to serious ADRs (see FDA drug labeling for
clopidogrel [which must be activated by CYP2C19] which
warns against concurrent omeprazole therapy). Similarly,
a GWAS analysis of clopidogrel’s cardiovascular response
yielded highly significant associations with CYP2C19
SNPs, linked to alleles *2 and *3.29 Currently, a genetic
CYP2C19 biomarker test is being introduced to guide the
selection of antiplatelet therapy. Because of the many drug–
protein interactions, a prediction of metabolizer phenotype
status from genotype data differs for each gene–drug pair, a
critical point to consider when applying genetic biomarkers
to drug therapy.
The CYP family of enzymes is arguably the most preva-
lent in drug metabolism, with a large majority of current
drugs affected (~65%). While there are ~65 CYP genes,
only a few are considered critical to drug metabolism
(Table 7.2). Among these, CYP2D6 metabolizes ~15% of
all drugs, but it is the most polymorphic. Approximately
7% of Caucasians are poor metabolizers (carrying two null
alleles), followed by intermediate and extensive metaboliz-
ers, with no clear boundary between the latter two, because
numerous variant alleles alter expression and protein activity
to different extents. Also, other factors play a role, such as
enzyme induction, disease state, etc. The uncertainty about
the distinction between intermediate and extensive metab-
olizer status may have impeded clinical implementation
1 0 4   •   P rincip l es o f G enomic M edicine
of a CYP2D6 biomarker test. Nevertheless, poor metab-
olizer status is well defined genetically and indeed has a
strong impact on the response to drugs that are primarily
CYP2D6 substrates. In particular, several antidepressant
and antipsychotics are mainly CYP2D6 substrates,30 with
mounting evidence that metabolizer status is clinically ac-
tionable. As with a few other genes, CYP2D6 has under-
gone duplication, leading to ultra-rapid metabolizer status
if the additional gene copies are fully functional. Prevalence
of ultra-rapid metabolizers is high in the Horn of Africa
and neighboring countries (up to 30% of the population),
but diminishes upon human migration out of Africa, with
a residual prevalence of under 5% in Europe and America.
Ultra-rapid metabolizers may not benefit from the drug,
or, as in the case of codeine, which must be converted by
CYP2D6 to morphine, they may suffer from overdosing.
There is a rich literature on CYP genetics and its impact
on drug therapy, in some instances viewed in the context
of membrane transporters that work hand-in-hand with
DMEs in determining drug fate in the body.31
Possibly the most pervasive drug-metabolizing enzyme,
affecting nearly half of all drugs, cytochrome P450 3A4
(CYP3A4) is one of the most abundant enzymes in the
liver, with activities in the liver varying over a 30-fold range
between individuals. The literature has been inconsistent in
assigning genetic and environmental factors to the highly
variable enzymatic activity, with some claiming that environ-
mental factors (e.g., enzyme induction) are largely respon-
sible. However, an individual’s basal level and inducibility
are thought to be strongly influenced by genetic factors,8
but no frequent polymorphisms had been reported that can
account for it. In 2011, we reported on the first relatively
frequent SNP, located in intron 6 of CYP3A4 (designated
CYP3A4*22, rs3559937), with allele frequencies of 4–8%,
affecting expression in the liver 1.7–6.2 fold.23 Expression
was also affected by transcription factors involved in en-
zyme induction, such as CAR and PXR, showing that both
genetic and environmental factors interact to determine an
individual’s CYP3A4 metabolizer status. As seen with other
regulatory variants (CYP3A4*22 is an srSNP) (Figure 7.3),
the *22 allele affects expression in the liver but not in the
intestines,23 a tissue selectivity important for understanding
genetic factors in overall drug disposition in the body.
Among conjugation reactions, glucuronidation by
UGT1A1 is a prominent example. A 9/10 repeat poly-
morphism in the promoter region affects gene expression,
with homozygous 9-repeat carriers suffering from Gilber’s
syndrome, caused by elevated bilirubin, a substrate of
UGT1A1. An FDA-approved biomarker test has been
shown to predict increased toxicity of the anticancer drug

p1 p2
1 2 3 4 5 6
IIF
Schematics of the mRNA isoforms generated from the NAT1 gene locus. Several transcription start sites, alternative splicing sites, and
Figure 7.6 
polyadenylation sites generate multiple mRNA isoforms with distinct properties, while the protein coding region is largely unchanged. NAT1 *10
is located in the 3′UTR, enhancing translation, while the *11 haplotype comprises multiple SNPs across the 3′ end of the coding region and the
3′UTR, altering usage of the poly-adenylation sites, thereby also enhancing translation.33
irinotecan (its active metabolite is eliminated by UGT1A1)
in 9-repeat carriers32; however, this effect may only be sig-
nificant if high irinotecan doses are given, so that clinical
utility is still undecided for UGT1A1-irinitecan. Another
common DME reaction is acetylation, catalyzed by
N-acetyltransferases 1 and 2 (NAT1 and NAT2). Showing
a bimodal distribution, low acetylator status enhances effi-
cacy of isonizide in the treatment of tuberculosis, but it can
also enhance toxicity. Rapid NAT2 acetylators may require
more frequent dosing to be effective in tuberculosis therapy.
Genetic variation in NAT1 was less certain, while the *10
and *11 allele had been reported to alter expression of the
enzyme by unknown mechanisms, with associations of risk
of certain cancers—acetylation can lead to both xenobiotic
inactivation or activation to more proximate carcinogens.
Relevant to local effects, NAT1 is expressed in many tissues
(NAT2 mostly in the liver) and metabolizes amines, includ-
ing sulfonamides such as sulfamethoxazole, substrates that
also interact with NAT2. We have determined that the
NAT1*10 and *11 alleles enhance protein expression by
regulating translation, thereby protecting against sulfa-
methoxazole-induced skin rash in HIV/AIDS patients—
an effect only observed in poor NAT2 metabolizers, a rare
example of a gene–gene–drug interaction in the pharmaco-
genomics literature.33 NAT1 also serves as an example of the
diverse mRNA isoforms typically generated from a single
gene locus, as a result of alternative transcription start sites,
alternative splicing events, and alternative polyadenylation
sites (Figure 7.6); indeed, we found that *11 alters polyad-
enylation site usage, there by favoring the mRNA transcript
with enhanced translation capacity.33
Transporter Genes
Numerous transporter genes are involved in pharmaco-
kinetics. These are divided into two main classes, solute
transporters (SLCs) and active ABC transporters (ATP
binding cassette). The former allow facilitated diffusion
along concentration gradients, or are secondarily driven by
P harmacogenomics — C ritica l C omponent o f G enomic M edicine   •   1 0 5
Major form
Short form polyA2
p3 polyA1 polyA3
Long form
7 8 9a 9
IIIF F1 F2 R2 R1
co-substrates (e.g., sodium ions and protons) that can drive a
drug against its concentration gradient. Typically, SLCs are
needed to enhance drug entry into the cell, whereas ABC
transporters are extrusion pumps and therefore can func-
tion as drug resistance factors, including MDR1 (ABCB1),
the first such chemoresistance factor found upregulated
in certain cancer tissues. Because MDR1 is highly preva-
lent in multiple tissues, including the blood–brain barrier,
much research has been devoted to its role and genetic
factors that might alter drug response. We have reported
that a synonymous SNP (C3435T) affects mRNA turn-
over, thereby lowering expression of the mature protein.34
However, the effect is rather small, and countless clinical
studies have yielded conflicting results. A robust effect may
occur in the treatment of HIV/AIDs with agents that are
substrates of MDR1, as the antivirals must enter into lym-
phocytes where MDR1 is expressed. Low expression associ-
ated with the 3435T allele appears to enhance drug efficacy
in some but not all studies.35 One therefore needs to distin-
guish between general PK effects of transporters on overall
clearance and distribution in the body, and specific effects
at target sites, such as lymphocytes and the blood–brain
barrier. MDR1 contributes to a formidable barrier for sub-
strates to enter the brain in appreciable quantities, a char-
acteristic exploited for peripheral use of the potent opioid
loperamide for treatment of diarrhea. Substantial genetic
differences in MDR1 activity in the blood–brain barrier
(BBB) would readily become apparent by unwanted ad-
verse opioid effects such as respiratory depression. The lack
of these events,36 even though loperamide is administered
to countless patients, suggests that low MDR1 transporter
status is rare, or alternatively, that secondary mechanisms
can substitute for MDR1-mediated efflux.
Because of the sheer number of membrane trans-
porter genes, often with low-affinity promiscuous sub-
strate interactions, one would not expect polymorphisms
in single genes to have significant impact on drug distribu-
tion and effect. Nevertheless, a number of examples have
emerged where low-activity variants dramatically alter PK
parameters and drug response. One such example, the inter-
action between SLCO1B1 and simvastatin with respect
to muscle toxicity, has already been discussed earlier.29 As
SLC01B1 is primarily and extensively expressed in hepa-
tocytes, the site of action for statins, two consequences
arise:  statin blood levels are elevated in subjects carrying
low-activity SLCO1B1, with less of the drug reaching the
site of metabolism, and less of the drug is reaching the site
of action, HMGCoA reductase, also in the hepatocyte. As
a result, the statin dose is titrated up to reach cholesterol re-
duction goals, enhancing risk of muscle toxicity. However,
because even small alterations of the chemical structure
change relative affinities to membrane transporters, one
cannot extrapolate these results to all statins without test-
ing each statin analogue individually. Indeed, PK char-
acteristics of several statins are unaffected by SLCO1B1
variants—a pertinent example that each gene–drug pair is
distinct and needs evaluation before becoming clinically
actionable. Clinically, one would recommend against using
simvastatin in low-activity SLCO1B1 carriers, or at least
avoid high doses. The entire drug transporter literature is
voluminous and cannot be reviewed here (see Chapter 31
for interactions with DMEs).
G E N ET I C FAC TO R S I N
P H A R M AC O DY NA M I C S
Drug response is characterized by potency (drug dosage
needed for an effect), efficacy (the extent of the maximum
response), and toxicity. Each of these parameters is sub-
ject to distinct genetic influences causing inter-individual
variability: potency is largely related to genes affecting PK
(see previous discussion), efficacy to genes encoding drug
targets and signaling or metabolic cascades, and toxicity
is subject to multiple factors, including off-target events.
Even after a century of drug discovery, poor efficacy results
in ineffective therapy of a majority of complex disorders in
a substantial portion of patients. Moreover, little progress
has been made in predicting treatment outcomes with suf-
ficient accuracy to compel clinical biomarker implementa-
tion—even though it likely that genetic factors do play a
significant role. In part, this failure results from the multi-
factorial nature of the drug response in a disease state,
which itself is of multigenic origin. On the other hand, we
suspect that a drug’s response involves only a portion of the
biological processes underlying complex disorder, because
the drug target is usually well known. Therefore, a patient’s
positive or poor response to drugs can serve to sort patients
with similar symptoms into subgroups with distinct disease
etiologies, potentially leading to biomarkers with greater
1 0 6   •   P rincip l es o f G enomic M edicine
predictive value for drug response compared to disease risk.
Considering the multigenic nature of the drug response,
however, we have yet to develop clinically robust pharma-
cogenetic biomarkers encompassing more than one gene.
An exception is the use of a two-gene biomarker panel to
optimize warfarin therapy, genotyping a promoter SNP in
the warfarin target VKORC1 combined with CYP2C9*2
and *3 alleles (CYP2C9 is the main CYP enzyme metabo-
lizing the active S-warfarin isoforms). Together with other
personal variables such as weight and age, the biomarker
panel accounts for ~60% of the variability in steady-state
warfarin dosages. Recent prospective clinical trials have
shown that inclusion of the genetic biomarker panel short-
ens the time needed to reach stable dosing and reduces the
rate of rehospitalization.37 Yet the test only addresses the
dosing schedule needed to reach the target blood-clotting
time, which itself is merely a biomarker for clinical out-
come (avoidance of bleeding or clotting events). Moreover,
the gene–gene interaction is considered to be additive ra-
ther than epistatic; as discussed earlier, one needs to test for
epistatic interactions to fully understand overall genetic in-
fluence in drug efficacy to search for more such biomarker
panels.
Drug targets may carry frequent regulatory variants that
could affect response; to test this, we have applied allelic
mRNA expression analyses to human target tissues (brain,
liver, heart, etc.), finding frequent rSNPs and srSNPs (see
Figure  7.3) in key drug receptors and downstream fac-
tors, including DRD2, DAT, TPH2, MAOA, 5-HTR2A,
VKORC1, CETP, and ACE. These regulatory variants did
not primarily contribute to disease risk per se, but they did
yield significant risk odds ratios in combination with other
factors, such as cocaine abuse, antidepressant therapy, and
concomitant cardiovascular risk factors such as high chol-
esterol levels. For example, two DRD2 SNPs in introns 5
and 6 affect splicing of the dopamine D2 receptor mRNA
to reduce formation of D2S (short form lacking exon 6) in
favor of D2L (long form).24 As D2S is mostly localized pre-
synaptically and functions as an autoreceptor, we tested the
effect of the splicing SNPs in a cohort of deceased cocaine
abusers and controls, from the Miami Dade County Brain
Endowment Bank.38 Remarkably, these splicing SNPs con-
ferred a threefold increased risk of death by cocaine over-
dose relative to the controls,38 remarkable in that this single
genetic variation could have affected such complex trait.
Since D2S physically interacts with the dopamine trans-
porter DAT, for which we had also determined regulatory
variants,25 we are now testing the dynamic interactions
between DRD2-DAT–cocaine, finding SNP combina-
tions with odds ratios of ~8, indicative of epistatic events
(unpublished). Further investigations along these lines may
reveal robust genetic effects on drug efficacy.
Molecularly targeted cancer chemotherapy has
emerged as a promising approach to effective treatments.
Upon discovery of driver mutations to which the cancer
cells are “addicted,” drugs targeting these driver genes
proved dramatically effective in reducing tumor burden.
The first such example was trastuzumab (Herceptin)
directed against the growth factor receptor HER-2
(ERBB2). Because trastuzumab is effective only against
breast tumors over-expressing HER-2, a biomarker test
determining the expression level became mandatory,
representing a so-called companion test (Table  7.3A).
Without this test, the overall efficacy of the drug would
have been too low, but directed against HER-2-positive
tumors (~20% of all subjects), the treatment proved ef-
fective. Anticancer companion tests are discussed further
below, representing a rapidly growing clinical application
in pharmacogenomics.
G E N ET I C FAC TO R S I N
I D I O SY N C R AT I C D RU G R E AC T I O N S
This type of adverse drug reaction is rare and at first was
deemed unpredictable. However, systematic genetic/
genomic studies have revealed contributions from genes
involved in drug metabolism, reactive oxygen species
(ROS) interactions, inflammatory processes, and immune
response. Typically, oxidative steps catalyzed by CYPs and
other redox enzymes generate reactive drug intermediates
in a process that also produces ROS, such as peroxides and
hydroxyl radicals. A  series of enzymes (e.g., those involv-
ing glutathione (GSH) reactions) then renders these ROS
molecules inactive, thereby protecting the cells. ROS reac-
tions result in oxidative cell damage, affecting mitochondria
and causing inflammation (cytokine mediators play a role).
Moreover, reactive drug intermediates can covalently bind
to proteins and other cell components to stimulate an im-
mune response, with specific HLA alleles implicated for
certain drugs. On the other hand, drug metabolism can
also protect against IDRs by rendering the drug inert. For
example, sulfamethoxazole can cause dose-limiting skin
rash, an IDR that is aggravated in HIV/AIDs patients,
presumably because of GSH depletion. Sulfamethoxazole
is either inactivated by NAT1 and NAT2, or transformed
to a reactive intermediate by CYP2C9, further producing
ROS counterbalanced by GSH-mediated detoxification.
We have found that the NAT1 *10 and *11 alleles repre-
sent gain-of-function variants that afford protection against
sulfamethoxazole IDR; however, this effect was detectable
only in slow NAT2 acetylator genotype carriers (an inter-
esting gene–gene interaction).33 Further clinical association
analyses indicated involvement of cytokine genes and HLA
alleles (unpublished), consistent with the various pathways
contributing to IDRs. Having clinical significance, HLA al-
lele effects can be dramatic, with certain drugs conveying
high risk of serious IDRs, to be discussed subsequently with
clinical biomarkers.

Table 7.3A  SELECT ENTRIES IN THE FOOD AND DRUG ADMINISTRATION’S TABLE OF PHARMACOGENOMIC
BIOMARKERS IN DRUG LABELS

URL:
DRUG THERAPEUTIC BIOMARKER LABEL SECTIONS WITH PHARMACOGENOMIC
AREA INFORMATION
Abacavir Antiviral HLA-B*5701 Boxed warning contraindications Warnings and
precautions
Patient counseling
Carbamazepine Neurology HLA-B*1502 Boxed warning
Cetuximab Oncology EGFR Indications and usage warnings and precautions
KRAS Indications and usage
Clopidogrel Cardiovascular CYP2C19 Boxed warning
Dasatinib Oncology Ph Chromosome/BCR-ABL Indications and usage
Doxepin Psychiatry CYP2D6 Precautions
Imatinib C-KIT, BCR-ABL, PDGFR, FIP1L1-PDGFRa Indications and usage
Maraviroc Antivirals CCR5 Warnings and precautions
Tamoxifen Oncology ER Indications and usage
Trastuzumab Oncology Her-NEU Indications and usage

SOURCE: Table of Pharmacogenomic Biomarkers in Drug labeling. http://www.fda.gov/Drugs/ScienceResearch/ResearchAreas/Pharmacogenetics/ucm083378.htm

P harmacogenomics — C ritica l C omponent o f G enomic M edicine   •   1 0 7

 VA L I DAT E D P H A R M AC O G E N O M I C therapy demonstrated a significant reduction (33%) in
B I O M A R K E R T E S TS A N D C L I N I C A L rehospitalizations.37 Such prospective trials are rare, and
I M P L E M E N TAT I O N one must often rely on published retrospective evidence
in clinical applications. However, robust data on perceived
With pharmacogenomics evolving, its translation into
improvement in therapy outcomes and cost–benefit ratios
clinical practice is rapidly moving forward. As a majority
are typically unavailable. Ad hoc single pharmacogenomic
of drug therapies proves effective in only a portion of the
biomarker tests administered when needed are expensive,
patient population, and with ADRs considered a leading
thereby inflating cost–benefit ratios.
cause of morbidity and mortality in the United States (see
With genotyping technology advancing, and with the
5
for references), enhancing drug efficacy and reducing tox-
introduction of electronic medical records (eMRs), it will
icity is an imperative, in light of the immense drug use across
become routine to test for large numbers of genetic variants
the entire population. In some indications, only 40–60% of
early, to safeguard the information in secure information
patients benefit from drug therapy, and a substantial portion
warehouses, and to transfer “actionable” information onto
suffer from ADRs, defined as adverse reactions that occur
the eMR of an individual. Once established, this reduces
upon prescribing the proper drug and dose for the intended
the cost of retrieving the genotype information and at-
diagnosis. Under this definition, ADRS may be considered
tendant therapeutic guidelines, and the cost of implementa-
unavoidable; however, when we surveyed the entire ADR
tion. This process will become standard procedure for drug
literature and overlaid this onto the pharmacogenomics lit-
therapies where a biomarker test provides significant advan-
erature, we found that a majority of the drugs accounting
tages:  a narrow therapeutic index, serious adverse effects,
for most ADRs are metabolized by polymorphic enzymes,
inability to measure efficacy (e.g., preventing future events
in particular CYP enzymes.5 Therefore, what appears to be
such as myocardial infarction), and heuristic choice of op-
“unavoidable” may indeed be manageable if one links gen-
timal therapy (e.g., with antipsychotics) when failure to re-
etic biomarkers with drug therapy. For a few drugs, a phar-
spond has long-lasting consequences.
macogenomic biomarker test has become obligatory before
Up-to-date information on these questions is compiled
the drug can be administered or is reimbursed by insurers
at the Pharmacogenomics Knowledge Base (PharmGKB)
(e.g., Herceptin and HER-2), a combination often referred
website (https://www.pharmgkb.org/index.jsp), provid-
to as “theranostic.”
ing information on drug-related genes and their variants,
In principle, biomarkers can serve to guide drug
including the evidence of underlying mechanisms and clin-
­selection or drug dosing, and to increase drug efficacy or
ical relevance. It is important that one consider each gene–
­decrease toxicity. For example, a poor CYP2D6 metabol-
drug pair individually, because even small modifications in
izer in need of antidepressant therapy should either not be
the drug’s molecular structure results in different affinities
given a CYP2D6 substrate such as doxepin, or only given
to drug metabolizing enzymes, transporters, and recep-
the drug in a much reduced dosage. The latter may be dif-
tors/targets. While the impact of genetic variants may be
ficult to ­titrate to effective levels without toxicity, so that
similar between closely related drugs, the clinical benefit of
a switch to other CYP2D6-independent drugs appears
a biomarker test may be reduced—for example, if one drug
indicated. Even though this thought process appears rather
is largely metabolized by CYP2D6 while a drug analogue
compelling, CYP2D6 genotyping is still rarely used in clin-
may also be recognized by CYP2C19, as is the case for some
ical practice, until biomarker companies begin to offer CYP
antidepressants.
genotyping panels directly to physicians and patients.
A prime example of a biomarker panel to optimize a
Pharmacogenomic Biomarker Tests
dosage regimen is used in anticoagulant therapy with war-
farin. Displaying a rather narrow therapeutic window, with With clear evidence of potential clinical utility, the FDA’s
either bleeding episodes or embolisms from clotting the Center for Drug Evaluation Research (CDER) has estab-
serious consequences of improper dosing, warfarin is one lished a process to evaluate biomarker tests, publishing
of the worst ADR offenders. A genotyping panel consist- their findings in the FDA Table of Pharmacogenomic
ing of SNPs in CYP2C9 and the drug target VKORC139 Biomarker Tests in Drug Labels, summarizing the cur-
predicts ~50% of dosing variability between patients, rent status of approved tests and clinical relevance; a few
enabling accelerated dose titration to a stable target level. examples are provided in Table  7.3A. The site provides
A large-scale trial measuring the benefits from prospective rich information resources on the drug, gene or gene panel
use of the biomarker panel upon initiation of warfarin (some tests are also phenotypical, such as high expression

1 0 8   •   P rincip l es o f G enomic M edicine

of the HER-2 protein), and clinical relevance, with links
to drug-labeling information that is publicly available to
therapists and patients. The intention is not to specify how
therapeutic decisions are to be made, but rather to provide
an assessment of current knowledge, updated regularly, and
to develop a regulatory framework in specific cases where
the evidence of a clinically actionable drug–gene link is
strong. In a growing number of cases, such information is
contained in a “boxed warning,” highlighting particularly
hazardous ADRs or other therapeutic risks (e.g., carba-
mazepine; Table 7.3B). Also, for an increasing number of
drugs, a prospectively applied biomarker test is mandated
before the drug can be prescribed, including trastuzumab
(Herceptin), maraviroc, and dasantinib (Table 7.3A). The
FDA Table has grown dramatically in the recent past, pre-
saging a time when use of biomarkers will become routine
in drug therapy. A  few examples will be discussed here,
while the reader is advised to consult the FDA Table itself
for details. Because of the growing importance of biomark-
ers in targeted cancer chemotherapy, this area will be dis-
cussed separately.
Antivirals Maraviroc and Abacavir
The HIV virus infects cells by binding to T-cell recep-
tors, but it also requires specific co-receptors for ef-
fective cell entry; different HIV stains require different
co-receptors—a targeting process called tropism. M-tropic
strains of HIV-1 recognize CCR5, expressed on mac-
rophages and CD4+ T cells. A  selective inhibitor of
the CCR5 receptor, maraviroc is only effective against
CCR5-tropic viruses; therefore, only patients infected
with CCR5-tropic HIV benefit from maraviroc, while
patients with CRR5X4-tropic viruses respond poorly.40
The tropism test exposes CCR5 or CXCR4 positive cell
cultures to the patient’s HIV particles, measuring infect-
ivity. As a result, it is critical to determine the patient’s
HIV tropism before prescribing maraviroc.
On the other hand, abacavir can cause severe hyper-
sensitivity reactions in a portion of subjects.41 This has
led to a boxed warning in the abacavir label, stating that
the HLA-B*5701 allele is strongly predictive of this IDR.
As a result, prospective HLA genotyping has been intro-
duced, thereby excluding fewer than 10% of patients while
reducing the incidence of IDRs substantially. The abacavir
example was one of the early cases where prospective geno-
typing was critical to propelling an otherwise efficacious
drug into broad clinical use.
Carbamazepine
A well-established antiepileptic drug, carbamazepine
(Tegretal), causes severe and sometimes fatal skin reac-
tion (Steven Johnson syndrome), an IDR rarely observed
in Caucasians but with tenfold greater frequency in
Asians. Therefore, the FDA has issued a boxed warning in
the drug label information, advising of this serious IDR,
which was shown to be associated with an HLA allele,
HLA-B*1502. Prospective genotyping in Asians is strongly
advised; by withholding carbamazepine from patients car-
rying the HLA-B*1502, the incidence of IDRs is drastically
reduced.15 However, the HLA locus is extremely heteroge-
neous, so that this allele biomarker may not be valid even in
Asian subpopulations. A GWAS study in Japanese subjects
has revealed a strong association of carbamazepine-induced
cutaneous IDRs with the HLA-A*3101 allele.42 Therefore,
care has to be taken in applying HLA allele associations
across ethnic groups.
C O M PA N I O N D I AG N O S T I C S
AND THER ANOSTICS IN
C A N C E R C H E M OT H E R A P Y
Genetic and genomic studies have revealed that certain
cancers are driven by oncogenic “driver mutations,” such

Table 7.3B  BOXED WARNING IN THE DRUG LABEL OF CARBAMAZEPINE AND HLA-B*1502.

THE LINK IS PROVIDED IN THE TABLE OF PHARMACOGENOMIC BIOMARKERS (TABLE 7.3A). THE TEXT IS ABBREVIATED.
SERIOUS DERMATOLOGIC REACTIONS AND HLA-B*1502 ALLELE
SERIOUS AND SOMETIMES FATAL DERMATOLOGIC REACTIONS, INCLUDING TOXIC EPIDERMAL NECROLYSIS
(TEN) AND STEVENS-JOHNSON SYNDROME (SJS),. . . . ESTIMATED TO OCCUR IN 1 TO 6 PER 10,000 NEW USERS
IN. . . . CAUCASIAN POPULATIONS, BUT THE RISK IN SOME ASIAN COUNTRIES IS ESTIMATED TO BE ABOUT 10
TIMES HIGHER. . . . STRONG ASSOCIATION BETWEEN THE RISK OF DEVELOPING SJS/TEN AND THE PRESENCE OF
HLA-B*1502, AN INHERITED ALLELIC VARIANT OF THE HLA-B GENE. HLA-B*1502 IS FOUND ALMOST EXCLUSIVELY
IN PATIENTS WITH ANCESTRY ACROSS BROAD AREAS OF ASIA. . . . PATIENTS WITH ANCESTRY IN GENETICALLY
AT-RISK POPULATIONS SHOULD BE SCREENED FOR THE PRESENCE OF HLA-B*1502 PRIOR TO INITIATING
TREATMENT WITH TEGRETOL. PATIENTS TESTING POSITIVE FOR THE ALLELE SHOULD NOT BE TREATED WITH
TEGRETOL. . . .

P harmacogenomics — C ritica l C omponent o f G enomic M edicine   •   1 0 9

that the transformed cell is considered “addicted” to the
oncogene’s permanent activation. This insight has led to the
development of “targeted chemotherapies,” with the anti-
body trastuzumab the first example (Table 7.4). Search for
small molecule inhibitors led to the discovery of imatinib
as highly effective inhibitor of BCR-ABL (imatinib), a per-
manently active fusion tyrosine kinase on the Philadelphia
chromosome.43 The majority of acute lymphoblastic leu-
kemias carry the Philadelphia chromosome—actually a
diagnostic marker detected via karyotyping, while specific
biomarker tests with reverse transcription polymerase chain
reaction (RT-PCR) and fluorescence in situ hybridization
(FISH) analysis may also be used. Indeed, highly sensitive
detection of residual BCR-ABL-positive cells in the circu-
lation can serve to assess response to therapy for any type of
BCR-ABL-positive cancers, an example where a genotype
assay is used as a response biomarker rather than a predictive
biomarker. However, after initial responses to imatinib,
resistance often develops, in part because of point muta-
tions that render the drug ineffective as an inhibitor.44 To
overcome the developing resistance, new drugs have been
designed to also inhibit the doubly mutated kinase:  for
example, dasatinib (Table  7.4). However, before selecting
dasatinib as the follow-up drug, it may be useful to apply
a genotyping test to determine whether there is a muta-
tion present in the imatinib-resistant tumor that renders
it sensitive to dasatinib, or whether it has acquired dual
resistance to dasatinib as well (e.g., F317).45 Yet dasatinib
also blocks a series of other kinases, and its action may not
be limited to BCR-ABL. Therefore, dasatinib may be pre-
scribed to any patients resistant to imatinib, or experiencing
ADRs from it, while the FDA label also permits first-line
use of dasatinib in chronic lymphocytic leukemia (CLL) or
BCR-ABL+ acute lymphocytic leukemia (ALL).
Meanwhile, a host of novel drugs has been developed
and brought into clinical use, mostly targeted against
activated or overexpressed oncogenic kinases (Table  7.4).
Typically, these agents are highly effective in the respective
“addicted” tumors, but companion biomarker tests to de-
termine the type of driver mutation are needed to optimize
an individual’s response. Being the first anticancer drug that
is highly effective in the treatment of breast cancer if com-
bined with a companion diagnostic, trastuzumab requires
testing for HER-2 protein overexpression or amplification
at the gDNA level.46 Since then, dramatic examples of highly
effective therapies have emerged, including treatment with
the BRAF inhibitor vemarufinib of melanomas carrying a
frequent BRAF mutation (V600E) present in >60% of mel-
anomas.47 Other examples include treatment with EGFR
inhibitors of NSCLC (non-small cell lung carcinoma)
carrying activating EGFR mutations, but while under spe-
cific drug-selection pressure, different activating mutations
render the kinase resistant to the drug, requiring a switch
to different inhibitors44. Also, testing for KRAS-activating
mutations in colon cancer has become standard practice,
because EGFR inhibitors fail when a downstream driver
mutation maintains oncogenic signaling.
While these targeted therapies have generated the
hope for highly effective therapies, resistance emerges in
most cases, either through amplification, mutations of
the driver gene to abrogate the inhibitor’s potency, and
numerous possible bypass processes in parallel or down-
stream. For example, resistance to vemurafinib has been
shown to emerge upon inappropriate dimerization of

Table 7.4  COMPANION BIOMARKER TESTS FOR TARGETED ANTICANCER AGENTS DIRECTED AGAINST
“DRIVER” ONCOGENE KINASES

ONCOGENE DRUG MUTATION REMARKS

HER2/ERBB2 Trastuzumab Overexpression Effective treatment of HER2-positive breast
(Herceptin) cancer
EGFR/ERBB1 EGFR inhibitors Activating mutations NSCLC*
ALK Crizotinib Gene rearrangements Late stage NSCLC
KRAS EGFR inhibitors: Activating mutations Lack of efficacy in metastatic colorectal cancer,
Cetuximab, panitumumab lung cancer
BRAF Vemurafinib V600E activating mutation in 60% V600E melanomas highly sensitive to
of melanomas vemurafinib
BCR-ABL Imatinib (Gleevec), dasatinib Fusion kinase Chronic lymphocytic leukemia
Other kinases and signaling molecules with potential use in diagnostics:
JAK2, CDK8, AURKA, MEK, PLK1, PIL3CA, FLT3, PI3K
NRAS, HRAS, CDKN2C, TMPRSS2-FAYSI, HPV16, MTOR; loss of function of multiple tumor suppressor genes

*NSCLC: non-small cell lung cancer

1 1 0   •   P rincip l es o f G enomic M edicine

BRAF-V600E,48 among other reported mechanisms. As
a result of bypass processes, other tumor types express-
ing BRAF-V600E may respond only marginally to vemu-
rafinib; for example, colon cancer, rendered resistant
through rapid upregulation of EGFR, a kinase already
expressed at a higher level than in melanomas.49 As dis-
cussed for BCR-ABL and dasatinib, the search is on to de-
termine which resistance mechanisms are likely to occur,
and discover how one can optimize treatment strategies
in anticipation of these mechanisms.
Synthetic lethality is yet another concept that leads to
novel anticancer therapies requiring companion diagnos-
tics. This idea implies the presence of a mutation that that
is nonlethal and in fact critical to tumor development; for
example, breast cancer gene (BRCA) mutations in breast
cancer, affecting repair of double-strand breaks.50 In this
scenario, a second nonlethal defect—for example, in PARP
playing a critical role in excision repair—effectively kills the
cancer cell while leaving normal cells unscathed. Applied to
any BRCA-deficient tumor, not just breast cancer, PARP
inhibitors thus acquire potent cytotoxic power; as a result,
PARP inhibitors are under clinical study against various
cancers, while genetic tests are needed to determine BRCA
deficiency, or any other defect in homologous recombin-
ation events.
This raises a question:  Why not sequence the entire
genome of normal and tumor tissue in each subject51,52?
Genomic medicine is moving in this direction as sequenc-
ing costs continue to be in free-fall. First attempts have
used exome sequencing with the expectation that a ma-
jority of cancer-relevant mutations generated under strong
selection pressure will prove to be non-synonymous SNPs,
or easily recognizable variants that yield inactive proteins.
Applied to 500 patients with NSLC, identifiable onco-
genic mutations were detectable in 22% of the tumors,
including KRAS (24%), EGFR (13%), ALK (5%), TP53
(5%), PK3Ca (4%), CTTNB1 (2%), BRAF (2%), NRAS
(1%), HER2 (1%), and IDH1 (1%).53 These patients
therefore can be matched to appropriate clinical trials
with targeted drugs53; however, it remains to be seen how
successful this strategy will be; moreover, exome sequenc-
ing is likely to miss many oncogenic mutations, including
chromosomal rearrangements, copy number variants, and
regulatory processes.
Another main impediment to successful targeted
chemotherapy is introduced with intra-tumor heterogen-
eity, reflecting the evolutionary history of the tumor and
metastases. Therefore, single tumor biopsy samples are likely
to underestimate tumor heterogeneity. Use of biopsies from
multiple sites in primary renal carcinomas revealed that
P harmacogenomics — C ritica l C omponent o f G enomic M edicine   •   1 1 1
more than 60% of all somatic mutations were not pres-
ent in all tumor regions.54 Testing the utility of diagnostic
mRNA expression panels, mRNA expression profiles in-
dicative of both good and poor prognoses were detected
in different regions of the same tumor.54 The authors con-
clude that these findings “may present major challenges to
personalized-medicine and biomarker development”; how-
ever, renal carcinomas are highly resistant to therapies, and
therefore, the results presented by Gehrlinger et al.54 may not
be representative for all tumors. In all these scenarios, mul-
tiple strategies will be needed to manage aggressive cancer,
rather than seek “cures.” Therefore, the current trend is to-
wards multidrug and multimodality therapies, still including
relatively nonspecific toxic anticancer drugs to avoid tumor
recurrence.
C O N C LU S I O N S
This chapter provides but a survey of the current status
and future direction of pharmacogenomics. We are clearly
far from understanding the genetic factors critical to drug
response and toxicity, let  alone the intricate interactions
between genetics and the environment, with epigenetic
processes contributing substantially. Yet our knowledge
has advanced sufficiently that pharmacogenomic principles
and tests can be applied in clinical practice. Broad adoption
by therapists will require education, documentation of the
cost–benefit ratios, clear guidelines as to what actions need
to be taken once a genotype is ascertained, and a system for
dissemination of the results. With respect to genetic bio-
markers, we anticipate large-scale assays to cover a majority
of relevant variants, or whole-genome sequencing, done
prospectively and with specific results retrievable instantly
when needed. With ever-increasing use of diagnostic tests
in medicine as a whole, an electronic health care system
with individual electronic medical records for each subject
appears to be an essential component of successful imple-
mentation of the promising therapeutic strategies emerging
from pharmacogenomics.
REFERENCES
1. Wooding S. Phenylthiocarbamide: a 75-year adventure in genetics
and natural selection. Genetics. 2006;172:215–223.
2. Omenn GS. Evolution in health and medicine Sackler collo-
quium:  evolution and public health. Proc Natl Acad Sci U S A.
2010;107(Suppl 1):1702–1709.
3. Weber WW. The legacy of pharmacogenetics and potential applica-
tions. Mutat Res. 2001;479:1–18.
 4. Meyer UA. Pharmacogenetics—five decades of therapeutic lessons
from genetic diversity. Nat Rev Genet. 2004;5:669–676.
5. Phillips DL, et al. The potential role of pharmacogenomics in reduc-
ing adverse drug reactions:  a systematic review. J Amer Med Assn.
2001;286:2270–2279.
6. Sadee W, et  al. Pharmacogenomics of the RNA world:  struc-
tural RNA polymorphisms in drug therapy. Clin Pharmacol Ther.
2011;89:355–365.
7. Hudson KL. Genomics, health care, and society. New Engl J Med.
2012;365:1033–1041.
8. Rahmioglu N, et al. Genetic epidemiology of induced CYP3A4 ac-
tivity. Pharmacogenet Genomics. 21:642–651, 2011
9. Khokhar JY, et  al. Pharmacogenetics of drug dependence:  role of
gene variations in susceptibility and treatment. Annu Rev Pharmacol
Toxicol. 2010;50:39–61.
10. Tsuang MT. The Harvard Twin Study of Substance Abuse: what we
have learned. Harv Rev Psychiatry. 2001;9:267–279.
11. Manolio TA, et al. Finding the missing heritability of complex dis-
eases. Nature. 2009;461:747–753.
12. Zuk O et  al:  The mystery of missing heritability:  genetic inter-
actions create phantom heritability. Proc Natl Acad Sci U S A.
2012;109:1193–1198.
13. Johnson AD, Wang D, Sadee W. Polymorphisms affecting gene
regulation and mRNA processing: broad implications for pharma-
cogenetics. Pharmacol Ther. 2005;106:19–38.
14. Linkblad Toh K, et al. A high resolution map of human evolutionary
constraints using 29 mammals. Nature. 2011;478:476–482.
15. Ganesan S, Hussain N. Question 2 Should phenytoin and carba-
mazepine be avoided in Asian populations with the HLA-B*1502
positive genetic variant? Arch Dis Child. 2011;96:104–106.
16. Ramsey LB, et  al. Rare versus common variants in pharmacogen-
etics:  SLCO1B1 variation and methotrexate disposition. Genome
Res. 2012;22:1–8.
17. The SEARCH Collaborative Group. SLCO1B1 variants and
statin-induced myopathy—a genome-wide study. New Engl J Med.
2008;359:789–799.
18. Carleton B. Demonstrating utility of pharmacogenetics in pediatric
populations: methodological considerations. Clin Pharmacol Ther.
2010;88:757–759.
19. Ge, B, et  al. Global patterns of cis variation in human cells
revealed by high-density allelic expression analysis. Nat Genet.
2009;41:1216–1222.
20. Daly AK. Genome-wide association studies in pharmacogenomics.
Nature Reviews Genetics. 2010;11:241–246.
21. Huang Y, et  al. Membrane transporters and channels:  role of the
transportome in cancer chemosensitivity and chemoresistance.
Cancer Res. 2004;64:4294–4301.
22. Johnson AD, et  al. Polymorphisms affecting gene transcription
and mRNA processing in pharmacogenetic candidate genes: detec-
tion through allelic expression imbalance in human target tissues.
Pharmacogen Genom. 18: 781–791, 2008.
23. Wang D. Intronic polymorphism in CYP3A4 affects hepatic
expression and response to statin drugs. Pharmacogenomics J.
2011;11:274–286.
24. Zhang, Y, et  al. Polymorphisms in human dopamine D2 re-
ceptor gene affect gene expression, splicing, and neuronal ac-
tivity during working memory. Proc Natl Acad Sci U S A.
2007;104:20552–20557.
25. Pinsonneault JK, Han DD, Burdick KE, M. Kataki, A. Bertolino,
Malhotra AK, Gu HH, Sadee W. Dopamine transporter gene
variant affecting expression in human brain is associated with bi-
polar disorder. Neuropsychopharmacology. 2011;8:1644–1655.
26. Lim J-E, Papp A, Pinsonneault J, Sadee W, Saffen D. Allellic expres-
sion of serotonin transporter (SERT) mRNA in human pons: lack
of correlation with the polymorphism SERTLPR. Mol Psychiatry.
2006;11:649–662.
1 1 2   •   P rincip l es o f G enomic M edicine
27. Johnson AD, Gong Y, Wang D, Langaee TY, Shin J, Cooper-DeHoff
RM, Schork NJ, Binkley P, Pepine CJ, Johnson JA, Sadee W.
Promoter polymorphisms in ACE (angiotensin-I converting en-
zyme) associated with clinical outcomes in hypertension. Clin
Pharmacol Ther. 2009;85:36–44.
28. Papp AC, Pinsonneault JK, Wang D, Newman LC, Gong Y,
Johnson JA, CJ. Pepine, Kumari M, A.D Hingorani, Talmud PJ,
Shah S, Humphries SE, Sadee W. Cholesteryl ester transfer protein
(CETP) polymorphisms affect mRNA splicing, HDL levels, and
sex-dependent cardiovascular risk. PLoS One. 2012;7:e31930.
29. Shuldiner AR, et al. Association of cytochrome P450 2C19 geno-
type with the antiplatelet effect and clinical efficacy of clopidogrel
therapy. JAMA. 2009;302:849–857.
30. Cacabelos R, Martínez-Bouza R. Genomics and pharmacogenomics
of schizophrenia. CNS Neurosci Ther. 2011;17:541–565.
31. Decleves X, et al. Interplay of drug metabolizing CYP450 enzymes
and ABC transporters in the blood-brain barrier. Curr Drug Metab.
2011;12:732–741.
32. Ratain MJ, Innocenti F. Individualizing dosing of irinotecan. Clin
Cancer Res. 2010;16:371–372.
33. Wang D, Para MF, Koletar SL, W. Sadee. Human N-acetyltransferase
1 (NAT1) *10 and *11 alleles increase protein expression via distinct
mechanisms and associate with sulfamethoxazole-induced hyper-
sensitivity. Pharmacogen Genomics. 2011;21:652–664.
34. Wang D, et al. Multidrug resistance polypeptide 1 (MDR1, ABCB1)
variant 3435C>T affects mRNA stability. Pharmacogen Genomics.
2005;15:693–704.
35. Saitoh A, et al. An MDR1-3435 variant is associated with higher
plasma nelfinavir levels and more rapid virologic response in HIV-1
infected children. AIDS. 2005;19:371–380.
36. Pauli-Magnus C, et  al. No effect of MDR1 C3435T variant on
loperamide disposition and central nervous system effects. Clin
Pharmacol Ther. 2003;74:487–498.
37. Epstein RS, et al. Warfarin genotyping reduces hospitalization rates
results from the MM-WES (Medco-Mayo Warfarin Effectiveness
study). J Am Coll Cardiol. 2010;55:2804–2812.
38. Moyer RA, et al. Intronic polymorphisms affecting alternative splic-
ing of human dopamine D2 receptor are associated with cocaine
abuse. Neuropsychopharm. 2011;19:76–83.
39. Wang D, et al. Regulatory polymorphism in vitamin K epoxide re-
ductase complex subunit 1 (VKORC1) affects gene expression and
warfarin dose requirement. Blood. 2008;112:1013–1021.
40. Latinovic O, et al. Pharmacotherapy of HIV-1 infection: focus on
CCR5 antagonist maraviroc. Clin Med Ther. 2009;1:1497–1510.
41. Hughes AR, et  al. Pharmacogenetics of hypersensitivity to aba-
cavir:  from PGx hypothesis to confirmation to clinical utility.
Pharmacogenomics J. 2008;8:365–374.
42. Ozeki T, et  al. Genome-wide association study identifies
HLA-A*3101 allele as a genetic risk factor for carbamazepine-induced
cutaneous adverse drug reactions in Japanese population. Hum Mol
Genet. 2011;20:1034–1041.
43. Duffaud F, Le Cesne A. Imatinib in the treatment of solid tumours.
Target Oncol. 2009;4:45–56.
44. Sequist LV, et  al. Genotypic and histological evolution of lung
cancers acquiring resistance to EGFR inhibitors. Sci Tranlat Med.
3:75ra26, 2011.
45. Intermesoli T, et al. Durable molecular response despite F317L and
E255K mutations:  successful treatment of chronic myeloid leu-
kemia with sequential imatinib, nilotinib and dasatinib. Leuk Res.
36:e10–11, 2012
46. Emens LA. Trastuzumab: targeted therapy for the management of
HER-2/neu-overexpressing metastatic breast cancer. Am J Ther.
2005;12:243–253.
47. Luke JJ, Hodi FS. Vemurafenib and BRAF inhibition:  a new
class of treatment for metastatic melanoma. Clin Cancer Res.
2012;18:9–14.
 48. Poulikakos PI, et  al. RAF inhibitor resistance is mediated by di-
merization of aberrantly spliced BRAF(V600E). Nature. 2011;
480:387–439.
49. Prahallad A, et  al. Unresponsiveness of colon cancer to
BRAF(V600E) inhibition through feedback activation of EGFR.
Nature. 2012;483:100–104.
50. Lord CJ, Ashworth A. The DNA damage response and cancer
therapy. Nature. 2012;481:287–294.
51. Corless CL. Personalized cancer diagnostics. Science. 2011;334:
1217–1218.
P harmacogenomics — C ritica l C omponent o f G enomic M edicine   •   1 1 3
52. Hayden EC. Targeted treatment tested as potential cancer cure.
Nature. 479:281, 2011.
53. Sequist LV, et  al. Implementing multiplexed genotyping of
non-small-cell lung cancers into routine clinical practice. Ann Oncol.
2011;22:2616–2624.
54. Gerlinger M, et  al. Intratumor heterogeneity and branched evo-
lution revealed by multiregion sequencing. New Engl J Med.
2012;366:883–892.
 8.
NEW DRUG DEVELOPMENT, DRUG RESPONSE,
AND PRECISION MEDICINES
Michelle Penny and Duncan McHale
INTRODUCTION
In its broadest sense, pharmacogenomics can be defined as
the investigation of variations of DNA and RNA character-
istics as related to drug response. The last decade has seen
a large increase in the amount of genomics data generated
and, with it, increased the expectations of how improved
understanding of disease will lead to the development
of more effective therapies and personalized medicines.
Despite the regular reports of novel genes being identified
in a range of disorders, by 2012 the much-heralded promise
of the human genome project has only started to materi-
alize. However, it is important to realize that this does not
represent a failure of the science to deliver as there are mul-
tiple clear examples of the predictability and clinical utility
of pharmacogenetics but is a reflection of the length of time
it takes to develop new drugs and implement changes in
healthcare. The 1980s and 1990s saw a boom time for the
pharmaceuticals industry producing many highly effective
new classes of drugs from statins to proton-pump inhibi-
tors and quinolone antibiotics. All were novel therapeutic
approaches offering significant benefit to individuals and
society. The science that drove many of these advances was
based on the greater understanding of biochemistry and
pharmacology that emerged during the 1970s and early
1980s. This 10- to 15-year time-lag from gaining scien-
tific knowledge to developing therapies is typical for the
pharmaceutical industry, and reflects the complexity of
drug discovery and the time required for preclinical and
clinical testing to ensure safety and efficacy. This chapter
will introduce the major concepts of drug discovery and
development and give a broad overview of how genetics
and genomics is used across the whole drug discovery and
development pipeline, from pre-target identification to
post-marketing surveillance to help discover and develop
improved medicines. It will describe some of the examples
of how pharmacogenetics has impacted the lives of patients.
114
I N T E G R AT I O N O F G E N O M I C S
I N TO T H E D RU G D I S C O VE RY
A N D D E VE L O PM E N T
Pharmaceutical companies have historically focused their
drug discovery and development programs on finding ther-
apies for broad use in large disease populations, the “block-
buster business model.” A blockbuster drug is usually defined
as one with peak annual sales of greater than $1 billion and
is generally developed for long-term use to treat common
complex chronic disorders in the general population. The
strategy to identify and develop blockbuster drugs has been
the response to the high costs of drug discovery and devel-
opment. A survey of the drug development costs of 68 new
compounds from 10 pharmaceutical companies estimated
that the cost to develop a new drug in 2000 was $802 mil-
lion (DiMasi et  al., 2003). The high costs of developing
drugs can be attributed to two main factors: the large size
and duration of the clinical trials required to provide the
data to show safety and efficacy of the compound, and the
high rate of attrition of compounds in clinical development;
fewer than 10% of compounds entering phase I clinical de-
velopment reach the market, the majority failing in clinical
development due to lack of efficacy in phase II. The lack of
recent research and development (R&D) success in finding
blockbuster drugs, combined with financial pressure due to
patent expiry and downward pressure on pricing, has led to
a shift in strategy for many companies in the biopharma-
ceutical industry. Companies are shifting towards the dis-
covery and development of stratified medicines. A stratified
medicine is one that is targeted at a subgroup of a tradition-
ally classified disease, such as Herceptin for the treatment
of Her2-overexpressing breast cancer. Stratified medicines
offer a significant opportunity: to the industry, as they have
an increased probability of success and the potential of
smaller programs; to the regulators, as the benefit–risk pro-
files of these medications are greater than with unselected
medications; to the payers, as they are more cost effective;
and most importantly, to patients, as they are more effective
and safer therapies. Genomics has a large role to play in the
development of stratified medicines, as many of the tools
used to stratify the patient populations are genomic, such as
selective epidermal growth factor receptor (EGFR) muta-
tion status and Gefitinib, K ras mutation status and Erbitux
and Vectibix, Alk4 mutation status and Crizotinib.
Pharmacogenomics—the investigation of variations
of DNA and RNA characteristics (germline or tumor) as
related to drug response in individual patients or groups of
patients—is one of a number of methods employed by the
pharmaceutical industry to stratify patient populations.
A major cause of the attrition of drugs for lack of effi-
cacy is the heterogeneity of the diseases we currently classify
as single entities. Most would be better referred to as syn-
dromes rather than single diseases. The disease classification
currently used is based on phenotypical consequences of
disease processes rather than on the underlying pathological
mechanisms. This has led to the clustering of heterogeneous
disease syndromes based on symptoms rather than based on
molecular pathology. Genomics will be an important tool
in reclassifying diseases into a new molecular taxonomy of
human disease. Oncology is one therapeutic area where this
is most advanced, as the scientific evidence base for tumor
etiology is more advanced than in other areas. The majority
of drug development programs in oncology are now strati-
fying patient populations based on molecular changes in
the tumor. During the period from 2005 to 2012, over 5
stratified medicines in oncology were approved (Table 8.1).
Most of the current drug development programs in on-
cology are using a stratified medicine approach linking the
target to the dysregulated disease pathways on the tumors
and only being used when the right pathway is driving
tumorigenesis. It is widely expected that this approach will
expand across other therapeutic areas as our understanding
of disease biology improves.
The Drug Discovery and Development
Process
The generation of an idea that a particular protein might be
a suitable therapeutic target for the treatment of a disease
sets in motion what is often depicted as a linear process
known as the “drug discovery and development pipeline,”
in which new medicines follow a set route from early dis-
covery and preclinical stages through a set of clinical devel-
opment processes to the marketplace (Figure 8.1). In reality,
the process is generally far from linear, but for the purposes
of describing the component parts, we will consider it a se-
quential process.
Candidate-Seeking
The ultimate aim of the drug discovery process is to find a
chemical (e.g., small molecule) or biological reagent, such
as an antibody, that has the potential to be a drug that can
be moved into preclinical and then clinical testing. In order
to start the process of identifying a potential drug, a bio-
logical assay testing interactions with the drug target must
be developed. This assay is often based on a cloned and
expressed form of the drug target and will be converted into
a format that will allow high-throughput testing, as millions
of chemicals may need to be screened in the assay. The need
to screen millions of chemicals means that it is usually only
feasible to screen one protein variant of the target in the
high-throughput screen. It is therefore vital to screen the
“right” variant. In the situation where there may be more
than one form of the protein that can be included in the
screen, it is important to know that the most biologically
relevant and/or the most common variant is being screened,
and it may be necessary to screen the chemical matter
against more than one form of the protein. This is not al-
ways the most common form of the protein—Verumafenib,
a novel drug for the treatment of malignant melanoma,
was identified by specifically screening against the V600E

Table 8.1  STRATIFIED MEDICINES IN ONCOLOGY

DRUG MECHANISM OF ACTION DISEASE DIAGNOSTIC TEST

Trastuzumab HER2 inhibitor Breast cancer HER2 gene amplification (FISH)
Imatinib Multiple tyrosine kinase inhibitor Gastrointestinal stromal tumor KIT mutation positive
Chronic myelogenous leukemia Philadelphia chromosome positive
Gefitinib Epidermal growth factor receptor Non–small cell lung cancer EGFR activating mutation positive
Crizotinib Alk4 inhibitor Non–small cell lung cancer Alk4 activating mutation positive
Vemurafenib B-RAF inhibitor Melanoma V600E B-RAF mutation positive tumors
Cetuximab Epidermal growth factor receptor Metastatic colorectal cancer K-RAS wild type tumors
Panitumumab Epidermal growth factor receptor Metastatic colorectal cancer K-RAS wild type tumors

N ew D rug D evelopment, D rug R esponse , and P recision M edicines   •   1 1 5

 Clinical development
Discovery Exploratory
development
Preclinical Phase I
Candidate
seeking
Figure 8.1  The drug discovery and development pipeline.
mutated form of the BRAF protein to ensure that it only
blocked signaling of the pathogenic form.
The high-throughput screens generally identify several
potential “hits,” which need to be tested in more rigorous
biological assays to determine the type of interaction and
the effects and then refined using medicinal chemistry.
Promising “leads” are then developed by a series of minor
chemical changes to the original lead, and the final candi-
date is chosen based on the selectivity and potency criteria
required for the drug candidate as well as the physicochemi-
cal properties of the molecule to ensure druglike properties.
This candidate is then taken forward into preclinical testing.
The final testing phase is usually based on in vivo test-
ing of the compound in animal models that have been dem-
onstrated to have some translatability to the target human
disease, or in a range of ex vivo models of human tissue that
recapitulate components of the disease. The predictability
and translatability of these models to humans varies with
different diseases and is the focus of biomedical research in
many therapeutic areas.
Preclinical Testing
Once a drug candidate has been made, it goes into a pre-
clinical toxicology testing that includes in vitro screening
tests to identify potential pharmacological effects at other
receptors that could lead to adverse events, and genetic toxi-
cology testing, which evaluates mutagenicity and clastogen-
icity. Only if these are satisfactory does animal testing begin.
The animal testing is done in two mammalian species and is
staged to ensure that as few animals as possible are used and
that major problems are picked up early. Toxicology studies
to evaluate long-term exposure, reproductive toxicological
effects, juvenile toxicity, and carcinogenicity are generally
only performed once the data have been obtained from
shorter-term human studies that support safety and efficacy.
To date, toxicology induced by new chemicals are identified
and classified by standard phenotypical and histological
changes. While this picks up the majority of potentially
1 1 6   •   principles of G enomic M edicine
Post approval
Full
development
Phase II Phase III Phase IV
Filing
toxic effects, it can be insensitive to subtle changes and can
identify species-specific effects that can be difficult to in-
terpret. A greater understanding of the molecular changes
following drug administration could identify more subtle
effects and species-specific effects. Similarly, the applic-
ability of animal models of a disease could be assessed by
evaluating molecular changes rather than phenotypical
similarities that can be misleading. Greater emphasis is now
being placed on molecular and biomarker changes that re-
sult in organ damage, where they are available (e.g. nephro-
toxicity biomarkers).
Adverse events can be due to unexpected consequences
of the primary pharmacology or to unexpected interactions
with off-target proteins. Understanding the mechanism of
the toxicological effects is important, as this allows a more
quantitative evaluation of the risk of the event happening
in humans. Genomics can be used to identify interactions
with off-target proteins as transcription changes induced
in the organ damaged by the compound can point to the
mechanism of the toxicity. This is often referred to as toxi-
cogenomics. Multiple consortiums (e.g., the Predictive Safety
Testing Consortium [PSTC] and the Safety in Science in
Medicines Education & Training [SAFESCIMET]) are
currently working to identify genomic biomarkers that are
more sensitive than current histopathological scores, allow-
ing early detection of toxicology and the demonstration of
species-specific toxic effects. Similarly, where specific organ
toxicity is expected due to the mechanism of action of the
compound or known off-target effects, then transcription
changes can offer a more sensitive assay to detect early organ
damage.
Clinical Development
Once the initial in vitro testing and acute animal toxicology
studies (which generally take 14 days) have been performed,
then it is possible to start testing the candidate in humans.
The human studies have traditionally been split into four
phases (phases I–IV), each with specific aims (Box 8.1).
 Box 8.1  HUMAN STUDIES HAVE
TR ADITIONALLY BEEN SPLIT INTO FOUR
PHASES (PHASES I–IV)
•
Phase I—Pharmacokinetic and safety profiles in
healthy volunteers
• Phase II—Safety and efficacy in patients, and the
establishment of the dose response
• Phase III—Safety and efficacy at the chosen dosage
• Phase IV—Post-approval studies to answer
specific safety or efficacy questions and to support
commercial strategies
PHASE I
The first time a novel compound (or biological therapy) is
tested in humans, a broad range of doses is tested, starting at
very low exposures to minimize any risks to the clinical trial
participants. Although these initial studies have generally
been performed on healthy volunteers, there is an increas-
ing trend towards incorporating patients as early as possible.
The dose is escalated over several weeks, starting at a point
between 10-fold and 100-fold below the expected pharma-
cological exposure levels, and rising to a maximum tolerable
level, or several-fold beyond the expected maximum clinical
dose (whichever is reached sooner). The aim is to identify
common adverse events and their relationship with plasma
exposure as well as to establish the basic pharmacokinetic
(PK) parameters of the therapeutic agent. As drug develop-
ment continues, more studies are performed to understand
the effects of multiple dosing, specific drug–drug interac-
tions, and food effects. The aim of these studies is to provide
a more comprehensive understanding of the pharmacokin-
etics of the drug and any significant causes of variability in
the pharmacokinetic profiles. Collections of DNA samples
for pharmacogenomic analysis in phase I clinical protocols
allow the assessment of the impact of known genetic varia-
tions on drug metabolism and transport.
There is a growing trend of performing some of these
very early studies in patients, and these are often referred to
as “phase Ib studies.” The primary intent of these studies is
still to establish safety and pharmacokinetic action of the
compound, but the use of patients allows early indicators
of target engagement and biomarkers of efficacy to provide
evidence that the compound is modulating the proposed
mechanism. Where it is possible to biopsy disease tissue
in these studies, transcription analysis can provide some
N ew D rug D evelopment, D rug R esponse , and P recision M edicines   •   1 1 7
evidence that the target pathway is being modulated. This
is generally restricted to some tumor types and dermato-
logical conditions like psoriasis, where it is possible to ob-
tain high-quality tissue samples.
PHASE II
Phase II is traditionally divided into phase IIa, where the aim
is to demonstrate the safety and PK parameters in patients,
and IIb, where the aim is to establish efficacy and delineate
the dose–response curve. However, most companies now
endeavor to generate some data in the phase IIa studies to
provide evidence of efficacy and confidence to progress into
the more expensive and larger phase IIb dose-ranging study.
This is a critical time in the development process, as up to
75% of all drug candidates will fail in phase II. If preclin-
ical data or data from translational medicine studies have
identified a patient population more likely to respond to the
mechanism (e.g., BRAF activating mutation positive mel-
anoma tumors for MEK inhibitors), then the studies can
be restricted to this patient population to increase the like-
lihood of seeing an efficacy signal. Even when there is no
strong a priori hypothesis, samples should be collected in
phase II studies for pharmacogenomic analysis, as they are
useful for testing less-validated hypotheses on the impact
of genetic variation with respect to drug response. These
studies are limited to detecting genetic variants with large
effects, as these studies comprise relatively small numbers
of patients (50–100). Samples for these pharmacogenomic
studies may be collected with specific consent for genotyp-
ing of named genes within the protocol, which can be cor-
related with clinical data collected in the trial. With the
ever-reducing cost of whole-exome and even whole-genome
sequencing, there is a growing trend for collecting samples
with broader consent to include sequencing studies.
PHASE III
Phase III trials form the basis of the regulatory approval
and are often termed pivotal or registration studies. They are
large studies evaluating the safety and efficacy of the can-
didate at the clinical dose and in the population where the
drug will ultimately be used. The cost of this phase of de-
velopment is significantly more than that of the others, so
failure at this point has a major impact on the company.
The larger numbers of patients included in these studies
provide more power for pharmacogenomic analysis. In add-
ition, these samples also provide a useful resource for more
disease-focused phenotype–genotype correlations, and
samples can be collected with broad consent for genotyping
that allows the investigation of many candidate genes.
The patient population studies in the phase III program
form the basis of the population approved to use the drug
once it is launched. Therefore, if a genetically defined patient
population is used in these studies, then the drug will only
be approved for use in that group of patients. Even if the
drug will only be used in a pharmacogenomically defined
population, it is often necessary to include at least one study
where all patient groups are included, to ensure that there
is not an unexpected benefit in the nonselected population
and also to provide a safety database for that group should
they be prescribed the drug once it is approved. The inclu-
sion of a prospectively stratified “all-comers” strategy also
allows a more robust evaluation of the positive and nega-
tive predictive value of the test and, importantly, enables
researchers to differentiate between a predictive pharmaco-
genomics test, where the test identifies subjects who differ-
entially respond to the drug, from a prognostic test, where
the test differentiated subjects with a more severe prognosis
from the disease, regardless of treatment paradigm.
PHASE IV
Drug testing does not stop with regulatory approval, and
phase IV studies are run after the drug has been approved.
Sometimes there are clinical studies required by regulatory
authorities as a post-approval commitment. These generally
test a specific question about safety and efficacy or are used
to generate data to support commercial strategies. Studies
conducted after the regulatory approval of the drug are an
excellent resource for the implementation of a pharma-
cogenomics strategy because of the availability of larger
sample sets. The potential to collect genomic samples from
thousands of individuals recruited into large phase IV clin-
ical studies presents the opportunity to link genomic data
to good-quality clinical data, biomarker data, and, in many
cases, long-term follow-up. An area where post-market
pharmacogenomic surveillance can have a great impact is
in addressing safety issues, thanks to the very large studies.
The availability of large numbers of patients on ac-
tive treatments not only provides the material to look for
pharmacogenomic effects but is also a valuable resource
for understanding the molecular basis for disease, which in
turn feeds back into idea-generation in the early-discovery
section of the pipeline.
The studies performed within drug development pro-
grams are still classified according to this system, but, in-
creasingly, companies are looking to generate potential
signals of efficacy data in the early phase I and IIa studies
1 1 8   •   principles of G enomic M edicine
(sometimes called the learn phase) to provide confidence
that the compound will work, before investing in the more
expensive phase IIb and III studies (sometimes called the
confirm phase).
A P P LY I N G G E N O M I C S TO D RU G
D I S C OVE RY
Choosing the Best Drug Targets
One key area where genetics has impacted the drug dis-
covery and development process is target selection.
Between 50% and 75% of compounds fail in development
due to lack of efficacy, and this is in large part because the
target, and hence the mechanism of action of the drug, are
not linked to the pathogenesis of the disease to which they
are directed. Taking the view that the more you know about
a drug target early in the discovery process, the less likely
it is to fail in development due to lack of confidence in ra-
tionale (CIR), many companies are now investing up front
in understanding the molecular genetics of the complex
diseases we treat, and using genetics to identify novel tar-
gets and prioritize target selection from candidate gene lists
for drug development programs. The advances in DNA
sequencing, bioinformatics, and genetic analysis are provid-
ing great opportunities to use human genetics to identify
novel targets.
Before 1990, pharmaceutical companies had worked
on approximately 500 potential drug targets, with around
100 of these mechanisms having produced marketed drugs
(Hopkins and Groom, 2002). Initial analysis of the final
draft of the human genome project suggested that the total
number of targets druggable with small chemicals might in-
crease to 5000 (Drews, 2000). However, not all of these tar-
gets will be relevant to disease; therefore, current estimates
are that there are 600–1500 drug targets in the human
genome (Hopkins and Groom, 2002). This expansion of
potential targets in concert with the rising costs of drug de-
velopment means that the choice of targets is increasingly
important. This number increases further when biological
approaches are included.
Given the length of time it takes to get from an idea, to
a compound, to the market, there are still only a few pro-
spective examples of marketed compounds where genomics
has provided a new drug target or supported its initial CIR;
thus, there are insufficient data to show that having gen-
etic or genomic CIR from complex traits has significantly
increased candidate survival in the drug-development
pipeline. Human genetics is a simple and effective way of
beginning to assess the molecular evidence and provide
the CIR for establishing a drug development program for
a particular target. It is possible to retrospectively identify
positive genetic associations between drug target and in-
cidence or severity of disease for drugs that are currently
widely prescribed; for example, angiotensin-converting
enzyme inhibitors and hypertension (Zee et  al., 1992;
Province et al., 2003), β-agonists and asthma (Turki et al.,
1995; Santillan et  al., 2003), and serotonin reuptake
inhibitors and depression (Ogilvie et al., 1996; Golimbet
et  al., 2004). Although this is not always the case, as the
proton-pump inhibitors, used to treat gastroesophageal re-
flux disease (GERD), are one of the most frequently pre-
scribed classes of drugs worldwide, but currently very little
is known about the molecular genetics of GERD, and no
reported association between the genes encoding the α and
β subunits of the drug target hydrogen/potassium adeno-
sine triphosphatase (ATPase) and the disease (Post et al.,
2005). Knockout mouse data also provide evidence relevant
to the function of target on the phenotype (Zambrowicz
and Sands, 2003). The CIR for the statins, one of the most
successful drug classes to be developed for the lowering of
low-density lipoprotein (LDL) cholesterol, was derived
from biochemistry. Interestingly the HMG-CoA reductase
knockout mouse is lethal, and there are very few published
genetic association studies on HMG-CoA reductase (Tong
et al., 2004).
C O M P L E X T R A IT G E N ET I C S
The ability to carry out large-scale whole genome studies in
well characterized populations extends the candidate gene
approach, and has increased the potential to identify novel
targets and new pathways that are relevant to disease. The
challenge with these broad approaches is linking the find-
ings back to our understanding of the disease process and
using that knowledge to select a target. Linkage studies have
had some success in identifying genetic variants associated
with complex diseases; examples include phosphodiesterase
4D and stroke (Gretarsdottir et al., 2003), osmoprotectants
taurine cyanate and nitrate (OTCN) cation transporter and
DLG5 (discs large [Drosophila] homologue 5) genes with
inflammatory bowel disease (Peltekova et  al., 2004; Stoll
et al., 2004), and 5-lipoxygenase-activating protein (FLAP)
and myocardial infarction and stroke (Helgadottir et  al.,
2004). To date, these studies have provided some support-
ing evidence for the link between potential drug targets and
disease, but only rarely are they the only evidence support-
ing this link. This is due to the fact that the reproducibility
of early genetic association studies was poor, with many
false positives reported; the identification of the causative
N ew D rug D evelopment, D rug R esponse , and P recision M edicines   •   1 1 9
variant is often challenging; hence making the prediction
of whether the genetic variant is casing an increase or de-
crease in protein function can be a challenge. Three major
advances have occurred in the last decade that have posi-
tively impacted the use of complex trait genetics. The first
was the publication of the Wellcome Trust Case Control
Consortium, which clearly demonstrated the need for
larger sample sizes and rigorous quality control (QC) pro-
cedures (Frazer et al., 2004; John et al., 2004, The Wellcome
Trust Case Control Consortium, 2007). The second ad-
vance has been the rapid development of DNA sequencing,
which in 2013 was reaching a point where it is possible to
sequence large cohorts of subjects, allowing the evaluation
of rare variants as well as the common variants covered by
the whole-genome association studies. The final advance
is the development of bioinformatics and genetic analysis,
which is allowing the combining of the genetic variations
into pathway maps looking for dysregulated pathways ra-
ther than just individual SNPs. This is allowing the identi-
fication of optimal intervention points in pathways and the
design of functional experiments that can confirm the dir-
ection of the dysregulation and hence whether an agonist
or antagonist approach is required. There is therefore a
renewed enthusiasm for the use of complex trait genetics to
inform target choice, and the success of this will play out
over the next five years.
SINGLE GENE DISORDER S
A N D T R A ITS
Although the use of complex-trait genetics has yet to show
real value, the use of rare genetic disorders has proven to
be successful, albeit in a small number of cases. This ap-
proach of using the genetics of rare syndromes to identify
drug targets with high confidence that pharmacological
approaches will mimic the human phenotype has a grow-
ing precedence. The last five years have seen the first cohort
of drugs to reach approval or late-stage clinical develop-
ment where human genetics either identified the target or
provided significant confidence in the approach. Examples
of these drugs are included in Table  8.2, and include
Maraviroc and chemokine receptor 5 (CCR5) (human
immunodeficiency virus [HIV]), tofacitinib and the Janus
kinases ( JAK) (RA), romasozumab and sclerostin (post-
menopausal osteoporosis), and vemurafenib and BRAF
(melanoma).
The identification of CCR5 as a potential therapeutic
target for HIV infection came from the discovery that
CCR5 was a coreceptor required for HIV infection, and
from a genetic study of individuals who, despite multiple
Table 8.2  APPROVED DRUGS BASED ON
GENOTYPE-PHENOTYPE CORRELATION
DRUG GENE PHENOTYPE
Maraviroc CCR5 HIV resistance
Tofacitinib JAK 3 Severe combined immunodeficiency
Romosozumab Sclerostin Sclerosteosis
Plavix P2yR Congenital bleeding
Alirocumab PCSK9 Hypercholesterolemia
In development Nav 1.7 Insensitivity to pain
high-risk exposures, did not become infected with the virus.
The genetic study demonstrated that individuals who were
homozygous for this mutation (CCR5Δ32) and therefore
had no functional CCR5 protein were apparently healthy
and resistant to infection by HIV (Samson et  al., 1996).
Subsequent candidate gene studies have shown that het-
erozygosity for the CCR5Δ32 mutation is associated with
slower progression to AIDS (Michael et al., 1997). Recent
data have shown that a genetic polymorphism in the pro-
moter of the CCR5 gene, resulting in increased CCR5
expression, is more common in individuals rapidly progress-
ing to AIDS (Salkowitz et  al., 2003). Thus, within seven
years of the publication of genetic evidence that CCR5
would be a valid target in HIV therapy, clinical validation
of this drug target was achieved with both Pfizer, Inc., and
Schering-Plough publishing data showing significant viral
load drops in patients with HIV infection treated with the
potent CCR5 antagonists Maraviroc and Schering C, re-
spectively (Feinberg, 2003).
The discovery of JAK and the identification of causa-
tive mutations in the JAK3 gene and severe combined im-
munodeficiency (SCID) highlighted the key role of this
target in cytokine signaling and lymphocyte development
and function, and provided CIR for the development of
a selective JAK3 antagonist for the treatment of rejection
in renal transplantation and rheumatoid arthritis. As with
CCR5 above, the fact that individuals with the mutations
only have the very specific effects of immunodeficiency and
no other apparent deleterious phenotype means that these
genetic data also provide confidence in safety (CIS) for the
therapeutic approach (O’Shea et al., 2004).
Sclerosteosis is a rare genetic condition with only
a small number of affected families in the world. A  key
aspect of the disease phenotype of sclerosteosis is bone
overgrowth. This bone overgrowth is seen in the het-
erozygotes when they have generalized increase in bone
density and mass, and the homozygotes when they have
increased bone growth and density, which can lead to
nerve-entrapment syndromes causing deafness and visual
1 2 0   •   principles of G enomic M edicine
problems. The gene for sclerosteosis was identified in
2005, and the disease is caused by the absence of a protein
called sclerostin. Sclerostin is a secreted protein that is
highly amenable to a biologics approach, and reduction in
circulating sclerostin will lead to increase in bone density.
This led to a collaboration between UCB Celltech and
Amgen to produce an antibody to sclerostin for the treat-
ment of postmenopausal osteoporosis. This antibody has
now been tested in phase IIb trials and has been shown to
increase bone mineral density to a greater extent than do
current therapies.
D RU G G A B L E TA RG ET S A P P ROAC H
An alternative strategy to the single-gene and
whole-genome approaches is to carry out association
studies in a subset of druggable target genes. Several
companies have taken this approach to explore genetic
associations with as many tractable targets as possible
in a wide range of indications. Oxagen is a biopharma-
ceutical company specializing in understanding the gen-
etic basis of common human diseases. One of the main
areas of interest for the company is in G-protein coupled
receptors (GPCRs); 20–30% of marketed drugs are tar-
geted to the products of this class of genes. There are over
750 GPCR genes, thus Oxagen applied a filtering pro-
cess to select the best targets for further analysis, based
on expression profiling, known biology, whether they
have a known drug targeted to them, or whether they
are likely to be chemically tractable, before embarking
on high-throughput genetic analysis (Allen and Carey,
2004). The Structural Genomics Consortium has focused
on kinases (the Kinome). This consortium is funded by
private and public sources and focuses on the identifica-
tion of crystal structures of novel kinases and then the de-
velopment of chemical tools. In concert with this, there
has been considerable effort to identify kinases and their
role in disease. Much of this has focused on the use of
genetic mutations of kinases in cancer and genetic asso-
ciations in conditions such as rheumatoid arthritis.
With the increasing use of genetics to drive
target-identification in well-defined patient populations
comes the dilemma of knowing which of all the targets
identified is the best to take forward. The application of
whole-genome technologies to understanding common
complex disease has also led to new potential targets if they
could be drugged. This increase in the number and type of
targets will provide unprecedented opportunity to fight
disease if we can choose the right targets and the right thera-
peutic approaches.
 Free ebooks ==> www.Ebook777.com
Effect of Genetic Variation on Compound
Screening
Regardless of the original source of the target, genetic analy-
ses are important in understanding how to move forward
in the drug discovery process. Undertaking a comprehen-
sive analysis of the genetic variation that exists in putative
drug targets will provide information that could have a
powerful impact on drug-discovery processes downstream.
In an internal study within Pfizer, Inc., comparing coding
SNP (cSNP) frequency, a selection of 111 genes encoding
potential druggable targets and 160 genes considered as
“non-druggable” targets found that 15% (26/111) of the pu-
tative targets were not polymorphic at the amino acid level,
while 40% (45/111) had one or two cSNPs. There are also
well-documented differences in the frequencies of specific
polymorphisms between ethnic groups. Prior knowledge
of any polymorphisms in a target can be incorporated into
target validation, lead optimization, and inform preclinical
projects supporting the development of the compound.
The effect of genetic variation can be assessed through in
vitro assays that incorporate a comparison of polymorphic
targets by using either cells or biological reagents obtained
from donors of known genotypes (where available), or by
site-directed mutagenesis. This will facilitate early assess-
ment of the potential impact of genetic variation on the
activity of compounds and offer the potential to choose
candidates that are the least likely to be influenced by the
target polymorphism (Penny and McHale, 2005).
Gaining an early understanding of the impact of gen-
etic variation can increase confidence in chemistry (CIC).
For example, CCR5 has been shown to be the second core-
ceptor required for primary HIV infection. As such, it was
a very attractive drug target for the treatment of HIV, as
blockade of CCR5 should reduce HIV entry into cells and
hence lower viral turnover. There have been multiple poly-
morphisms reported in the CCR5 gene, and some of these
have been associated with effects on HIV infection rates
and/or progression from infection to AIDS. A key question
that had to be asked was, What were the functional effects
of these polymorphisms, and would they would impact the
effectiveness of the therapy? Preclinically, it was possible to
demonstrate that the predominant effect of the functional
polymorphisms was to alter receptor expression rather than
structure; hence, the variability could be managed by iden-
tifying a dose that could effectively inhibit viral entry across
a wide range of receptor expression levels.
The pharmacogenomic studies included in the preclin-
ical phase of drug discovery that provide CIR and CIC and
support nomination of a candidate for development are
N ew D rug D evelopment, D rug R esponse , and P recision M edicines   •   1 2 1
www.Ebook777.com
not intended to replace any of the clinical studies required
for exploratory drug development or predict response in
patient populations. The preclinical strategy will produce
data to inform the pharmacogenomic plan for compounds
in exploratory and full development. The challenge facing
pharmacogenomics specialists in the pharmaceutical in-
dustry is to use the available genomic data to improve the
efficiency of clinical trials.
A P P LY I N G G E N O M I C S TO D RU G
D EVE L O PM E N T
Pharmacogenetics
There are several definitions of pharmacogenetics in the lit-
erature, but the term was originally used in 1959 by Vogel
to describe the inter-individual differences in drug response
due to variations in DNA (Vogel, 1959). Although this
is the origin of the term, the concept of inherited differ-
ences in biochemical attributes dates back much further,
with Garrod describing the inheritance of alcaptonuria and
phenylketonuria in 1902, and Snyder in 1932 describing
the inherited ability to taste (or not) phenylthiocarbamide
(Garrod, 1902; Snyder, 1932). The article by Motulsky in
1957 was the first serious attempt to understand the basis
of inherited inter-individual response to drug therapies,
with descriptions of the effects of glucose-6-phosphate
dehydrogenase (G6PD) deficiency and primaquine in
African-American soldiers (Motulsky, 1957). During World
War II, scientists from the University of Chicago observed
that approximately 10% of black American soldiers and
(rarely) some of the white soldiers developed hemolytic an-
emia of varying severity when given conventional doses of a
then-new antimalarial drug, primaquine. Further investiga-
tion revealed that this was due to the lack of the G6PD en-
zyme in red cells, which was the same genetic defect that had
been shown to be responsible for the development of hemo-
lytic anemia in susceptible individuals following the inges-
tion of fava beans. This was one of the first descriptions of a
Mendelian (X-linked) pharmacogenetic trait. Also, in 1957,
Kalow and Genest described an autosomal recessive pharma-
cogenetic trait (Kalow and Genest, 1957). Approximately 1
in 2000 subjects undergoing anesthesia develop a prolonged
pharmacodynamic effect of succinyl choline due to a defi-
ciency in the enzyme pseudocholinesterase. This autosomal
recessive trait has since been recognized in a wide variety of
ethnic populations, and although the enzyme deficiency was
identified in 1957, it was a further 30 years before the causa-
tive genetic mutations responsible for these reactions were
identified (McGuire et al., 1989).
 Pharmacogenetics remained a relatively small field until
the 1990s, due to the fact that although it was well recog-
nized that all drugs exhibited significant inter-individual
variability in response, the genetic tools to examine this
variability were not available. Apart from a few standard
approaches (e.g., renal impairment studies and gender
differences), there was limited investigation of this phe-
nomenon during drug development. The approach of the
drug companies and regulators alike was to ensure that all
compounds had a sufficiently good therapeutic index that
the average benefit significantly outweighed the potential
risk. This has led to the withdrawal or termination of de-
velopment of a number of compounds with good efficacy
but an insufficient population-based safety profile, which
can often be driven by a small number of potentially serious
adverse events. These events can be categorized into those
that are expected based on an understanding of the pharma-
cological action of the drug (type A), and those that cor-
relate with plasma exposure levels or idiosyncratic (type B)
(Rawlins and Thompson, 1991). The mechanisms of idio-
syncratic reactions are generally unknown and do not have
a clear dose–response relationship.
Pharmacokinetic Variability
Inter-individual variation in drug metabolism is now
a well-documented phenomenon, but it was not until
Mahgoub et al, (1977 Lancet 2[8038]:854–856) described
Reduced response
Rapid metabolisers
Excess CYP2D6 activity
Patients
xxx xxxx x x
x xx x x
CYP2D6 metaboliser phenotype
distribution in white Caucasians
Figure 8.2  Individual variation in drug metabolism.
1 2 2   •   principles of G enomic M edicine
the polymorphic metabolism of debrisoquin that significant
interest grew in the genetic contribution. The cytochrome
P450 (CYP) enzyme family protects the body from xeno-
biotic agents and is the major route of metabolism of many
drugs (Danielson, 2002). Several of these enzymes (e.g.,
cytochrome P450 2D6, 2C9, and 2C19) are known to have
functional genetic polymorphisms that result in significant
reductions or increases in function (Lee et al., 2002; Shimizu
et  al., 2003). Genetic variation in cytochrome P450 2D6
(CYP2D6) is well characterized, and approximately 10% of
Caucasians make no CYP2D6 enzyme. Experiments with
the antihypertensive agent debrisoquin yielded the first
proven examples of a pharmacogenetic effect. Debrisoquin
is metabolized by the CYP2D6 enzyme. An individual who
makes no CYP2D6 and takes a standard dose of debriso-
quin will suffer a profound hypotensive event resulting
from high plasma exposure levels due to an inability to me-
tabolize the drug (Idle et al., 1978). Approximately 20% of
all drugs are metabolized by CYP2D6, and subjects who are
unable to make this enzyme are at increased risk of devel-
oping adverse events when taking one of these compounds
(Cascorbi, 2003) (Figure 8.2).
The incorporation of genetic testing for CYP2D6 or
related enzymes in clinical trials has the potential to iden-
tify, prospectively, subjects who are likely to have adverse
events due to poor metabolism, or those who may have
limited response through inadequate exposure because of
ultra-rapid metabolism.
Increased frequency of
adverse events
Normal Poor metabolisers
CYP2D6 No CYP2D6 activity
activity
xxx x
xxx xxx
Plasma drug level
8%
10% 80%
2%
No CYP2D6 activity
Below normal CYP2D6 activity
Normal CYP2D6 activity
Excess CYP2D6 activity
 Many drug-metabolizing enzymes have genetic vari-
ants leading to reduced or increased function, with
consequent impact on the PK variability. Despite this
knowledge, there are few drugs for which pharmacogenetic
tests are routinely applied, and only recently has it become
accepted best practice to test for the presence of variation
in the gene encoding the thiopurine methyltransferase
(TPMT) enzyme before prescription of azathioprin and
6-mercaptopurine.4,5 Approximately 1 in 300 individu-
als is homozygous for mutations in the gene encoding the
TPMT (Evans, 2004). If treated with a standard dose of
azathioprin (6-mercaptopurine), these individuals have a
substantially increased risk of developing the potentially
fatal complication of red cell aplasia (Evans, 2004). Suitable
dose reduction decreases this risk. The recent decision by
the Clinical Pharmacology division of the FDA to recom-
mend that subjects be tested for TPMT enzyme status (ei-
ther phenotypically or genotypically) before dosing with
6-mercaptopurine is evidence of the increasing awareness
of the value of understanding inter-individual variation
in drug metabolism. Similarly, the recently approved drug
Strattera from Eli Lilly provides safety data for poor and ex-
tensive metabolizers of CYP2D6, and the availability of a
suitable test to distinguish these two groups is also included
in the label, although there is currently no recommenda-
tion about using the test and adjusting the dose according
to genotype.
As the clinical value of these tests becomes established
and is translated into practice, so will the acceptability of
requiring a metabolizing enzyme diagnostic before dis-
pensing a drug. Clear demonstration of the advantages of
prospectively using a diagnostic test versus clinical manage-
ment of drug dosing will be vital if these tests are to be used
in clinical practice. This will also allow the development of
chemicals with narrow therapeutic windows and predomin-
antly metabolized by a polymorphic enzyme. Many of these
compounds have historically been terminated, as the risk of
adverse events due to high plasma exposures outweighed the
potential benefit. A clinically acceptable way of managing this
risk would make the safe use of these compounds possible.
Pharmacodynamic Variability
The importance of being able to predict drug response is
highlighted by the fact that it has been estimated that ap-
proximately 30% of prescriptions written do not benefit
the patient, and even in highly controlled environments,
such as clinical trials, it is rare to get response rates sig-
nificantly above 70% (Silber, 2000). If we assume that
subjects take the medication in the prescribed manner,
N ew D rug D evelopment, D rug R esponse , and P recision M edicines   •   1 2 3
then lack of efficacy may result from inadequate exposure
to the drug (PK variability), an inability to respond to
the therapy due to genetic variation in the target and/
or downstream effectors (pharmacodynamic [PD] vari-
ability), or because the pharmacological intervention
does not alter the underlying pathophysiological process
(disease heterogeneity). While some commentators have
suggested that differences in disease genetics (disease
heterogeneity) should be considered as separate from
pharmacogenetics, at a practical level, understanding this
genetic variation will result in the same outcome—for
example, understanding the increased or decreased likeli-
hood of response to therapy. Therefore, this group will be
included in the PD variability subgroup.
There are now multiple examples of the use of
pharmacogenetics to predict drug response. The ma-
jority of these are in oncology, where tumor mutations
have been shown to drive pharmacodynamic response
in multiple areas. The best known examples of this are
Herceptin and Gleevec. In the case of Herceptin, amp-
lification of the Her2 gene leads to up regulated Her2
protein expression in approximately 25% of all breast
cancers. These tumors are responsive to Herceptin,
whilst tumors with lower levels of expression of Her2
do not respond. Imatinib is a treatment for Philadelphia
chromosome positive chronic myeloid leukemia specif-
ically designed to target the BCR-ABL fusion protein
generated from this chromosomal translocation. It also
is active in tumors with mutated KIT genes (e.g., GIST).
Table 8.3 contains a list of anti-tumor therapies aimed at
genotypically defined tumors.
Vemurafenib is a very exciting example, as this com-
pound was screened using the common V600E mutation
of the BRAF gene. This mutation is present in approxi-
mately 60% of melanoma tumors. A  counter-screen of
non-mutated BRAF was also run, ensuring the identified
compound was specific for the mutated allele. This drug is
Table 8.3  ANTI-TUMOR THERAPIES FOR
­G ENOTYPICALLY DEFINED TUMORS
DRUG INDICATION GENE
Herceptin Breast cancer HER2NEU
Gleevec GIST KIT
Gefitinib Non–small cell lung cancer EGFR
Erlotinib Non–small cell lung cancer EGFR
Cetuximab Colorectal cancer KRAS
Pannitumumab Colorectal cancer KRAS
Crizotinib Non–small cell lung cancer Alk4
Vemurafanib Melanoma BRAF
highly effective in V600E-positive tumors and has a very Predicting Type A Adverse Events
good safety profile, as it does not bind to the non-mutated
Adverse drug reactions (ADRs) are a major cause of mor-
protein, hence only working within the tumor cells.
bidity, leading to approximately 5% of all hospital admis-
Although the majority of examples are in oncology,
sions, and severe adverse drug reactions are a leading
there are exemplars in other therapeutic areas as well.
cause of death in young adults. Despite initial optimism,
One of the clearest examples is in the treatment of hepa-
pharmacogenetics has had limited impact in reducing this
titis C. Subjects who have the AA polymorphism in their
morbidity and mortality. There is, however, evidence that
interferon gene have a greater chance of responding to
genetic variation can influence our risk of developing type
interferon therapy than do individuals who are AT or
1 adverse events by either increasing our exposure to the
TT. Other examples exist, particularly in the rare disease
active agent or altering the pharmacodynamics effects of
field, where therapies are directed at specific genetic dis-
the drug. Warfarin is one of the best understood examples
orders, and in this case, it is disease genetics rather than
of how genetic variation can influence the risk of adverse
pharmacogenetics.
events. Bleeding events on warfarin are among the com-
Despite the success stories described over the last few
monest adverse events resulting in significant morbidity.
years, most therapies tested to date do not appear to have a
Underlying genetic variation accounts for at least 50% of
clear pharmacogenetic signature. It may be that the current
the risk of developing a bleeding event. This risk is predom-
approaches are unable to identify the correct genetic vari-
inantly driven by two key genes; the drug-metabolizing
ation or (more likely) the combination of variants that can
enzyme cytochrome P450 2C19, and the gene encod-
predict response, or it may be that genetic variation is not a
ing the vitamin K receptor. Studies by Pirmohamad et al.
major cause of the heterogeneity of drug response.
have shown that poor metabolizer status of cytochrome
P450 2C19 have a Y-fold increase in plasma exposure of
P R E D I C T I N G S A FET Y
S-warfarin (the active moiety). The increase in exposure
Predicting Type B Adverse Events results in a Z-fold increase in bleeding risk due to pharma-
cokinetic variability. The vitamin K receptor is the target
The last few years have demonstrated that pharmacogenetics
for warfarin and is required for the production of vitamin
can be used to predict some rare adverse events. Extreme
K–dependent clotting factors. A  common variant in this
pharmacodynamics adverse responses to drugs have been
receptor results in a decrease in vitamin K receptor func-
described in the past, such as malignant hyperthermia
tion. Whilst this normally causes no significant sequelae,
and inhaled anesthetics, succinyl choline deficiency, and
it does affect response to warfarin. Individuals who are
prolonged paralysis. More recently, an immunogenetic
homozygous for the rare allele have an increase in bleeding
explanation for rare hypersensitivity reactions was dis-
risk of Y when taking warfarin. By combining the results of
covered. Abacavir was a key drug in highlighting the role
these genotypes it is possible to refine an individual’s risk
of HLA variation and drug hypersensitivity. Two retro-
of developing a bleeding adverse event if they are given a
spective studies have identified the HLA-B*5701 allele of
standard dose of warfarin. Prospective trials are now on-
the major histocompatibility complex (MHC) class  I  B
going to determine the utility of using genotype results to
gene as a genetic determinant of hypersensitivity to abacavir
adjust the starting dose of warfarin.
(Ziagen) (Hetherington et  al., 2002; Mallal et  al., 2002).
The availability of a relatively large patient population led
Individualized Therapy—An Integrated Response
to the identification of the HLA-B*5701-Hsp70-Hom
variant haplotype in 94.4% of cases compared to only 0.4% In real life, the response of an individual is based on both
of controls. Analysis in different ethnic groups, however, the plasma exposure and how that affects the various physio-
showed that HLA-B*5701 alone would not be sufficiently logical processes in the target organs. Evans and Relling
predictive of hypersensitivity in diverse patient populations, generated a hypothetical graph representing the PK and PD
suggesting that other genetic determinants of hypersensi- variation in concert (Evans and Relling, 1999).
tivity remain to be identified. Additional HLA associations Variation in drug-metabolizing enzymes can dramatic-
with adverse drug reactions have been described. Chung ally impact plasma exposure levels (see left-hand column in
et al. in 2004 described an association between HLA B1502 Figure 8.3). However, it is not until we integrate this with
and Stevens Johnson syndrome in the Han Chinese popu- variation in genes affecting PD response (in the right-hand
lation. Again, this association appears to be confined to the column) that we start to get a real understanding of the
Han Chinese. impact on response for the individual. It is important to

1 2 4   •   principles of G enomic M edicine

 Pharmacokinetic Pharmacokinetic pharmacodynamic realize that dose-related adverse events are observed in ex-
effect integrated effect tensive metabolizers as well as poor metabolizers, but the
(A)
Drug conc.
100
wt/wt
100
wt/wt incidence is dependent upon the frequency of variation
in the genes affecting PD response. As the frequency of

Effect (%)
50 50 wt/m
30 variation in genes affecting PD response approaches 0.5,
m/m
0
0 24 hr
0
0 50 100 the predictive power of a test solely looking at drug me-
tabolism decreases. Similarly, the predictive power of a
(B) 100
wt/m
100
wt/wt test evaluating variation in genes impacting PD response
Drug conc.

Effect (%)
50 50 wt/m will vary depending upon PK variability. Most pharma-
65 cogenetic studies that have been published to date have
m/m
0 0
0 24 hr 0 50 100 concentrated on single genes or small numbers of can-
didate genes, which are likely to affect either PK or PD
100
m/m
100
wt/wt variability. It is unsurprising that these studies fail to dem-
Drug conc.

Effect (%)

50 99 50 wt/m onstrate high positive or negative predictive information

m/m for drug response, as it is generally due to a combination of
0 0
0 24 hr 0 50 100 both of these factors. As we move forward, a more holistic
Time Drug concentration approach to the examination of genetic factors impacting
Figure 8.3 
Drug response due to pharmacokinetic (PK) and drug response should lead to the identification of sets of
pharmacodynamic (PD) interactions. The impact of genetic variation SNPs with higher predictive values, leading to improved
leading to altered plasma exposures depends on the variation in the prescribing (Table 8.4).
genes leading to the effector mechanisms of the drug.

Table 8.4  EXAMPLES OF DRUG RESPONSE MODIFICATION ASSOCIATED WITH GENETIC POLYMORPHISMS

IN “DISEASE-MODIFYING” OR “TREATMENT-MODIFYING” GENES

GENE OR GENE PRODUCT DISEASE OR DRUG EFFECT MEDICATION INFLUENCE OF

POLYMORPHISM
Adducin Hypertension Diuretics Myocardial infarction or
stroke
Apolipoprotein E (APOE) Atherosclerosis, ischemic cardio Statins (simvastatin) Enhanced survival
­vascular events
Apolipoprotein E (APOE) Alzheimer’s disease Tacrine Clinical improvement
HLA Toxicity Abacavir Hypersensitivity reaction
Cholesterol ester transfer protein Progression of atherosclerosis Statins (e.g., pravastatin) Slowing of atherosclerosis
(CETP)
Ion channels (HERG, KvLQT1, Congenital long-QT syndrome Erythromycin, cisapride, terfena- Increased risk of drug,
Mink MiRP1 dine, clarithromycin, quinidine) induced torsade de pointes
Methylguanine methyltransferase Glioma Carmustine Response of glioma to
(MGMT) carmustine
Parkin Parkinson’s disease Levodopa Clinical improvement and
levodopa-induced dyskinesias
Prothrombin and factor V Deep-vein thrombosis and cerebral Oral contraceptives Increased risk of deep-vein
vein thrombosis and cerebral-vein thrombosis
with oral contraceptives
Stromelysin-I Atherosclerosis progression Statins (pravastatin) Reduction in cardio-vascular
events—death, ­myocardial
infarction, stroke, ­angina;
­reduction in risk of
angioplasty

Adapted from Evans WE, McLeod HL. Pharmacogenomics—drug disposition, drug targets, and side effects. N Engl J Med. 2003 Feb 6;348(6):538–549.

N ew D rug D evelopment, D rug R esponse , and P recision M edicines   •   1 2 5

 Improving Disease Classification: Stratified
Medicines
The need to accurately and precisely characterize the disease
under investigation has important implications in drug de-
velopment. The current disease classification system has
changed little in the last two hundred years and is based on
the phenotypical clustering of symptoms. That is, diseases
that present with similar symptoms have been classified
as the same condition. These diseases are therefore more
like syndromes and do not necessarily reflect a common
underlying pathology. Similarly, there may be conditions
with similar pathological mechanisms that are currently
classified as different diseases, as the phenotypical features
are not similar enough. A  very clear example of this is in
oncology, where many mechanisms are represented in sub-
sets of organ-classified tumors:  for instance, EGFR muta-
tions are present in multiple tumor types. The knowledge
from the outset of a drug discovery program that there are
molecular subtypes of a disease means that appropriate
preclinical experiments can be developed early to predict
the likelihood of a pharmacogenomic effect, and this in-
formation can be used advantageously in the drug devel-
opment program. Combining genotype data with other
genomic data provides valuable information about the
disease subtype. Integration of genotyping data with gene
expression, for example, has identified subtypes of obesity
phenotypes in a mouse model (Schadt et al., 2005). Using
similar approaches and including microRNA, epigenetic,
proteomic, and metabonomic analyses in well-defined pa-
tient cohorts will provide powerful tools to aid the dissec-
tion of the phenotype of disease in humans in order to drive
the development of targeted therapies based on molecular
sub-classification of diseases (disease stratification). This
reclassification of disease has become the focus of several
cross-academic/industry consortiums, and the next decade
could see the development of new disease taxonomies
reflecting the true molecular mechanisms of the patholo-
gies, rather than their consequences.
One therapeutic area where using genetic and genomic
technologies has undoubtedly had a major and measur-
able impact on understanding the molecular subtypes of
disease is oncology. The advances in understanding the mo-
lecular mechanisms predisposing to cancer have seen the
number of oncology compounds in clinical development
rise from 10 to over 400 in a 10-year period. The majority
of the new compounds now being tested are classed as “tar-
geted biotech medicines.” Imatinib mesylate (Gleevec) and
trastuzumab (Herceptin) were the first two such targeted
compounds approved. Herceptin is a therapy targeting the
1 2 6   •   principles of G enomic M edicine
HER2/neu receptor in breast cancer. The rationale for this
therapy was based on a sound understanding of the underly-
ing molecular pathology. It was known that only 20–30%
of breast tumors overexpress this protein, and it was dem-
onstrated in the drug development program that response
to Herceptin was limited to subjects whose tumors overex-
pressed the target (Vogel et al., 2002). Similarly, Gleevec is a
therapy targeting the fusion protein product resulting from
the Philadelphia chromosomal translocation observed in
most cases of chronic myeloid leukemia (CML) (Deininger
et  al., 1997). This therapy provided dramatic efficacy in
cases of CML with the chromosomal translocation, and it
was rapidly approved by the Food and Drug Administration
(FDA).
Following the rapid approval and success of Gleevec
and Herceptin, many other targeted cancer therapies have
entered clinical trials, thus highlighting the absolute require-
ment to continue to investigate and understand the underly-
ing molecular mechanisms that are associated with disease.
Gefitinib (Iressa) was the first in class selective EGFR in-
hibitor to receive accelerated approval based on preliminary
data from phase II studies in non–small cell lung carcinoma
(NSCLC) patients. Activating mutations and overexpres-
sion of EGFR were known to occur in many cancers, provid-
ing CIR for development of an EGFR-inhibitor for cancer
treatment. Inactivation of the EFGR gene in mice did not
cause any major phenotypical effects, which fact in turn
provided CIS with respect to pharmacological inhibition of
this target (Wong, 2003). However, initial tumor response
to treatment in the clinical trials of subjects with non–small
cell lung cancer was only observed in 9–19% of patients.
Subsequent analysis to predict factors that would indi-
cate good response to Iressa identified that female gender,
nonsmoking status, and a specific histological subtype
of tumor were associated with better response to therapy.
Investigation of biological and markers of response failed to
show an association with EGFR expression levels. However,
somatic mutations in the ATP-binding site of the tyrosine
kinase domain of EGFR were observed more frequently in
the tumors of patients who responded to Iressa. The EGFR
mutations are located close to the putative binding site for
compounds like Iressa and lead to increased signaling in the
growth factor pathway; therefore, tumors harboring these
mutations are more susceptible to treatment with an EGFR
inhibitor (Lynch et al., 2004). This highlights the import-
ance of defining the molecular subtypes of disease and
understanding the impact on response to therapy. Had the
molecular profile of NSCLC been identified before testing
in humans, it may have been possible to design preclinical
cell-based assays to determine whether the genetic profile
of the tumor would influence response to therapy and then
inform clinical trial design.
The majority of oncology programs now in develop-
ment are focusing on stratified populations based on gen-
etic or genomic classifications of tumor type.
P H A R M AC O G E N O M I C S
A N D M A R K ET E D D RU G S
Pharmacovigilance
In a recent study of adverse drug reactions (ADRs), 5% of
hospital admissions in the United Kingdom were identified
as being due to ADRs. Over 70% were considered avoid-
able, and while drug interactions accounted for the ma-
jority of the ADRs, and older drugs were implicated in the
hospital admission, there is still a need to understand the
underlying causes of all ADRs (Pirmohamed et al., 2004).
It is difficult to detect rare adverse events in the confines of
a clinical trial, due to the relatively small number of subjects
in the study, and the current system for monitoring ADRs
has been suggested to be “too disparate.” A move to a more
comprehensive epidemiological approach to monitoring
drug safety has been proposed. The inclusion of pharma-
cogenomic analyses within this approach would allow the
systematic assessment of the contribution of genetic deter-
minants to ADRs. Pharmacogenomic surveillance in large
phase IV trials of approved compounds has the potential to
have a great impact in addressing safety issues.
One therapeutic area where detailed pharmacosurveil-
lance, including pharmacogenomic analyses, post-approval,
is not new, is in the antiretroviral treatment of HIV infec-
tion. Viral resistance and drug toxicity are common and
often lead to treatment failure. HIV genetic sequences are
determined, and the viral load is constantly monitored
to assess viral resistance to highly active antiretroviral
therapy (HAART). Polymorphisms in drug transporters
and drug-metabolizing enzymes have also been monitored
in HIV therapy. Two retrospective studies have identified
the HLA-B*5701 allele of the major histocompatibility
complex (MHC) class  I  B gene as a genetic determinant
of hypersensitivity to abacavir (Ziagen) (Hetherington
et al., 2002; Mallal et al., 2002). The availability of a rela-
tively large patient population led to the identification of
the HLA-B*5701-Hsp70-Hom variant haplotype in 94.4%
of cases compared to only 0.4% of controls. Analysis in dif-
ferent ethnic groups, however, showed that HLA-B*5701
alone would not be sufficiently predictive of hypersensitivity
in diverse patient populations, suggesting that other genetic
determinants of hypersensitivity remain to be identified.
N ew D rug D evelopment, D rug R esponse , and P recision M edicines   •   1 2 7
Implementation of pharmacogenetics post-approval will
have a role in increasing the CIS of new products.
P H A R M AC O G E N O M I C S I N C L I N I C A L
P R AC T I C E
There are two clear areas where pharmacogenomics has
impacted clinical practice and is now being widely used. The
first is oncology, where genetic profiling of non–small cell
lung cancer, colorectal cancer, and breast cancer is routinely
used to drive treatment choice. This has been a relatively
rapid change in practice with the approval of multiple tar-
geted drugs (Table 8.3) and is likely to grow further as our
increasing understanding of tumor biology is matched with
new targeted therapies (e.g., Crizotinib and Cetuximab).
The second area is in infection, where we have seen routine
testing become established for the HIV therapy abacavir
and the hepatitis C therapy interferon C.  Both of these
areas are used to complex prescribing, which has enabled
the more rapid integration of testing into the treatment
paradigm.
However, this success has not been seen in all areas.
Despite extensive knowledge of the genetics of CYP2D6
and related enzymes, and their involvement in the metab-
olism of many commonly used drugs, drug-metabolizing
PG has had little impact in the clinic. Multiple case-control
studies have implicated the role of genetic variants in these
enzyme systems and the risk of adverse events (Brockmoller
et  al., 2002; Rau et  al., 2004; Steimer et  al., 2005). These
studies have investigated both specific compounds and
adverse event rates in drugs metabolized by polymorphic
enzymes compared with non-polymorphic pathways. The
failure to implement genetic testing for these variants in
the clinic and appropriate adjustments in dosing is due to
a number of factors. The lack of appropriately designed
prospective trials demonstrating the clinical benefits of
this approach, and inconsistency of results in some of the
retrospective case-control studies, are often cited as reasons
for the lack of clinical usage. Additional factors include the
need for rapid, easy testing and increased genetic education
for many healthcare groups: such as physicians, nurses, and
pharmacists.
The degree to which pharmacogenetics is incorporated
into mainstream clinical practice depends not only on the
science but also on the regulatory and societal environment.
To date, there has been little impact in the clinic, but the
available tests have had limited predictive value, and there
have been few good prospective studies performed. As the
science progresses, the regulatory and societal factors will
become more important. The nations’ regulatory authorities
are responsible for ensuring that all drugs licensed for use
have an appropriate risk–benefit ratio. However, this ratio
is an average based on efficacy across the total treated popu-
lation and on the adverse event rate across the same popula-
tion. Approval can be, and has been, refused for drugs that
offer significant benefit but have serious adverse effects in a
few subjects. The ability to detect the subjects at increased
risk of these adverse events would allow these drugs to be
used safely. The number of drugs withdrawn in the last 5 to
10 years reflects the increasingly risk-averse regulatory en-
vironment. The potential of preventing these withdrawals
in the future by identifying at-risk subjects has stimulated
significant interest from the regulatory authorities. It is
likely that, in the future, the identification or confirmation
of these adverse events following a drug’s launch will stimu-
late research into the precise mechanism of the event and
strategies to identify subjects at risk, rather than an imme-
diate withdrawal of the drug. While studies to understand
the mechanisms of ADRs have always been attempted
during drug development, pharmacogenetics offers the po-
tential not only to understand why a reaction has occurred
but also to identify who is at risk of it, before administra-
tion of the drug. The regulatory authorities have been a key
driver of the use of pharmacogenetics to improve the safety
profiles of drugs.
An improved efficacy profile for a compound is im-
portant in the context of gaining drug approval, but it can
be vital when drug reimbursement and use are considered.
The use of drugs is primarily driven by the physicians who
prescribe them and the healthcare infrastructures that re-
imburse their costs, such as the National Health Service
(NHS) in the United Kingdom or health maintenance
organizations (HMOs) in the United States. In order for
the use of newer drugs to be justified, there needs to be
significant benefit over existing therapies, which may be
generic and have proven safety profiles. It is possible to use
pharmacogenetics to improve efficacy profiles by identify-
ing subjects who are likely to respond well (or those likely
to get minimal benefit) and targeting the therapies accord-
ingly. This use of pharmacogenetics is driven by the payers
for the therapies, as the increasing pressure on healthcare
budgets means that paying for more expensive branded
therapies can only be justified for patients likely to gain sig-
nificant benefit.
While the role of the regulators and healthcare payers
in driving forward the use of pharmacogenetics is already
emerging, it is unclear what the role of the patient will be,
although it is clear that this could be significant. The risks
of taking a medication must always be placed in context
with the potential benefits and not treated in isolation. It
1 2 8   •   principles of G enomic M edicine
may be perfectly acceptable to license and use a drug with
significant risks if the potential benefits are substantial, as
in cancer therapies. Meanwhile, in other situations, very
little risk of adverse events can be tolerated (e.g., erectile
dysfunction). The indication being treated and the current
available therapies are the key determinants of the level
of adverse events that would be tolerable for the efficacy
observed. As the science becomes more sophisticated and
the prediction gets better, it will then be possible to pro-
vide more refined risk–benefit ratios for each individual. It
is unclear how this range of risk benefits will be managed.
Traditionally the regulatory authorities have, in conjunc-
tion with independent experts, determined what is an ac-
ceptable population-based risk–benefit ratio. This average
risk–benefit ratio may soon become a range of risks and
benefits, and the acceptability of an individual risk–benefit
ratio will become a question for the patient and his or her
physician rather than the regulators. As the patients’ role in
drug selection becomes more central to the prescribing pro-
cess, so will their influence on drug licensing and the use of
pharmacogenetics.
S U M M A RY
Pharmacogenomics offers great promise to all stakeholders
in the healthcare community. To industry, it offers the po-
tential of improving the efficiency of drug development by
reducing the current high failure rate through better choice
of targets and improved understanding of drug response
early in development. To the healthcare providers, it offers
the potential to reduce the burden of adverse events by
identifying the subjects at increased risk and offering them
alternative therapies, as well as targeting their resources to
use newer, more expensive treatments on subjects who will
derive most benefit. Finally, and most importantly, it offers
to the patient the opportunity, with their physician, to
identify from the range of available therapeutic options the
one most suited to them. While pharmacogenetic testing is
unlikely to be able to guarantee that the therapy will work
and will not cause an adverse event, it will increase the prob-
ability that a drug will work and reduce uncertainty about
adverse events, and provide a rational way of choosing be-
tween therapies.
As our understanding of genomics improves, so will our
ability to determine key factors involved in variability of
drug response. The quest for precision medicines will start
at the beginning of the drug discovery process, with more
comprehensive understanding of the molecular basis of the
disease, patient stratification, and the role of the drug target
in the pathological process. Significant variability in PKs
will be explained by systematic evaluation of all the relevant
metabolizing enzymes and transport proteins. The drug
candidates will only be tested in patients with suitable vari-
ants of the drug target. Drugs will be approved with variable
dosage levels dependent upon underlying genotypes affect-
ing PKs and variation at the drug target. Finally, pharmaco-
genetics will not stop with a drug’s approval: post-marketing
research will endeavor to identify the causes of rarer adverse
events, leading to continuous refinement of how we use
drugs throughout their lifecycle.
REFERENCES
Allen MJ, Carey AH (2004). Target identification and validation
through genetics. Drug Discov Today. 3(5):183–191.
Brockmoller JJ, Kirchheiner, et al. (2002). The impact of the CYP2D6
polymorphism on haloperidol pharmacokinetics and on the outcome
of haloperidol treatment. Clin Pharmacol Ther. 72(4):438–452.
Cascorbi I (2003). Pharmacogenetics of cytochrome p4502D6: genetic
background and clinical implication. Eur J Clin Invest. 33(Suppl
2):17–22.
Chung WH, Hung SI, Hong HS, et al. Medical genetics: a marker for
Stevens–Johnson syndrome. Nature. 2004;428:486. (PubMed)
Danielson PB (2002). The cytochrome P450 superfamily:  biochem-
istry, evolution, and drug metabolism in humans. Curr Drug Metab.
3:561–597.
Deininger MW, Goldman JM, et al. (1997). The tyrosine kinase inhibitor
CGP57148B selectively inhibits the growth of BCR-ABL-positive
cells. Blood. 90(9):3691–3698.
DiMasi JA, Hansen RW, et al. (2003). The price of innovation: new esti-
mates of drug development costs. J Health Econ. 22(2):151–185.
Drews J (2000). Drug discovery:  a historical perspective. Science.
287(5460):1960–1964.
Evans WE (2004). Pharmacogenetics of thiopurine S-methyltransferase
and thiopurine therapy. Ther Drug Monit. 26(2):186–191.
Evans WE, Relling MV (1999). Pharmacogenomics:  translat-
ing functional genomics into rational therapeutics. Science.
286(5439):487–491.
Feinberg J (2003). Meeting notes from the 43rd Interscience Conference
on Antimicrobial Agents and Chemotherapy (ICAAC). New
CCR5 antagonist shows antiretroviral effect. AIDS Clin Care.
15(11):94–95.
Frazer K, Seymour AB, et al. (2004). Identification of the genetic basis
of individual differences in human serum high-density lipoprotein
cholesterol (HDL-C) concentration. Proceedings-American Society
of Human Genetics, Los Angeles, California, USA.
Garrod A (1902). The incidence of alcaptonuria: a study in chemical in-
dividuality. Lancet. 2:1616–1620.
Golimbet VE, Alfimova MV, et al. (2004). Serotonin transporter poly-
morphism and depressive-related symptoms in schizophrenia. Am J
Med Genet. 126B(1):1–7.
Gretarsdottir S, Thorleifsson G, et  al. (2003). The gene encoding
phosphodiesterase 4D confers risk of ischemic stroke. Nat Genet.
35(2):131–138.
Helgadottir A, Manolescu A, et  al. (2004). The gene encoding
5-lipoxygenase activating protein confers risk of myocardial infarc-
tion and stroke. Nat Genet. 36(3):233–239.
Hetherington S, Hughes AR, et  al. (2002). Genetic variations in
HLA-B region and hypersensitivity reactions to abacavir. Lancet.
359(9312):1121–1122.
N ew D rug D evelopment, D rug R esponse , and P recision M edicines   •   1 2 9
Hopkins AL, Groom CR (2002). The druggable genome. Nat Rev Drug
Discov. 1(9):727–730.
Idle JR, Mahgoub A, et al. (1978). Hypotensive response to debrisoquine
and hydroxylation phenotype. Life Sci. 22(11): 979–983.
John S, Shephard N, et  al. (2004). Whole-genome scan, in a complex
disease, using 11,245 single-nucleotide polymorphisms: comparison
with microsatellites. Am J Hum Genet. 75(1):54–64.
Kalow, W, Genest K (1957). A method for the detection of atypical
forms of human serum cholinesterase; determination of dibucaine
numbers. Can J Biochem Physiol. 35(6):339–346.
Lee CR, Goldstein JA, et al. (2002). Cytochrome P450 2C9 polymor-
phisms:  a comprehensive review of the in-vitro and human data.
Pharmacogenetics. 12(3):251–263.
Lynch TJ, Bell DW, et al. (2004). Activating mutations in the epidermal
growth factor receptor underlying responsiveness of non–small cell
lung cancer to gefitinib. N Engl J Med. 350(21):2129–2139.
Mallal S, Nolan D, et  al. (2002). Association between presence of
HLA-B*5701, HLA-DR7, and HLA-DQ3 and hypersensi-
tivity to HIV-1 reverse-transcriptase inhibitor abacavir. Lancet.
359(9308):727–732.
McGuire MC, Nogueira CP, et al. (1989). Identification of the structural
mutation responsible for the dibucaine-resistant (atypical) variant
form of human serum cholinesterase. Proc Natl Acad Sci U  S  A.
86(3):953–957.
Michael NL, Louie LG, et al. (1997). The role of CCR5 and CCR2 poly-
morphisms in HIV-1 transmission and disease progression. Nat Med.
3(10):1160–1162.
Motulsky AG (1957). Drug reactions, enzymes, and biochemical gen-
etics. J Am Med Assoc. 165(7):835–837.
Ogilvie AD, Battersby S, et  al. (1996). Polymorphism in serotonin
transporter gene associated with susceptibility to major depression.
Lancet. 347(9003):731–733.
O’Shea JJ, Husa M, et  al. (2004). Jak3 and the pathogenesis of severe
combined immunodeficiency. Mol Immunol. 41(6–7):727–737.
Peltekova VD, Wintle RF, et al. (2004). Functional variants of OCTN
cation transporter genes are associated with Crohn disease. Nat
Genet. 36(5):471–475.
Penny MA, McHale D (2005). Pharmacogenomics and the drug discovery
pipeline: when should it be implemented? Am J Pharmacogenomics.
5(1):53–62.
Pirmohamed M, James S, et al. (2004). Adverse drug reactions as cause of
admission to hospital: prospective analysis of 18,820 patients. BMJ.
329(7456):15–19.
Post JC, Ze F, et al. (2005). Genetics of pediatric gastroesophageal reflux.
Curr Opin Allergy Clin Immunol. 5(1):5–9.
Province MA, Kardia SL, et al. (2003). A meta-analysis of genome-wide
linkage scans for hypertension:  the National Heart, Lung and
Blood Institute Family Blood Pressure Program. Am J Hypertens.
16(2):144–147.
Rau T, Wohlleben G, et al. (2004). CYP2D6 genotype: impact on ad-
verse effects and nonresponse during treatment with antidepres-
sants—a pilot study. Clin Pharmacol Ther. 75(5):386–393.
Rawlins, M, Thompson J (1991). Mechanisms of adverse drug reac-
tions. In: D Davies, ed. Textbook of Adverse Drug Reactions. Oxford,
UK: Oxford University Press. :18–45
Roses AD, Burns DK, et  al. (2005). Disease-specific target selec-
tion; a critical first step down the right road. Drug Discov Today.
10(3):177–191.
Salkowitz JR, Bruse SE, et al. (2003). CCR5 promoter polymorphism
determines macrophage CCR5 density and magnitude of HIV-1
propagation in vitro. Clin Immunol. 108(3):234–240.
Samson M, Libert F, et  al. (1996). Resistance to HIV-1 infection in
Caucasian individuals bearing mutant alleles of the CCR-5 chemo-
kine receptor gene. Nature. 382(6593):722–725.
Santillan AA, Camargo CA Jr, et  al. (2003). Association between
beta2-adrenoceptor polymorphisms and asthma diagnosis among
Mexican adults. J Allergy Clin Immunol. 112(6):1095–1100.
Schadt EE, Lamb J, et al. (2005). An integrative genomics approach to
infer causal associations between gene expression and disease. Nat
Genet. 37(7):710–717.
Shimizu T, Ochiai H, et  al. (2003). Bioinformatics research on
inter-racial difference in drug metabolism. I. Analysis on frequencies
of mutant alleles and poor metabolizers on CYP2D6 and CYP2C19.
Drug Metab Pharmacokinet. 18(1):48–70.
Silber BM (ed.) (2000). Pharmacogenomics, biomarkers and the promise
of personalised medicine. Pharmacogenetics-Pharmacogenomics.
Snyder L (1932). Studies in human inheritance. IX. The inheritance of
taste deficiency in man. Ohio J Sci. 32:436–468.
Steimer W, Zopf K, et al. (2005). Amitriptyline or not, that is the ques-
tion: pharmacogenetic testing of CYP2D6 and CYP2C19 identifies
patients with low or high risk for side effects in amitriptyline therapy.
Clin Chem. 51(2):376–385.
Stoll M, Corneliussen B, et  al. (2004). Genetic variation in DLG5
is associated with inflammatory bowel disease. Nat Genet. 36(5):
476–480.
Thompson JF, Durham LK, et al. (2005). CETP polymorphisms asso-
ciated with HDL cholesterol may differ from those associated with
cardiovascular disease. Atherosclerosis. 181(1):45–53.
Thompson JF, Lira ME, et al. (2003). Polymorphisms in the CETP gene
and association with CETP mass and HDL levels. Atherosclerosis.
167(2):195–204.
1 3 0   •   principles of G enomic M edicine
Tong Y, Zhang S, et al. (2004). 8302A/C and (TTA)n polymorphisms in
the HMG-CoA reductase gene may be associated with some plasma
lipid metabolic phenotypes in patients with coronary heart disease.
Lipids. 39(3):239–241.
Turki J, Pak J, et  al. (1995). Genetic polymorphisms of the beta
2-adrenergic receptor in nocturnal and non-nocturnal asthma.
Evidence that Gly16 correlates with the nocturnal phenotype. J Clin
Invest. 95(4):1635–1641.
Vogel CL, Cobleigh MA, et al. (2002). Efficacy and safety of trastuzumab
as a single agent in first-line treatment of HER2-overexpressing meta-
static breast cancer. J Clin Oncol. 20(3):719–726.
Vogel F (1959). Moderne probleme der humangenetik. Ergeb Inn Med
Kinderheilkd. 12:52–125.
Wong RW (2003). Transgenic and knock-out mice for deciphering the
roles of EGFR ligands. Cell Mol Life Sci. 60(1):113–118.
Zambrowicz BP, Sands AT (2003). Knockouts model the 100
best-selling drugs—will they model the next 100? Nat Rev Drug
Discov. 2(1):38–51.
Zee RY, Lou YK, et  al. (1992). Association of a polymorphism of the
angiotensin I-converting enzyme gene with essential hypertension.
Biochem Biophys Res Commun. 184(1):9–15.
Wellcome Trust Case Control Consortium. Genome-wide association
study of 14,000 cases of seven common diseases and 3,000 shared
controls. Nature. 7 June 2007;447:661–678.
 Free ebooks ==> www.Ebook777.com
9.
MITOCHONDRIAL GENETICS AND GENOMICS
IN CLINICAL MEDICINE
Agnès Rötig and Dhavendra Kumar
INTRODUCTION
A number of hereditary disorders present with complex
genetic inheritance that do not follow the conventional
principles of inheritance as outlined in Chapter 1. There is
now overwhelming evidence that some hereditary diseases
follow unusual mechanisms that suggest nontraditional
inheritance. The mechanisms involve epigenetics (epig-
enomics), phenotypical effects of genome-wide copy num-
ber variation, structural genomic aberrations resulting from
non-homologous allelic recombination involving unstable
genomic regions flanked by low copy number repeats, and
mutations within the mitochondrial DNA (mtDNA)
molecule.
Mitochondria are central to cellular energy production
and maintenance mediated through oxidative phosphoryla-
tion (OXPHOS) in the respiratory chain (RC) that results
in oxygen consumption and the production of adenine
triphosphate (ATP). These are unevenly distributed across
all organs and body systems except red blood cells. Organs
that require rapid and high energy turnover would evi-
dently possess a relatively higher concentration of physi-
ologically active mitochondria; for example, brain, eyes,
the vestibulo-cochlear part of the auditory system, endo-
crine glands, the heart, the liver, the pancreas, the kidneys,
and the gastrointestinal tract. Qualitative and quantitative
defects in the mitochondrial system are known to be asso-
ciated with multi-organ and multi-system manifestations
(Figure 9.1).
The mitochondrial RC includes several proteins
encoded by a complex set of DNA sequences distributed
across the nuclear and mitochondrial genomes. This chapter
outlines the basic components of the mitochondrial genet-
ics that are relevant to systemic clinical medicine. Emphasis
is given to complex disorders related to respiratory chain
defects. Description and delineation of all mitochondrial
diseases are beyond the scope of this chapter; an interested
131
www.Ebook777.com
reader or student is encouraged to consult other resources
(see “Further Reading”).
R E S P I R ATO RY C H A I N
The mitochondrial respiratory chain (RC) catalyzes the
oxidation of fuel molecules by oxygen, and the concomitant
energy transduction into adenosine triphosphate (ATP) via
five complexes embedded in the inner mitochondrial mem-
brane1 (Figure 9.2). Complex I (CI, dihydronicotinamide
adenine dinucleotide dehydrogenase [NADH]-coenzyme
Q reductase) carries reducing equivalents from NADH
to coenzyme Q (CoQ, ubiquinone) and comprises more
than 40 different polypeptides. Complex II (CII, succinate-
CoQ reductase) carries reducing equivalents from FADH2
to CoQ and comprises four polypeptides, including the
flavine adenine dinucleotide (FAD)-dependent succinate
dehydrogenase and iron-sulfur proteins. Complex III (CIII,
reduced CoQ-cytochrome c reductase) carries electrons
from CoQ to cytochrome c, and has 11 subunits, while
complex IV (CIV, cytochrome c oxidase or COX), the ter-
minal oxidase of the respiratory chain, catalyzes the transfer
of reducing equivalents from cytochrome c to molecular
oxygen. It is composed of two cytochromes (a and a3), two
copper atoms, and 13 different protein subunits.
During the oxidation process, electrons are transferred
to oxygen via the energy-transducing complexes of the respi-
ratory chain:  CI, CIII, and CIV for NADH-producing
substrates; CII, CIII, and CIV for succinate; and CIII and
CIV for FADH2 derived from the β-oxidation pathway via
electron transfer flavoprotein (ETF) and the ETF–CoQ
oxidoreductase system. CoQ, a highly hydrophobic qui-
none, and cytochrome c, a low-molecular-weight hemopro-
tein, act as “shuttles” between complexes. The free energy
generated from the redox reactions is converted into a
transmembrane proton gradient. Protons are pumped
 Eye
Heart Optic neuropathy
Conduction disorder Ophthalmoplegia
Wolft-Parkinson-White Retinopathy
Skeletal muscle
Weakness Cardiomyopathy
Fatigue
Myopathy
Neuropathy Liver
Hepatopathy

Brain
Seizures Kidney
Myoclonus Fanconi's syndrome
Ataxia Glomerulopathy
Stroke
Dementia
Migraine
ATP
Nuclear DNA
Pancreas
subunits OX PHOS
mt DNA Diabetes mellitus
O2 H O
2

Defects in intergenomic
communication
Multiple mtDNA deletions and
mt DNA depletion

Blood
Pearson's syndrome

Inner ear
Colon
Sensorineural
Pseudo-obstruction
hearing loss
OX PHOS = OXIDATIVE PHOSPHORYLATION

Figure 9.1  Organs involved in mitochondrial disorders.

CI CII CIII CIV CV

Outer membrane

H+ H+ H+ H+

c
Q
Inner
membrane Q

Matrix

NADH Succinate O2

Electron flux
Proton flux

ATP ADP

Nuclear encoded subunit Mitochondrial encoded subunit

Figure 9.2  The mitochondrial respiratory chain. CI–CV: complexes I to V. Q: ubiquinone.

1 3 2   •   P rincip l es o f G eno m ic Medicine

through CI, CIII, and CIV of the respiratory chain, which and little intergenic noncoding DNA. Some genes even
creates a charge differential. Complex V (ATP synthase) overlap. Mitochondrion makes a large reticular network
allows protons to flow back into the mitochondrial matrix and contains several molecules of mtDNA. Each molecule
and uses the released energy to synthesize ATP. Three ATP contains 37 genes encoding one large and one small ribo-
molecules are produced from the oxidized NADH. somal RNA (12S rRNA and 16S rRNA, respectively), 22
transfer RNAs (tRNA), and 13 key respiratory chain sub-
units (Figure  9.3).3 ND1–ND6 are subunits of CI, cyto-
M I TO C H O N D R I A L G E N ET I C S chrome b is the only mitochondrially encoded subunit of
CIII, COXI–COXIII are subunits of CIV, and ATP6 and
The bulk of the eukaryotic cellular DNA is contained in ATP8 are subunits of ATPase (complex V). A short segment
the nucleus (nuclear DNA, or nDNA). Apart from nDNA, of the genome is triple-stranded. The displacement-loop
a small amount of DNA is contained in the mitochondria (D-loop) contains the only significant amount of noncod-
(mtDNA). Mitochondrial RC is made up of about 100 dif- ing DNA in the mitochondrial genome. Perhaps because of
ferent proteins, only 13 of which are encoded by mitochon- this, it is the location of many of the DNA polymorphisms
drial genes; the rest are encoded by nuclear genes. All of the that are such useful tools for anthropologists research-
RC complexes, except CII, have a double genetic origin, ing the origins of human populations. Because there is no
and one to seven subunits of the complexes are mitochon- recombination among mitochondrial DNAs, complete
drially encoded. In addition, several hundreds of nuclear haplotypes of polymorphisms are transmitted through the
genes are needed for various RC functions. The result is that generations, modified only by recurrent mutation, making
the number of mitochondrial proteins represents more than mtDNA a highly informative marker of ancestry, at least
3% of all cellular proteins. along the maternal line.
The mitochondrion has independent replication,
transcription, and translation systems. The mitochon-
M ITO C H O N D R I A L D NA
drial genome is replicated in two stages. Replication
The mitochondrial genome, in many respects, has more in starts at the heavy-strand replication origin (OH) in
common with bacterial genomes than with the eukaryotic the D-loop and extends clockwise around the mtDNA.
nuclear genome (Table 9.1). This is consistent with the idea When the light-strand replication origin (OL) is exposed
that mitochondria originated as endosymbiotic bacteria as a single strand, the second strand is then replicated
within some ancestral eukaryotic cell. If this theory is cor- in the opposite direction, starting from OL.4 Thus, rep-
rect, then over the years the mitochondria have gradually lication is bidirectional, but asynchronous. Recently, a
transferred more and more of their functions to the nucleus. new model of mtDNA replication has been proposed in
Human mitochondrial DNA (mtDNA) is a 16,569 mammals. Replication of mtDNA arises from multiple
base-pair, closed circular molecule.2 There are no introns, origins and proceeds via a strand-coupled mechanism.5
The two mtDNA strands are transcribed from specific
promoters into two large polycistronic RNAs, which are
Table 9.1  COMPARISON OF NUCLEAR AND further processed into ribosomal RNAs (rRNA), transfer
MITOCHONDRIAL GENOMES RNAs (tRNA), and messenger RNAs (mRNA). All the
protein components of the translation machinery are
NUCLEAR MITOCHONDRIAL
GENOME GENOME nuclear-encoded, but the rRNAs and tRNAs are exclu-
Size 3 × 109 bp 16,659 bp
sively mitochondrially encoded, and these use a coding
scheme slightly different from the otherwise universal
Topology 23 linear molecules 1 circular molecule
code. There are two stop codons—UAG and UAA. UGA
No. of genes Ca. 24,000 37
encodes tryptophan, and AUA is methionin and not iso-
% coding sequence Ca. 1.1% 93%
(incl. genes for func- leucine. Presumably, with only 13 protein-coding genes,
tional RNAs) the mitochondrial system could tolerate mutations that
Average gene density Ca. 1 per 125 kb 1 per 0.45 kb modified the coding scheme in a way the main genome
(variable) could not.
Introns Average 8 per gene None During cell division, mitochondria are randomly parti-
(variable) tioned into daughter cells (mitotic segregation). Usually, all
Repetitive DNA Ca. 50% None mtDNA molecules are identical, but, occasionally, a mixture

Mitochondria l G enetics and G eno m ics in C l inica l Medicine   •   1 3 3

 OH PH

D-loop

NA Phe 16S
7S D Val
Thr 23S
CYB
Pro Leu
H
STRAND PL NDI

Glu IIe
ND6 f-Met
Gln
ND2
ND5 Ala
OL.... Asn
Cys Trp
Tyr
Leu
Ser
His CO1
Ser
ND4

Asp
L ND4L
STRAND CO2
ND3
CO3 Lys
Arg ATPase8
Gly
ATPase6

Figure 9.3 
The heavy (H) and light (L) strands of the circular 16,659 bp mtDNA double helix are shown. Protein-coding genes are shaded; transfer
RNA genes are shown as short lines with the name of the amino acid. There are no introns; the heavy arrows indicate the origins and directions
of replication of the two strands; the light arrows show the promoters and direction of transcription of the two multicistronic transcripts that are
subsequently cleaved into individual mRNAs. (Adapted with permission from Figure 9.1 of Strachan & Read, Human Molecular Genetics, 3rd edition, London and New York: Garland; 2004).

of wild-type and mutant mtDNA is encountered. This is NUCLEAR GENES

called “heteroplasmy,” whereas “homoplasmy” refers to
the occurrence of only one type of mtDNA. In heteroplas-
mic cells, however, the mtDNA genotype can shift during
cell replication. Consequently, some lineages drift toward
wild-type mtDNA and become homoplasmic, while others
remain heteroplasmic.
The mitochondrial genome is maternally transmitted.
Abundant amounts of maternal mitochondria are con-
tained in the ovum cytoplasm, whilst the sperm tail con-
tains the complete male mitochondrial content. At the time
of fertilization, all maternal mitochondrial content, includ-
ing the mutated mtDNA, is passed on wholly in the ovum
cytoplasm to the embryo. The sperm-derived paternal mito-
chondria enter the oocyte cytoplasm after fertilization, but
the paternal mtDNA or the paternal mitochondria them-
selves are then degraded by a variety of mechanisms in order
to prevent paternal mtDNA transmission.6 The mother
transmits her mtDNA to all her progeny, males and females,
and her daughters, in turn, will transmit their mtDNA to
the next generation. Theoretically, males never transmit
their mtDNA.
Genes Encoding Respiratory Chain Subunits
The majority of RC proteins are encoded by nuclear genes.
The nuclear-encoded proteins are translated in cytosol
and transported across mitochondrial membranes. These
nuclear genes are spread out across all human chromosomes
on both autosomes and sexual chromosomes. For example,
33 genes of CI subunits have been mapped to various auto-
somes:  one to the X chromosome and seven to mtDNA.
Several of these nuclear genes have one or more pseudo-
genes (non-expressed copies) that can complicate mutation
screening in patients.
G E N E S I N VO LVE D I N R E S P I R ATO RY
C H A I N A S S E M B LY
The large number of RC proteins and their double genetic
origin indicate tightly regulated communication between
mitochondria and nuclear compartments. Therefore, in
addition to the structural components of the RC, many
nuclear-encoded proteins are involved in the assembly and

1 3 4   •   P rincip l es o f G eno m ic Medicine

maintenance of complexes. Most of these genes were first
identified in yeast, a model organism for mitochondrial
function and dysfunction. Indeed, analysis of yeast muta-
tions resulting in abnormal RC assembly has led to the
identification of many of the nuclear products involved in
protein folding, stabilization, quality control, membrane
translocation, and cofactor addition.7 To date, at least 350
such genes are known in yeast.
G E N E S I N VO LVE D I N MT D NA
M ETA B O L I S M A N D M A I N T E NA N C E
Mitochondria possess specific replication, transcription,
and repair mechanisms. All of the proteins involved in
these mechanisms are encoded by nuclear genes, trans-
lated in the cytosol, and then translocated to mitochon-
dria. Only the two rRNAs (12S rRNA and 16S rRNA)
and the 22 tRNAs are mitochondrially encoded. Thus far,
over 100 genes in yeast are known to result in mtDNA
loss when defective.8,9 The proteins involved in mamma-
lian mtDNA maintenance are those directly involved in
mtDNA processing, such as DNA polymerase γ (POLG),
Twinkle helicase, and mitochondrial transcription fac-
tor (TFAM). It has long been claimed that mitochondria
have no repair mechanisms, but recent evidence suggests
that specific DNA repair processes are present in these
organelles.9
G E N E S I N VO LVE D I N M ITO C H O N D R I A L
P ROT E I N T R A NS L AT I O N
Mitochondrial translation requires both mitochondrial and
nuclear genes. Mitochondrial DNA encodes rRNA and
tRNA, whereas nuclear genes encode ribosomal proteins,
aminoacyl tRNA synthetases, tRNA modification enzymes,
and elongation and termination factors. Altogether, several
hundreds of proteins are involved in the translation of the
13 proteins encoded by the mitochondrial genome, empha-
sizing the considerable investment required to maintain the
mitochondrial genetic system.10
In addition, a large number of other nuclear genes
encode proteins that are not directly related to RC assembly
or mtDNA maintenance, but that may interact with them.
Mutations in these genes can, therefore, give rise to abnor-
mal RC. Among them are chaperones, proteases, proteins
involved in mitochondrial inheritance or morphology, anti-
oxidant enzymes, and various carriers of iron, phosphate,
and citrate. As it is commonly assumed that mitochondria
contain roughly 1000 different proteins,11 their genes are all
possible candidates for mitochondrial disorders.
Mitochondria l G enetics and G eno m ics in C l inica l Medicine   •   1 3 5
M I TO C H O N D R I A L R E S P I R ATO RY
CHAIN DISORDER S
Oxidative phosphorylation is a ubiquitous metabolic path-
way that supplies most organs and tissues with energy.
Consequently, RC deficiency can theoretically give rise to
any symptom in any organ or tissue, and at any age with
any mode of inheritance, due to the twofold genetic ori-
gin of respiratory enzymes (nuclear DNA and mtDNA).
Mitochondrial diseases associated with mtDNA point
mutations or large rearrangements have now been charac-
terized, but it should be emphasized that mtDNA muta-
tions represent only a fraction of the genetic causes of these
disorders. Indeed, mtDNA encodes only 13 RC subunits,
2 rRNAs, and 22 tRNAs, but several hundreds of proteins
involved in various RC function and maintenance are
encoded by nuclear genes.
These disorders are characterized by a vast clinical het-
erogeneity, suggesting high genetic heterogeneity. Yet,
although the common feature is RC deficiency, the condi-
tions are due to different enzyme or protein deficiencies.
Moreover, the age of onset is highly variable (ranging from
the neonatal period to late adulthood), and the deficiency
can result in isolated organ deficiency or multivisceral
involvement. In the past few years, it has become increas-
ingly clear that genetic defects of oxidative phosphoryla-
tion account for a wide variety of clinical symptoms in
childhood.12 In general, the diagnosis of RC deficiency is
difficult to make on the basis of a single initial symptom,
but becomes easier when two or more seemingly unrelated
symptoms are observed.
M ETA B O L I C S C R E E N I N G F O R RC
DISORDER S
The current screening for RC deficiency includes determi-
nations of plasma lactate, pyruvate, and ketone bodies and
their molar ratios as indexes of oxidation/reduction status
in cytoplasm and mitochondria, respectively. Persistent
hyperlactatemia (>2.5 mM), with elevated lactate/pyruvate
(L/P, >20) and ketone body molar ratios, is highly sugges-
tive of RC deficiency (particularly in the post-absorptive
period). When basal screening tests are inconclusive, other
tests need to be carried out.12,13
When screening tests are negative, RC deficiency may
be misdiagnosed. For this reason, the investigation of
patients at risk of RC deficiency should include systematic
screening of all possible target organs and tissues, regardless
of the onset symptom, as multiple-organ involvement is an
important diagnostic clue to RC deficiency.
 B I O C H E M I C A L D E FEC TS I N RC
DISORDER S
Diagnostic tests include polarographic and spectrophoto-
metric studies; each provides an independent clue towards
the diagnosis of RC deficiency.
Polarographic studies consist of the measurement of
oxygen consumption by mitochondria-enriched frac-
tions, using a Clarke electrode in the presence of various
oxidative substrates (malate + pyruvate, malate + gluta-
mate, succinate, palmitate, etc.).13 The only limitation
of these techniques is the absolute requirement for fresh
material: no polarographic studies are possible using fro-
zen material.
Spectrophotometric studies involve isolated or com-
bined respiratory enzyme assays, using specific electron
donors and acceptors. These do not require the isolation
of mitochondrial fractions and can be carried out on tis-
sue homogenates. For this reason, the amount of material
required for enzyme assays (1–20 mg) is very small and
can easily be carried out using needle biopsies of liver or
kidney, endomyocardial biopsies, or from a pellet of lym-
phocytes or cultured skin fibroblasts. Samples should be
immediately frozen and kept dry in liquid nitrogen (or
at –80°C).13
The question of which tissue should be investigated
merits particular attention. In principle, the relevant tissue
is the one that clinically expresses the disease. Whatever the
affected organ, it is mandatory to take skin-biopsy patients
(even post mortem) for subsequent investigations using cul-
tured fibroblasts.
RC deficiency associating more than one complex (mul-
tiple RC deficiency) is the most frequent cause of mito-
chondrial disorders in our series of patients (55%). Isolated
CI and CIV deficiencies are the second and third causes of
these disorders (19% and 13% respectively). A  variety of
neuromuscular and non-neuromuscular symptoms may be
observed, although trunk hypotonia, growth retardation,
cardiomyopathy, encephalopathy, and liver failure are the
most frequent symptoms. Similar clinical presentations
of mitochondrial dysfunction can result from various RC
deficiencies or gene mutations, thereby hampering easy clas-
sification of these diseases. For example, Leigh syndrome,
a subacute necrotizing encephalomyopathy resulting in
a devastating encephalopathy, characterized by recurrent
attacks of psychomotor regression, with pyramidal and
extrapyramidal symptoms, leukodystrophy, and brain-stem
dysfunction, may be associated with deficiency of any of the
RC complexes and be the result of mutations in either mito-
chondrial or nuclear genes. Moreover, the mutant genes
1 3 6   •   P rincip l es o f G eno m ic Medicine
can encode structural proteins of RC complexes, proteins
involved in the assembly of these complexes, or proteins
involved in the process of mitochondrial translation. On
the other hand, in most cases, mutations of a specific gene
lead to a relatively homogeneous clinical presentation. Thus,
elucidating the genetic bases of RC is essential for both the
genetic diagnosis of patients and our fundamental knowl-
edge of these disorders, a prerequisite of any therapy-based
research.
G E N ET I C S O F M I TO C H O N D R I A L
R E S P I R ATO RY C H A I N
DISORDER S
Mitochondrial RC disorders are genetically heteroge-
neous. Indeed, the RC is made up of around 100 polypep-
tides encoded by as many different genes. These genes are
either nuclear or mitochondrial. In addition, the biogen-
esis and assembly of all these polypeptides require several
dozens of nuclear genes, some of which are found only in
humans. Mitochondrial RC disorders can result from a
mutation of any one of these hundreds of genes. Thus, all
modes of inheritance may be encountered in mitochondrial
RC disorders: sporadic cases due to mtDNA mutations or
nuclear gene mutations; maternal transmission in the case
of mtDNA mutations; and autosomal recessive, autosomal
dominant, or X-linked inheritance. Unfortunately, in only
a few patients (25% in our series) have the disease-causing
mutations been identified (Figure  9.4). Nevertheless,
mtDNA mutations have been systematically excluded in
most patients with unknown mutations, suggesting that
in these cases the mitochondrial disease is due to a nuclear
gene mutation.
700
600
500
400
Nuclear
mtDNA
300
Unknown
200
100
0
CI CII CIII CIV CV Multiple
Figure 9.4 
Distribution of mtDNA and nDNA mutations in RC
deficiency in Necker Hospital (Paris).
 M ITO C H O N D R I A L D NA MU TAT I O N S
Pathological alterations of mtDNA fall into three major
categories:  point mutations; deletion-duplications; and
copy-number mutations (depletions). In most cases,
mtDNA mutations are heteroplasmic, as both normal and
mutant mtDNA are present. In cells harboring both mutant
and wild-type molecules, the phenotype is a reflection
of the proportion of mutant mtDNA molecules and the
extent to which the cell type relies on mitochondrial func-
tion. Point mutations include amino-acid substitutions and
protein-synthesis mutations (tRNA, rRNA) (Table 9.2).
Mitochondrial protein-Synthesis point mutations
The A3243G mutation in the tRNALeu gene is responsi-
ble for the MELAS (mitochondrial encephalomyopathy
with lactic acidosis and stroke-like episodes) syndrome.14
MELAS is characterized by onset in childhood with
intermittent hemicranial headache, vomiting, proximal
limb weakness, recurrent neurological deficit resembling
stroke (hemiparesis, cortical blindness, hemianopsia),
lactic acidosis, and, occasionally, ragged red fibers on
muscle biopsy. Computed tomography (CT) brain imag-
ing shows low-density areas (usually posterior) that may
involve both white and gray matter, but that do not always
correlate with clinical symptoms or vascular territories
The pathogenesis of stroke-like episodes in MELAS has
been ascribed to either cerebral blood flow disruption or
acute metabolic decompensation in biochemically defi-
cient areas of the brain.
The A3243G mutation also causes the maternally
inherited insulin-dependent diabetes mellitus and deaf-
ness (MIDD) that may also complicate with dilated
cardiomyopathy.15
The A8344G missense mutation in the mt tRNALys gene
accounts for 80% of cases of MERRF (myoclonus epilepsy
with ragged red fibers).16 This disease is characterized by
encephalomyopathy with myoclonus, ataxia, hearing loss,
muscle weakness, and generalized seizures.
Also, several other mutations of either tRNA or rRNA
genes have been reported in non-related families.3
Mutations in protein-Coding genes
The most frequent mutations in mitochondrial genes
encoding structural proteins have been reported in LHON
(Leber hereditary optic neuropathy) and in NARP (neuro-
genic, ataxia, and retinitis pigmentosa)/Leigh syndrome.3
These mutations are recurrent mutations, as numerous
Mitochondria l G enetics and G eno m ics in C l inica l Medicine   •   1 3 7
non-related patients harbor these mutations. Other muta-
tions have also been described in a number of genes, but
these are most often restricted to a specific patient or family.
NARP, and variable sensory neuropathy, seizures, and
mental retardation, are due to an amino-acid change in the
ATPase6 gene (T8993G).17
LHON is associated with rapid bilateral central vision loss
due to optic-nerve death. Cardiac dysrhythmia is frequently
associated with the disease, but no evidence of skeletal mus-
cle pathology or gross structural mitochondrial abnormal-
ity has been documented. The median age for vision loss is
20–24 years, but it may occur at any time between adolescence
and late adulthood. Expression among maternally related indi-
viduals is variable, and there is a bias towards affecting males.
To date, 90% of LHON cases harbor one of the three pri-
mary mutations (G11778A, T14484C, G15257A), although
several other missense mutations—called “secondary muta-
tions”—in the mtDNA can act autonomously or in associa-
tion with each other to cause the disease.3
CI deficiency is the most frequent cause of mitochon-
drial disorders, having been found in more than 30% of
patients. Trunk hypotonia, antenatal and postnatal growth
retardation, encephalopathy, and liver failure are the main
clinical features.18,19 Systematic sequencing of the mitochon-
drial CI genes has shown that around 20% of CI-deficient
patients harbored point mutations in one of these genes—
at least, in our series.
In contrast, CIII deficiency is a relatively rare cause of
respiratory enzyme dysfunction. Indeed, in our experience,
of all RC enzyme-deficient patients, only 7% had a CIII defi-
ciency.20 However, the clinical presentation of CIII-deficient
patients is highly heterogeneous, including myopathy,
encephalomyopathy, multi-organ disorders, cardiomyopa-
thy, tubulopathy, and intrauterine growth retardation.20,21
This complex contains 11 subunits, and only one, cyto-
chrome b (cytb), is of mitochondrial origin. So far, 12 cytb
mutations have been described in association with various
clinical presentations. Interestingly, in most patients (8/12),
the predominating presenting feature was severe exercise
intolerance, sometimes including muscle weakness and/or
myoglobinuria.22 Two other patients presented with cardio-
myopathy, another patient had encephalomyopathy, and a
further patient had MELAS and an akinetic rigid syndrome.
Cytochrome c oxidase (COX) deficiency is a frequent
cause of RC disorder in childhood and is clinically heteroge-
neous, with phenotypes such as encephalomyopathy, Leigh
syndrome,23 fatal and benign infantile myopathy, hepatic
failure,24 and myoglobinuria.25 In fact, mtDNA mutations
have been identified in patients with various clinical presen-
tations. These mutations have been described in the three
mitochondrial COX genes (COXI, COXII, COXIII),3 and
have always been associated with single pedigrees and are,
therefore, private mutations. Most patients with mtDNA
COX mutations have muscle-related or neuromuscular
symptoms. However, mutations of COXI have also been
reported in two cases of acquired sideroblastic anemia.26
Most of these mutations are maternally inherited
and heteroplasmic, and associated with a striking variety
of clinical phenotypes, depending on the proportion of
mutant mtDNA inherited among the maternal relatives.
Within one particular pedigree, clinical presentations can
range from migraines and attention-deficit disorders to the
full MELAS syndrome. Maternal relatives of patients are
generally healthy as long as they have no more than 85%
mutant mtDNA. Once the percentage of mutant mtDNA
rises beyond this level, there are increasingly serious conse-
quences for the clinical phenotype, thus highlighting the
sharp threshold in protein-synthesis mutations.
Large-Scale mtDNA rearrangements
The second category of mtDNA diseases is deletion of the
mitochondrial genome. Although the size and position of
the deletion differ markedly among patients, they usually
encompass several encoding and tRNA genes. They are usu-
ally sporadic, heteroplasmic, and unique, and frequently
occur between directly repeated sequences, suggesting that
they are caused by de novo rearrangements arising during
oogenesis or early development.
The Kearns–Sayre syndrome (KSS) is a multisystem
disorder characterized by a non-variable triad: onset before
age 20; progressive external ophthalmoplegia; and pig-
mentary retinal degeneration, plus at least one of the fol-
lowing:  complete heart block, cerebrospinal fluid (CSF)
protein >100 mg/dL, and/or cerebellar ataxia. Large-scale
heteroplasmic mtDNA deletions are frequently detected in
skeletal muscle.27
Pearson syndrome comprises refractory sideroblastic
anemia, with variable neutropenia and thrombocytopenia,
vacuolization of marrow precursors, and exocrine pancreatic
dysfunction. Severe transfusion-dependent macrocytic ane-
mia begins in early infancy (before age one year) and is fatal
by three years of age in 62% of cases. The patients who survive
spontaneously recover from their myelodysplasia, but usually
develop KSS. Large-scale heteroplasmic mtDNA deletions
are present in all tissues, with the ratio of normal-to-deleted
genomes being related to expression of the disease.28
Progressive external ophthalmoplegia (PEO) is a mito-
chondrial myopathy with progressive muscle weakness and
external ophthalmoplegia. Ataxia, episodic ketoacidotic coma,
1 3 8   •   P rincip l es o f G eno m ic Medicine
and early death have also been reported. Large-scale mtDNA
deletions are found in the skeletal muscle of the patients.29
Few patients with syndromic diabetes also present with
mtDNA deletions. In most cases, the deletion is maternally
inherited and always heteroplasmic. Diabetes is frequently
associated with deafness, but may also be part of a multi-organ
disorder.30–33 It is worthwhile noting that the most common
deletion (4977 bp), found in 30% of patients harboring a
unique deletion, flanked by 13-bp direct repeats, has been
described in both Pearson syndrome and KSS, and subse-
quently also reported in PEO. Thus, no correlation can be
found between clinical presentation and the nature or extent
of the rearrangements. Finding progressive organ involvement
should prompt the suspicion of a diagnosis of mtDNA rear-
rangement, and lead to long-range polymerase chain reaction
(PCR) or Southern blot analyses on total DNA.
Quantitative (depletion) and qualitative (multiple dele-
tion) mtDNA anomalies may also be due to mutations of
nuclear genes involved in mtDNA maintenance. In such
cases, the disease is inherited as an autosomal dominant or
recessive trait (see below).
So far, no mitochondrial disease has been exclusively
associated with any population-specific subgroup. However,
it is likely that mtDNA mutations and/or polymorphisms
are important in disease causation when they occur in asso-
ciation with either heterozygous or homozygous nuclear
DNA mutations.
N U C L E A R D NA MU TAT I O N S
The number of disease-causing mutations in nuclear genes
is steadily growing, and these mutations probably underlie
the vast majority of RC deficiencies. It should be borne in
mind that mtDNA deletions and mutations account for
no more than 15–20% of cases, at most, among pediatric
patients. Thus, in most cases, nuclear gene defects are those
most likely to be responsible for RC deficiency.
Indeed, proper RC functioning requires not only the
presence of various subunits of each complex, but also ancil-
lary proteins at different stages of holoenzyme biogenesis,
including transcription, translation, chaperoning, addition
of prosthetic groups, and assembly of proteins, as well as
various enzymes involved in mtDNA metabolism.
D E F E C T S I N S T RU C T U R A L
R E S P I R ATO RY C H A I N G E N E S
The various nuclear genes encoding RC subunits have all
been identified and mapped, and mutations of some of these
Table 9.2  COMMON CLINICALLY RECOGNIZABLE MITOCHONDRIAL DISORDERS ASSOCIATED WITH
MTDNA MUTATIONS

DISORDER MAJOR CLINICAL FEATURES TYPE OF GENE MITOCHONDRIAL DNA

Chronic progressive external oph-
thalmoplegia (CPEO)
Kearns-Sayre syndrome (KSS)
Pearson syndrome
Diabetes and deafness
Leber's hereditary optic neuropathy
(LHON)
Neurogenic ataxia with retinitis
­pigmentosa (NARP)
Leigh syndrome (LS)
Exercise intolerance and
myoglobulinuria
Mitochondria encephalomyopathy
with lactic acidosis and strokelike
episodes (MELAS)
Myoclonic epilepsy with ragged-red
fibers (MERRF)
Cardiomyopathy
Infantile myopathy/encephalopathy
Nonsyndromic sensorineural
deafness
Aminoglycoside-induced
­nonsyndromic deafness
MUTATION
External ophthalmoplegia, bilateral ptosis, tRNA A3243G, T8356C rearrange-
mild proximal myopathy ment (deletion/duplication)
PEO onset <20 years, pigmentary Rearrangement (deletion/
­retinopathy, cerebellar ataxia, heart block, duplication)
CSF ­protein >1g/L
Sideroblastic anemia of childhood, Rearrangement (deletion/
­pancytopenia, renal tubular defects, duplication)
­exocrine ­pancreatic deficiency
Diabetes mellitus, sensorineural hearing loss tRNA A3243G, C12258A rearrange-
ment (deletion/duplication)
Subacute painless bilateral visual loss, age Protein encoding G11778A, T14484C, G3460A
of onset 24 years, males > females (-4:1),
­dystonia, cardiac pre-excitation syndromes
Late-childhood or adult-onset peripheral Protein encoding T8993G/C
neuropathy, ataxia, pigmentary retinopathy
Subacute relapsing encephalopathy, cerebellar Protein encoding T8993G/C
and brainstem signs, infantile onset
Exercise-induced myoglobulinuria Protein encoding cytb mutations
Strokelike episodes before 40 years, s­ eizures tRNA A32343G, T3271C, A3251G
and/or dementia, ragged-red fibers and/
or lactic acidosis, diabetes mellitus,
­cardiomyopathy (HCM/DCM), deafness,
cerebellar ataxia
Myoclonus, seizures, cerebellar ataxia, tRNA A8344G, T8356C
myopathy, dementia, optic atrophy, bilateral
deafness, peripheral neuropathy, spasticity,
multiple lipomata
Hypertrophic cardiomyopathy (HCM) pro- tRNA A3243G, A4269G
gressing to dilated cardiomyopathy (DCM)
Early-onset progressive muscle weakness with tRNA T14709C, Al2320G, G1606A,
developmental delay T10010C
Early-onset progressive bilateral moderate to rRNA A7445G
severe sensorineural hearing loss
Early-onset nonprogressive sensorineural rRNA A1555G
deafness secondary to aminoglycoside
administration

genes have been found in patients. Interestingly, most of the
mutations encountered have been in CI genes; whereas,
despite the concerted efforts of various teams of researchers,
very few mutations in other genes encoding other complex
subunits have been found.
The first mutation in a gene encoding an RC subunit
was reported in 1995 in two sisters with Leigh syndrome
and CII deficiency. The pathogenic mutation was in the
SDHA gene encoding the flavoprotein of CII.34 Mutations
in the same gene were subsequently reported in another
patient who also presented with Leigh syndrome.35,36
However, only a few patients with CII deficiency have been
molecularly characterized, despite systematic study of the
four genes encoding the CII subunits, suggesting that these
deficiencies could be due to mutations in assembly proteins.
Nevertheless, mutations of the three other genes encoding
subunits B, C, and D of CII have been reported in heredi-
tary paraganglioma and phaeochromocytoma, suggesting
that such “housekeeping” genes may be involved in carci-
nogenesis.37 It is hypothesized that succinate dehydrogenase
(SDH) mutations cause an accumulation of succinate and
reactive oxygen species (ROS), which could act as down-
stream signaling molecules to activate hypoxia-inducing
pathways.37

Mitochondria l G enetics and G eno m ics in C l inica l Medicine   •   1 3 9

 The pioneering work by Smeitink and colleagues in G E N E S I N VO LVE D I N RC A S S E M B LY
the Netherlands identified the first molecular bases of CI
The various RC complexes contain four (CII) to 45 sub-
deficiencies.19 This complex is the largest RC complex, and
units (CI) each. Therefore, a functional complex requires
consists of seven mitochondrial-encoded and at least 35
structural integrity and tight regulation of each of its sub-
nuclear-encoded proteins. CI deficiency is one of the most
units and cofactors. Alterations in any of the mechanisms
common causes of mitochondrial disease. Screening of the
allowing structural integrity in these complexes may result in
various structural CI genes by several teams has allowed
catalytic dysfunction or instability of the complex assembly.
the identification of mutations in conserved subunits of
These mechanisms include architectural assembly of com-
this complex, and the findings show that around 40% of
plex subunits, incorporation of cofactors, translation of spe-
CI deficiencies are related to mutations in those genes.
cific subunits, and heme, iron, or copper assembly. Several
Most of the patients presented with Leigh or Leigh-like
of these assembly factors have been identified in humans by
syndrome, although cardiomyopathy has also been
homology with their yeast counterparts. Also, searching for
reported (Figure 9.5).
disease-causing genes in patients has led to the identification
Mutations in only two genes encoding CIII subunits,
of new human genes involved in these mechanisms.
UQCRB and UQCRQ, have been reported,38,39 as have
CI, the largest RC complex, has several assembly factors
the two genes encoding CIV subunits, COX4I240 and
(NDUFAF1, NDUFAF2, C6orf66, C8orf38, and C20orf7)
COX6B1.41 However, no mutation of any of the nuclear
that have been identified in humans as a result of studying
genes in complex V has been described, whereas several
patients.19 As yet, however, the exact function of these genes
mutations in its mitochondrial genes have been reported
is unknown (Figure 9.5).
(Figure 9.5).

Genes encoding subunits of the respiratory chain

NDUFV1, NDUFS8
NDUFS7, NDUFS1 mtCOXI-III
NDUFA8, NDUFB6 HUMQPC COX4AI2
NDUFS4, NDUFV2 cytb COX6B1 ATP6
mtND1-6
SDH-FP

CI
CIII CIV CV

QQ

CII

NDUFAF2 SDHAF1 PDSS1 BCS1 SURF1

NDUFAF1 PDSS2 TTC19 SCO2
C6orf66 COQ2 COX10
C8orf38 CABC1 COX15
C20orf7 COQ9 SCO1 ATP12
ACAD9 LPPRC TMEM70
FOXRED1 FASTKD2
NUBPL TACO1
C12ORF62
Genes encoding proteins involved in respiratory chain assembly

Figure 9.5 
Gene mutations resulting in specific respiratory chain complexes. Genes encoding for the subunits of complexes and assembly factors are
shown in the upper and lower registers, respectively. Genes in boldface are mutations found in several patients or families.

1 4 0   •   P rincip l es o f G eno m ic Medicine

 CII deficiency represents a rare cause of mitochon-
drial disorders, although, recently, two genes involved in
its assembly have been found in humans. By homozygos-
ity mapping and candidate-gene analyses, mutations in the
SDHAF1 gene were identified in patients with infantile leu-
koencephalopathy and isolated CII deficiency.42 SDHAF1
mutations lead to reduced amounts of the complex. SDH5
is a mitochondrial protein gene required for flavination of
the SDH1 subunit, and mutations of the gene have been
found in paraganglioma.43
Only two genes involved in CIII assembly are known
in humans. The gene BCS1L allows the assembly of the
iron-sulphur protein subunit in the complex. BCS1L muta-
tions have been identified in three clinical entities associated
with CIII deficiency: one group of patients presents with
tubulopathy and hepatic failure44; another group has the
GRACILE (growth retardation, aminoaciduria, cholesta-
sis, iron overload, lactacidosis and early death) syndrome45;
and the final group presents with the Björnstad syndrome
(sensorineural hearing loss and pili torti)46 (Figure 9.5). The
TTC19 gene encodes tetratricopeptide 19, a new assembly
factor of CIII. Mutations of this gene results in severe neu-
rological abnormalities.47
Using different approaches in patients with COX defi-
ciency, such as gene-mapping, functional complementation,
or candidate gene studies, several assembly genes have been
identified as disease-causing. SURF1 represents a major
gene for Leigh syndrome associated with COX deficiency,
with 25–75% of Leigh–COX patients having SURF1
mutations.48 It has also recently been shown that SURF1 in
bacteria is a heme-binding protein that may be involved in
heme insertion in cytochrome oxidase.49 COX10 encodes
heme A:farnesyltransferase, catalyzing the first step in the
conversion of protoheme to the heme A prosthetic group
required for cytochrome c oxidase activity, while COX15
allows the hydroxylation of heme O to form heme A.
However, few patients present with mutations of these two
genes, thereby hindering any genotype–phenotype correla-
tion. COX15 mutations lead to cardiomyopathy or Leigh
syndrome,50–52 whereas COX10 mutations are associated
with tubulopathy and leukodystrophy.53,54 Mutations in
SCO155 and SCO2 genes,56 both of which are involved in
mitochondrial copper maturation and synthesis of subunit
II of COX,57 also give rise to hepatopathy and ketoacidotic
coma (SCO1) and cardiomyopathy (SCO2). Mutations
in LRPPRC cause Leigh syndrome, French-Canadian
type.58 The LRPPRC protein is thought to be involved
in the translation or stability of the mRNA of subunits I
and III of COX.59 FASTKD2 mutations, reported in only
one kindred, cause encephalomyopathy and convulsions.60
Mitochondria l G enetics and G eno m ics in C l inica l Medicine   •   1 4 1
Although the gene is not directly involved in the assembly
of CIV, it may play a role in the regulation of mitochon-
drial apoptosis (Figure 9.5). TACO1 is a recently identified
gene of COX deficiency that encodes a translational activa-
tor of the COX1 subunit encoded by mtDNA. Mutations
of this gene result in late-onset Leigh syndrome.61 Finally,
C12orf62 is required for coordination of the early steps of
COX assembly and mutations in this gene result in fatal
neonatal lactic acidosis.62
Most complex V deficiencies are associated with
mtDNA mutations (ATP6 and ATP8 gene mutations), and
only two complex V nuclear mutations have been reported.
The ATP12 gene encodes a protein required for assembly
of the alpha and beta subunits, and mutations of this gene,
reported in one patient, resulted in dysmorphic features,
neurological involvement, and methylglutaconic aciduria.63
A  large kindred of Gipsy origin with isolated complex V
deficiency and neonatal encephalocardiomyopathy were
reported to present with mutations in TMEM70, a gene
encoding a transmembrane mitochondrial protein of yet
unknown function involved in the assembly of complex V64
(Figure 9.5).
Several proteins of the RC are iron-sulphur proteins, and
deficiencies in the assembly of the iron-sulphur cluster have
been reported to result in dysfunction of CI, CII, and CIII,
which contain iron-sulphur proteins. Indeed, Friedreich’s
ataxia, one of the most common forms of autosomal reces-
sive ataxia, associated with hypertrophic cardiomyopa-
thy and diabetes in 10% of cases, is due to a mutation of
frataxin, a mitochondrial protein involved in iron-sulphur
protein biogenesis.65 Mutations on the iron-sulphur cluster
scaffold protein (ISCU), which interacts with frataxin in
iron-sulphur cluster biosynthesis, lead to myopathy, with
exercise intolerance and myoglobinuria.66,67
Electron transfer along the RC depends on a quinone
pool synthesized in the mitochondria. Deficiencies in sev-
eral enzymes or proteins (PDSS1, PDSS2, COQ2, COQ9,
CABC1) involved in this biosynthesis pathway are reported
to be associated with various clinical presentations68–70
(Figure 9.5).
G E N E S I N VO LVE D I N MT D NA
S TA B I L I T Y
Mitochondrial DNA is packaged into protein–DNA com-
plexes called “nucleoids.” Nucleoids in cultured mammalian
cells contain 5 to 10 mtDNA molecules and appear to be
tethered to the inner mitochondrial membrane. Although
their protein composition remains controversial, it is
now well established that nucleoids contain the mtDNA
replisome.71 Crucial proteins or enzymes involved in
either mtDNA replication (mtDNA polymerase γ, mtSSB,
Twinkle helicase) or transcription (TFAM) are also compo-
nents of nucleoids. In theory, defects in any of the proteins
involved in mtDNA replication can affect mtDNA copy
number. Such replication is also highly dependent on the
mitochondrial deoxyribonucleotide triphosphate (dNTP)
supply, suggesting that mutations in several genes involved
in mitochondrial dNTP synthesis may, therefore, result in
mtDNA anomalies (Figure 9.6).
Autosomal-dominant external ophthalmoplegia
(adPEO) is associated with multiple mtDNA deletions,72
which are restricted to muscle tissue. With an onset in adult-
hood, adPEO includes progressive weakness of the extraoc-
ular muscles as a cardinal feature; patients have ptosis and
limited eye movements, with additional features varying
from one family to another. Most cases of adPEO associ-
ated with multiple mtDNA deletions are due to mutations
in POLG1 (mtDNA polymerase γ),73 POLG2,74 ANT1
(mitochondrial ADP/ATP translocator),75 PEO1 (Twinkle
helicase)76 and OPA1, a dynamin-related GTPase involved
in mitochondrial fusion77(Figure 9.6). In rare cases, the dis-
ease is autosomal-recessive.
Mitochondrial DNA depletion syndrome (MDS)
was initially described as “congenital myopathy” or “hep-
atopathy.”78 Since then, however, many patients have
demonstrated different clinical presentations, includ-
ing hepatocerebral, myopathic, and encephalomyopathic
forms.79,80 In such patients, there is a marked (usually
tissue-specific) deficiency in mtDNA levels, with a residual
amount of mtDNA that is often less than 10% of normal
values. Such depletion leads to deficiencies of multiple RC
complexes, while nuclear-encoded components such as
CII are mostly expressed normally. Depletion is related to
abnormal mtDNA replication, especially defects in dNTP
supply and replication mechanisms. Mutations in POLG
and PEO1 encoding the Twinkle helicase, two replication
factors, are associated with severe and often fatal hepato-
cerebral forms81 and/or Alpers’ syndrome,82 characterized
by psychomotor retardation, intractable epilepsy, and liver

mtDNA replication mitochondrial dNTP salvage pathway

POLG dNTP DGUOK TK2

PEO1 synthesis
mtSSB cytosolic dNTP de novo synthesis
Twinkle TP RRM2B
helicase T C
A G

POLG
MPV17
TFAM TFAM CI CIII CIV CV
TOPO

CII

TRMU AARS2 DARS2 HARS2

PUS1 RARS2 SARS2 YARS2

tRNALeu
aminoacyl–tRNA
tRNA synthetases
tRNA processing
EFTu

MRPS22 EFG1
mtDNA rRNA
MRPS16 TSFM
MTFMT MRPL3 C12ORF65

MTPAP Translation factors

mRNA
Ribosomal
proteins Mitochondrial protein translation

Figure 9.6 
Mitochondrial DNA replication and translation of proteins encoded by mtDNA. The gene mutations resulting in mitochondrial diseases
and multiple respiratory chain deficiencies are shown in the black ovals.

1 4 2   •   P rincip l es o f G eno m ic Medicine

failure in infants and young children. Hepatocerebral forms
are also associated with the DGUOK gene,83 which encodes
the mitochondrial deoxyguanosine kinase involved in the
salvage pathway of dNTP for mtDNA synthesis, or the
MPV17 gene, which encodes a protein of unknown func-
tion.84 The TK2 gene encodes mitochondrial thymidine
kinase, which is also involved in mitochondrial dNTP
salvage. Mutations of this gene are associated with severe
infantile myopathy85 and, recently, with motor neuron dis-
ease resembling spinal muscular atrophy,86 encephalopa-
thy or seizures, cardiomyopathy, or dystrophic changes in
muscle.87 Synthesis of mtDNA requires not only sufficient
mitochondrial dNTP pools, but also balanced cytosolic
dNTP synthesis. Indeed, mutations of RRM2B encoding
a small subunit of cytosolic ribonucleotide reductase, the
enzyme that catalyzes dNDP synthesis from nucleotide
diphosphate (NDP), results in severe muscle mtDNA
depletion and neonatal trunk hypotonia, with hyperlac-
tatemia.88 Mutations of two subunits of the succinyl-CoA
synthase (SUCLA2, SUCLG1) have been reported in
patients with mild mtDNA depletion. SUCLA2 patients
present with psychomotor retardation, muscle hypotonia,
hearing impairment, and seizures,89 whereas those with
SUCLG1 mutations have encephalomyopathy and methyl-
malonic aciduria.90 Although both these mutations lead to
methylmalonic aciduria, the pathogenesis of this condition
is poorly understood.
Mitochondrial neurogastrointestinal encephalopa-
thy (MNGIE) syndrome is a multisystem disorder clini-
cally characterized by onset between the second and fifth
decades of life, ptosis, PEO, gastrointestinal dysmotility,
diffuse leukoencephalopathy, peripheral neuropathy, and
myopathy. Patients may have multiple mtDNA deletions
and/or mtDNA depletion. The disease-causing gene (TP)
encodes thymidine phosphorylase.91 Mutations of this
gene can affect the balance of the intramitochondrial
dNTP pool and lead to mtDNA deletions and depletion.80
Interestingly, it is becoming more and more evident
that mutations of a same gene can be associated with vari-
ous clinical presentations and result in either quantitative
or qualitative mtDNA anomalies. In addition, the inheri-
tance of the associated diseases may be either dominant or
recessive. Indeed, POLG and PEO1 mutations can lead to
either adPEO with multiple mtDNA depletion, or a fatal,
recessive, hepatocerebral form with severe mtDNA deple-
tion. On the other hand, RRM2B mutations, which usu-
ally cause severe muscle mtDNA depletion in neonates,
have been recently associated with adPEO, with multiple
mtDNA deletions,92 and with MNGIE syndrome in adult
patients.93
Mitochondria l G enetics and G eno m ics in C l inica l Medicine   •   1 4 3
Genes involved in translation of
mtDNA-Encoded proteins
Mitochondria contain separate protein-synthesis mecha-
nisms allowing the synthesis of polypeptides encoded by
mtDNA. The process not only requires tRNA and rRNA
encoded by mtDNA, but also hundreds of nuclear genes
encoding ribosomal proteins, aminoacyl-tRNA synthetases,
tRNA modification enzymes, rRNA base-modification
enzymes, and elongation and termination factors10
(Figure 9.6). Mutations in several of these factors have been
reported in patients with multiple RC deficiencies and
various clinical presentations, such as myopathy and sid-
eroblastic anemia (PUS1, YARS2)94,95; leukoencephalopa-
thy (DARS2)96; pontocerebellar hypoplasia (RARS2)97;
cardiomyopathy (AARS2)98; hyperuricemia, pulmonary
hypertension, and renal failure (SARS2)99; Perrault syn-
drome (HARS2)100; encephalomyopathy (TSFM)101;
hypertrophic cardiomyopathy (TSFM)101; hepatoencepha-
lopathy (GFM1)102; infantile encephalopathy (EFTu)103;
fatal neonatal lactic acidosis (MRPS16)104; multivisceral
involvement (MRPS22)105; hypertrophic cardiomyopathy
(MRPL3)106; hepatic failure (TRMU)107; Leigh syndrome
(MTFMT)108; and spastic ataxia (MTPAP).109 Considering
the large number of patients with multiple RC deficiencies
without mtDNA anomalies, it may be hypothesized that
such cases may be related to abnormal translation due to
nuclear gene deficiency.
G E N ET I C C O U N S E L I N G
A N D   P R E N ATA L D I AG N O S I S
Prenatal diagnoses of mitochondrial disorders fall in
two categories, depending on the nature of the identified
mutation. If the disease is related to a nuclear gene muta-
tion, mutation screening using a sample of chorionic villi
at 10 weeks of gestation offers early and reliable prenatal
diagnosis. Identification of an mtDNA mutation in the
proband should always prompt examination and testing of
the patient’s maternal relatives for the mutation. In cases of
maternal inheritance of mtDNA mutations (or deletions),
there is no risk for the progeny of an affected male. The risk
is high, however, for the progeny of carrier females. In this
case, prenatal diagnosis based on chorionic villi or amni-
otic cells represents a rational approach to the prevention
of these serious disorders. Nevertheless, prevention is cur-
rently hampered by our incomplete knowledge the actual
proportion of mutant mtDNA, its relationship to disease
severity, and its random tissue distribution and selection in
the affected population during development, which may PRONUCLEAR TRANSFER HOW IT WORKS
also be related to various metabolic activities. Indeed, it PARENT’S DONOR’S
EMBRYO EMBRYO
would be difficult to make an accurate prediction of the age
of onset and the phenotype in any offspring of an affected
mother. The pedigree (Figure 9.7) illustrates variable sever-
ity of MELAS in terms of myopathy, myoclonic epilepsy,
deafness, and dementia. According to the data so far, having UNHEALTHY
a percentage of mutant mtDNA less than 30% or greater MITOCHONDRIA

than 80% is predictive of a reasonable chance of a good or HEALTHY

MITOCHONDRIA
bad prognosis, respectively.110 Any results in between these
limits would be of even less certain predictive value. But INTENDED
PARENTS’
whatever the results, any studies aimed at prenatal diagnosis PRONUCLEI

or predictive genetic guidelines require careful validation,

as the proportions of mutant mtDNA may change not DONOR
only between fetal life and infancy, but also throughout PRONUCLEI
REMOVED
adulthood.
Recent reports surrounded by intense ethical and moral
debate have led to acceptance of the challenging approach for
preventing recurrence of a mitochondrial disease through
allowing the embryo to be nurtured in wild mitochon-
drial environment.109 This method, essentially described as
“the three-parent offspring,” involves creating the embryo RECONSTRUCTED EMBRYO

through in vitro fertilization of the ovum from an affected Figure 9.8 

Prevention of transmission of mutated maternal mtDNA
woman with her partner’s sperm (or any other source). The through transfer of in vitro embryo into an enucleated ovum with wild
embryo is then removed and transferred to an enucleated mtDNA.
ovum from another woman (Figure  9.8). Essentially, the
resulting embryo would contain the complete set of bio-
logical parents’ nuclear DNA and mtDNA from another
1 woman (see Chapter  46, on reproductive medicine, for
I
further details). The technique, undoubtedly exciting and
technically feasible, has led to intense public and medial
debate. The United Kingdom’s Parliament is considering
1 2 3 4
an amendment to the existing Human Embryology and
II Fertilization Act to make this method legal.

1 2 3 4 5 6 7
S U MM A RY
III
Mitochondrial disorders are due to respiratory chain defi-
ciency. Oxidative phosphorylation—ATP synthesis by the
MITOCHONDRIAL oxygen-consuming respiratory chain (RC)—supplies most
MYOPATHY DEAFNESS organs and tissues with a readily usable energy source and is
already fully functioning at birth. This means that, in theory,
RC deficiency can give rise to any symptom in any organ
MYOCLONIC EPILEPSY, DEMENTIA
ABNORMAL EEG or tissue at any age and with any mode of inheritance, due
to the twofold genetic origin of RC components (nuclear
BLACK = SEVERE RED = MILD DNA and mitochondrial DNA). It has long been errone-
ously believed that RC disorders originate from mutations
Figure 9.7 
Segregation of variable clinical phenotypes in a family with
multi-system mitochondrial disorder similar to myoclonic epilepsy of mtDNA, as, for some time, only mutations or deletions
lactic acidosis with strokelike symptoms (MELAS). of mtDNA could be identified. However, the number of

1 4 4   •   P rincip l es o f G eno m ic Medicine

known disease-causing mutations in nuclear genes is now
steadily growing. These genes not only encode the various
subunits of each complex, but also the ancillary proteins
involved in the different stages of holoenzyme biogenesis,
including transcription, translation, chaperoning, addition
of prosthetic groups, and assembly of proteins, as well as the
various enzymes involved in mtDNA metabolism.
The increasing number of abnormal mitochondrial
functions and genes leading to RC deficiency continues to
shed light on the clinical heterogeneity of mitochondrial
disorders. The identification of the disease-causing genes
is important not only for genetic counseling and prenatal
diagnosis, but also for a greater understanding of the patho-
physiology of these disorders. However, the ever-increasing
numbers of genes found to be involved in mitochondrial
functions and possibly mitochondrial diseases require the
development of large-scale, high-throughput technologies
that can increase our insight into these highly complicated
pathologies.
REFERENCES
1. Hatefi Y. The mitochondrial electron transport and oxidative phos-
phorylation system. Annu Rev Biochem. 1985;54:1015–1069.
2. Anderson S, Bankier AT, Barrell BG, et al. Sequence and organi-
zation of the human mitochondrial genome. Nature. 1981;290:
457–465.
3. Ruiz-Pesini E, Lott MT, Procaccio V, et  al. An enhanced
MITOMAP with a global mtDNA mutational phylogeny. Nucleic
Acids Res. 2007;35:D823–D828.
4. Clayton DA. Mitochondrial DNA replication:  what we know.
IUBMB Life. 2003;55:213–217.
5. Yang MY, Bowmaker M, Reyes A, et  al. Biased incorporation of
ribonucleotides on the mitochondrial L-strand accounts for appar-
ent strand-asymmetric DNA replication. Cell. 2002;111:495–505.
6. Luo SM, Sun QY. Autophagy is not involved in the degrada-
tion of sperm mitochondria after fertilization in mice. Autophagy.
2013;9:2156–2157.
7. Grivell LA, Artal-Sanz M, Hakkaart G, et al. Mitochondrial assem-
bly in yeast. FEBS Lett. 1999;452:57–60.
8. Contamine V, Picard M. Maintenance and integrity of the mito-
chondrial genome: a plethora of nuclear genes in the budding yeast.
Microbiol Mol Biol Rev. 2000;64:281–315.
9. Bogenhagen DF. Repair of mtDNA in vertebrates. Am J Hum
Genet. 1999;64:1276–1281.
10. Jacobs HT, Turnbull DM. Nuclear genes and mitochondrial transla-
tion: a new class of genetic disease. Trends Genet. 2005;21:312–314.
11. Pagliarini DJ, Calvo SE, Chang B, et  al. A mitochondrial pro-
tein compendium elucidates complex I  disease biology. Cell.
2008;134:112–123.
12. Munnich A, Rustin P. Clinical spectrum and diagnosis of mito-
chondrial disorders. Am J Med Genet. 2001;106:4–17.
13. Rustin P, Chretien D, Bourgeron T, et al. Biochemical and molecu-
lar investigations in respiratory chain deficiencies. Clin Chim Acta.
1994;228:35–51.
14. Goto Y, Nonaka I, Horai S. A mutation in the tRNA(Leu)(UUR)
gene associated with the MELAS subgroup of mitochondrial
encephalomyopathies. Nature. 1990;348:651–653.
Mitochondria l G enetics and G eno m ics in C l inica l Medicine   •   1 4 5
15. van den Ouweland JM, Lemkes HH, Ruitenbeek W, et al. Mutation
in mitochondrial tRNA(Leu)(UUR) gene in a large pedigree with
maternally transmitted type II diabetes mellitus and deafness. Nat
Genet. 1992;1:368–371.
16. Shoffner JM, Lott MT, Lezza AM, Seibel P, Ballinger SW, Wallace
DC. Myoclonic epilepsy and ragged-red fiber disease (MERRF) is
associated with a mitochondrial DNA tRNA(Lys) mutation. Cell.
1990;61:931–937.
17. Holt IJ, Harding AE, Petty RK, Morgan-Hughes JA. A new mito-
chondrial disease associated with mitochondrial DNA hetero-
plasmy. Am J Hum Genet. 1990;46:428–433.
18. Benit P, Chretien D, Kadhom N, et  al. Large-scale deletion
and point mutations of the nuclear NDUFV1 and NDUFS1
genes in mitochondrial complex I  deficiency. Am J Hum Genet.
2001;68:1344–1352.
19. Distelmaier F, Koopman WJ, van den Heuvel LP, et al. Mitochondrial
complex I deficiency: from organelle dysfunction to clinical disease.
Brain. 2009;132:833–842.
20. von Kleist-Retzow JC, Cormier-Daire V, de Lonlay P, et al. A high
rate (20%–30%) of parental consanguinity in cytochrome-oxidase
deficiency. Am J Hum Genet. 1998;63:428–435.
21. Valnot I, Kassis J, Chretien D, et al. A mitochondrial cytochrome b
mutation but no mutations of nuclearly encoded subunits in ubiqui-
nol cytochrome c reductase (complex III) deficiency. Hum Genet.
1999;104:460–466.
22. Andreu AL, Hanna MG, Reichmann H, et al. Exercise intolerance
due to mutations in the cytochrome b gene of mitochondrial DNA.
N Engl J Med. 1999;341:1037–1044.
23. Rahman S, Blok RB, Dahl HH, et  al. Leigh syndrome:  clinical
features and biochemical and DNA abnormalities. Ann Neurol.
1996;39:343–351.
24. Cormier V, Rustin P, Bonnefont JP, et al. Hepatic failure in disor-
ders of oxidative phosphorylation with neonatal onset. J Pediatr.
1991;119:951–954.
25. DiMauro S. Mitochondrial myopathies. Curr Opin Rheumatol.
2006;18:636–641.
26. Gattermann N, Retzlaff S, Wang YL, et  al. Heteroplasmic point
mutations of mitochondrial DNA affecting subunit I of cytochrome
c oxidase in two patients with acquired idiopathic sideroblastic ane-
mia. Blood. 1997;90:4961–4972.
27. Holt IJ, Harding AE, Morgan-Hughes JA. Deletions of muscle
mitochondrial DNA in patients with mitochondrial myopathies.
Nature. 1988;331:717–719.
28. Rotig A, Cormier V, Blanche S, et  al. Pearson’s marrow-pancreas
syndrome. A multisystem mitochondrial disorder in infancy. J Clin
Invest. 1990;86:1601–1608.
29. Moraes CT, DiMauro S, Zeviani M, et  al. Mitochondrial DNA
deletions in progressive external ophthalmoplegia and Kearns-Sayre
syndrome. N Engl J Med. 1989;320:1293–1299.
30. Souied EH, Sales MJ, Soubrane G, et al. Macular dystrophy, diabe-
tes, and deafness associated with a large mitochondrial DNA dele-
tion. Am J Ophthalmol. 1998;125:100–103.
31. Rotig A, Bessis JL, Romero N, et al. Maternally inherited duplica-
tion of the mitochondrial genome in a syndrome of proximal tubu-
lopathy, diabetes mellitus, and cerebellar ataxia. Am J Hum Genet.
1992;50:364–370.
32. Nicolino M, Ferlin T, Forest M, et al. Identification of a large-scale
mitochondrial deoxyribonucleic acid deletion in endocrinopathies
and deafness: report of two unrelated cases with diabetes mellitus
and adrenal insufficiency, respectively. J Clin Endocrinol Metab.
1997;82:3063–3067.
33. Paquis-Flucklinger V, Vialettes B, Vague P, et  al. Importance of
searching for mtDNA defects in patients with diabetes and hearing
deficit. Diabetologia. 1998;41:740–741.
34. Bourgeron T, Rustin P, Chretien D, et  al. Mutation of a nuclear
succinate dehydrogenase gene results in mitochondrial respiratory
chain deficiency. Nat Genet. 1995;11:144–149.
 35. Parfait B, Chretien D, Rotig A, Marsac C, Munnich A, Rustin P.
Compound heterozygous mutations in the flavoprotein gene of
the respiratory chain complex II in a patient with Leigh syndrome.
Hum Genet. 2000;106:236–243.
36. Horvath R, Abicht A, Holinski-Feder E, et  al. Leigh syndrome
caused by mutations in the flavoprotein (Fp) subunit of suc-
cinate dehydrogenase (SDHA). J Neurol Neurosurg Psychiatry.
2006;77:74–76.
37. Baysal BE. Clinical and molecular progress in hereditary paragan-
glioma. J Med Genet. 2008;45:689–694.
38. Haut S, Brivet M, Touati G, et al. A deletion in the human QP-C
gene causes a complex III deficiency resulting in hypoglycaemia and
lactic acidosis. Hum Genet. 2003;113:118–122.
39. Barel O, Shorer Z, Flusser H, et al. Mitochondrial complex III defi-
ciency associated with a homozygous mutation in UQCRQ. Am J
Hum Genet. 2008;82:1211–1216.
40. Shteyer E, Saada A, Shaag A, et al. Exocrine pancreatic insufficiency,
dyserythropoeitic anemia, and calvarial hyperostosis are caused by a
mutation in the COX4I2 gene. Am J Hum Genet. 2009;84:412–417.
41. Massa V, Fernandez-Vizarra E, Alshahwan S, et  al. Severe infan-
tile encephalomyopathy caused by a mutation in COX6B1, a
nucleus-encoded subunit of cytochrome c oxidase. Am J Hum
Genet. 2008;82:1281–1289.
42. Ghezzi D, Goffrini P, Uziel G, et  al. SDHAF1, encoding a LYR
complex-II specific assembly factor, is mutated in SDH-defective
infantile leukoencephalopathy. Nat Genet.2009
43. Hao HX, Khalimonchuk O, Schraders M, et  al. SDH5, a gene
required for flavination of succinate dehydrogenase, is mutated in
paraganglioma. Science. 2009;325:1139–1142.
44. de Lonlay P, Valnot I, Barrientos A, et al. A mutant mitochondrial
respiratory chain assembly protein causes complex III deficiency in
patients with tubulopathy, encephalopathy and liver failure. Nat
Genet. 2001;29:57–60.
45. Visapaa I, Fellman V, Vesa J, et  al. GRACILE syndrome, a lethal
metabolic disorder with iron overload, is caused by a point mutation
in BCS1L. Am J Hum Genet. 2002;71:863–876.
46. Hinson JT, Fantin VR, Schonberger J, et  al. Missense mutations
in the BCS1L gene as a cause of the Bjornstad syndrome. N Engl J
Med. 2007;356:809–819.
47. Ghezzi D, Arzuffi P, Zordan M, et al. Mutations in TTC19 cause
mitochondrial complex III deficiency and neurological impairment
in humans and flies. Nat Genet. 2011;43:259–263.
48. Sue CM, Karadimas C, Checcarelli N, et al. Differential features of
patients with mutations in two COX assembly genes, SURF-1 and
SCO2. Ann Neurol. 2000;47:589–595.
49. Bundschuh FA, Hannappel A, Anderka O, Ludwig B. SURF1, asso-
ciated with Leigh syndrome in humans, is a heme-binding protein in
bacterial oxidase biogenesis. J Biol Chem. 2009;284:25735–25741.
50. Antonicka H, Mattman A, Carlson CG, et al. Mutations in COX15
produce a defect in the mitochondrial heme biosynthetic pathway,
causing early-onset fatal hypertrophic cardiomyopathy. Am J Hum
Genet. 2003;72:101–114.
51. Oquendo CE, Antonicka H, Shoubridge EA, Reardon W, Brown
GK. Functional and genetic studies demonstrate that mutation
in the COX15 gene can cause Leigh syndrome. J Med Genet.
2004;41:540–544.
52. Bugiani M, Tiranti V, Farina L, Uziel G, Zeviani M. Novel muta-
tions in COX15 in a long surviving Leigh syndrome patient with
cytochrome c oxidase deficiency. J Med Genet. 42, e28.2005
53. Valnot I, von Kleist-Retzow JC, Barrientos A, et  al. A muta-
tion in the human heme A:farnesyltransferase gene (COX10)
causes cytochrome c oxidase deficiency. Hum Mol Genet. 2000;9:
1245–1249.
54. Antonicka H, Leary SC, Guercin GH, et al. Mutations in COX10
result in a defect in mitochondrial heme A biosynthesis and account
for multiple, early-onset clinical phenotypes associated with isolated
COX deficiency. Hum Mol Genet. 2003;12:2693–2702.
1 4 6   •   P rincip l es o f G eno m ic Medicine
55. Valnot I, Osmond S, Gigarel N, et  al. Mutations of the SCO1
gene in mitochondrial cytochrome c oxidase deficiency with
neonatal-onset hepatic failure and encephalopathy. Am J Hum
Genet. 2000;67:1104–1109.
56. Papadopoulou LC, Sue CM, Davidson MM, et  al. Fatal infantile
cardioencephalomyopathy with COX deficiency and mutations in
SCO2, a COX assembly gene. Nat Genet. 1999;23:333–337.
57. Leary SC, Sasarman F, Nishimura T, Shoubridge EA. Human SCO2
is required for the synthesis of CO II and as a thiol-disulphide oxi-
doreductase for SCO1. Hum Mol Genet. 2009;18:2230–2240.
58. Mootha VK, Lepage P, Miller K, et al. Identification of a gene caus-
ing human cytochrome c oxidase deficiency by integrative genomics.
Proc Natl Acad Sci U S A. 2003;100:605–610.
59. Xu F, Morin C, Mitchell G, Ackerley C, Robinson BH. The role of
the LRPPRC (leucine-rich pentatricopeptide repeat cassette) gene
in cytochrome oxidase assembly: mutation causes lowered levels of
COX (cytochrome c oxidase) I  and COX III mRNA. Biochem J.
2004;382:331–336.
60. Ghezzi D, Saada A, D’Adamo P, et al. FASTKD2 nonsense muta-
tion in an infantile mitochondrial encephalomyopathy associ-
ated with cytochrome c oxidase deficiency. Am J Hum Genet.
2008;83:415–423.
61. Weraarpachai W, Antonicka H, Sasarman F, et  al. Mutation in
TACO1, encoding a translational activator of COX I, results in
cytochrome c oxidase deficiency and late-onset Leigh syndrome.
Nat Genet. 2009;41:833–837.
62. Weraarpachai W, Sasarman F, Nishimura T, et  al. Mutations in
C12orf62, a factor that couples COX I synthesis with cytochrome
c oxidase assembly, cause fatal neonatal lactic acidosis. Am J Hum
Genet. 2012;90:142–151.
63. De Meirleir L, Seneca S, Lissens W, et al. Respiratory chain complex
V deficiency due to a mutation in the assembly gene ATP12. J Med
Genet. 2004;41:120–124.
64. Cizkova A, Stranecky V, Mayr JA, et al. TMEM70 mutations cause
isolated ATP synthase deficiency and neonatal mitochondrial
encephalocardiomyopathy. Nat Genet. 2008;40:1288–1290.
65. Rotig A, de Lonlay P, Chretien D, et al. Aconitase and mitochon-
drial iron-sulphur protein deficiency in Friedreich ataxia. Nat Genet.
1997;17:215–217.
66. Mochel F, Knight MA, Tong WH, et  al. Splice mutation in the
iron-sulfur cluster scaffold protein ISCU causes myopathy with
exercise intolerance. Am J Hum Genet. 2008;82:652–660.
67. Olsson A, Lind L, Thornell LE, Holmberg M. Myopathy with lac-
tic acidosis is linked to chromosome 12q23.3–24.11 and caused by
an intron mutation in the ISCU gene resulting in a splicing defect.
Hum Mol Genet. 2008;17:1666–1672.
68. Rotig A, Mollet J, Rio M, Munnich A. Infantile and pediatric quinone
deficiency diseases. Mitochondrion. 2007 Jun;7 Suppl:S112–S121.
69. Duncan AJ, Bitner-Glindzicz M, Meunier B, et  al. A nonsense
mutation in COQ9 causes autosomal-recessive neonatal-onset
primary coenzyme Q10 deficiency: a potentially treatable form of
mitochondrial disease. Am J Hum Genet. 2009;84:558–566.
70. Mollet J, Delahodde A, Serre V, et al. CABC1 gene mutations cause
ubiquinone deficiency with cerebellar ataxia and seizures. Am J
Hum Genet. 2008;82:623–630.
71. Bogenhagen DF, Rousseau D, Burke S. The layered struc-
ture of human mitochondrial DNA nucleoids. J Biol Chem.
2008;283:3665–3675.
72. Spinazzola A, Zeviani M. Disorders of nuclear-mitochondrial
intergenomic communication. Biosci Rep. 2007;27:39–51.
73. Van Goethem G, Dermaut B, Lofgren A, Martin JJ, Van
Broeckhoven C. Mutation of POLG is associated with progressive
external ophthalmoplegia characterized by mtDNA deletions. Nat
Genet. 2001;28:211–212.
74. Longley MJ, Clark S, Yu Wai Man C, et al. Mutant POLG2 disrupts
DNA polymerase gamma subunits and causes progressive external
ophthalmoplegia. Am J Hum Genet. 2006;78:1026–1034.
 75. Kaukonen J, Juselius JK, Tiranti V, et al. Role of adenine nucleotide
translocator 1 in mtDNA maintenance. Science. 2000;289:782–785.
76. Spelbrink JN, Li FY, Tiranti V, et al. Human mitochondrial DNA
deletions associated with mutations in the gene encoding Twinkle, a
phage T7 gene 4-like protein localized in mitochondria. Nat Genet.
2001;28:223–231.
77. Amati-Bonneau P, Valentino ML, Reynier P, et al. OPA1 mutations
induce mitochondrial DNA instability and optic atrophy “plus”
phenotypes. Brain. 2008;131:338–351.
78. Moraes CT, Shanske S, Tritschler HJ, et al. mtDNA depletion with
variable tissue expression: a novel genetic abnormality in mitochon-
drial diseases. Am J Hum Genet. 1991;48:492–501.
79. Sarzi E, Bourdon A, Chretien D, et al. Mitochondrial DNA deple-
tion is a prevalent cause of multiple respiratory chain deficiency in
childhood. J Pediatr. 2007;150:531–534, 534.
80. Rotig A, Poulton J. Genetic causes of mitochondrial DNA deple-
tion in humans. Biochim Biophys Acta. 2009;1792:1103–1108.
81. Ferrari G, Lamantea E, Donati A, et  al. Infantile hepatocerebral
syndromes associated with mutations in the mitochondrial DNA
polymerase-gammaA. Brain. 2005;128:723–731.
82. Naviaux RK, Nguyen KV. POLG mutations associated with
Alpers’ syndrome and mitochondrial DNA depletion. Ann Neurol.
2004;55:706–712.
83. Mandel H, Szargel R, Labay V, et  al. The deoxyguanosine kinase
gene is mutated in individuals with depleted hepatocerebral mito-
chondrial DNA. Nat Genet. 2001;29:337–341.
84. Spinazzola A, Viscomi C, Fernandez-Vizarra E, et  al. MPV17
encodes an inner mitochondrial membrane protein and is mutated
in infantile hepatic mitochondrial DNA depletion. Nat Genet.
2006;38:570–575.
85. Saada A, Shaag A, Mandel H, Nevo Y, Eriksson S, Elpeleg O. Mutant
mitochondrial thymidine kinase in mitochondrial DNA depletion
myopathy. Nat Genet. 2001;29:342–344.
86. Mancuso M, Salviati L, Sacconi S, et al. Mitochondrial DNA deple-
tion: mutations in thymidine kinase gene with myopathy and SMA.
Neurology. 2002;59:1197–1202.
87. Gotz A, Isohanni P, Pihko H, et al. Thymidine kinase 2 defects can
cause multi-tissue mtDNA depletion syndrome. Brain.2008
88. Bourdon A, Minai L, Serre V, et al. Mutation of RRM2B, encod-
ing p53-controlled ribonucleotide reductase (p53R2), causes severe
mitochondrial DNA depletion. Nat Genet. 2007;39:776–780.
89. Elpeleg O, Miller C, Hershkovitz E, et  al. Deficiency of the
ADP-forming succinyl-CoA synthase activity is associated with
encephalomyopathy and mitochondrial DNA depletion. Am J Hum
Genet. 2005;76:1081–1086.
90. Ostergaard E, Christensen E, Kristensen E, et al. Deficiency of the
alpha subunit of succinate-coenzyme A ligase causes fatal infantile
lactic acidosis with mitochondrial DNA depletion. Am J Hum
Genet. 2007;81:383–387.
91. Nishino I, Spinazzola A, Hirano M. Thymidine phosphorylase gene
mutations in MNGIE, a human mitochondrial disorder. Science.
1999;283:689–692.
92. Tyynismaa H, Ylikallio E, Patel M, Molnar MJ, Haller RG,
Suomalainen A. A heterozygous truncating mutation in RRM2B
causes autosomal-dominant progressive external ophthalmople-
gia with multiple mtDNA deletions. Am J Hum Genet. 2009;85:
290–295.
93. Shaibani A, Shchelochkov OA, Zhang S, et al. Mitochondrial neu-
rogastrointestinal encephalopathy due to mutations in RRM2B.
Arch Neurol. 2009;66:1028–1032.
Mitochondria l G enetics and G eno m ics in C l inica l Medicine   •   1 4 7
94. Patton JR, Bykhovskaya Y, Mengesha E, Bertolotto C,
Fischel-Ghodsian N. Mitochondrial myopathy and sideroblastic
anemia (MLASA):  missense mutation in the pseudouridine syn-
thase 1 (PUS1) gene is associated with the loss of tRNA pseudouri-
dylation. J Biol Chem. 2005;280:19823–19828.
95. Riley LG, Cooper S, Hickey P, et al. Mutation of the mitochon-
drial tyrosyl-tRNA synthetase gene, YARS2, causes myopathy,
lactic acidosis, and sideroblastic anemia--MLASA syndrome. Am
J Hum Genet. 2010;87:52–59.
96. Scheper GC, van der Klok T, van Andel RJ, et al. Mitochondrial
aspartyl-tRNA synthetase deficiency causes leukoencephalopathy
with brain stem and spinal cord involvement and lactate elevation.
Nat Genet. 2007;39:534–539.
97. Edvardson S, Shaag A, Kolesnikova O, et  al. Deleterious muta-
tion in the mitochondrial arginyl-transfer RNA synthetase gene
is associated with pontocerebellar hypoplasia. Am J Hum Genet.
2007;81:857–862.
98. Gotz A, Tyynismaa H, Euro L, et al. Exome sequencing identifies
mitochondrial alanyl-tRNA synthetase mutations in infantile mito-
chondrial cardiomyopathy. Am J Hum Genet. 2011;88:635–642.
99. Belostotsky R, Ben-Shalom E, Rinat C, et  al. Mutations in the
mitochondrial Seryl-tRNA synthetase cause hyperuricemia,
pulmonary hypertension, renal failure in infancy and alkalosis,
HUPRA syndrome. Am J Hum Genet. 2011;88:193–200.
100. Pierce SB, Chisholm KM, Lynch ED, et  al. Mutations in mito-
chondrial histidyl tRNA synthetase HARS2 cause ovarian dysgen-
esis and sensorineural hearing loss of Perrault syndrome. Proc Natl
Acad Sci U S A. 2011;108:6543–6548.
101. Smeitink JA, Elpeleg O, Antonicka H, et al. Distinct clinical phe-
notypes associated with a mutation in the mitochondrial transla-
tion elongation factor EFTs. Am J Hum Genet. 2006;79:869–877.
102. Coenen MJ, Antonicka H, Ugalde C, et al. Mutant mitochondrial
elongation factor G1 and combined oxidative phosphorylation
deficiency. N Engl J Med. 2004;351:2080–2086.
103. Valente L, Tiranti V, Marsano RM, et al. Infantile encephalopathy
and defective mitochondrial DNA translation in patients with
mutations of mitochondrial elongation factors EFG1 and EFTu.
Am J Hum Genet. 2007;80:44–58.
104. Miller C, Saada A, Shaul N, et al. Defective mitochondrial trans-
lation caused by a ribosomal protein (MRPS16) mutation. Ann
Neurol. 2004;56:734–738.
105. Saada A, Shaag A, Arnon S, et al. Antenatal mitochondrial disease
caused by mitochondrial ribosomal protein (MRPS22) mutation.
J Med Genet. 2007;44:784–786.
106. Galmiche L, Serre V, Beinat M, et al. Exome sequencing identifies
MRPL3 mutation in mitochondrial cardiomyopathy. Hum Mutat.
2011;32:1225–1231.
107. Zeharia A, Shaag A, Pappo O, et  al. Acute infantile liver failure
due to mutations in the TRMU gene. Am J Hum Genet. 2009;85:
401–407.
108. Tucker EJ, Hershman SG, Kohrer C, et al. Mutations in MTFMT
underlie a human disorder of formylation causing impaired mito-
chondrial translation. Cell Metab. 2011;14:428–434.
109. Crosby AH, Patel H, Chioza BA, et al. Defective mitochondrial
mRNA maturation is associated with spastic ataxia. Am J Hum
Genet. 2010;87:655–660.
110. Bouchet C, Steffann J, Corcos J, et  al. Prenatal diagnosis of
MELAS syndrome: contribution to understanding mitochondrial
DNA segregation during human embryo fetal development. J Med
Genet. 2006;43:788–792.
 10.
GENOMICS TECHNOLOGY IN CLINICAL DIAGNOSTICS
Kevin White and Jeremy Segal
INTRODUCTION
During the last two decades, and especially since the com-
pletion of the first human genome draft sequence in 2001,1,2
we have witnessed an unprecedented expansion of our
molecular genetics capabilities for both discovery and diag-
nostics. While certain advances arose out of the demands
of the Human Genome Project itself, others have only been
made possible because of its successful outcome. With
the initial sequence in hand, a series of ambitious projects
has enabled the in-depth investigation of variation in the
human genome (e.g., 1,000 Genomes Project),3 the identi-
fication of functional elements encoded within the human
genome (Encyclopedia of DNA Elements—ENCODE—
Project),4 and mapping of the genetic architectures of com-
mon cancers (e.g., The Cancer Genome Atlas—TCGA).5–8
Each of these projects represents much more than just
an achievement or a milestone in human genetics. These
human genome sequences have become one of the most
important and frequently used tools available to human
genetics researchers and diagnosticians. Many of the newer
genomics technologies described in this chapter relied on
genomics data for their creation, and others (particularly
next-generation sequencing) are dependent on the fruits of
the Human Genome Project to perform basic analyses.
New molecular biology techniques are always initially
adopted by research laboratories, and it is in this setting
that recent genomics technologies have so far been the most
transformative. In many ways, research in genetics is almost
unrecognizable compared with the state of the science even
five or ten years ago. In the clinical setting, there is more at
stake when it comes to replacing traditional, proven tech-
nologies with novel applications, so the pace of change is
understandably more cautious. Still, the integration of next
generation technologies into the clinical laboratory has
begun in earnest, and it is only accelerating.
The number of applications of human genomic analy-
sis in modern clinical practice is vast and is constantly
148
expanding due to new discoveries and technological break-
throughs. However, clinical applications today generally
focus on two main types of genomic variation:
• Inborn (or constitutional) variation: Constitutional
genetic variants are present from the point of fertiliza-
tion, and thus are found in the genome of every cell in
the individual. As a result, essentially any cellular sample
type may be acceptable for testing. Constitutional vari-
ants are commonly heterozygous or homozygous (i.e.,
present at allelic fractions of 50% or 100%, respectively),
meaning that they are comparatively easy to detect and
may be effectively assayed using lower sensitivity meth-
ods. Mitochondrial genetics are an obvious exception, as
are sex chromosome genetics and other rare conditions
such as chimerism.
• Somatic variation: Somatic DNA alterations occur in
individual cells of the body through a variety of means,
including DNA damage and replication errors as seen
in both normal aging and neoplasia, as well as in normal
cellular processes such as immune cell variable-diverse-
joining gene segments (VDJ) recombination and somatic
hypermutation. In somatic mutation testing, s­ uccess
depends on proper sample selection. For example, in
cancer diagnostics, pathological analysis of each ­sample
should be conducted prior to testing to determine
whether a sufficient number/proportion of tumor cells
are present that may harbor mutations, in order to avoid
false negative results. Due to anticipated sample hetero-
geneity and the potential for mutations with low allelic
percentage, high-sensitivity methods may be required.
The full length of the haploid human genome is approx-
imately 3 billion base pairs. Remarkably, clinically relevant
genetic anomalies can be of any size, from the smallest
single base substitutions to macroscopic chromosomal
defects, altered chromosomal numbers, and even altered
 Performance of Genomic Technologies Across Size Scales
100 101 102 103 104 105 106 107 108 109 bp

PCR-based Methods

MLPA

Southern Blotting

CGH/SNP Arrays

Fluorescence In Situ Hybridization (FISH)

Spectral Karyotyping

Cytogenetics

Next Generation Sequencing (NGS)

Point Mutation Small Insertion/ Larger Duplication/ Trisomy/Monosomy Altered Ploidy

Deletion Deletion

Figure 10.1 
Depicted are the major genomic analysis technologies discussed in this chapter with the approximate genomic size scales to which they are
best suited for detection of anomalies. It should be noted that certain anomalies, such as balanced translocations, require separate consideration. For
example, such translocations may be detected by cytogenics but not by CGH/SNP arrays.

copy number of the entire genome (e.g., triploidy or tet- T R A D I T I O N AL G E N ET I C A N D

raploidy in partial hydatidiform molar pregnancies9). This
size range of possible anomalies, covering nine orders of
magnitude, is akin to the difference between the length of
one of your fingernails and the circumference of the earth.
This represents a remarkable challenge from the point of
view of genetic analysis and is the reason why so many dif-
ferent genetic analysis technologies and strategies exist.
Nearly every tool is best suited to interrogate anomalies of
a certain size range, and it is critical to keep in mind the
size scale of expected anomalies and their anticipated loca-
tion when planning any genetic investigation or diagnos-
tic. Figure 10.1 shows some of the most common genetic
analysis technologies and the size scales to which they are
best suited. This should serve as a useful reference for the
following discussion.
A full discussion of all genetic and genomic analysis
techniques is beyond the scope of any one chapter, so here
we will focus on a sampling of the most common technolo-
gies in use or in development in the diagnostic setting. We
will begin by covering more traditional methods of genetic
analysis before moving to a discussion of today’s modern
genomics tools.
G E N O M I C A N ALYS I S T E C H N I Q U ES
Though it is somewhat of an arbitrary distinction, let us
state that small genetic anomalies are those that are less than
approximately 1000 base pairs. These include mainly single
base substitutions and small insertions, deletions, or inser-
tion/deletions (with both loss and gain of DNA). There are
numerous ways that these small anomalies can be detected
(including newer genomics technologies), but tradition-
ally the most common methods in both the research labo-
ratory and clinic are based on polymerase chain reaction
(PCR). Larger anomalies cover the remaining six orders of
magnitude, and a variety of techniques is utilized in their
detection.
D ET EC T I O N O F S M A L L G E N ET I C VA R I A N TS
The invention of PCR in 1983 stands as one of the most sig-
nificant developments in the history of biology, as it was the
first technique to make targeted genetic analysis (including
sequence analysis) practical and straightforward.10 Because
of the size of the genome, bulk genomic DNA contains only

G e no m ic s T e c h no l ogy in C l inic a l D i agno s tic s   •   1 4 9

a vanishingly small proportional amount of any individual
sequence of interest. Before the introduction of PCR, this
represented a nearly insurmountable “signal-to-noise” prob-
lem for the investigation of genomic loci. Essentially the
only available direct method not based on fragment cloning
was Southern blotting (described below).
Briefly, PCR requires the design of short oligonucle-
otide primers that flank a sequence of interest. Primers,
free nucleotides, and polymerase enzyme are added to the
target DNA. The target DNA strands are separated (dena-
tured) by increasing the temperature, and upon cooling,
the target sequences are bound by the primers, which are
then extended by the polymerase to copy the template. The
cycle can be repeated simply by repeating the temperature
changes. Because amplification proceeds in an exponential
fashion, 40 rounds of PCR theoretically can produce a 240
(trillion)–fold amplification from as little as one original
template molecule. Thus, the procedure essentially trans-
forms a dilute sample of heterogeneous genomic DNA
into a concentrated clonal solution of amplicon copies of
the desired target sequence. This amplified product is then
highly amenable to a large variety of downstream analytics,
including Sanger sequencing (Figure 10.2) to determine its
sequence and identify mutations or variants. The progress
of a PCR reaction itself can also be measured via fluorescent
markers, from which data the amount of initial template in
the reaction can be inferred. This application, quantitative
PCR (qPCR) is used routinely in diagnostics laboratories
for a variety of purposes, including viral load testing.
A Labeled
AT Products
AT G
A TG A T A CG TG
AT G A
AT G AT
AT G ATA
Electrophoresis
AT G ATAC
AT G ATAC G
AT G ATAC G T
AT G ATAC G TG
A T G A T A C G T G …..
Template DNA Sequence Trace
Figure 10.2 
Sanger sequencing. Following PCR amplification, many
methods are available to derive length and sequence information from
a target amplicon. One common and important method is Sanger
sequencing. A sequencing primer binds the amplified template DNA
and is extended by a polymerase in the presence of fluorescent dead-end
(terminator) nucleotides, producing labeled fragments of different size.
These are separated by capillary electrophoresis to produce characteristic
sequencing plots, from which the DNA sequence may be read.
1 5 0   •   P rincip l e s o f G e no m ic M e dicin e
Although it is possible to perform PCR on DNA seg-
ments longer than a few thousand bp, the technical chal-
lenges increase beyond this point and ultimately require
specialized polymerases and protocol modifications to pre-
vent the amplification of non-specific products.11 Clinical
procedures seldom include amplicons greater than 1kb.
Today, PCR is used in too many clinical diagnostic applica-
tions to easily count. Many are described in other chapters,
including such applications as:
• Assaying inborn disease mutations
• Identifying cancer mutations and cancer-related
translocations
• Identity testing, for both patient and forensic samples
• Identifying/quantifying viral and bacterial genomes for
infectious disease diagnostics
• Analyzing and quantifying mRNA expression
D ET EC T I O N O F L A RG E R S C A L E A N O M A L I E S
There is a variety of traditional approaches for the analysis
of larger genomic anomalies, the selection of which again
depends critically the exact size scale of the expected anom-
aly, the predictability of its location, as well as the particular
application or question. When looking at these techniques,
there is a general trend from molecular biology–based
approaches at the smaller orders of magnitude, moving ulti-
mately towards straightforward microscopy as anomalies
become large enough to detect with the assisted eye.
Cytogenetics
For the largest scale anomalies, classical cytogenetics is a
powerful traditional approach for carrying out a genome-
wide scan at a relatively low cost.12 It requires first culturing
cells and arresting them in metaphase (when chromosomes
are condensed after replication) using a mitotic inhibitor
such as colcemid, which blocks the microtubule polymer-
ization necessary for forming the mitotic spindle appara-
tus. Cells are then exposed to hypotonic solution causing
them to swell, and dropped onto glass slides, where they
burst and locally scatter their chromosomes into “spreads.”
Chromosomes may be counted at this stage, but typically
they are stained via any of a number of different tech-
niques, which serve to reveal structural details and allow
individual identification.13 Giemsa banding (G-banding) is
the most common method, which produces characteristic
 Free ebooks ==> www.Ebook777.com
chromosomal bands.14 By this method, late-replicating,
transcriptionally quiet, and A/T-rich DNA staining is more
intense (G-positive) and early-replicating, transcriptionally
active, and relatively G/C-rich DNA is stained more lightly
(G-negative) (Figure 10.3). The highest quality G-banded
preparations are able to yield up to 850 bands across the
genome.15 Microscopic analysis of banded chromosomes
assists with their individual identification and can reveal loss
or gain of material down to a few million base pairs (mega-
bases or Mb) in size, though exactly how small depends on
the location of the defect and the resolution of the bands. It
can also reveal events such as translocations and inversions,
even when no net gain or loss of material occurs (compare
with comparative genomic hybridization [CGH] and single
nucleotide polymorphisms [SNP] arrays, see below). Other
banding methods are available that can reveal complemen-
tary information about the structure and organization of the
genome at the chromosomal level. One particular method,
Q-banding, involves a fluorescent dye such as quinacrine,
DAPI, or Hoechst 33258, that binds preferentially to A/T-
rich sequences, producing a banding pattern comparable
to G-banding that can be used during fluorescence in situ
hybridization (FISH) experiments (see below).16
Though developed many decades ago, cytogenetics
is still widely used in clinical practice today for both con-
stitutional and oncology diagnostics. It is a first-line test
for children with developmental abnormalities and for
fetal samples obtained via amniocentesis or chorionic vil-
lus sampling, when there is reason to suspect a structural
or numerical chromosomal abnormality.17,18 There is also
a long and rich history of cytogenetic analysis of hemato-
logical malignancies, going back to the discovery of the
Philadelphia chromosome in chronic myelogenous leuke-
mia in 1960 (the first observation of a recurrent genetic
anomaly in cancer),19 followed soon after by the elucidation
of the t(9;22) translocation from which it arises.20 Today,
cytogenetics is still heavily relied upon in the diagnostic
work-up of these diseases, particularly the leukemias and
myelodysplastic syndromes (MDS). There are many cytoge-
netic signatures for these diseases, including some that are
disease-defining or major diagnostic criteria, and others that
provide therapy-related or prognostic information.21,22,23
For instance, the 5q minus syndrome is a specific subtype of
MDS showing deletions of the long arm of chromosome 5
(often involving 5q31-5q32) that is associated with an over-
all favorable prognosis and a high likelihood of response to
lenalidomide therapy.24
In contrast to the case in heme malignancies, there is
essentially no routine clinical utility of cytogenetics for solid
tumors. Solid tumors tend to have many more cytogenetic
G e no m ic s T e c h no l ogy in C l inic a l D i agno s tic s   •   1 5 1
www.Ebook777.com
anomalies, and chromosomes prepared from these tumors
are generally more condensed and thus have lower banding
resolution, interfering with interpretation. Though modern
genomics technologies are now beginning to be applied to
these tumors, diagnostic analyses for many large rearrange-
ments have typically been performed by FISH.
Fluorescence In Situ Hybridization (FISH)
In situ hybridization techniques, in particular FISH, have
allowed study of the structure of the genome at a level of
detail greater than that seen by conventional banding tech-
niques.25,26,27 However, unlike cytogenetics, FISH is a tar-
geted assay requiring foreknowledge of the expected genetic
lesion. The method depends on the specific hybridization
of a probe DNA sequence to its complementary sequence
in the genome. Labeling the probe with a fluorescent dye
allows its location to be revealed by fluorescence micros-
copy. The probes used in FISH experiments are typically
derived from human sequences cloned into bacterial arti-
ficial chromosomes (BACs), with sizes of approximately
100kb. The creation of an extensive human BAC library
(containing approximately 32,000 BACs tiled across the
genome) was actually the first critical step of the Human
Genome Project itself, effectively breaking the genome into
“bite-sized” pieces that could be individually sequenced,
with subsequent assembly to create the final sequence.28
This library now serves as a main source of DNA for FISH
probes, thus there are only very few regions of the genome
not amenable to FISH experiments.
FISH can be performed either on metaphase chromo-
some spreads or on preparations of cells in interphase.29
When applied to metaphase spreads, specific signal from
sites of probe binding can be evaluated in the context of
Q-banding data. Thus, metaphase FISH allows the count-
ing of probe binding sites, determination of the identity
of chromosomes showing probe staining, as well as the
sub-chromosomal location of binding events (Figure 10.4).
This is a wealth of useful information, but as with cytoge-
netics, this technique relies on the growth of cells in culture,
which is expensive and can take days to weeks. Not all speci-
men types may grow reliably, and in certain scenarios, this
process cannot be completed in a clinically relevant time
frame. For example, in cases of suspected acute promyelo-
cytic leukemia, rapid identification of the t(15;17) trans-
location producing the PML-RARa fusion oncogene is
important in order to inform decisions regarding treatment
with all trans retinoic acid (ATRA).30 In practice, this is typ-
ically performed either via reverse transcription of RNA to
DNA followed by PCR (RT-PCR) or via interphase FISH.
 Interphase FISH can be readily performed on essen-
tially any sample that contains nucleated cells, with no
requirement for culture. This includes smears of any cel-
lular bodily fluid, tissue touch preparations and frozen sec-
tions, and even formalin-fixed, paraffin embedded (FFPE)
tissue sections. Aside from some differences in the prepa-
ration of cells and target DNA, the underlying concept
is the same as for metaphase FISH. The difference lies in
the state of the chromatin of non-metaphase cells, which
is uncondensed and dispersed throughout the nucleus.
However, though the probe lengths may be on the order of
100kb, this is short enough to produce staining in discrete
spots within the nucleus. Of course, by this technique, no
contextual chromosomal information is produced. This is
a limiting factor, yet the process still has many important
applications:

(A)
Case:

A 2 y old female with developmental delay, short stature and microcephaly, otherwise not dysmorphic. Normal MRI brain scan.

Found to have an apparently balanced translocation between chromosomes 7 and 13 (46,XX,t(7;13)(q21.2;q12.3)de novo), as shown by the two
arrows on the karyotype.

Subsequent array CGH analysis, using a whole genome array of 0.1 Mb BAC clones spaced ~1 Mb apart, has shown that in the regions of
the breakpoints on chromosomes 7 and 13 that there has actually been loss of DNA (indicated in red on the diagram of the chromosomes;
Fig 7.2), and the translocation is therefore not balanced. Loss of 0.2 Mb of chromosome 13 and 8.42 Mb of chromosome 7 is consistent with the
phenotype and its severity.

(Courtesy of Sian Morgan, Cytogenetics Laboratory, Institute of Medical Genetics, Cardiff, UK)

(B)

1 2 3 4 5

6 7 8 9 10 11 12

13 14 15 16 17 18

19 20 21 22 X Y

Figure 10.3 
Conventional cytogenetic analysis by G-banding. Depicted is a chromosomal spread from a two-year-old female with developmental delay,
short stature, and microcephaly. Cytogenetic analysis revealed a translocation between chromosomes 7 and 13 (46XX, t(7:13)(q21.2;q12.3)), as
shown by the two arrows on the karyotype. A deletion of chromosomal material at the breakpoint region might be expected based on the clinical
scenario, yet the translocation appears to be balanced. However, such is the sensitivity of cytogenetics that even megabase-scale anomalies may not
be readily detectable. (Courtesy of Sian Morgan, Cytogenetics Laboratory, Institute of Medical Genetics, Cardiff, UK)

1 5 2   •   P rincip l e s o f G e no m ic M e dicin e
 (C)
0
16
32
48
63
i) Chromosome 7 79
95
111
127
143
159
–2.00 –1.60 –1.20 –0.80 –0.40 0.00 0.40 0.80 1.20 1.60 2.00
Chromosome 7 Log2 Ratio Ch1/Ch2
01100200_top - 01100200_bottom.bsn - 14/06/2006

0
11
23
34
46
57
ii) Chromosome 68
13 80
91
103
114
–2.00 –1.60 –1.20 –0.80 –0.40 0.00 0.40 0.80 1.20 1.60 2.00
Chromosome 13 Log2 Ratio Ch1/Ch2
01100200_top - 01100200_bottom.bsn - 14/06/2006

Figure 10.3 Continued

• Chromosome counting: Individual chromosomes can be Multicolor FISH and Spectral Karyotyping

counted using probes targeting specific chromosomes,
though with the caveat that structural data about the
intactness of the entire chromosome will not be avail-
able. A common example is rapid aneuploidy screening
on amniocentesis samples using centromeric probes for
chromosomes 13, 18, 21, X, and Y (Figure 10.5a).
• Deletion/duplication analysis: Interphase FISH allows
the counting of gene or target region dosage; for exam-
ple, using probes to look for Her2 gene amplifications in
breast cancer that predict a favorable response to trastu-
zumab therapy.31
• Translocation analysis: Translocations and large inversions
can be detected with high sensitivity using a common
technique called “break apart FISH,” which uses two dif-
ferently colored probes targeted immediately upstream
and downstream of a gene of interest. The two probes
produce essentially overlapping spots in the normal state,
but show physical separation if a copy of the gene has
been involved in a translocation event (Figure 10.5b).
In many types of cancer, large numbers of cytogenetic alter-
ations (translocations, etc.) can frequently interfere with
definitive chromosomal identification by classical cytoge-
netics. Marker chromosomes (the term for those that are
unidentifiable) can be very complex assemblies of multiple
chromosomal parts. Technically, multiple individual FISH
experiments could be performed to attempt to identify
component parts of marker chromosomes, but in practice
this may be infeasible. Another approach is to perform
FISH using probes generated from individual whole chro-
mosomes (chromosome painting), which would light up,
not only all of the target chromosome, but also any part
of a marker chromosome derived from it. If all 24 differ-
ent chromosome paints are applied, the technique becomes
even more sophisticated and is known as multicolor FISH
(M-FISH), which was further developed as spectral karyo-
typing (SKY).32,33 This technique, though mostly used in
the research setting, produces wonderfully detailed images
of chromosome spreads (Figure 10.6), and can help resolve

G e no m ic s T e c h no l ogy in C l inic a l D i agno s tic s   •   1 5 3


Deleted chromosome 22
Normal chromosome 22
Figure 10.4 
Metaphase FISH of a patient with DiGeorge syndrome
(22q11.2 deletion syndrome). Two differently colored probes are
applied: a control probe targeted to the distal portion of the long arm
of chromosome 22 (green) and a test probe targeted to the DiGeorge
region (red). Two copies of chromosome 22 are present, but only one
contains the sequence matching the DiGeorge test probe, indicating the
presence of a deletion on the other chromosome. (Courtesy of Dr. Peter
Thompson, Cytogenetics Laboratory, Institute of Medical Genetics,
Cardiff, UK)
even some of the most complex marker chromosome
rearrangements.34,35
Southern Blotting
Southern blotting involves restriction digestion of genomic
DNA into reliable fragments, size separation by gel elec-
trophoresis, and transfer to a membrane. The membrane is
then hybridized with a specific probe to detect a particular
genomic fragment, with the probe labeled with a radioac-
tive isotope or other similar system to produce sufficient
signal amplification. Labeled bands are then analyzed
to determine if they are of the expected size.36 Southern
blotting can show insertions or deletions if they are large
enough to affect the migration of the probed band, but very
small-scale variants may only be identified if they destroy or
create a restriction enzyme site, producing a new fragment
size (i.e., a restriction fragment length polymorphism, or
RFLP). There is significant labor associated with Southern
blotting, and for many applications it is of only limited util-
ity. Today it is used clinically for a few anomalies too large
for PCR and too small for FISH. For example, in particu-
lar tri-nucleotide repeat disorders (e.g., CGG repeats in the
FMR1 gene in fragile X syndrome), the repeat stretches can
grow longer than 1000 bp, leading to failed amplification
and false-negative PCR results. In these cases, Southern
blotting is often necessary to ensure the detection of long
repeats.37
1 5 4   •   P rincip l e s o f G e no m ic M e dicin e
Multiple Ligation-Dependent Probe
Amplification (MLPA)
MLPA is a recently developed technique that enables low
cost, targeted copy number analysis.38 Each MLPA reaction
relies on two oligonucleotide probes designed to hybridize
side-by-side on the target. Probes that hybridize success-
fully are enzymatically ligated together, thus the amount of
ligated product is proportional to the amount of template
in the sample. Subsequently, successfully ligated probes can
be amplified by PCR to enable quantification. Use of tagged
sequences at the ends of the probes allows the amplification
of many probe-sets using a single pair of primers, support-
ing greater multiplexing than does multiplex PCR (which
uses many different primer pairs). Other similar assays
exist, such as the molecular inversion probe (MIP) method,
which relies on enzymatic extension/ligation to circularize
(and thus protect from degradation) a specially designed
probe upon binding to the appropriate target sequence.39
However, MLPA is the most frequently used in the clini-
cal setting. Common applications include testing for dele-
tions in cancer genes and Mendelian disease genes.40,41
Because the assay essentially provides pinpoint analysis, it
can be designed with closely spaced probes to detect small
deletions (e.g., one probe-set per exon of a gene), or with
probes more spread out to infer the presence of larger chro-
mosomal deletions/duplications. In this way it behaves like
a small microarray (described below).
N EW E R G E N O M I C A N ALYS I S
P LAT F O R MS
Over the last two decades, a host of newer genomic analy-
sis platforms has been developed that allows for the simul-
taneous, parallelized interrogation of millions or billions
of targets genome-wide. In many ways it is appropriate to
think of these as “digital” versions of older “analogue” analy-
ses: for example, in comparison to the visual data produced
by cytogenetics and FISH, microarray or next-generation
sequencing (NGS) technologies yield discrete results across
quantized locations, and can be thought of as having resolu-
tion and amplitude range much like a digital image. This
analogy is fitting because these systems produce so much
digital data that whole fields of expertise in programming
and computation have grown up around them, collectively
termed genome bioinformatics. With this technology now
migrating in wholesale fashion into clinical medicine, the
potential ramifications for patient treatment and outcomes,
care practices, and health records are enormous.
 (A)

(B)

Normal (overlapping) Mutant (separated)

Normal (overlapping)

Figure 10.5 
Interphase FISH. A) For rapid prenatal diagnosis of common aneuploidies, FISH can be performed on cells in interphase obtained at
amniocentesis. In this example, staining with a probe for the centromeric region of chromosome 21 is notable for three signals in every cell, which
is diagnostic for Down syndrome. Unlike conventional cytogenetic analysis, this technique only indicates the number of copies of the probe region,
which does not necessarily equate to the number of copies of whole chromosomes. B) Translocations and other rearrangements with specific
breakpoints can be detected using multicolor “break-apart” FISH. Shown are interphase cells from a 36-year-old patient with newly diagnosed acute
myeloid leukemia (AML). The cells are stained with two probes that target the CBFB gene: a red probe that binds just upstream (5′) of the gene,
and a green probe that binds just downstream (3′). In a normal cell (left), the probes produce essentially overlapping spots, which can appear orange.
In a cell harboring an inversion of chromosome 16 (right) which produces the CBFB-MYH11 fusion gene, one copy of the gene shows physical
separation of the probe spots, indicating a chromosomal breakage between the probe binding sites.(Courtesy of Dr. Peter Thompson, Cytogenetics
Laboratory, Institute of Medical Genetics, Cardiff, UK)(Courtesy of Dr. Gordana Raca, Cancer Cytogenetics Laboratory, Department of Medicine,
University of Chicago, Illinois)

C O M PA R AT I VE G E N O M I C CGH, the genomic DNA to be tested is labeled with a fluo-

H Y B R I D I Z AT I O N A R R AYS
Comparative genomic hybridization is what its name sug-
gests: instead of preparing chromosomes and staining them
with probes as in FISH, the genomic DNA itself is labeled
and hybridized onto a solid surface dotted with individual
probes to which it can hybridize.42,43 Each probe spot con-
tains DNA from a specific genomic locus, such as individ-
ual BAC clones, spaced across the genome. In traditional
rescent dye of one color, and a second “normal” reference
sample is labeled with a different color. Upon hybridization
to the array, the relative binding by the two different sam-
ples is assessed at each probe location. At a given locus, if the
test sample has more or less DNA than the reference (due
to either amplification or deletion), the discrepancy will be
revealed by a proportional color imbalance at that probe
location. The result is a genome-wide assessment of local
copy number, with the resolution determined by the probe

G e no m ic s T e c h no l ogy in C l inic a l D i agno s tic s   •   1 5 5

 1 2 3 4 5

6 7 8 9 10 11 12

13 14 15 16 17 18

19 20 21 22 X Y

Figure 10.6 
FISH probes covering entire chromosomes can be tagged with chromosome-specific fluorescent dye signatures and hybridized to
metaphase spreads to produce M-FISH or spectral karyotyping (SKY) images. The example shown is an analysis of the colon cancer cell line
SW480. This allows for the in-depth characterization of complex “marker” chromosomes, including one that contains material from chromosomes
3, 10, and 12 (seen in the chromosome 10 box). This type of determination would be impossible with conventional cytogenetic banding techniques.
(Courtesy of George Poulogiannis, Department of Pathology, University of Cambridge, England)

density (Figure 10.7). It should be noted that array-based
techniques such as CGH can only detect unbalanced chro-
mosomal rearrangements where material is lost or gained.
Balanced inversions and translocations are not detectable
because the total amount of DNA at each locus remains
unchanged.
S N P C Y TO G E N O M I C S A R R AYS
In recent years, BAC-based CGH arrays have largely been
replaced by newer array systems that use short oligonucle-
otide probes, which are easier to mass produce and can be
made with higher probe density. Array probe counts have
reached the millions, offering as low as 1kb resolution.
Additionally, many platforms are now based either entirely
or in part on probes that interrogate genomic SNPs.44 One
complicating factor when discussing more recent genomics
technologies is that, for each technology, platforms may be
available from a variety of different companies, each with
different chemistries and workflows designed to avoid intel-
lectual property entanglements. Such is the case with SNP
arrays as well as next-generation sequencing platforms (dis-
cussed below), among others. However, suffice it to say that
each of the major SNP array platforms can monitor copy
number at each probed locus based on binding affinity,
while establishing the zygosity of the SNP call at each loca-
tion. Thus, they offer the same type of locus-specific dosage
data as CGH arrays, but provide the added benefit of up
to millions of SNP genotyping calls genome-wide, which is
valuable for a number of different applications.
Such arrays have been the workhorses of genome-wide
association studies (GWAS), which are efforts to uncover
genetic underpinnings of complex, multifactorial human
diseases or phenotypes. Assaying for many SNPs across the
genome of many affected and normal individuals allows
investigators to statistically link the phenotype with par-
ticular SNP markers that may lie in close proximity to the
responsible genetic factor. This strategy takes advantage
of the concept of linkage disequilibrium (LD), whereby
closely adjacent genetic markers will tend to co-segregate
in families rather than being inherited independently, as
would distantly separated loci; for example, those on sepa-
rate chromosomes. These types of studies have been used to
identify SNPs that predispose to some amount of elevated
risk for a variety of conditions (macular degeneration,
heart disease, diabetes, etc.).45,46 Though there are disagree-
ments regarding the clinical utility of this information and
the applicability of the data across ethnic groups, certain

1 5 6   •   P rincip l e s o f G e no m ic M e dicin e
companies now offer clinical testing using these array plat- Technical Overview
forms, providing individualized risk assessments based on
One of the reasons NGS is so exciting is that it is the first
the results provided by GWA studies. We stress that these
technology with the prospect of detecting genetic varia-
results should be considered and interpreted cautiously, and
tion of every possible size scale. As with array platforms, a
only under the guidance of appropriate experts trained in
variety of NGS platforms exist that share many of the same
clinical genetics and genetic counseling.
important underlying features. As a group, these technolo-
For traditional diagnostic testing, however, these arrays
gies circumvent the signal-to-noise problem of genomic
are proving valuable as a means of surveying the whole
DNA (or any other heterogeneous DNA sample type) in
genome for deletions, duplications, and other anomalies
a fundamentally different way than PCR:  by performing
that are too small to be visible by cytogenetics (Figure
independent analyses of many individual DNA molecules
10.8). While the absence of signal from one individual
from a large pool in a massively parallel manner. This is
probe spot may represent noise, a cluster of absent signals
made possible by a critical first step whereby molecules are
can provide statistical confidence of a local genomic dele-
spatially separated from each other, essentially transform-
tion. Thus, depending on overall and local probe density,
ing a complex pool into a collection of isolated single mol-
these arrays can confidently call deletions approximately
ecules. The details of the separation are not critical for this
50kb or less, up to 100 times smaller than those visible
discussion, but (for example) may involve separating DNA
by cytogenetics. The SNP genotyping data can add confi-
molecules into individual microbubbles (emulsion PCR) or
dence to such copy number identification, because a het-
spreading fragments across a surface covered with a “lawn”
erozygous deletion will show all “homozygous” SNP calls
of complementary oligonucleotides. In order to produce
in that region. Additionally, SNP results can detect other
enough signal from each molecule, local PCR amplification
copy-neutral anomalies such as uniparental disomy (which
is performed, creating a population of local amplified clus-
shows a normal number of DNA copies but the absence
ters. Once the individual clonal groups are generated, enor-
of heterozygous SNP calls) or mosaicism (which also has
mous numbers of individual starting DNA molecules can
normal copy number but can show a variety of bizarre SNP
be sequenced simultaneously, with separate data produced
allelic ratio patterns).47 Today, these arrays are widely used
for each.52
in the clinical setting for children with dysmorphic fea-
Nearly any sample of DNA is compatible with NGS
tures and other developmental abnormalities, particularly
analysis. The DNA does not even have to be particularly
when cytogenetics results are normal and when other tar-
intact, because NGS systems require that input DNA
geted testing options (using FISH or PCR, e.g.) have been
take the form of small fragments, typically a few hundred
exhausted.48
base pairs or less. Typically, this is achieved by ultrasonic
fragmentation, or alternative methods based on mul-
N E X T- G E N E R AT I O N S EQ U E N C I N G tiplex PCR, etc. Fragments also typically require that
specific oligonucleotide adapter sequences be incorpo-
Perhaps the most consequential development in biol-
rated onto their ends to make them compatible with the
ogy since the introduction of PCR is the advent of
sequencer. This process is called library preparation, and
next-generation sequencing (NGS).49 NGS represents a
only sequencer-compatible libraries may be applied to the
truly transformational technological shift, because it offers
instruments. After focal amplification of individual input
for the first time the prospect of fast and inexpensive full
library DNA molecules, sequencing primers bind to the
genomic sequence analysis. The implications for research
adapter sequences, and sequencing proceeds inwards into
have been enormous, not just in clinical research but in
the cloned fragment. Some platforms produce only one
virtually every field of biomedicine and basic biology. As
sequence read (single-end sequencing), while others allow
of this writing, the full genome sequences of hundreds
for a second read from the opposite end of each fragment
of animal species have been completed (and much larger
(paired-end sequencing). The details of the sequencing
numbers of bacterial and viral genomes), each offering
reactions are also platform-specific.
new possibilities for discoveries within and across spe-
cies.50 The technique can be used to probe any aspect of
biology related to nucleic acids, from chromatin structure,
Sequencing Strategies
genetics, and epigenetics, to transcription, RNA process-
ing, and more, and can be applied to any species, whether Because of the flexibility of NGS, a myriad of different
well-characterized or novel.51 wet-lab library preparation techniques have been developed

G e no m ic s T e c h no l ogy in C l inic a l D i agno s tic s   •   1 5 7

 (A)

1 2 3 4 5

6 7 8 9 10 11 12

13 14 15 16 17 18

19 20 21 22 X Y

(B)
0
16
32
48
i) Chromosome 7 63
79
95
111
127
143
159
–2.00 –1.60 –1.20 –0.80 –0.40 0.00 0.40 0.80 1.20 1.60 2.00
Chromosome 7 Log2 Ratio Ch1/Ch2
01100200_top - 01100200_bottom.bsn - 14/06/2006

0
11
23
34
48
57
ii) Chromosome 13 68
80
91
103
114
–2.00 –1.60 –1.20 –0.80 –0.40 0.00 0.40 0.80 1.20 1.60 2.00
Chromosome 13 Log2 Ratio Ch1/Ch2

01100200_top - 01100200_bottom.bsn - 14/06/2006

Figure 10.7 
Low-density CGH array analysis of the two-year-old patient discussed in Figure 10.3. Patient DNA (dark) and control DNA (grey) are
labeled and hybridized together onto an array spotted with BAC clones. Chromosomes 7 and 13 are shown (the chromosomes involved in the
translocation). Most array positions show equivalent binding (normal). However, at the positions corresponding to the breakpoint regions of
both chromosomes are regions of reduced patient signal, indicating loss of chromosomal material. Thus, despite its appearance by conventional
cytogenetics, the translocation is not balanced. It can be seen from the result that approximately 0.2 Mb of chromosome 13 and 8.42 Mb of
chromosome 7 are lost, which is consistent with the patient’s presentation. (Courtesy of Sian Morgan, Cytogenetics Laboratory, Institute of Medical
Genetics, Cardiff, UK)
 Copy Number

Genotype

Normal (diploid) Duplication Deletion

Figure 10.8 
SNP array analysis of copy number and genotype. Newer cytogenomics array platforms may contain millions of individual probes, and
can assay the genome at a resolution above one probe per 1000 bp. Many commercial platforms are available, and many options exist with respect
to array composition. Unlike traditional CGH, most modern arrays do not rely on comparative hybridization against a control sample. Instead,
only the test sample is applied to the array, and the data are informatically compared against historical normal samples to derive copy number
and genotype data. Arrays can be used to capture genotype calls genome-wide for GWA studies, etc., or viewed in plots such as the one shown to
evaluate for chromosomal abnormalities. In this example, tumor DNA from a pediatric patient with neuroblastoma is analyzed. Only data from
chromosome 11 are shown. At the end of the p-arm (left), the copy number assessment is normal, as is the genotype pattern (three signal ratios
indicating AA, AB, and BB genotypes). Around the centromere (middle), the copy number is elevated, indicating duplication. The genotype pattern
is complicated by the addition of an extra copy of this chromosomal segment. On the right, a deletion of most of the q-arm can be seen. The copy
number plot shows decreased signal, and the associated genotype plot is consistent with reduced copy number, with each SNP positive for only the
A or B genotype. (Courtesy of Dr. Gordana Raca, Cancer Cytogenetics Laboratory, Department of Medicine, University of Chicago, Illinois)

to focus sequencing efforts on different aspects of genomic
biology. Among other applications, ingenious methods
have been developed to elucidate the complex domain
structure of unwound chromatin in interphase nuclei, to
reveal genome-wide patterns of transcription factor bind-
ing and epigenetic modifications, and to study the spectrum
of RNA/RNA binding protein interactions.53–57 However,
a full description of these techniques is beyond the scope
of this chapter. From the standpoint of clinical medicine,
today, most assay designs are geared towards identification
of patient-specific genetic variants, either constitutional or
somatic, via either a whole-genome or (most frequently) a
targeted sequencing approach. As with any other diagnos-
tic technique, careful planning is required to ensure that an
NGS assay will have the power to detect the full spectrum
of sizes and locations of desired genetic features.
Despite its bioinformatic complexity, whole genome
sequencing (WGS) is perhaps the simplest sequencing
application to perform, and the creation of a whole genome
library is actually a first step for a variety of applications. As
described earlier, after the basic steps of fragmentation and
adaptor ligation, genomic DNA fragments are essentially
ready for sequencing. The newest NGS platforms enable the
complete sequencing of an entire human genome in only
one day, at a cost of a few thousand dollars, a remarkable
advance reducing cost by several orders of magnitude over
just the past decade. WGS has been an invaluable research
tool and is beginning to have a clinical impact.58 As costs
continue to decrease, it should see expanded or even routine
clinical use.
However, because of the current expense associated
with WGS, various targeted sequencing approaches are
gaining traction in research and clinical laboratories. These
utilize a variety of wet-lab chemistry approaches to create
sequencing libraries enriched for DNA sequences of inter-
est. Targeted sequencing can be readily performed, either
by fishing out fragments of interest from a whole genome
library using capture hybridization, or by preparing targeted
sequencing libraries directly from genomic DNA via multi-
plex PCR or other methods (Figure 10.9). Library prepa-
ration strategies can be devised to address essentially any
question related to nucleic acid biology, giving laboratories
a great deal of freedom to design assays in order to meet the
needs of their clinician and patient populations.
One important targeted sequencing approach is whole-
exome sequencing (WES), in which capture baits are used
to select protein-coding sequence fragments from whole-
genome libraries.59 The resulting sequence data are heav-
ily weighted towards this exonic coding sequence, which
represents only about 2% of the genome, and data from
these loci can be attained at a fraction of the cost of
whole-genome sequencing. It has been estimated that as
much as 80% of Mendelian disease can be explained by
mutations in protein-coding DNA. This makes for a very
favorable cost–benefit calculation for WES and explains
why it is growing in popularity as a diagnostic choice for

G e no m ic s T e c h no l ogy in C l inic a l D i agno s tic s   •   1 5 9

patients with unexplained diseases of presumably genetic
origin.59,60,61
In preparing samples for NGS, libraries from multiple
samples can be mixed together and run in a highly multi-
plexed fashion on a single instrument run, if each individ-
ual library is appropriately “barcoded.” This is achieved by
using slightly different adaptor molecules for each patient
that contain individualized sequences. Upon ligation, every
DNA molecule from a patient is then tagged with its own
unique patient-specific sequence, and each DNA molecule
is then sequenced together with its barcode. After sequenc-
ing, each patient’s data can then be extracted from the
pooled data on the basis of these barcode sequences. This is a
tremendously valuable feature that is used frequently in the
clinical laboratory to provide targeted data on many patients
at once, thereby reducing per-patient sequencing costs.
Bioinformatics
The eventual output of NGS is a simple list of the sequences
(reads) of all of the clonal clusters that were produced from
the submitted sequencing library. Each base that is read is
also paired with a quality score, indicating the instrument’s
statistical confidence in that particular base assignment.
Paired-end reads, where two opposite ends of the same
DNA molecule are sequenced, share linked identifiers.
Despite the simplicity of the data type, analysis of NGS
data can be significantly challenging due in large part to the
sheer volume of sequences produced. Today’s most power-
ful sequencing instruments can process over a half a billion
individual clonal clusters in one day, yielding approximately
150 billion bases (gigabases or Gb) of data with paired-end
150 bp sequencing reads.
For most NGS applications, the data must first be put
into context in order to derive meaning. A single fragment
sequence by itself is not particularly informative: it can be
individually compared to the human reference sequence, but
any discrepancy could be the result of either a true genetic
variant or a false sequencing error. Conversely, a normal
read may reflect only one normal allele and reveal nothing
about a second variant allele. Therefore, almost universally,
the first step in NGS bioinformatics analysis is alignment, a
computational process by which all of the collected reads are
compared with the reference genome sequence and mapped
to the most likely correct position. Many software applica-
tions (both publicly available “freeware” and commercial
software) are available to perform this function, though it
should be noted that this process can only be readily per-
formed with an appropriate reference genome sequence to
which reads can be mapped.62,63 Once the data are aligned,
1 6 0   •   P rincip l e s o f G e no m ic M e dicin e
sequence variants can be identified by comparing the reads
at each base position against the reference genome sequence
(Figure 10.10). The greater the number of reads covering a
position (sequencing depth), the greater their statistical
power for detecting anomalies. Likewise, increased depth
also minimizes the likelihood that rare random errors will
be interpreted as variant sequences. For constitutional
whole-genome sequencing, an average of 30x depth is the
currently accepted standard. However, if expected variants
are rarer than, for example, a SNP at 50% allelic frequency,
greater depth is required to produce the same degree of con-
fidence. Suppose for example that only one out of 100 frag-
ments of DNA corresponding to codon 12 of the KRAS
gene was mutated in a particular tumor sample: with only
30 samplings, we would be unlikely to detect a single variant
molecule! This is a critical factor in cancer analysis, because
tumor heterogeneity and interference by normal cells can
very easily lead to low mutational allelic percentage in can-
cer specimens, requiring much higher read depth to attain
adequate sensitivity (Figure 10.10).64
Larger insertions/deletions and translocations require
special informatics approaches, because reads spanning
a breakage or translocation point may not “map” effec-
tively. Mapping is based on the degree of unique match-
ing between a read and the reference sequence. Alignment
software is forgiving to a point, but a read crossing a trans-
location point, for example, is likely to fail to map under
ordinary conditions. Care must be taken to assess and per-
haps reprocess the alignment with these variants in mind,
including reevaluation of sequencing depth and mapping
of reads and read pairs.65,66 An important weakness of cur-
rent NGS technologies is the production of only short
sequence reads. This precludes straightforward analyses
of long repetitive or duplicative sequences (such as CGG
repeats in FMR1 or genetic typing of the pseudogene-rich
HLA region), without the application of complex wet-lab
preparations designed to circumvent these problems.67
Copy number variations may also be identified by
NGS testing via analysis of coverage depth, producing data
analogous to that of a CGH array. For example, an ideal-
ized sample with a deleted chromosomal segment would be
expected to show roughly half the read depth at that locus if
the deletion were heterozygous, and absent coverage if the
deletion were homozygous.68 Taking this idea a step further,
counting reads on a per-chromosome basis can help reveal
the presence of extra or missing chromosomes. The most
striking examples of this concept in clinical testing today are
new non-invasive assays for the detection of fetal trisomies.
In pregnancy, approximately ten percent of cell-free DNA
may be derived from the fetus.69 A fetus with a trisomy will
 Genomic DNA Fragment/repair
Alternate Targeted
Library Prep

Targeted Library Adaptor Ligation

Target Capture
Whole Genome Library

NGS
Instrument

PCR Amplification

Targeted Library

Figure 10.9 
NGS library preparation. There are many ways to produce sequencer-compatible libraries from genomic DNA. The choice of method
is heavily dependent on the types and sizes of genomic features being investigated, as well as other factors such as cost, sequencing platform,
etc. Typically, genomic DNA sequencing proceeds along a few lines. For WGS and some targeted sequencing approaches, DNA is first sheared
by ultrasonic fragmentation into small pieces (<1kb). After enzymatic end-repair, adapter sequences are ligated to the ends to produce a
sequencer-ready whole-genome library. If this library is applied to the sequencer, the result will be WGS. If instead certain sequences of interest are
purged from this library prior to sequencing, the sequencing will be targeted (i.e., either exome or other targeted capture sequencing). Alternatively,
genomic DNA may be processed in a variety of different ways to produce targeted libraries. For example, targeted libraries can be produced via
multiplex PCR or restriction digest–based methods to produce short fragment libraries with appropriate barcoded adaptors. Every targeted method
can be easily customized to concentrate the sequencing effort on the genomic territory of interest.

contribute extra copies of that chromosome to the mother’s
blood, which can be detected simply by counting and ana-
lyzing the number of NGS reads that map to each individual
chromosome.70 This technique is already having an impact
in prenatal care, helping to detect these anomalies with
high sensitivity and specificity, thus reducing unnecessary
amniocentesis and chorionic villus sampling procedures.71
It should be noted that RNA is also quite amenable to
NGS analysis following reverse transcription into cDNA.
RNA-seq, as it is termed, can reveal which genes are actively
transcribed in a sample, providing important contributory
data to other genomics studies.72 Gene expression levels can
also be compared between samples via analysis of depth
of coverage, and evaluation of read mapping can uncover
information about the expression of specific transcripts
and alternative mRNA splicing. Additionally, RNA analy-
sis can be used to uncover particular cancer translocations
that produce fusion genes.73 Such gene fusion events typi-
cally take place via breakpoints within intronic sequences.
As introns are frequently orders of magnitude larger than
exons, hunting for fusion breakpoints using genomic DNA
requires covering large areas. Breakpoint introns also often
contain repetitive sequences that can create mapping prob-
lems. The process of transcription and mRNA splicing
removes intronic sequences, making the mature mRNA
a more straightforward target for analysis. Traditional
PCR-based translocation assays also routinely utilize RNA
for the same reason.
Clinical Integration of NGS
The computational processing associated with NGS anal-
ysis can be quite demanding. Processing the data from
just one instrument involves the production of multiple

G e no m ic s T e c h no l ogy in C l inic a l D i agno s tic s   •   1 6 1

 100
90
80
Detection sensitivity (%)
70
60
50
40
30
20
10
0 text of well patients.
0 100 200 300 400
Sample median exon coverage
Figure 10.10 
Increased NGS read depth is needed to call low allelic
percentage mutations with high confidence. In cancer specimens, which
often show significant sub-clonality and admixture with normal cells,
clinically relevant mutations can be present at very low mutant allelic
frequencies (MAF). This may be true in other clinical scenarios as
well (e.g., mitochondrial disease). In the cancer setting, it is clinically
desirable to be able to reliably detect mutations at 5% MAF. Here,
samples with MAF between 5–10% required at least 250x read depth
for high-sensitivity detection (compare with 30x read depth, which
is the gold standard for typical inherited genetics). Below 5% MAF,
extremely high read depths are required. Specificity suffers as well at
very low MAF, because it becomes difficult to discriminate very low
percentage mutations from very low percentage sequencing errors.
(Courtesy of Foundation Medicine, Cambridge, Massachusetts)80
intermediary files potentially equaling hundreds of giga-
bytes, and can take days to process on even some of the
fastest computers. The era of big data has arrived, and
we face large hurdles in the coming years as this technol-
ogy continues to move into the clinic.74,75 For each labo-
ratory to install and maintain sufficient computational
infrastructure to support NGS applications would be
prohibitively expensive, and not enough bioinformat-
ics expertise currently exists to allow every laboratory to
participate. For this reason, many groups are developing
cloud-computing resources to support NGS operations,
with the hope that economies of scale and resource shar-
ing (informatics pipelines, etc.) will allow smaller clinics
and laboratories to utilize NGS testing in the care of their
patients.76
Clinical interpretation of NGS tests is fundamen-
tally no different from that of more traditional assays. As
always, proper controls, documentation, and adherence to
strict workflows are required in order to ensure high con-
fidence in the primary data before it may be interpreted
in the context of the clinical scenario. For findings that
are well documented and understood within the clinical
context, interpretation is quite straightforward. For exam-
ple, a truncating mutation in the dystrophin gene is eas-
ily interpretable in the context of a patient with findings
characteristic of Duchenne muscular dystrophy (DMD).
1 6 2   •   P rincip l e s o f G e no m ic M e dicin e
However, interpretation can be made more challenging by
a few factors, alone or in combination, including:
• Atypical clinical scenario: A mutation of high expected
MAF > 10%
severity may be difficult to interpret if the patient’s pre-
sentation is atypical for the disease. It should be noted
MAF 5–10%
that “wellness” is an atypical clinical scenario for every
MAF < 5
known disease. This raises the substantial problem of
how to interpret and act upon genomic data in the con-
500 600 700
• Indeterminate mutational effects: Even in a patient with
a classical presentation, mutations with unknown effects
on protein function may be difficult to interpret, even if
they are on the “correct” gene. This is particularly true of
substitution mutations. Many effects-prediction software
algorithms are available to help predict protein-function
impact, but this is an imperfect science.
• Indeterminate gene effects: Similarly, interpretation of
mutations in genes with little or no known association
with the presumed disease process may be difficult or
impossible to interpret as clinically pathogenic, even
if they are predicted to severely affect the gene. Often
this requires perusal of the scientific literature to help
postulate mechanistic links, though this is most fre-
quently insufficient to produce a clear determination of
pathogenicity.
In some ways, the application of NGS to clinical medi-
cine is a double-edged sword, because as a direct by-product
of its power and scope, it raises the likelihood of all of the
above issues and thus the chance of producing indetermi-
nate results. Fortunately, many tools are emerging to help
us sort through the data. Many public and private databases
are available that contain disease–gene associations and pre-
viously documented disease mutations, both for constitu-
tional genetic disease (e.g., Online Mendelian Inheritance
in Man, ClinVar, Human Gene Mutation Database, etc.)
and cancer (The Cancer Genome Atlas, Catalogue of
Somatic Mutations in Cancer, etc.). Similarly, to help avoid
confusion between common inherited SNPs and potential
disease mutations, other critical sources provide compen-
diums of both common and rare inherited variants (e.g.,
Exome Variant Server, The 1000 Genomes Project, dbSNP,
HapMap Project, etc.).
While these resources are tremendously valuable, in
practice there is a frequent need for additional informa-
tion, which is often unavailable. When presented with a
variant of uncertain significance (VUS), whether somatic
or constitutional, the only ways to develop confidence in
its clinical importance are either to document its biological
effect in a laboratory, or to produce corroborating evidence
from another patient or affected family. What is currently
missing in our health systems is a way to share this type of
information broadly between laboratories to help identify
similar cases. With the appropriate exchange of knowledge,
the same VUS identified at two different laboratories could
instead become two successful diagnoses. Today, a num-
ber of groups are working to create programs and systems
to enable such sharing, and it is hoped that cloud comput-
ing and other communal resources can help support such
endeavors.75,77 Clearly, there are important privacy concerns
surrounding this type of data, and different rules govern-
ing such communications in each country, and any reforms
or agreements must be made in a careful and responsible
fashion.78,79
CURRENT AND FUTURE
D I R E C T I O N S I N C L I N I C AL
GENOMICS
Due to ongoing technological upheaval, the field of clini-
cal genomics is in an exciting state of flux. In many cases,
there is still minimal consensus regarding what is the most
appropriate way to apply this new technology, particularly
NGS, to individualized patient care. Technical capabili-
ties are changing so quickly that best practices are in many
ways a moving target. One clear trend is the current rapid
replacement of traditional Sanger-based gene-sequencing
tests with NGS assays. Full gene-sequencing by Sanger, par-
ticularly for large genes implicated in constitutional genetic
disease, is a laborious process requiring separate PCRs and
sequencing reactions for each exon for each patient. In con-
trast, NGS enables the rapid and inexpensive sequencing
of panels of many genes, increasing the likelihood of suc-
cessful diagnoses and helping prevent diagnostic odysseys.
Similarly, nearly any assay relying on PCR to uncover small
anomalies can be performed better and for less cost via
NGS, though some smaller mainstay assays continue to rely
on PCR.
In cancer diagnostics laboratories, most traditional
assays for detecting small-scale mutational events (includ-
ing both oncogene and tumor-suppressor mutations) are
now being rapidly retired in favor of NGS profiling assays.
In many academic and commercial laboratories, routine
targeted examination of tens to hundreds of different
cancer-related genes is now routine, with the goal of pro-
viding individualized treatment recommendations based
on each patient’s mutational spectrum.80 Other clinical
G e no m ic s T e c h no l ogy in C l inic a l D i agno s tic s   •   1 6 3
laboratories are adopting the strategy of even wider analysis,
performing exome and even genome sequencing in order
to uncover treatable targets. Some are convinced that this
should be the ultimate strategic goal for all cancer patients,
while others believe that data from whole genome cancer
sequencing studies should be distilled so that clinical diag-
nostics can focus on the relevant (recurrently mutated) tar-
gets for each tumor.
With respect to large anomalies, particularly in can-
cer, certain types of variation are more amenable to NGS
analysis than others. For example, NGS is well suited for
targeted translocation detection because the identification
of any cancer-specific fusion sequence is diagnostic. In con-
trast, copy number analyses are more difficult to perform in
a superior manner via NGS or SNP arrays compared with
FISH or cytogenetics. This is because FISH and cytoge-
netics operate on a single-cell basis, providing data from
each individual cell analyzed. Thus, FISH analysis for Her2
amplification in a breast cancer biopsy would be able to
detect significant amplification in a small number of tumor
cells, even if the biopsy was heavily contaminated with nor-
mal cells. In contrast, NGS and SNP arrays only provide
an average result of all sampled cells. In order to match
the sensitivity of traditional methods for heterogeneous
samples (particularly when tumor cell proportion is low),
single-cell analysis or specific microdissection/sorting may
be required. Single-cell sequencing is currently performed
in the research setting, but as of today it is still too expensive
and error-prone to be used as a clinical diagnostic. Thus, for
the foreseeable future, FISH and cytogenetics will continue
to retain an important role in the work-up of certain tumors.
Looking to the future, there is vast potential for novel
and expanded uses of NGS technology in clinical medi-
cine. While clinical laboratories are beginning to get a
handle on DNA sequence analysis, many of the sequencing
approaches that are routinely used in research laboratories
are still uncharted territory for diagnostics. For example,
there are essentially no clinically certified assays performed
today that are based on RNA-seq. In clinical laboratories,
RNA is typically analyzed either to investigate expression
or to look for clinically relevant fusion (translocation)
transcripts. With respect to gene expression, most clinical
laboratories’ needs are met by oligonucleotide microarrays,
which have become quite standardized and inexpensive.81
In contrast, gene-expression analysis by RNA-seq is still
regarded as informatically complex. Therefore, it is most
likely that RNA-seq will first find widespread adoption in
clinical laboratories as a tool for translocation detection,
as oligonucleotide arrays are incapable of detecting these
anomalies.
 As discussed, many other wet-lab library preparation
methods exist to interrogate various aspects of genome biol-
ogy, but technical challenges, shortages of expertise, and a
lack of key clinical questions serve to slow the adoption of
these methods into clinical laboratories. Examples include
methylation sequencing or any of the methods revolving
around protein pull-down to investigate protein–nucleic
acid interactions.82,83
In addition to allowing more powerful analyses of
traditional specimens, the freedom provided by NGS to
explore essentially any sample type is opening up entirely
new ways of thinking about health monitoring and diag-
nostics. Analysis of non-human genomes, such as those of
parasites, bacteria, and viruses, is expanding, and recent dis-
coveries relating health and the microbiome are beginning
to change our understanding of self and wellness.84 For can-
cer and many other diseases, NGS and other technologies
offer much potential for early screening and detection via
assays to detect scant nucleic acid signatures in blood and
other tissues (e.g., circulating tumor DNA). Maternal blood
trisomy testing has opened the doors to many new possibili-
ties for prenatal diagnostics. It is now possible to sequence
the entire genome of a fetus from maternal plasma, raising
both hope for this technology as well as potential ethical
concerns.85
Many NGS applications, particularly WGS, are still
quite expensive. However, the cost of sequencing has fallen
precipitously over the last decade, and this trend seems
likely to continue. If it does, applications like WGS may
reach a point at which they are routinely affordable in the
clinic. When that happens, it will place great pressure on
our electronic health systems, as well as on the pathologists
and geneticists who will face the task of interpreting all of
the data. This may be compounded by the emergence of
even more powerful technologies that may supplant NGS
in the coming years. Single-molecule (or third-generation)
sequencing systems are currently available that can operate
directly on individual DNA molecules without requiring
on-instrument amplification. As a result, they are not lim-
ited to short read lengths and can process DNA in a fraction
of the time compared with NGS instruments.86,87 While
they have not yet had an impact on the clinical diagnostics
landscape, they offer an intriguing view of what awaits us
over the horizon.
R EFE R E N C ES
1. Lander ES, et  al. Initial sequencing and analysis of the human
genome. Nature. 2001;409(6822):860–921.
1 6 4   •   P rincip l e s o f G e no m ic M e dicin e
2. Venter JC, et  al. The sequence of the human genome. Science.
2001;291(5507):1304–1351.
3. Genomes Project Consortium, et al. A map of human genome varia-
tion from population-scale sequencing. Nature. 2010;467(7319):
1061–1073.
4. Consortium EP, et al. An integrated encyclopedia of DNA elements
in the human genome. Nature. 2012;489(7414):57–74.
5. Cancer Genome Atlas Network. Comprehensive molecular por-
traits of human breast tumours. Nature. 2012;490(7418):61–70.
6. Cancer Genome Atlas Research Network Comprehensive genomic
characterization defines human glioblastoma genes and core path-
ways. Nature. 2008;455(7216):1061–1068.
7. Cancer Genome Atlas Research Network. Integrated genomic anal-
yses of ovarian carcinoma. Nature. 2011;474(7353):609–615.
8. Cancer Genome Atlas Research Network. Comprehensive molec-
ular characterization of clear cell renal cell carcinoma. Nature.
2013;499(7456):43–49.
9. Atkin NB, et  al. The superfemale mole. Lancet. 1962;280(7258):
727–728.
10. Saiki RK, et  al. Enzymatic amplification of beta-globin genomic
sequences and restriction site analysis for diagnosis of sickle cell ane-
mia. Science. 1985;230(4732):1350–1354.
11. Cheng S, et al. Effective amplification of long targets from cloned
inserts and human genomic DNA. Proc Natl Acad Sci U S A.
1994;91(12):5695–5699.
12. Tjio JH, et al. The chromosome number in man. Heriditas. 1956;
42:1–6.
13. Craig JM, et al. Genes and genomes: Chromosome bands—flavours
to savour. Bioessays. 1993;15:349–354.
14. Drets ME, et al. Specific banding patterns of human chromosomes.
Proc Natl Acad Sci U S A. 1971;68(9):2073–2077.
15. Shaffer LG, McGowan-Jordan J, Schmid M. An International
System for Human Cytogenetic Nomenclature. Recommendations
of the International Standing Committee on Human Cytogenetic
Nomenclature. Published in collaboration with ‘Cytogenetic and
Genome Research’. Plus fold-out: ‘The Normal Human Karyotype G-
and R-bands’. 2013. http://www.karger.com/Book/Home/257302
16. Rowley JD, et al. Relationship of centromeric heterochromatin to
fluorescent banding patterns of metaphase chromosomes in the
mouse. Nature. 1971;231(5304):503–506.
17. Blakemore KJ, et  al. A method of processing first-trimester cho-
rionic villous biopsies for cytogenetic analysis. Am J Hum Genet.
1984;36(6):1386–1393.
18. Ferguson-Smith MA. Cytogenetics and the evolution of medical
genetics. Genet Med. 2008;10(8):553–559.
19. Nowell PC, et  al. Chromosome studies on normal and leukemic
human leukocytes. J Natl Cancer Inst. 1960;25:85–109.
20. Rowley JD. A new consistent chromosomal abnormality in chronic
myelogenous leukaemia identified by quinacrine fluorescence and
Giemsa staining. Nature. 1973;243(5405):290–293.
21. Vardiman JW, et  al. The 2008 revision of the World Health
Organization (WHO) classification of myeloid neoplasms and acute
leukemia: rationale and important changes. Blood. 2009;114(5):
937–951.
22. Rowley JD, et  al. 15/17 translocation, a consistent chromosomal
change in acute promyelocytic leukaemia. Lancet. 1977;1(8010):
549–550.
23. Sakurai M, et al. 8–21 translocation and missing sex chromosomes
in acute leukaemia. Lancet. 1974;2(7874):227–228.
24. List A, et  al. Lenalidomide in the myelodysplastic syndrome
with chromosome 5q deletion. N Engl J Med. 2006;355(14):
1456–1465.
25. Pinkel D, et  al. Cytogenetic analysis using quantitative,
high-sensitivity, fluorescence hybridization. Proc Natl Acad Sci U S
A. 1986;83(9):2934–2938.
26. Trask BJ. Fluorescence in situ hybridization: applications in cytoge-
netics and gene mapping. Trends Genet. 1991;7(5):149–154.
 27. van Ommen GJ, et al. FISH in genome research and molecular diag-
nostics. Curr Opin Genet Dev. 1995;5(3):304–308.
28. Osoegawa K, et  al. A bacterial artificial chromosome library
for sequencing the complete human genome. Genome Res.
2001;11(3):483–496.
29. Trask BJ, et al. Fluorescence in situ hybridization to interphase cell
nuclei in suspension allows flow cytometric analysis of chromosome
content and microscopic analysis of nuclear organization. Hum
Genet. 1988;78(3):251–259.
30. Miller WH Jr., et al. Reverse transcription polymerase chain reac-
tion for the rearranged retinoic acid receptor alpha clarifies diag-
nosis and detects minimal residual disease in acute promyelocytic
leukemia. Proc Natl Acad Sci U S A. 1992;89(7):2694–2698.
31. Wolff AC, et  al. Recommendations for human epidermal growth
factor receptor 2 testing in breast cancer:  American Society of
Clinical Oncology/College of American Pathologists clinical prac-
tice guideline update. J Clin Onc. 2013;31(31):3997–4013.
32. Schrock E, et al. Multicolor spectral karyotyping of human chromo-
somes. Science. 1996;273(5274):494–497.
33. Schrock E, et al. Spectral karyotyping refines cytogenetic diagnos-
tics of constitutional chromosomal abnormalities. Hum Genet.
1997;101(3):255–262.
34. Karpova MB, et  al. Combined spectral karyotyping, compara-
tive genomic hybridization, and in vitro apoptyping of a panel
of Burkitt’s lymphoma-derived B cell lines reveals an unexpected
complexity of chromosomal aberrations and a recurrence of
specific abnormalities in chemoresistant cell lines. Int J Oncol.
2006;28(3):605–617.
35. Veldman T, et al. Hidden chromosome abnormalities in haemato-
logical malignancies detected by multicolour spectral karyotyping.
Nat Genet. 1997;15(4):406–410.
36. Southern EM. Detection of specific sequences among DNA frag-
ments separated by gel electrophoresis. J Mol Biol. 1975;98(3):
503–517.
37. Monaghan KG, et  al. ACMG standards and guidelines for frag-
ile X testing:  a revision to the disease-specific supplements to the
standards and guidelines for clinical genetics laboratories of the
American College of Medical Genetics and Genomics. Genet Med.
2013;15(7):575–586.
38. Schouten JP, et  al. Relative quantification of 40 nucleic acid
sequences by multiplex ligation-dependent probe amplification.
Nucleic Acids Res. 2002;30(12):e57.
39. Hardenbol P, et  al. Multiplexed genotyping with sequence-tagged
molecular inversion probes. Nat Biotechnol. 2003;21(6):673–678.
40. Willis AS, et  al. Multiplex ligation-dependent probe ampli-
fication (MLPA) and prenatal diagnosis. Prenat Diagn.
2012;32(4):315–320.
41. Hömig-Hölzel C, et al. Multiplex ligation-dependent probe ampli-
fication (MLPA) in tumor diagnostics and prognostics. Diagn Mol
Pathol. 2012;21(4):189–206.
42. Pinkel D, et al. High resolution analysis of DNA copy number varia-
tion using comparative genomic hybridization to microarrays. Nat
Genet. 1998;20(2):207–211.
43. Solinas-Toldo S, et al. Matrix-based comparative genomic hybridiza-
tion: biochips to screen for genomic imbalances. Genes Chromosomes
Cancer. 1997;20(4):399–407.
44. Sapolsky RJ, et al. High-throughput polymorphism screening and
genotyping with high-density oligonucleotide arrays. Genet Anal.
1999;14(5–6):187–192.
45. Klein RJ, et al. Complement factor H polymorphism in age-related
macular degeneration. Science. 2005;308(5720):385–389.
46. The Wellcome Trust Case Control Consortium. Genome-wide
association study of 14,000 cases of seven common diseases and
3,000 shared controls. Nature. 2007;447(7145):661–678.
47. Conlin LK, et al. Mechanisms of mosaicism, chimerism and unipa-
rental disomy identified by single nucleotide polymorphism array
analysis. Hum Mol Genet. 2010;19(7):1263–1275.
G e no m ic s T e c h no l ogy in C l inic a l D i agno s tic s   •   1 6 5
48. Miller DT, et  al. Consensus statement:  chromosomal microar-
ray is a first-tier clinical diagnostic test for individuals with devel-
opmental disabilities or congenital anomalies. Am J Hum Genet.
2010;86(5):749–764.
49. Brenner S, et al. Gene expression analysis by massively parallel sig-
nature sequencing (MPSS) on microbead arrays. Nature Biotech.
2000;18(6):630–634.
50. Alföldi J, et al. Comparative genomics as a tool to understand evolu-
tion and disease. Genome Res. 2013;23(7):1063–1068.
51. Koboldt DC, et al. The next-generation sequencing revolution and
its impact on genomics. Cell. 2013;155(1):27–38.
52. Shendure J, et al. Next-generation DNA sequencing. Nat Biotechnol.
2008;26(10):1135–1145.
53. Dixon JR, et al. Topological domains in mammalian genomes identi-
fied by analysis of chromatin interactions. Nature. 2012;485(7398):
376–380.
54. ENCODE Project Consortium. An integrated encyclopedia of DNA
elements in the human genome. Nature. 2012;489(7414):57–74.
55. Lister R, et al. Human DNA methylomes at base resolution show wide-
spread epigenomic differences. Nature. 2009;462(7271): 315–322.
56. Robertson G, et al. Genome-wide profiles of STAT1 DNA associa-
tion using chromatin immunoprecipitation and massively parallel
sequencing. Nat Methods. 2007;4(8):651–657.
57. Licatalosi DD, et  al. HITS-CLIP yields genome-wide insights into
brain alternative RNA processing. Nature. 2008;456(7221): 464–469.
58. Bainbridge NM, et  al. Whole-genome sequencing for optimized
patient management. Sci Transl Med. 2011;3(87): 87re3.
59. Ng SB, et al. Targeted capture and massively parallel sequencing of
12 human exomes. Nature. 2009;461(7261):272–276.
60. Worthey EA, et al. Making a definitive diagnosis: successful clinical
application of whole exome sequencing in a child with intractable
inflammatory bowel disease. Genet Med. 2011;13(3):255–262.
61. Bamshad MJ, et al. Exome sequencing as a tool for Mendelian dis-
ease gene discovery. Nat Rev Genet. 2011;12(11):745–755.
62. Langmead B, et  al. Ultrafast and memory-efficient alignment
of short DNA sequences to the human genome. Genome Biol.
2009;10(3):R25.
63. Li H, et al. Fast and accurate short read alignment with Burrows–
Wheeler transform. Bioinformatics. 2009;25(14):1754–1760.
64. Ulahannan D, et al. Technical and implementation issues in using
next-generation sequencing of cancers in clinical practice. Br J
Cancer. 2013;109(4):827–835.
65. Ruibin X, et  al. Detecting structural variations in the human
genome using next generation sequencing. Brief Funct Genomics.
2010;9(5–6):405–415.
66. Jiang Y, et  al. PRISM:  pair-read informed split-read mapping for
base-pair level detection of insertion, deletion and structural vari-
ants. Bioinformatics. 2012;28(20):2576–2583.
67. Wang C, et al. High-throughput, high-fidelity HLA genotyping with
deep sequencing. Proc Natl Acad Sci U S A. 2012;109(22): 8676–8681.
68. Baslan T, et al. Genome-wide copy number analysis of single cells.
Nature Protocols. 2012;7(6):1024–1041.
69. Lo YM, et al. Presence of fetal DNA in maternal plasma and serum.
Lancet. 1997;350(9076):485–487.
70. Chiu RW, et  al. Noninvasive prenatal diagnosis of fetal chromo-
somal aneuploidy by massively parallel genomic sequencing of
DNA in maternal plasma. Proc Natl Acad Sci U S A. 2008;105(51):
20458–20463.
71. Mersy E, et al. Noninvasive detection of fetal trisomy 21: systematic
review and report of quality and outcomes of diagnostic accuracy
studies performed between 1997 and 2012. Hum Reprod Update.
2013;19(4):318–329.
72. Jongeneel CV, et  al. An atlas of human gene expression from
massively parallel signature sequencing (MPSS). Genome Res.
2005;15(7):1007–1014.
73. Edgren H, et al. Identification of fusion genes in breast cancer by
paired-end RNA-sequencing. Genome Biol. 2011;12(1):R6.
 74. Tucker T, et al. Massively parallel sequencing: the next big thing in
genetic medicine. Am J Hum Genet. 2009;85(2):142–154.
75. Grossman RL, et al. A vision for a biomedical cloud. J Intern Med.
2012;271(2):122–130.
76. Schatz MC, et al. Cloud computing and the DNA data race. Nat
Biotechnol. 2010;28(7):691–693.
77. Baker M. One-stop shop for disease genes. Nature. 2012;
491(7423):171.
78. Lucassen A, et  al. Consent and confidentiality in clinical genetic
practice: guidance on genetic testing and sharing genetic informa-
tion. Clin Med. 2012;12(1):5–6.
79. Creating a Global Alliance to Enable Responsible Sharing of
Genomic and Clinical Data- the Global Genome Alliance. https://
www.broadinstitute.org/files/news/pdfs/GAWhitePaper June3.pdf
80. Frampton GM, et al. Development and validation of a clinical can-
cer genomic profiling test based on massively parallel DNA sequenc-
ing. Nat Biotechnol. 2013;31(11):1023–1031.
1 6 6   •   P rincip l e s o f G e no m ic M e dicin e
81. Arpino G, et al. Gene expression profiling in breast cancer: a clinical
perspective. Breast. 2013;22(2):109–120.
82. Laird PW. Principles and challenges of genome-wide DNA meth-
ylation analysis. Nat Rev Genet. 2010;11(3):191–203.
83. Park PJ. ChIP-seq: advantages and challenges of a maturing technol-
ogy. Nat Rev Genet. 2009;10(10):669–680.
84. The Human Microbiome Project Consortium. Structure, func-
tion and diversity of the healthy human microbiome. Nature.
2012;486(7402):207–214.
85. Fan HC. Non-invasive prenatal measurement of the fetal genome.
Nature. 2012;487(7407):320–324.
86. Eid J. Real-time DNA sequencing from single polymerase mol-
ecules. Science. 2009;323(5910):133–138.
87. Howorka S, et  al. Sequence-specific detection of individual DNA
strands using engineered nanopores. Nat Biotechnol. 2001;19(7):
636–639.
 11.
MICROBIAL GENOMICS: TARGETED ANTIMICROBIAL
THERAPY AND GENOME VACCINES
Immaculada Margarit and Rino Rappuoli
INTRODUCTION
Since its pioneering introduction in 1796 by Edward Jenner,
vaccination has been revealed as the most effective medi-
cal intervention for the prevention of human infections,
greatly contributing to increased life expectancy(1). Formal
vaccine development started only one century later, when
it became clear that infections were caused by microbes,
and Louis Pasteur proposed to “isolate, inactivate and inject
the microorganism.” This practice established the basis for
further key interventions by Jonas Salk and Albert Sabin
leading to the eradication of poliovirus infections, and by
Maurice Hilleman, who developed vaccines against measles,
mumps, and rubella(2). At that time, vaccination approaches
were mainly based on the use of crude inactivated or attenu-
ated whole microorganisms.
In the first half of the twentieth century, Glenny,
Ramon, Pappenheimer, and others pioneered the isolation
and partial purification of bacterial or viral culture compo-
nents, paving the way for the development of subunit vac-
cines like those against diphtheria, tetanus, and influenza.
Fifty years later, the vaccine field greatly benefited from the
introduction of new technologies such as antigen produc-
tion by recombinant DNA approaches, chemical conjuga-
tion of proteins to polysaccharide antigens, and the use of
novel adjuvants. New vaccines against important patho-
gens like Neisseria meningitidis, Streptococcus pneumoniae,
Haemophilus influenza, and more effective adjuvanted vac-
cines were developed. In less than a century, vaccines based
on Pasteur’s principle had an enormous impact on global
public health by globally eliminating some of the most dev-
astating infectious diseases(3). Table 11.1 lists the vaccines
licensed to date. However, it also became apparent that
new technologies were required to defeat a large number
of diseases still causing high morbidity to mankind, includ-
ing tuberculosis, malaria, HIV, hepatitis C, Group A and
B streptococcal infections, emerging diseases like pandemic
167
influenza, as well as a growing list of infections caused by
antibiotic-resistant bacteria. In 1995, the genomics revolu-
tion was about to start with the completion of the genome
sequence of Haemophilus influenzae(4), opening a new era in
vaccine development.
D I S C O VE RY A N D D E VE L O PM E N T
O F G E N O M I C VAC C I N E S
The recent advances in microbial and human genomics have
greatly accelerated the development of novel tools for the
diagnosis, monitoring, prevention, and treatment of human
infectious diseases (Figure 11.1). During the twentieth cen-
tury, the invention and development of vaccines have been
the major goals of microbiology and preventive medicine,
encompassing the whole field of vaccinology. While histori-
cally most vaccines were developed using the target micro-
organism itself, the current focus is utilizing the microbial
genomes for developing the targeted vaccine by an approach
called “reverse vaccinology.”
T H E FI R S T G E N O M E -BA S E D
VAC C I N E : M E N I N G O C O C C US B
Once the full set of genes of a bacterial pathogen could
be made available, the possibility emerged to identify
vaccine targets by computer-facilitated predictions of
antigen surface exposure and immunogenicity, without
the need of cultivating the pathogen. The genes encoding
potentially suitable vaccine targets could be expressed and
purified by high-throughput methods in non-pathogenic
hosts, and tested in preclinical models for their immuno-
genicity and their ability to neutralize the original infec-
tious agent.
The first successful application of this novel reverse
vaccinology strategy (Figure 11.2) came from the work by
Table 11.1  LICENSED VACCINES FOR HUMAN USE

TARGET DISEASE CAUSING AGENT TYPE OF VACCINE COMMENTS

Anthrax Bacillus anthracis Subunit cell-free extract of the bacteria, The only anthrax vaccine for human use.
adsorbed on alum
Diphtheria Corynebacterium diphtheriae Subunit toxoid, alum adsorbed Used mainly in combination with tetanus toxoid and acellular pertussis
­(diphtheria, pertussis, tetanus: DPT).
Cervical cancer Human papilloma virus (HPV) Recombinant DNA vaccine Quadrivalent vaccine of several viral strains.
Cholera Vibrio Cholera 1. Whole organism inactivated The B subunit of cholera toxin serves also as the basis for a vaccine against toxic
2. Subunit—B subunit of cholera toxin E. coli.
Hib-induced disease, Haeomophilus influenzae type B Conjugate vaccine, the Hib ­polyribosylribitol Two proteins are used in the licensed conjugate vaccines—meningococcal
pneumonia meningitis phosphate (PRP) capsular polysaccharide outer membrane protein (OMP) or tetanus toxoid (TT), produced by different
­conjugated to a protein manufacturers.
Hepatitis type A Hepatitis A virus 1. Whole virus inactivated Formalin-inactivated virus of several strains.
2. Whole virus live attenuated Prepared by serial passage in cell culture.
Hepatitis type B Hepatitis B virus 1. Subunit surface antigen (HBsAg) Approved but used only in Third World countries.
blood-derived The first recombinant vaccine used also in combination with inactivated hepatitis
2. Recombinant yeast-derived HBsAg A vaccine.
Influenza Influenza virus 1. Whole virus inactivated trivalent type A and Seasonal vaccine—the three strains H1N1, H3N2, and B are determined each
type B vaccine year by the World Health Organization (WHO).
2. Subunit trivalent vaccine Seasonal vaccine, trivalent as above, comprising a complex of HA and NA.
3. Whole virus H5N1 monovalent inactivated Prepared only in the USA (for national stockpile).
4. Whole virus H1N1 2009 monovalent Prepared against pandemic flu of 2009, by many vaccine producers, with or
5. Live attenuated nasal vaccine FluMist ­without adjuvant.
6. Subunit proteosome (OMP)—Conjugate Trivalent, for intranasal use.
FluInsure Trivalent, for intranasal use.
Japanese encephalitis Japanese encephalitis virus 1. Whole virus inactivated vaccine, alum Vaccine produced in Japan, and licensed worldwide, including the USA.
( JEV) adsorbed This vaccine is produced and was tested only in China and is not licensed in the
2. Live attenuated USA or Europe.
Measles Measles virus Live attenuated strains Schwarz or Used alone or in combination with: Mumps virus vaccine, with mumps and
E2-19 of Measles with mumps and rubella rubella (MMR), or with mumps, rubella, and varicella live virus vaccines.
(MMR)
Meningitis Neisseria meningitidis 1. Subunit capsular polysaccharide (CP) The vaccines consist of different combinations of purified high-molecular weight
2. Conjugate vaccine—CP with Diphteria CP from serogroups A, C, Y, and W-135 combined.
Toxin (DT) The above subunit CP vaccine conjugated to diphtheria toxoid.
3. Subunit outer membrane protein (OMP) The vaccine containing group B/C OMP, produced and licensed only in Cuba.
4. Subunit, protein based Recently approved, 1st genome reverse vaccinology vaccine.
Mumps Mumps virus Live attenuated strains Jeryl Lynn or Rubini This vaccine is rarely used as such, but only in combination with measles, mumps,
and rubella (MMR).
Plague Yersinia pestis Whole bacteria inactivated Very rarely used since disease in humans was practically eradicated.
Pneumonia Streptococcus pneumoniae 1. Subunit M protein-based Polyvalent vaccine. The repeat sequence of the pneumococcal M protein of
2. Conjugate vaccine—cell wall ­several serotypes.
polysaccharides conjugated to a protein Polysaccharides of several strains 7—valent conjugated to either diphtheria
(CRM 197 protein) or tetanus toxoid.
Polio Polio virus 1. Whole virus inactivated (Salk vaccine) Known as oral polio vaccine (OPV) and very widely used worldwide.
2. Live attenuated virus (Sabin vaccine, types 1,
2, and 3)
Rabies Rabies virus Live attenuated virus Used only post-exposure to the virus.
Rotavirus infections Rotaviruses Live attenuated oral Vaccine is safe in infants for protection against diarrhea and other infections that
(mainly diarrhea) could be fatal.
Rubella Rubella virus Live attenuated virus (Strain RA27/3) Vaccine is used either as such, or (mainly) in combination with measles and
mumps vaccines (MMR).
Smallpox Vaccinia virus 1. Live-attenuated dried calf lymph As smallpox was declared by WHO an eradicated disease, the routine vaccination
2. Live attenuated virus with vaccinia vaccine was discontinued worldwide.
Tetanus Clostridium tetanii Subunit-toxoid, alum adsorbed The tetanus toxoid is used as such, mainly as a booster vaccine after injury, but
is routinely used as a part of a combined pediatric vaccine DPT (diphtheria,
­pertussis, tetanus).
Tuberculosis Mycobacterium tuberculosis Whole bacteria live BCG (Bacillus Calmette Three main strains of the BCG are available and in use: Grown under different
-Guérin) conditions, they vary in their characteristics. BCG is used also as an adjuvant and
immunopotentiator.
Typhus Salmonella typhi, 1. Whole cell inactivated phenol-preserved New generation of genetically attenuated S. typhi vaccines is currently underway,
Salmonella typhimurium 2. Live attenuated oral Ty21a strain as candidate live oral vaccines.
3. Subunit Vi capsular poly-saccharide
Varicella/Zoster Varicella Zoster virus (VZV) Live attenuated virus (Oka strain) The Oka strain was isolated in Japan from human embryonic lung (HeL) cells by
numerous passages.
Whooping cough Bordetella pertussis Subunit Acellular pertussis vaccine The acellular pertussis vaccine is composed of 5 components of the bacteria,
including pertussis toxin. This vaccine is routinely used in combination with
diphtheria and tetanus toxoids (DPT).
Yellow fever Yellow fever virus Live attenuated whole virus Strain 17D Attenuation is achieved by repeated passages.

SOURCE: Horizon Scientific Press/Caister Academic Press UK

 Vaccinology Clinical
Microbiology

Personalized Therapeutics
medicine

Figure 11.1 
Impact of the recent advances in microbial and human genomics on the development of novel tools for the diagnosis, monitoring,
prevention, and treatment of human infectious diseases.

Pizza and coworkers on Neisseria meningitidis serogroup its 2,158 open reading frames (ORFs). Based on predic-
B (MenB)(5,6). MenB, causing 50% of the meningococ- tion algorithms, 570 ORFs were expected to encode sur-
cal meningitis worldwide, had been refractory to vaccine face-exposed or secreted proteins that might be accessible
development due to the identity of its capsular polysaccha- to the immune system. Further steps towards the selection
ride to a human self-antigen and to the extreme variability of the best vaccine candidates comprised expression of the
of its major outer membrane proteins. The MenB genom- predicted antigens as recombinant proteins in Escherichia
ics vaccine project started with full DNA sequencing of coli (350 ORFs), assessment of their high exposure on the
the virulent strain MC58 and bioinformatics analysis of meningococcal surface (91 candidates), and testing of their
ability to elicit antibodies mediating MenB killing by serum
bactericidal assays and/or protection against lethal challenge
in an animal infection model (28 selected candidates). After
screening these 28 antigens on a panel of diverse isolates to
Epidemiology determine whether their sequence was well conserved, a
multi-component vaccine was finally selected for develop-
ment. This genome-based MenB vaccine consists of three
Production of
recombinant proteins recombinant proteins representing five MC58 antigens,
plus outer membrane vesicles derived from another MenB
isolate(7). Notably, the main vaccine antigens were identified
Genome sequence, as important virulence factors(8). Factor H binding protein
antigen prediction (fHbp) binds a key inhibitor of the complement alternative
pathway enabling the meningococcus to evade killing by the
innate immune system; the Neisserial heparin binding anti-
In vitro/in vivo
testing
gen (NHBA) also plays a role in serum resistance; and the
40
Neisserial adhesin A (NadA) mediates bacterial adhesion to
20
host cells.
0 0
10 102 104
Further molecular epidemiological investigations
Vaccine testing revealed a certain degree of sequence variability in the vac-
in humans
cine antigens expressed by MenB isolates obtained from dif-
ferent patients. Antibody cross-reactivity was demonstrated
between the NHBA variants and also between the three
main NadA variants identified in hyper-virulent strains.
Little cross-protection was instead observed between
Figure 11.2 
Reverse vaccinology approach for the identification of vaccine the fHbp variant present in the MenB vaccine and the
candidates starting from the genome sequences of the pathogen. other two fHbp variant groups. Additionally, the level of

1 7 0   •   P rinciples of G enomic M edicine

 Free ebooks ==> www.Ebook777.com
expression of all MenB antigens was shown to vary between
strains. For these reasons, a new typing system (the menin-
gococcal antigen typing system, or MATS) was developed
to predict potential vaccine coverage among infective iso-
lates in different geographical settings(9). MATS is a sand-
wich enzyme-linked immunosorbent assay (ELISA) that
measures the amount of each target antigen expressed by a
strain and its immunological cross-reactivity with the pro-
tein variant present in the MenB vaccine. The data obtained
with MATS correlate with the killing of strains in a serum
bactericidal activity assay, and allowed the prediction of
78% coverage of the European MenB isolates. Results from
clinical trial studies have shown safety and robust immune
responses in both adults and infants(10), forming the basis
for the recent licensure of this vaccine by the European
Medicines Agency.
ADDRESSING ANTIGEN
VA R I A B I L IT Y BY M I C RO B I A L
C O M PA R AT I VE G E N O M I C S
Emerging technologies have greatly accelerated genome
sequencing during the past decade, leading to a further evo-
lution of reverse vaccinology to incorporate comparative in
silico analysis of multiple genomes from different strains of
the same species. This approach allows the selection of vac-
cine antigen candidates, while taking into account the intra-
species antigen diversification stratagem adopted by many
pathogenic species to escape the immune system. Intra-
species diversity is generated by a variety of mechanisms,
including mutation, horizontal gene transfer through
mobile genetic elements and recombination(11).
Multigenome reverse vaccinology was first applied to
Streptococcus agalactiae (Group B streptococcus, or GBS), a
Gram-positive microorganism that colonizes the ano-geni-
tal tract of 20–30% of healthy women and is a major cause
of neonatal sepsis and meningitis. The pathogen can also
cause severe invasive infections in the elderly, in pregnant
women, and in patients with underlying disease(12). There
are ten GBS serotypes distinguished by their capsular poly-
saccharide, and the amount of maternal antibodies directed
against each polysaccharide type is inversely proportional
to the risk of neonatal infection with strains of that spe-
cific serotype. This observation established the basis for the
development of vaccines based on capsular polysaccharides
conjugated to carrier proteins, which induce long-lasting
immune responses(13).
Parallel efforts to identify protective protein antigens
capable of conferring wide coverage were also under-
taken, given that protection by GBS polysaccharides
M icrobial G enomics   •   1 7 1
www.Ebook777.com
is serotype-specific and that non-typeable isolates not
expressing any capsule cannot be protected against by
­polysaccharide-based vaccines. Analysis of the full genome
of eight different GBS strains by Tettelin et al. revealed
novel genes to be added to the species gene pool after each
strain was sequenced. This observation highlighted GBS
intra-species diversity and introduced the concept of the
“pan-genome,” which comprises “core” genes shared by all
strains, and “dispensable” genes present only in one or a few
strains(14). Maione and colleagues applied the pan-genome
notion to design a universal vaccine against GBS(15). By
computational analysis of the eight sequenced genomes,
they predicted 589 surface-exposed proteins, 396 of which
encoded in core genes and 193 in dispensable genes. Of
these 589, 312 were successfully expressed as recombinant
proteins in E. coli and evaluated for their ability to mediate
protection in a mouse ‘maternal immunization–neonatal
pup challenge’ model. A four-antigen combination proved
protective against a large panel of strains. Three of these
protective antigens were encoded in dispensable genes, and
would not have been identified if only a single genome had
been screened. Interestingly, these three proteins were seen
to assemble into previously undescribed long filamentous
pilus-like structures extending outside the bacterial surface
that were shown to play an important role during bacterial
infection. Subsequent genomic analysis using a wider col-
lection of strains revealed three different pilus islands, PI-1,
PI-2a, and PI-2b, at least one of which was present in a wide
panel of strains. More interestingly, a vaccine incorporating
one component of each pilus variant was shown to provide
a high level of mouse protection against virulent isolates
representing all GBS serotypes(16).
A similar comparative genomics approach was success-
fully applied to Streptococcus pneumoniae (pneumococcus),
another major human pathogen causing sepsis, meningitis,
pneumonia, otitis media, and sinusitis, which accounts for
more than 10% of the mortality worldwide in children under
five years old(17). Pneumococcus can be classified into more
than 90 capsular serotypes, and the recently introduced
polysaccharide-conjugate vaccines have proven highly effec-
tive in preventing pneumococcal infections against their
represented serotypes(18). Pneumococcal protein antigens
have additionally been evaluated for their use in universal
serotype-independent vaccines to face variable regional dis-
tributions of serotypes, the occurrence of serotype replace-
ment after vaccination, as well as the complexity and cost
of conjugate vaccines(19). The availability of multiple pneu-
mococcal genome sequences, combined with an increased
understanding of pili in GBS and in other Gram-positive
pathogens, led to the discovery of pneumococcal pilus
proteins eliciting high protection in mouse infection mod-
els as potential components of a broad-coverage vaccine
combination(20).
A further step in multi-genome reverse vaccinology
introduced an additional criterion for the selection of
antigens specific to pathogenic strains and absent in com-
mensal strains of the same species, with the aim of reducing
the potential impact of a vaccine on the commensal flora.
Comparative analysis of the genomes of two E. coli strains
causing meningitis, five uro-pathogenic strains, one avian
strain, and the non-pathogenic K12, identified 230 surface
antigens present in the extra-intestinal pathogenic E.  coli
but absent (or poorly conserved) in the non-pathogenic
isolate. Nine potential vaccine antigens were able to induce
protection in a mouse-challenge sepsis model, some of
which also present in intestinal pathogenic E.  coli, show-
ing promise for a broad-coverage vaccine against different
pathogenic E. coli(21).
Genomic reverse vaccinology approaches have now
been applied to the discovery of new vaccines against
many other pathogens, including Chlamydia pneumoniae,
Bacillus anthracis, Porphyromonas gingivalis, Mycobacterium
tuberculosis, Helicobacter pylori, hepatitis C virus, the coro-
navirus responsible for severe acute respiratory syndrome
(SARS), and the malaria parasite Plasmodium falciparum(8).
Promising results towards defeating malaria have recently
been obtained by vaccinating infants and children with
the circumsporozoite protein 1 (CSP-1) fused with the
hepatitis-B surface antigen(22).
I N T EG R AT I N G G E N O M I C S ,
P ROT EO M I C S , A N D I M MU N O M I C S
F O R VAC C I N E D I S C OVE RY
A common drawback of genome-based approaches for
vaccine discovery is the need to screen a large number of
candidates by laborious and time-consuming in vivo and/
or in vitro assays, in order to select a limited number of
antigens conferring effective protection. Therefore, several
pre-screening strategies aimed at reducing the number of
antigens for further biological testing have been attempted.
Based on the observation that bacterial vaccines induc-
ing protective antibodies are mainly constituted by highly
expressed surface-exposed antigens and/or secreted tox-
ins, proteomic-based methods have been used to selec-
tively identify these categories of proteins. In a pioneering
approach, Rodriguez-Ortega et al. analyzed the surface of
Streptococcus pyogenes (Group A streptococcus, or GAS), a
severe human pathogen for which a vaccine is not yet avail-
able, by digestion of live bacteria with different proteases,
1 7 2   •   P rinciples of G enomic M edicine
followed by mass spectrometry (MS) analysis of gener-
ated peptides, and identified proteins highly expressed on
the bacterial surface and thus accessible to antibodies(23).
Similarly, antigens on the surface of Gram-negative bacte-
ria were identified by MS analysis of membrane fragments
released by the bacteria upon genetic modifications to
weaken their outer membrane(24).
An alternative approach aimed at reducing the number
of antigens in the pre-selection stage consists of interrogat-
ing the entire antigenic repertoire of a particular pathogen
by using representative libraries of recombinant peptides
that can be displayed on the surface of bacteria or bacte-
rial phages, or spotted onto microarrays. These libraries can
then be screened with sera from infected individuals who
recovered from infection for the presence of specific anti-
bodies, leading to the identification of a discrete number of
antigen targets(25).
In a recent study, Bensi et al. devised a strategy that incor-
porates quantification of bacterial surface proteins using
antibodies raised against recombinant surface-predicted
antigens, MS proteomic analysis, and high-throughput
screening of human sera, for the rapid selection of a limited
number of vaccine candidates prior to biological testing.
By applying this combined approach to GAS, highly selec-
tive identification of few protective antigens was achieved,
which allowed the definition of a multi-protein formula-
tion conferring consistent protection against multiple GAS
serotypes in mouse models of infection(26).
The new information derived from the growing list
of protective antigens from different microbial species is
expected to bring an additional improvement to the predic-
tion of vaccine candidates by bioinformatics tools. Indeed,
several curated databases have been established, based on
the information obtained from experimentally validated
antigens. Ultimately, improved algorithms are expected to
be developed that will allow, not only better prediction of
surface localization, but also the identification of common
signatures among protective antigens, that will guide the
identification of novel vaccine candidates.
R AT I O NA L D E S I G N O F E FFEC T I VE
VAC C I N E S BY S T RU C T U R E -BA S E D
A P P ROAC H E S
Recent advances in X-ray crystallography and nuclear
magnetic resonance (NMR) spectroscopy have greatly
accelerated structural studies on vaccine antigens and their
epitopes, opening the path for the structural design of
novel and improved vaccines(27). This type of approach can
be applied to address the high variability of many antigen
targets of protective antibodies that, as discussed above, is
exploited by many pathogens to evade the human immune
system.
A structural analysis of MenB fHbp and of the epitopes
recognized by protective monoclonal antibodies against the
three fHbp variants guided the construction of a chimeric
protein with broad protective capacity. The chimera was
built by incorporating in variant 1 of fHBP key amino acids
from variant 2 and 3 epitopes, while strictly maintaining
the three-dimensional structure of the native molecule. To
preserve folding, amino acid substitutions were introduced
only in residues with side chains well exposed to solvent,
leaving the internal core of the protein unaltered.
A similar structural approach was used to obtain a fusion
protein covering six variants of the Group B Streptococcus
pilus protein 2a. In this case, structural analysis revealed a
similar four-domain organization where domain 3 was the
main target of protective antibodies. The domain 3 regions
from each of the six variants could be fused in a single mol-
ecule exhibiting cross-protective properties against strains
expressing the different variants.
Structural vaccinology has also been applied to the
respiratory syncytial virus (RSV), a virus that infects the
lower respiratory tract of most infants and children, and
is often associated with hospitalization. RSV is a diffi-
cult target for which live attenuated vaccines have been
unsuccessful to date, and subunit vaccines, mainly relying
on the F glycoprotein antigen, have proven biochemically
challenging to develop. F is a trimeric protein that exists in
different structural forms; that is, a pre-fusion displayed
on infectious virions, a transient intermediate extended
structure, and a post-fusion state with detergent-like
properties that mediates host cell entry. The F pre-fusion
form is the target of most RSV-neutralizing antibodies in
human sera, but its instability has hindered its use for a
vaccine.
Structural insights guided the engineering of the RSV F
post-fusion to a more hydrophilic molecule with increased
solubility, where the best-characterized neutralizing epi-
topes between the pre- and post-fusion forms were well
preserved(28). This engineered antigen is now approaching
clinical trials for a novel vaccine against RSV. In a second
successful study, highly effective prefusion-specific antibod-
ies were identified and used to obtain a co-crystal structure
of the antibody in complex with the F glycoprotein locked
in its pre-fusion state(29).
The above-described examples clearly indicate that many
linear peptides that are usually called epitopes because they
are recognized in vitro by antibodies are imperfect mimics
of the real surfaces recognized by antibodies in immunized
M icrobial G enomics   •   1 7 3
or infected hosts, as they lack a single, defined conforma-
tion. Indeed, most often, linear peptides bind functional
antibodies with low affinities compared to the native pro-
tein or its folded domains, and they might preferentially
elicit immune responses against different structural aspects
than those recognized by protective antibodies. Therefore,
structural vaccinology will be an excellent tool for engineer-
ing antigenic surfaces on domains to avoid the use of iso-
lated epitopes.
The development of effective vaccines against the
human immunodeficiency virus (HIV) has been ham-
pered by its high antigenic diversity(30). Short peptides
representing a wide number of potential T-cell epitopes
were selected by computer-based methods and used
for the construction of polyvalent vaccine antigens(31).
Furthermore, some epitopes targeted by broad neutral-
izing protective antibodies (bnAbs) towards the HIV
envelop protein gp41 have been identified(32). Alternative
strategies for eliciting protective B cell responses to HIV
focus on trimeric spike protein variants of gp120 that
may closely resemble the native spike on infectious viri-
ons, although this trimeric structure has not yet been
obtained(33). Surprisingly, partial success was attained in
a recent HIV vaccine trial in which neither bnAbs, nor
potent T cell responses were induced(34).
S Y N T H ET I C G E N O M I C S F O R
F U T U R E VAC C I N E S
Another champion of immune system evasion is the influ-
enza virus, capable of rapidly evolving into thousands of dif-
ferent strains that make necessary the production of a new
flu vaccine every year. Most alarmingly, some of the variants
can switch hosts, generating new human pandemics.
Finding common universal flu epitopes has been the
object of continuous search, and epitopes targeted by
cross-reactive human antibodies against hemmaglutinin
(HA) were recently identified. These conserved epitopes
were found in the stem part of HA, while the variable
epitopes used in present vaccines are mainly located
in the more exposed head of the protein(30). However,
immuno-dominance of variable epitopes remains a key
challenge. Deep analysis of the human B cell reper-
toire and increasingly efficient physical and structural
epitope-mapping techniques will help understand the struc-
tural basis for epitope immune dominance and improve
vaccine design.
While these envisaged results may soon be attained,
of capital importance remains the acceleration of the
steps leading to the development of each new version of
the vaccine necessary to cope with unceasingly arising
flu variants. This is particularly true in the case of pan-
demics, as vaccine prevention tools are required well
before the appearance of the peak of disease incidence.
Genomic-based approaches have come to our aid in this
case also, in the form of “synthetic vaccinology.” In a sim-
ulated response to a pandemic with a H7N9  “bird flu,”
Dormitzer et al. utilized the viral DNA sequence informa-
tion to synthesize the genes coding for the necessary HA
and neuraminidase (NA) antigen variants(35). The genes
were built enzymatically in cell-free reactions that included
a critical error correction step. Co-transfection of canine
kidney cells with the synthetic genes cloned into optimized
expression vectors and plasmid DNA containing improved
viral backbone genes produced a combination resulting in
high yields of vaccine antigens. The authors could demon-
strate the potential of this new procedure to save weeks
off the time needed for vaccine manufacture to respond
quickly to a sudden pandemic. The study also offered proof
of concept for the potential of synthetic vaccinology for
the generation of tailor-made microorganisms optimized
for the expression of newly designed vaccines.
M O N I TO R I N G I N F E C T I O N
E M E R G E N C E , T R A N S M I S S I O N, A N D
PAT H O G E N E VO LU T I O N BY N E X T-
G E N E R AT I O N S E Q U E N C I N G
In this section, we will discuss how genomic-based tech-
nologies can be used in clinical microbiology both for the
diagnosis and the management of individual infections, as
well as for monitoring the emergence and epidemiology of
infectious diseases.
GENOMICS IN THE CLINICAL
M I C RO B I O L O GY L A B O R ATO RY
The main task of the clinical microbiology laboratory is
the rapid management of individual infections by isolat-
ing pathogens from clinical samples, identifying species
for diagnostic purposes, and testing for antimicrobial sus-
ceptibility. Many basic microbiology practices to accom-
plish these tasks were developed over decades, and are
time-consuming and labor-intensive. Most often the
pathogen has to be cultured before isolation, and complex
selective media are required to treat samples contaminated
with colonizing flora. Moreover, diagnostic characteriza-
tion depends on a wide range of biochemical testing path-
ways that are often species-specific. This multiple-step
1 7 4   •   P rinciples of G enomic M edicine
process often takes several days, or even weeks in the case
of slow-growing microorganisms. The “omics” revolution is
expected to deeply transform routine clinical microbiology
laboratory practices in the coming years by the progressive
substitution of many of these complex multifaceted proce-
dures with genome-based technologies(36,37).
High-density pan-microbial microarrays containing
nucleic acid probes specific for various pathogen sequences
now allow the rapid screening of a large number of patho-
gens: nucleic acids from a clinical specimen can be ampli-
fied randomly and then hybridized to the chips for species
identification(38).
A proteomics-based approach that has further accel-
erated species identification consists of comparing mass-
spectrometry profiles of pure microbial suspensions with
available databases(39). Polymerase chain reaction (PCR)
assays can also assist in the rapid identification and typing
of bacterial and viral pathogens, and in some cases they
have become an integral part of the standard of care. For
instance, DNA-sequencing analysis of the HIV genotype
for the presence of mutations conferring resistance to anti-
retroviral drugs, and PCR-based measurement of viral
loads, can be used to choose medication and help predict
the responses to therapy(40), and similar approaches have
been applied to other viral diseases like hepatitis B and C
and influenza. As far as testing of antimicrobial susceptibil-
ity in concerned, checking for the presence or absence of
antimicrobial resistance genes by PCR can accelerate effec-
tive treatment of infected patients.
Application of next-generation full-genome sequencing
will soon be sufficiently fast, accurate, and cheap to be rou-
tinely used by the clinical laboratory for improved patient
care. The major advantage of whole-genome sequencing is that
comprehensive DNA information can be obtained in a single
rapid step, providing all necessary data for diagnostic and typ-
ing needs. Nevertheless, substantial challenges will need to be
overcome before microbial full-genome sequencing can be
fully embraced in a clinical laboratory setting and gradually
replace present-day methodologies. Success will depend on the
development and implementation of reliable and user-friendly
bioinformatics tools to rapidly extract the most meaningful
genetic information from the generated complex data sets.
T R AC I N G I N FEC T I O N O U T B R E A K S
A N D T R A N S M I S S I O N EVE N TS
Microbial genomes are much smaller than eukaryotic
genomes but much more diverse, as up to 40% of their
DNA may consist of dispensable sequences that are not
shared by all members of the same species. The analysis MO N I TO R I N G T H E E FFEC T O F M E D I C A L
of single-nucleotide polymorphisms (SNPs) by DNA I N T E RVE N T I O N O N PAT H O G E N
deep-sequencing allows discriminating between closely EVO LU T I O N A N D D ET EC T I N G WIT H I N-
related strains of the same microbial species, and has H O S T M I C RO B I A L VA R I AT I O N
enabled epidemiological studies aimed at reconstruct-
A deeper understanding of the microbial population struc-
ing the source of infection outbreaks and the routes of
ture may also help researchers monitor the effect of pub-
person-to-person transmission. This type of information
lic health interventions, such as antibiotic use and vaccine
can help in managing contemporary threats and prevent-
introduction, on pathogen evolution.
ing future outbreaks(41,42). By SNP analysis, the geographi-
The recent discovery showing that genome-wide muta-
cal origin of historical infections like plague, tuberculosis,
tion rates in latent tuberculosis infection are similar to
or leprosy could be traced back in China, India, and East
those occurring in active disease may explain why mono-
Africa, respectively, and the sites of their initial spread
therapy for patients with latent infection is a risk factor for
leading to global dissemination were identified. In a more
selecting isoniazid-resistant strains(43). Emergence of the
recent example, genome analysis of Mycobacterium leprae
community-acquired, methicillin-resistant Staphylococcus
isolated from armadillos identified those animals as a pos-
aureus (CA-MRSA) USA 300 clone in the United States
sible source of zoonosis for isolated leprosy cases occur-
and throughout the world is a major public health con-
ring in the United States.
cern. By analyzing the genome sequences of 10 CA-MRSA
By understanding the genetic basis of infection out-
isolates recovered from diverse regions, Kennedy and col-
breaks, investigators will be able to validate diagnostic
laborators demonstrated a single clonal lineage under-
tests to rapidly design appropriate therapies. In addition,
going expansion and diversification(44). The data suggest
future outbreaks can be anticipated, and therefore con-
that the CA-MRSA clone will continue evolving under
tained, before becoming widespread. For instance, a panel
host-selective pressure and that higher-virulence clones
of 1225 informative SNPs was designed for the rapid typ-
may arise, further emphasizing the need for a preventa-
ing of entero-hemorragic E. coli O157:H7 isolates during
tive vaccine. Harris et  al. used population genomics to
outbreaks.
trace person-to-person transmission of CA-MRSA strains
Different microbial isolates of the same species can vary
within a hospital, confirming the potential of this approach
in their capacity to cause disease according to the presence
to identify unrecognized transmission chains for nosoco-
and expression of genes encoding toxins, adhesins, and
mial infection, determine the point source, and precisely
drug resistance, often carried by mobile genetic elements
guide infection-control activities.
like prophages. Intra-species whole-genome sequenc-
The emergence of non-vaccine strains in Streptococcus
ing allows discriminating between isolates with different
pneumoniae has been a concern since the introduction of
pathogenic potential, and evaluating their genetic rela-
the heptavalent conjugate polysaccharide vaccine in 2000,
tionships to infer the sequence of events driving pathogen
as it protects against many, but not all, serotypes. Genomic
evolution(41,42).
studies have found evidence for capsular switching, in
Longitudinal genomic studies performed on differ-
which hybrid strains arising through recombination and
ent isolates of Streptococcus pyogenes demonstrated that
expressing non-vaccine capsular types increased in their fre-
strains causing invasive disease are tightly related to those
quency due to vaccine selective pressure(45).
causing mild oro-pharyngeal infections, and differences in
Many human pathogens are common constituents of the
virulence, disease phenotype, and epidemic behavior are
normal flora, and their evolution during colonization may
probably due to genes encoded on mobile elements rather
trigger a transition from healthy carriage to invasive disease.
than on the core chromosome. Furthermore, evidence for
Whole-genome sequencing in populations of bacteria colo-
adaptation in genes involved in virulence regulation sup-
nizing a single individual is shedding light on the microbial
ported a model in which mutation in vivo plays an impor-
evolution dynamics within the host(41). Genetic variation,
tant role in progression from mild to severe S.  pyogenes
including SNPs, short insertions and deletions (indels), and
invasive disease. The authors could indeed demonstrate
mobile elements, has been discovered in single human hosts
that single-nucleotide mutations affecting the production
colonized by species as disparate as Mycobacterium tuber-
of a secreted protease implicated in tissue destruction and
culosis, Salmonella enterica, and Staphylococcus aureus. In
dissemination could significantly change the necrotizing
a study on a long-term carrier of S. aureus who developed
fasciitis capacity of particular subclones, thus offering new
a bloodstream infection, the genomes of invasive bacteria
targets for therapy and vaccine design(42).

M icrobial G enomics   •   1 7 5
were found to possess an excess of mutations that truncated
proteins, including a transcriptional regulator implicated
in pathogenicity. Another study, of a 16-year outbreak of
chronic Burkholderia dolosa infection, revealed evidence
for parallel adaptive evolution of 17 genes across 14 cystic
fibrosis patients.
M ETAG E N O M I C S A N D T H E HUM A N
M I C RO B I O M E
The different sites of the human body are populated by
complex microbial communities, which have become the
subject of a new field in microbiology aimed at defining
the “microbiome” composition, its interactions with the
human host, and its role in human health(46). Given the
impossibility of cultivating most of the bacteria compos-
ing the human microbiome, broadly applicable techniques
for analyzing massive amounts of DNA sequence data
have been developed, which have contributed significantly
to the growing field of metagenomics. These studies have
demonstrated great variations between host individuals
and confirmed that substantial alterations in the human
microbiome are important for a variety of disease states,
including psoriasis, sexually transmitted infections, Crohn
disease, gastroesophageal reflux disease, and others. Recent
studies have also addressed the role of the gut microbiome
in the development of the immune system. The established
links between microbial communities and the etiology of
human disease will inform future design of better vaccines
and therapeutics.
C O N T R O L L I N G B AC T E R I A L
I N F E C T I O N BY G E N O M E -B A S E D
ANTIMICROBIALS
Most of the antibiotics in use today originated many decades
ago as natural products isolated from bacteria and fungi.
The antimicrobial industry has excelled at fine-tuning these
natural molecules to improve their spectrums, efficacy,
and safety, especially by means of semi-synthetic chemis-
try approaches. Nonetheless, the growing emergence of
antibiotic-resistant bacterial strains and the public health
threat of pandemic viral infections have recently raised
a renewed interest in the discovery and development of
novel non-toxic and fast-acting antimicrobial drugs. The
Infectious Disease Society of America estimates that 70% of
hospital-acquired infections in the United States are resis-
tant to one or more antibiotics.
1 7 6   •   P rinciples of G enomic M edicine
G E N O M I C S A N D TA RG ET-BA S E D
A N T I M I C RO B I A L D I S C O VE RY
Genome-based approaches applied at the early stage of the
drug discovery process have generated a valuable inventory
of genes and cellular processes from which to further test
and validate novel antibacterial targets(47). Most of these
targets are selected for their essential role during in vitro
growth, usually by means of genetic manipulation (e.g., gene
knockout) of the relevant bacteria. An alternative approach
for target selection focuses on virulence factors required by
specific bacteria to cause disease, like toxin delivery or cell
adhesion, and has the potential advantages of better pre-
serving the host microbiome and of decreasing the proba-
bility of antibiotic resistance(48). Novel antimicrobial targets
can also be selected by “metabolomics” approaches—that is,
analysis of metabolite production by host cells by NMR or
Mass Spectrometry(49). For instance, changes in the meta-
bolic flux of human cytomegalovirus–infected cells were
used to identify metabolic pathways upregulated by viral
infection, and potential targets for novel antivirals aimed at
blocking viral replication.
By comparative genomics, target genes can be selected
for narrow- or large-spectrum therapeutic solutions on the
basis of their conservation profiles across species. Once vali-
dated, these target genes can be cloned and sequenced, and
their protein products expressed in an optimized expression
system (e.g., Pichia pastoris, Baculovirus, E. coli). Targets
are often screened by high-throughput methods against
large libraries of combinatorial chemistry-derived com-
pounds(50,51). Each specific molecular target or pathway of
interest is combined systematically with each possible drug
compound by automated platforms, and positive results,
or “hits,” are subsequently characterized with respect to
potency, mechanism of inhibition, spectrum, and selectiv-
ity(36,52). An example of an antimicrobial target identified
by a genomics-driven approach is the product of the def
gene, which is present in all pathogenic bacteria and does
not share a functionally equivalent gene in mammalian
cells. The gene encodes a peptide deformylase belonging to
the matrix metallo-protease family of enzymes, and a selec-
tive inhibitor could be identified by screening a library of
metallo-enzyme inhibitors.
An important condition that dictates the efficiency of
antimicrobial compounds is their need to cross the micro-
bial membrane barrier and to avoid subsequent extrusion by
multidrug-resistance efflux pumps. One strategy that takes
into account permeability and efflux issues consists of com-
bining genomics with classic whole-cell screening methods
by using genetically modified microorganisms that can
respond in a measurable manner when a target of interest is
inhibited. The response can be determined as growth inhi-
bition (absorbance) or induction of a linked reporter gene
(e.g., luminescence or fluorescence)(52).
A promising novel group of anti-infectives is repre-
sented by the family of lysins naturally produced by bac-
terial viruses (bacteriophage) to digest the Gram-positive
bacterial cell wall for phage progeny release. A large variety
of phage enzymes with different specificities was identified
by genome analysis and successfully used in animal models
to control antibiotic-resistant bacteria on mucosal surfaces
and in blood. The advantages over other antibiotics reside
in pathogen specificity without disturbing the normal flora,
low chance of bacterial resistance to lysins, and their ability
to kill colonizing pathogens on mucosal surfaces(53).
The incorporation of automation, computational
methods, and nanotechnology has allowed for greatly
increased efficiency in the development of both combinato-
rial libraries and high-throughput screening of potentially
useful drugs. Nevertheless, to date, only a few candidates
against genetically validated bacterial drug targets derived
from these types of screens have attained clinical testing.
Structural characterization of inhibitor–target complexes
is expected to further assist the design of higher-affinity
drugs with improved pharmacological properties. We are
currently witnessing an explosion in technological and
computational advances in structural genomics, with pro-
tein structures of hundreds or thousands of medically rel-
evant targets from infectious disease organisms likely to
be available over the next few years. This new information
is expected to provide an unprecedented opportunity to
accelerate the development of new and improved chemo-
therapeutic antimicrobial agents.
A N T I-I N FEC T I VE MO N O C L O NA L
ANTIBODIES
Human genomics now enables the examination of the full
epitope repertoire of antibodies in infected and in vacci-
nated individuals, which can facilitate the development of
therapeutic monoclonal antibodies (mAbs). The strategy
of displaying human antibody fragments on phage surfaces
has produced several mAbs with potential therapeutic
applications against agents of infectious disease, including
influenza A virus, Clostridium difficile, HIV, viral hepati-
tis, rabies, Pseudomonas aeruginosa, methicillin-resistant
S. aureus, and Bacillus anthracis(54). Of these, the mAb tar-
geting the protective antigen of B. anthracis, raxibacumab,
has met all criteria for approval by the U.S. Food and Drug
Administration(55).
M icrobial G enomics   •   1 7 7
A very promising approach for the discovery of novel
therapeutic antibodies and vaccine targets consists of iso-
lating single B cells from patients having recovered from a
particular infection, followed by deep sequencing of single
B cell clones and production of human monoclonal anti-
bodies with neutralizing or opsonizing capacities(56).
Interrogation of the entire human B cell response to
infection or vaccination has also been applied to the identi-
fication of cross-reactive protective epitopes that may repre-
sent structural determinants of broadly protective vaccines
to overcome high viral mutation rates. This process has been
named “analytical vaccinology,” and it was made possible by
several recently developed methods for generating human
monoclonal antibodies from blood samples(57), which are
tested for broad neutralization in high-throughput func-
tional assays(58).
HUM AN GENOMICS
AND PER SONALIZED MEDICINE
TO C O M B AT I N F E C T I O U S
DISEASES
As for other fields of medicine, the recent advances in
human genome sequencing and computational biology
are expected to revolutionize the treatment and prophy-
laxis of infectious diseases. Applications of personalized
genome-based medicine to the prediction of individual
factors that can predispose or affect the response to certain
infections as well as individual responses to therapeutic and
prophylactic measures are already becoming a reality(36).
The first evidence that certain infectious diseases have a
genetic predisposition initially came from a study showing
an increased risk of mortality in children born to parents
who also died from an infection(59). Since then, defects in
genes encoding effectors of the immune response have been
associated with an increased susceptibility to certain infec-
tions, like those caused by S.  aureus and M.  tuberculosis.
High-density DNA arrays capturing human genome-wide
variation can be used to analyze the association of certain
SNP with susceptibility to different bacterial and viral
diseases(60).
Genome-wide association studies (GWAS) have also
identified a series of markers associated with the quality of
individual responses to disease treatment(36). For example,
a SNP located upstream of the gene encoding the type III
interferon was found to be associated with differences in
the extent of response to an anti-HCV (human C virus)
drug treatment, while IL28B gene polymorphisms were
associated with the spontaneous clearance of acute HCV.
Large-scale profiling of the full RNA of peripheral blood
mononuclear cells (PBMCs) by microarrays or deep RNA
sequencing approaches has been applied to the identifica-
tion of patterns of gene expression characteristic of cer-
tain human infections like tuberculosis, dengue, influenza,
S.  aureus and salmonellosis, among others. Data derived
from these approaches will assist the diagnosis and progno-
sis of disease, the design of antiviral and antibiotic therapies,
and the development of genetic tests to predict adverse reac-
tions to antimicrobial drugs.
In the vaccines field, novel tools of systems biology
can be applied to the analysis of human immunological
response patterns to vaccines in order to uncover molecu-
lar signatures of vaccine efficacy and guide the design and
evaluation of new vaccines(61). This “systems vaccinology”
strategy has been applied to examine the initial molecular
signatures in individuals vaccinated against yellow fever, or
after administration of the trivalent inactivated influenza
virus vaccine. Similar approaches have been used to study
immune responses to Brucella melitensis and fungal infec-
tions. The obtained data will ideally lead to the design of
vaccines capable of inducing optimal immune responses
without toxic effects, thus improving vaccine safety profiles.
Integration of increasingly complex high-throughput
data into descriptive and predictive equations for immune
responses to vaccines is expected to drive faster and
more accurate ways of screening vaccine candidates for
their effectiveness. Pulendran et  al.(62) have predicted the
development of a vaccine chip microarray, similar to the
MammaPrint prognostic chip that was developed for
breast cancer, which will be able to predict the immunoge-
nicity of any vaccine.
Like in other genomic fields, the successful clinical
application of these novel technologies will depend on the
development of translational research approaches capable
of dealing with the enormous amount and different types
of generated information and with the uncertainty that is
typical of common clinical scenarios.
REFERENCES
1. R. Rappuoli, H. I. Miller, S. Falkow. Medicine. The intangible value
of vaccination. Science 297, 937 (Aug 9, 2002).
2. D. K.  Kaushik, D. Sehgal. Developing antibacterial vaccines in
genomics and proteomics era. Scandinavian Journal of Immunology
67, 544 ( Jun, 2008).
3. S. A.  Plotkin. Vaccines:  past, present and future. Nature Medicine
11, S5 (Apr, 2005).
4. R. D.  Fleischmann et  al. Whole-genome random sequencing and
assembly of Haemophilus influenzae Rd. Science 269, 496 ( Jul 28,
1995).
1 7 8   •   P rinciples of G enomic M edicine
5. M. Pizza et al. Identification of vaccine candidates against serogroup
B meningococcus by whole-genome sequencing. Science 287, 1816
(Mar 10, 2000).
6. H. Tettelin et al. Complete genome sequence of Neisseria menin-
gitidis serogroup B strain MC58. Science 287, 1809 (Mar 10, 2000).
7. M. M. Giuliani et al. A universal vaccine for serogroup B meningo-
coccus. Proceedings of the National Academy of Sciences of the United
States of America 103, 10834 ( Jul 18, 2006).
8. K. L. Seib, X. Zhao, R. Rappuoli. Developing vaccines in the era of
genomics: a decade of reverse vaccinology. Clinical Microbiology and
Infection: the Official Publication of the European Society of Clinical
Microbiology and Infectious Diseases 18 Suppl 5, 109 (Oct, 2012).
9. J. Donnelly et  al. Qualitative and quantitative assessment of
meningococcal antigens to evaluate the potential strain coverage
of protein-based vaccines. Proceedings of the National Academy of
Sciences of the United States of America 107, 19490 (Nov 9, 2010).
10. A. R.  Gorringe, R. Pajon. Bexsero:  a multicomponent vaccine
for prevention of meningococcal disease. Human Vaccines and
Immunotherapeutics 8, 174 (Feb, 2012).
11. T. T.  Binnewies et  al. Ten years of bacterial genome sequenc-
ing:  comparative-genomics-based discoveries. Functional and
Integrative Genomics 6, 165 ( Jul, 2006).
12. A. Schuchat. Group B streptococcus. Lancet 353, 51 ( Jan 2, 1999).
13. C. J.  Baker, M. S.  Edwards. Group B streptococcal conjugate vac-
cines. Archives of Disease in Childhood 88, 375 (May, 2003).
14. H. Tettelin et  al. Genome analysis of multiple pathogenic iso-
lates of Streptococcus agalactiae:  implications for the microbial
“pan-genome.” Proceedings of the National Academy of Sciences of the
United States of America 102, 13950 (Sep 27, 2005).
15. D. Maione et al. Identification of a universal Group B streptococcus
vaccine by multiple genome screen. Science 309, 148 ( Jul 1, 2005).
16. I. Margarit et al. Preventing bacterial infections with pilus-based vac-
cines: the group B streptococcus paradigm. The Journal of Infectious
Diseases 199, 108 ( Jan 1, 2009).
17. K. L. O’Brien et al. Burden of disease caused by Streptococcus pneu-
moniae in children younger than 5  years:  global estimates. Lancet
374, 893 (Sep 12, 2009).
18. K. L. Moffitt, R. Malley. Next generation pneumococcal vaccines.
Current Opinion in Immunology 23, 407 ( Jun, 2011).
19. T. M. Wizemann et al. Use of a whole genome approach to identify
vaccine molecules affording protection against Streptococcus pneu-
moniae infection. Infection and Immunity 69, 1593 (Mar, 2001).
20. J. L. Telford, M. A. Barocchi, I. Margarit, R. Rappuoli, G. Grandi.
Pili in Gram-positive pathogens. Nature Reviews. Microbiology 4,
509 ( Jul, 2006).
21. D. G.  Moriel et  al. Identification of protective and broadly con-
served vaccine antigens from the genome of extraintestinal patho-
genic Escherichia coli. Proceedings of the National Academy of Sciences
of the United States of America 107, 9072 (May 18, 2010).
22. S. T.  Agnandji et  al. First results of phase 3 trial of RTS,S/AS01
malaria vaccine in African children. The New England Journal of
Medicine 365, 1863 (Nov 17, 2011).
23. M. J. Rodriguez-Ortega et al. Characterization and identification of
vaccine candidate proteins through analysis of the group A strepto-
coccus surface proteome. Nature Biotechnology 24, 191 (Feb, 2006).
24. F. Berlanda Scorza et  al. Proteomics characterization of outer
membrane vesicles from the extraintestinal pathogenic Escherichia
coli DeltatolR IHE3034 mutant. Molecular and Cellular
Proteomics: MCP 7, 473 (Mar, 2008).
25. C. Giefing et al. Discovery of a novel class of highly conserved vac-
cine antigens using genomic scale antigenic fingerprinting of pneu-
mococcus with human antibodies. The Journal of Experimental
Medicine 205, 117 ( Jan 21, 2008).
26. G. Bensi et al. Multi high-throughput approach for highly selective
identification of vaccine candidates:  the Group A  streptococcus
case. Molecular and Cellular Proteomics:  MCP 11, M111 015693
( Jun, 2012).
 27. P. R.  Dormitzer, G. Grandi, R. Rappuoli. Structural vaccinology
starts to deliver. Nature Reviews. Microbiology 10, 807 (Dec, 2012).
28. K. A. Swanson et al. Structural basis for immunization with postfu-
sion respiratory syncytial virus fusion F glycoprotein (RSV F) to elicit
high neutralizing antibody titers. Proceedings of the National Academy
of Sciences of the United States of America 108, 9619 (Jun 7, 2011).
29. J. S.  McLellan et  al. Structure of RSV fusion glycoprotein trimer
bound to a prefusion-specific neutralizing antibody. Science 340,
1113 (May 31, 2013).
30. S. K. Grimm, M. E. Ackerman. Vaccine design: emerging concepts and
renewed optimism. Current Opinion in Biotechnology, (Mar 7, 2013).
31. D. H.  Barouch et  al. Mosaic HIV-1 vaccines expand the breadth
and depth of cellular immune responses in rhesus monkeys. Nature
Medicine 16, 319 (Mar, 2010).
32. G. Ofek et  al. Elicitation of structure-specific antibodies by epit-
ope scaffolds. Proceedings of the National Academy of Sciences of the
United States of America 107, 17880 (Oct 19, 2010).
33. J. M. Kovacs et al. HIV-1 envelope trimer elicits more potent neu-
tralizing antibody responses than monomeric gp120. Proceedings of
the National Academy of Sciences of the United States of America 109,
12111 ( Jul 24, 2012).
34. B. F. Haynes et al. Immune-correlates analysis of an HIV-1 vaccine
efficacy trial. The New England Journal of Medicine 366, 1275 (Apr
5, 2012).
35. P. R.  Dormitzer et  al. Synthetic generation of influenza vaccine
viruses for rapid response to pandemics. Science Translational
Medicine 5, 185ra68 (May 15, 2013).
36. J. M.  Fontana, E. Alexander, M. Salvatore. Translational research
in infectious disease:  current paradigms and challenges ahead.
Translational Research:  The Journal of Laboratory and Clinical
Medicine 159, 430 ( Jun, 2012).
37. X. Didelot, R. Bowden, D. J.  Wilson, T. E.  Peto, D. W.  Crook.
Transforming clinical microbiology with bacterial genome sequenc-
ing. Nature Reviews. Genetics 13, 601 (Sep, 2012).
38. D. Wang et al. Microarray-based detection and genotyping of viral
pathogens. Proceedings of the National Academy of Sciences of the
United States of America 99, 15687 (Nov 26, 2002).
39. E. Carbonnelle et  al. MALDI-TOF mass spectrometry tools for
bacterial identification in clinical microbiology laboratory. Clinical
Biochemistry 44, 104 ( Jan, 2011).
40. R. M. Gulick. Antiretroviral treatment 2010: progress and contro-
versies. Journal of Acquired Immune Deficiency Syndrome 55 Suppl 1,
S43 (Dec, 2010).
41. D. J. Wilson. Insights from genomics into bacterial pathogen popu-
lations. PLoS Pathogens 8, e1002874 (Sep, 2012).
42. R. J. Olsen, S. W. Long, J. M. Musser. Bacterial genomics in infec-
tious disease and the clinical pathology laboratory. Archives of
Pathology and Laboratory Medicine 136, 1414 (Nov, 2012).
43. C. B. Ford et al. Use of whole genome sequencing to estimate the
mutation rate of Mycobacterium tuberculosis during latent infection.
Nature Genetics 43, 482 (May, 2011).
44. A. D.  Kennedy et  al. Epidemic community-associated
methicillin-resistant Staphylococcus aureus: recent clonal expansion
M icrobial G enomics   •   1 7 9
and diversification. Proceedings of the National Academy of Sciences
of the United States of America 105, 1327 ( Jan 29, 2008).
45. N. J. Croucher et al. Population genomics of post-vaccine changes
in pneumococcal epidemiology. Nature Genetics 45, 656 ( Jun,
2013).
46. D. A. Relman. Microbial genomics and infectious diseases. The New
England Journal of Medicine 365, 347 ( Jul 28, 2011).
47. F. Arigoni et al. A genome-based approach for the identification
of essential bacterial genes. Nature Biotechnology 16, 851 (Sep,
1998).
48. A. E.  Clatworthy, E. Pierson, D. T.  Hung. Targeting virulence:  a
new paradigm for antimicrobial therapy. Nature Chemical Biology
3, 541 (Sep, 2007).
49. J. C. Lindon, E. Holmes, J. K. Nicholson. Metabonomics techniques
and applications to pharmaceutical research and development.
Pharmaceutical Research 23, 1075 ( Jun, 2006).
50. W. R.  Galloway, A. Bender, M. Welch, D. R.  Spring. The discov-
ery of antibacterial agents using diversity-oriented synthesis. Chem
Commun (Camb), 2446 (May 14, 2009).
51. D. A.  Pereira, J. A.  Williams. Origin and evolution of high
throughput screening. British Journal of Pharmacology 152, 53
(Sep, 2007).
52. S. D.  Mills. The role of genomics in antimicrobial discovery. The
Journal of Antimicrobial Chemotherapy 51, 749 (Apr, 2003).
53. V. A. Fischetti. Bacteriophage lytic enzymes: novel anti-infectives.
Trends in Microbiology 13, 491 (Oct, 2005).
54. C. F. Barbas, 3rd, A. S. Kang, R. A. Lerner, S. J. Benkovic. Assembly
of combinatorial antibody libraries on phage surfaces: the gene III
site. Proceedings of the National Academy of Sciences of the United
States of America 88, 7978 (Sep 15, 1991).
55. T. S. Migone et al. Raxibacumab for the treatment of inhalational
anthrax. The New England Journal of Medicine 361, 135 ( Jul 9,
2009).
56. D. Corti, F. Sallusto, A. Lanzavecchia. High throughput cellular
screens to interrogate the human T and B cell repertoires. Current
Opinion in Immunology 23, 430 ( Jun, 2011).
57. F. Sallusto, A. Lanzavecchia, K. Araki, R. Ahmed. From vaccines to
memory and back. Immunity 33, 451 (Oct 29, 2010).
58. H. X.  Liao et  al. High-throughput isolation of immunoglobulin
genes from single human B cells and expression as monoclonal anti-
bodies. Journal of Virological Methods 158, 171 ( Jun, 2009).
59. T. I.  Sorensen, G. G.  Nielsen, P. K.  Andersen, T. W.  Teasdale.
Genetic and environmental influences on premature death in adult
adoptees. The New England Journal of Medicine 318, 727 (Mar 24,
1988).
60. F. O. Vannberg, S. J. Chapman, A. V. Hill. Human genetic suscep-
tibility to intracellular pathogens. Immunological Reviews 240, 105
(Mar, 2011).
61. A. Bernstein, B. Pulendran, R. Rappuoli. Systems vaccinomics: the
road ahead for vaccinology. Omics: A Journal of Integrative Biology
15, 529 (Sep, 2011).
62. B. Pulendran, S. Li, H. I. Nakaya. Systems vaccinology. Immunity
33, 516 (Oct 29, 2010).
 12.
NUTRITIONAL GENOMICS
Zhenglong Gu, Kaixiong Ye, and Patrick J. Stover
INTRODUCTION
Nutrients and other food components are amongst the
most persistent, variable (both in terms of the nature and
abundance of the food supply), and essential environmen-
tal exposures for all life forms. The need for organisms to
balance constant nutrient “needs” with intermittent nutri-
ent “availability” has driven the evolution of sophisticated
yet distinct strategies to sense, store, and utilize individual
nutrients to achieve internal homeostasis. Indeed, epidemi-
ological, whole animal, and tissue- and cell-culture studies
validate nutrition’s pivotal role as an exogenous determinant
of health. Most common human chronic diseases, including
diabetes (type 2), metabolic syndrome, cardiovascular and
neurological disease, and many cancers are initiated and/
or accelerated by nutrient/food exposures. In the absence
of adaptation, nutrient deficiencies can impair the function
of transcriptional and metabolic networks, while nutrient
excesses can exceed their capacity and/or overwhelm the
buffering capability of the associated signaling pathways
that maintain homeostasis.
Nutritional genomics is a field that has emerged at the
interface of nutrition and genomics.1–3 Genomics is defined
as the “study of the functions and interactions of all the genes
in the genome, including their interaction with environmen-
tal factors.”4 Nutrients are essential environmental factors for
organismal survival; the term nutrient was defined as a
fully characterized (physical, chemical, physiologi-
cal) constituent of a diet, natural or designed, that
serves as a significant energy yielding substrate or
a precursor for the synthesis of macromolecules or
of other components needed for normal cell dif-
ferentiation, growth, renewal, repair, defense, and/
or maintenance or a required signaling molecule,
cofactor, or determinant of normal molecular struc-
ture/function and/or promoter of cell and organ
integrity.5
180
The interactions among nutrients and the genome are
seamless and essential features of organismal evolution.
They are fundamental for virtually all life processes; these
interactions affect both the primary sequence of DNA
(the genetic code), as well as the expression of the code
(Figure  12.1). Individual nutrients influence DNA muta-
tion rates in somatic cells, and more recently have been
shown to influence both the generation and the propaga-
tion of DNA mutations in the germline and thereby facili-
tate the generation of human genetic variation. Genomic
“signatures” can be found within DNA primary sequences
that validate the role of dietary components as selective
pressures throughout human evolution. The influence of
nutrients on the genome is not limited to DNA primary
sequences. Nutrients and metabolites function as signal-
ing molecules that enable networks to sense and respond
to their internal and external environments. In this regard,
nutrients can elicit transient alterations in gene expression
and/or influence more permanent and potentially heritable
whole-genome reprogramming events. The genome, in turn,
influences diet (Figure  12.1). Human genetic variation,
including variations in primary sequence and variations in
epigenetic programming, affects nutrient absorption and
utilization and thereby confers differences in food toler-
ances and, potentially, in nutrient requirements among
human individuals and human populations.
Nutritional genomics is a multidisciplinary field that
draws upon an extensive and rich foundation of knowledge
in nutritional anthropology, population genetics, nutri-
tional biochemistry, human clinical nutrition and metabo-
lism, human genetics and development, and nutritional
toxicology, among other disciplines. This new field emerged
as the sequence of human genomes and the genomes
of other organisms became available. Metabolic- and
nutrition-related disorders are complex traits with multiple
interacting environmental and genetic determinants. Both
diet and the genetic background of the individual modify
these disorders’ onset and severity. Therefore, nutritional
 Free ebooks ==> www.Ebook777.com
Food Intolerances Dietary Requirements
Dietary
components Human Genome
Genome evolution/Selection
Genome mutation rate
In-Utero Genome viability
Genome programming
Gene expression
Figure 12.1  Genome–nutrient interactions.
genomics, like most “-omics” fields, is focused on the biol-
ogy of the individual, but is distinguished by its unique
potential to advance our understanding of disease preven-
tion and healthy aging through manipulation of gene–diet
interactions. In addition, nutritional genomics has thera-
peutic applications through the rational design of dietary
interventions to manage chronic disease.
Advances in nutritional genomics research are antici-
pated to illuminate the mechanisms underlying the acute
and long-lasting diet/nutrition–genome interactions that
promote health and revolutionize both clinical and public
health nutrition practice, and culminate in: 1) genetically
informed nutrient- and food-based dietary guidelines for
disease prevention and healthy aging, 2) improved and/or
individualized nutritional therapeutic regimes for disease
management, and 3)  better targeted public health nutri-
tion interventions (e.g., micronutrient fortification and
supplementation) that maximize benefit and minimize
adverse outcomes within human populations. These objec-
tives will be met only with further development of the basic
science that underpins nutritional genomics, and effective
translation of this knowledge into nutrition practice in the
following areas:
1) Nutritional genetics, the identification, classification,
and characterization of human genetic variation that
modifies nutritional requirements and food tolerances;
2) Nutritional epigenetics, the modification of chromatin
structure (and hence gene expression) by diet, through
post-replication and post-translational modification
of DNA and protein, respectively, which serves to
program or reprogram biological networks, with
multigenerational consequences;
N utrition a l G e no m ic s   •   1 8 1
www.Ebook777.com
3) Nutritional systems biology, the application of systems
biology approaches in nutritional studies. It integrates
global data at multiple levels (genome, transcriptome,
proteome, metabolome, interactome, etc.) and utilizes
systemic analysis and model-based computational
simulation to study biological networks (e.g.,
regulatory, signaling, and metabolic networks) under
various physiological and nutritional states.
This chapter summarizes the underlying scientific prin-
ciples and preliminary advances in these three primary focus
areas of nutritional genomics.
N U T R I T I O N AL G E N ET I C S
Human populations and individuals within populations dif-
fer in their sensitivities to nutrient deficiencies and excesses.
Human genetic variation contributes to differences in physi-
ological responses to diet. The United Nations Educational,
Scientific and Cultural Organization (UNESCO) recog-
nized both the influence of human genetic variation on
human nutrition and the concept that nutrient utilization
and efficacy are unique characteristics of individuals, in the
Universal Declaration on the Human Genome and Human
Rights, Section A, Article 3, which states:
The human genome, which by its nature evolves, is
subject to mutations. It contains potentialities that
are expressed differently according to each individ-
ual’s natural and social environment including the
individual’s state of health, living conditions, nutri-
tion and education.
—UNESCO Document 27 V / 45, adopted by the Thirty-First
General Assembly of UNESCO, Paris, November 11, 1997
Indeed, human evolution is a continuous, albeit irregular,
process made manifest through the generation and expan-
sion of DNA mutations that permit survival amidst erratic
and unpredictable environmental exposures. Changes in
DNA primary sequences enable human evolution through
the generation of adaptive genetic variants that alter an
organism’s response to environmental challenges and hence
to its fitness. Human DNA primary sequence was origi-
nally estimated to differ by approximately 0.2–0.4% among
humans, mostly due to single-nucleotide polymorphisms.6,7
It was recently estimated that 1–3% of human genomes are
different considering various types of structural variations.8
These observed variations are partly products of historical
interactions among humans and their environment, includ-
ing dietary patterns.
T Y P E S O F G E N ET I C VA R I AT I O N
Differences in DNA primary sequence constitute a primary
molecular basis for human phenotypical variation, includ-
ing metabolic efficiency and disease susceptibility. Genomic
polymorphisms emerge through the sequential processes of
DNA mutation and expansion of the mutation within a
population. Environmental exposures can accelerate both
processes. Polymorphisms are classified according to the
origin and nature of the genomic mutation, and include
single-nucleotide polymorphisms (SNPs), micro- and
macrosatellite repeat sequences and structural variations,
such as repetitive element insertions and copy number
variations.9,10 SNPs are common nucleotide base-pair dif-
ferences in the primary sequence of DNA. As of 2011,
there were more than 50  million SNPs submitted to the
National Center for Biotechnology Information (NCBI)
dbSNP database (dbSNP build 137), about 38 million of
which had been validated.11 SNPs can be single base-pair
insertions, deletions, or substitutions of one base pair for
another. Nucleotide substitutions are the most common
polymorphism, whereas insertion/deletion mutations
occur at one-tenth that frequency.12 SNPs, like other poly-
morphisms, are usually defined as genetic variants that have
a frequency of at least 1% in human populations.13
Genetic variations can also result from the integration
and/or transposition of retroviral DNA.14 Approximately
50% of noncoding human DNA originates from transpos-
able elements that are highly mobile and contain repetitive
sequences.15–17 Retrotransposons are classified by size and
include long interspersed nuclear elements (LINEs) and
short interspersed nuclear elements (SINEs). About 10% of
the human genomic sequence consists of 280 base-pair Alu
SINE elements. Over 1,200 Alu elements integrated into
the human genome following early human migrations out
of Africa.18 Today, there are an estimated 1.4 million such
elements in the human genome, with a new Alu insertion
event occurring every 200 births.17 Alu elements are believed
to be catalysts for organismal evolution.17,19 They display
promoter activity, but their transcripts lack an open read-
ing frame and therefore are not translated into protein. Alu
insertions can alter genome stability and/or the function
of a single gene function near their site of integration. Alu
elements contain splice acceptor sites; therefore, their inte-
gration within an intron can lead to the generation of new
proteins through alternative splicing of gene transcripts.20
Because of their repetitive sequence and the high degree of
1 8 2   •   P rincip l e s o f G e no m ic M e dicin e
sequence similarities among family members, Alu insertions
can serve as nucleation sites for unequal intrachromosomal
and interchromosomal homologous recombination events
that result in chromosomal aberrations such as deletion and
translocation events.
Viral elements can confer new types of regulation to
existing genes, including regulation by essential nutrients,
but they may also interrupt genes or create new genes. Alu
elements can function as transcriptional silencers or activa-
tors that are responsive to nutrient status. Some Alu ele-
ments have retinoic acid (vitamin A) response elements
that bind nuclear receptor transcription factors and there-
fore confer retinoic acid responsiveness to genes that neigh-
bor the insertion site. Alu elements also contain multiple
CpG “islands,” a nucleotide sequence that attracts the DNA
methylation of the cytosine (C)  base. DNA methylation
typically serves to suppress transcription at that locus, and
the degree of methylation can be sensitive to dietary intake
of vitamins involved in one-carbon metabolism, including
folic acid.21
Alu insertions are also associated with metabolic dis-
ease. New Alu insertions may account for up to 0.1% of
human genetic disorders, including Apert syndrome, cho-
linesterase deficiency, and breast cancer.22 Alu-mediated
unequal homologous recombination events are respon-
sible for 0.3% of inherited human genetic diseases, such as
type 2 insulin-resistant diabetes and familial hypercholes-
terolemia.14,17 Such recombination events are rare because
Alu-mediated unequal homologous recombination events
are usually inhibited by CpG methylation of the element.
Copy number variations (CNV) involve changes of copy
number for DNA segments that are ~1 kb or longer.23,24
Usually, the insertion or deletion of transposable elements is
not classified as a CNV.24 The estimated genome-wide muta-
tion rate of CNV ranges from 1.7 x 10–6 to 1.0 x 10–4 per
locus per generation, which is 100 to 10,000 times higher
than nucleotide substitution rates.25 Due to the length of
DNA involved, CNV usually affects more nucleotides per
genome than SNPs.23,25 Because CNVs can modify gene
dosage, interfere with proper splicing, disrupt the coding
region, and alter regulation of a nearby gene, they can have
a significant functional impact on a genome, and therefore
are subject to selection.23,26 CNVs can be subject to puri-
fying (negative) selection by disrupting proper gene func-
tions, and to directional (positive) selection by contributing
to regional adaptation. Association studies have identified
CNVs that contribute to phenotypical diversity, such as dis-
ease pathology, drug sensitivity, steroid hormone, xenobiotic
metabolism, prostate cancer, nicotine metabolism, regula-
tion of food intake and body weight, neurodevelopment
and neurological disorders, colonic Crohn’s disease, toxin
resistance, coronary heart disease risk, Alzheimer’s disease,
HIV infection, and AIDS progression.10,24,27–29 Variations in
the copy number of the amylase gene resulted from dietary
adaptation during human evolution.30
D NA MU TAT I O N R AT E S A N D D I ET
The human genome is assembled from deoxynucleotide
monomers that are synthesized de novo from metabolic pre-
cursors derived from energy, amino acid, and one-carbon
metabolism. DNA mutations arise as a consequence of the
inherent chemical instability of DNA bases, errors associ-
ated with the fidelity of DNA replication and recombina-
tion, exposure to chemical oxygen radicals that are generated
during oxidative metabolism, as well as by numerous
genotoxic xenobiotics that are present in the food supply.
Therefore, some DNA mutations are unavoidable; however,
their deleterious impact is minimized by DNA repair sys-
tems that detect and correct most mutation events.
Environmental exposures such as nutrient deficiencies
or excesses, factors that increase cellular oxidative stress, or
genetic variations that modify the metabolism of dietary
components can accelerate DNA mutation rates by inducing
DNA modification reactions and/or by accelerating DNA
polymerase error rates. Many of the essential B-vitamins
(e.g., folate, niacin, flavin, vitamin B6, vitamin B12) and
minerals (e.g., zinc, iron) are required for nucleotide biosyn-
thesis; nutritional deficiencies can impair the synthesis of
nucleotide precursors and lower DNA synthesis rates (rates
of mitosis) and/or the fidelity of DNA replication (DNA
mutation rates). For example, folate deficiency inhibits
thymidylate (dTMP) synthesis, which increases incorpora-
tion of uridylate (dUTP) into DNA, resulting in increased
frequencies of DNA strand breaks.31–35 Furthermore, defi-
ciency of dietary antioxidants that scavenge chemical radi-
cals, or excesses of pro-oxidant nutrients such as iron, may
increase mutation rates.36–38 Other dietary components
affect DNA mutation rates by altering cellular redox states
or functioning as genotoxic radicals that chemically modify
purine and pyrimidine bases. Certain aflatoxins, a com-
mon class of natural xenobiotics found in soil molds that
contaminate certain foods, increase DNA mutation rates,
leading to the transformation of somatic cells and localized
cancer epidemics.39 However, only mutations that occur in
the germline contribute to a species’ heritable genetic varia-
tion. Mutations that have no functional consequences are
anticipated to be phenotypically silent and therefore selec-
tively neutral. The level of nucleotide diversity is a function
of the DNA mutation rate (estimated to be 2.5 x 10–8 on
N utrition a l G e no m ic s   •   1 8 3
average for autosomes in regions of the genome presumed
to be nonfunctional, including intronic and intergenic
regions).12,40
D I ET A N D MU TAT I O N E X PA N S I O NS
I N HUM A N P O P U L AT I O NS
There is increasing evidence that dietary challenges have
increased human genetic variation by driving the expansion
of rare gene variants in isolated human populations in iso-
lated geographic regions. Mutations that promote survival
in challenging dietary environments expand within popula-
tions through the generation of environmentally adaptive
gene alleles. Mutations that expand and reach a specific fre-
quency within a population contribute to genetic variation
as polymorphisms, and this expansion within human popu-
lations is the molecular basis for the evolution of genomes.
Germline mutation alone is necessary, but not sufficient, for
establishing genetic variation.
Not all genes “evolve” or change at the same rate. The
neutral theory of evolution (DNA mutation in the absence
of selection) does not account for the extent of amino
acid substitutions observed in mammalian genomes.41–44
Natural selection, which is the differential contribution of
genetic variants to future generations, is the only evolu-
tionary force that has adaptive consequences.45 Darwinian
selection favors the conservation and expansion of favor-
able mutations (by positive or balancing selection), and the
elimination of mutations that are deleterious (referred to
as “negative” or “purifying” selection). Not all genes evolve
at the same rate, because positive selection only acceler-
ates mutation fixation at defined loci within the genome.
Mutations that confer reproductive and/or survival advan-
tage within a single environmental context expand within
populations at higher rates than neutral mutations and
replace a population’s preexisting variation.
Mutations that alter physiological processes are under
constraint and subject to positive, balancing, or negative
selection. Patterns of genetic variation across the human
genome are affected by demographic history, mutation,
recombination, and, in some cases, selection.14 Although
protein-coding sequences are conserved among mammals
in general, rates of amino acid substitution vary markedly
among proteins compared to rates of synonymous substitu-
tion among genes (changes in the coding region of genes
that do not affect protein sequence).42 The proportion of
amino acid substitutions that result from positive selection
is estimated to be 35–45%.42 Mutations that alter amino
acid sequence, which affects protein structure and function,
can have physiological consequences that may be beneficial,
deleterious, or neutral, and thereby influence an organ-
ism’s fitness in specific environmental and dietary contexts.
Likewise, mutations that affect protein expression can alter
biological network outputs, leading to altered physiology.
Mutations can expand in the absence of selection and
contribute to metabolic disease. The rate of mutation fixa-
tion is a function of the effective population size, popula-
tion demographic history, and the effect of the mutation
on an organism’s fitness.14 Polymorphisms can expand and
become fixed within a population through the processes of
genetic drift or natural selection. Drift is a stochastic process
resulting from the random assortment of chromosomes at
meiosis. Because only a fraction of all possible zygotes are
generated and survive to reproduce, mutations can expand
through many generations by the random sampling of gam-
etes in the absence of selection.14 Drift is expected to have a
greater influence on genetic variation in small populations
expanding rapidly; drift in large, static populations is not
usually as significant. Genetic drift becomes relevant in large
populations undergoing “bottlenecks” (massive reductions
in population) or in founding events that have occurred dur-
ing human migrations; for instance, population groups that
include the Old Order Amish, Hutterites, and Ashkenazi
Jews.14 In such populations, rare disease alleles can expand
rapidly and increase the incidence of diseases, including
breast cancer, Tay-Sachs, Gaucher, Niemann-Pick, and
familial hypercholesterolemia.14 Although most human
genetic variations arose as a result of the neutral processes
of mutation and genetic drift, variation resulting from drift
rarely has physiological consequences in static environ-
ments. However, environmental shifts like alterations in the
food supply can challenge biological systems and convert
otherwise physiologically “silent” genetic variations into
functional gene variants. Relevant examples are discussed
below.
I D E N T I FI C AT I O N O F NU T R IT I O NA L LY
A N D E N VI RO N M E N TA L LY S E NS I T I VE
HUM A N A L L E L E S
Candidate gene approach
The vast majority of known functional polymorphisms that
contribute to food intolerances and metabolic disorders
were first identified as highly penetrant disease alleles from
epidemiological or clinical studies (Table 12.1). Candidate
genes were analyzed for genetic variation; their selection
as candidate genes was based on existing knowledge of
metabolic pathways and inference that their impairment
could result in metabolic phenotypes associated with a
1 8 4   •   P rincip l e s o f G e no m ic M e dicin e
particular disease state, or could affect biomarkers associ-
ated with that disease. Model organisms, including yeast,
Drosophila, Caenorhabditis elegans, and mice, have been
excellent resources to identify potential candidate genes
and to confirm their contribution to a metabolic pheno-
type. Other advancements, including the availability of
high-density SNP maps of the human genome have accel-
erated the identification of human disease alleles, including
low-penetrant alleles that may make relatively small con-
tributions to the initiation and/or progression of complex
disease.14 Furthermore, haplotype maps of human genetic
variation offer advantages for disease-associational studies
because they are simpler than SNP maps,46 but their util-
ity may be limited because of the variability in haplotype
diversity across candidate genes.47 The candidate gene
approach, while successful in identifying alleles underlying
monogenic traits,5,48 is limited by incomplete knowledge
of gene function, incomplete knowledge of transcriptional
and metabolic networks that suggest candidate genes for
analyses, and the multifactorial nature of most complex
human metabolic chronic diseases (most are polygenic
traits with multiple environmental components, which,
in isolation, make relatively minor contributions to the
disease phenotype). This is witnessed by the many incon-
sistent findings that have emerged within the nutritional
and genetic epidemiological literature, especially for the
involvement of low-penetrant genetic alleles in chronic
metabolic disease.49
Linkage analysis and association studies
Linkage analysis and association studies are two primary
methods to map causal alleles for human traits, including
diseases. Linkage analysis compares genetic markers across
the genome in normal and affected individuals from the
same family and determines whether certain markers are
inherited along with the trait. The tool has limited power
for research in complex diseases because sample sizes are
usually small, due to the limited number of meiosis events
within families.17,74,75 The genome-wide association studies
(GWAS) genotype a set of genetic markers, usually SNPs,
in a group of affected individuals and unaffected control
individuals to detect an association between a particular
genomic region and the specific traits of interest. Facilitated
by the HapMap project and the availability of large-scale
genotyping platforms, this approach has been widely used
to examine genome-wide genetic markers in samples with
thousands of or even tens of thousands of subjects. The list
of traits under GWAS study has been categorized in a pub-
lic database, and the new candidate genes generated from
 Table 12.1  CANDIDATE HUMAN DISEASE ALLELES THAT AFFECT THE
UPTAKE OR METABOLISM OF DIETARY COMPONENTS

FOOD COMPONENT GENE POLYMORPHIC REFERENCE NO.

ALLELE
Vitamins
Folate MTHFR A222V 76,258

CBS 844ins68 259

GCPII H475Y 260,261

Vitamin B12
MTR N919G 258

MTRR I22M 258

Vitamin D VDR many 262

Minerals
Iron HFE C282Y 154,263

Sodium CIC-Kb T481S 264,265

Lipids
APOB many 266,267

APOC3 many 268

APOE many 68

Alcohol
ADH/ALDH2 many 80,82

Carbohydrate
Lactose LCT promoter 64

Fructose Aldolase B many 66

Detoxification/Oxidative Stress
NAT1/NAT2 many 269,270

PON1 Q192R;L55M 271

Mn-SOD Ala(-9)Val 272,273

these studies provide novel hypotheses for disease initiation
and progression.
Evolutionary analysis
Genes that have undergone accelerated changes or evo-
lution display genomic signatures that can be identified
computationally. The identifiable genomic signatures
include the presence of an excess of rare variants within
a population (which can be indicative of a selective
sweep), large allele frequency differences among popula-
tions, and a common haplotype that remains intact over
long distances.19,45,50,51 The identification of polymorphic
alleles that have arisen as a result of historical selection
resulting from nutritional challenges offers the oppor-
tunity to identify genes that contribute to monogenic
metabolic disorders, as well as low-penetrant alleles
that contribute to complex metabolic disease.14,51 The
common disease–common variant hypothesis states that
disease-susceptibility alleles arose before humans migrated
out of Africa and therefore exists at high frequency across
all human populations.52,53 However, both “single-gene”
disorders, including cystic fibrosis and hemochromato-
sis, as well as complex diseases, can be associated with
geographically restricted populations because the alleles
arose after migrations out of Africa.7,19,51,54–56 Therefore,
although 85–90% of all human genetic variation is found
within populations and presumably arose prior to human
migrations, some of the 10–15% of variation among pop-
ulations probably arose from recent selective pressures
that contributed to both simple and complex disease.57,58
Comparison of genomic sequence divergence among
mammalian species enables the identification of ancient
selection throughout the process of speciation and genetic
divergence (Table 12.2). These approaches can identify
single genes or pathways within biological networks that

N utrition a l G e no m ic s   •   1 8 5
Table 12.2  DIET-RELATED GENES THAT DISPLAY Bitter-taste receptor
GENOMIC SIGNATURES OF ADAPTIVE EVOLUTION
Recognition of bitter taste may have conferred a selective
GENE SPECIES/FUNCTION REFERENCE NO.
advantage by deterring the consumption of plant toxins
lysozyme langur monkey 42,274,275
that often elicit the sensation of bitterness.61 The TAS2R16
ribonuclease langur monkey 42,275
gene encodes a G protein-coupled receptor that is activated
Cox4 Primates 276
by salicin in fruit, amygdalin in almonds, and many com-
LCT human lactose metabolism 63
mon β-glucopyranosides that elicit cyanogenic toxicity.
ADH1B human ethanol metabolism 81 The K172V polymorphism increases the receptor’s sensi-
ALDH2 human ethanol metabolism 83 tivity to cyanogenic glycosides and displays signatures of
HFE human iron homeostasis 77 positive selection. The adaptive allele arose in the Middle
PPARγ human nuclear receptor 45 Pleistocene era prior to human migrations out of Africa.61
PTC human bitter-taste receptor 99

TAS2R16 human bitter-taste receptor 61

Lactose and calcium metabolism
KEL human protein metabolism 51
Lactose metabolism requires the expression of lactase-
TRPV5 human calcium transport 51
phlorizin hydrolase, an enzyme encoded by the LCT gene.
TRPV6 human calcium transport 51
In most humans and other mammals, LCT expression
ABO human protein metabolism 51
declines after weaning, resulting in primary lactose intoler-
ACE2 human protein metabolism 51
ance. In some human populations, including those of north-
CYP1A2 human arylamine metabolism 277
west European descent and nomads of the Afro-Arabian
G6PD human NADP metabolism 92
desert region, LCT expression persists into adulthood
AGXT glyoxylate metabolism 85
and confers the ability to effectively digest dairy products.
SLC23A1 vitamin C transport 87 A  SNP was identified 14 kilobases upstream of the LCT
transcriptional initiation site in a cis-acting transcriptional
element. This SNP is enriched in individuals of northern
enabled adaptation. Similarly, analyses of human genomic European descent and displays genomic signatures of posi-
diversity among human populations can identify genetic tive selection.62–64 Its prevalence correlates with, but does
selection within the human species that occurred prior not fully account for, the persistence of LCT expression and
to and following human migrations out of Africa. These resistance to primary lactose intolerance throughout adult-
complementary approaches have permitted the identi- hood.64 The adaptive evolution of these polymorphisms in
fication of genes that have undergone accelerated and/ these populations may have been driven by the benefits of
or adaptive evolution (Table 12.2).41,51 Rapidly evolving milk consumption in cattle-herding populations, both as a
genes are inferred to have enabled adaptation and thus source of liquid in arid regions, and by prevention of rickets
became fixed in populations by positive selection, or are and osteomalacia in regions of low solar irradiation.50,63,65
subject to balancing selection, which maintains an allele at The requirement for efficient calcium absorption may also
an equilibrium frequency.59 Adaptive genes originate from have driven alleles for TRPV5 and TRPV6 to fixation in
region-specific selective factors and therefore are expected these same populations (Table 12.2).51
to concentrate in specific geographic regions where the
selection occurred.14 The geographic origins of an individ-
Fructose metabolism
ual can be predicted from genomic signatures of positive
selection to the degree that different selective pressures Hereditary fructose intolerance (HFI) is an autosomal
are operative across populations, but do not always cor- recessive disorder of fructose metabolism resulting from
respond to specific ethnic or racial groups, because races low fructose-1,6- aldolase activity, resulting in an accu-
are not homogenous.7,60 Many of the human alleles mulation of metabolic intermediate fructose-1-phosphate,
known to affect metabolism, food tolerances, or optimal which deregulates glycolysis. Twenty-five allelic variants of
nutrient intakes display signatures of positive selection aldolase B, the human liver isozyme, have been identified
(Table 12.2). Above is a summary of gene variants that are that impair enzyme activity by altering the catalytic proper-
known to affect diet, many of which display signatures for ties of the enzyme and/or protein stability.66 The accumula-
positive selection. tion of fructose-1-phosphate inhibits glycogen breakdown

1 8 6   •   P rincip l e s o f G e no m ic M e dicin e
and glucose synthesis, resulting in severe hypoglycemia
following ingestion of fructose. Individuals carrying poly-
morphic variants of aldolase B are asymptomatic in the
absence of fructose or sucrose consumption and can avoid
the recurrence of symptoms by remaining on a fructose- and
sucrose-free diet. Chronic fructose ingestion in infants ulti-
mately leads to hepatic and/or renal failure and death. The
prevalence of these variants differs throughout Europe; the
L288 delta C frameshift mutation is restricted to Sicilian
subjects. These aldolase B variants probably emerged in
populations through random drift; fructolysis is not an
essential metabolic pathway for humans, and fructose has
not been an abundant dietary component throughout most
of human history. However, the incidence of HFI intoler-
ance has increased since the widespread use of sucrose and
fructose as nutrients and sweeteners, providing an excel-
lent example of an environmental shift that has resulted in
the apparent conversion of normally nonpenetrant “silent”
aldolase B alleles into HFI disease alleles.67
Lipid metabolism
Apolipoprotein E (apoE) is a polymorphic protein that
functions in lipid metabolism and cholesterol transport.68
All human populations display apoE polymorphism.
There are the three common allelic variants, ε2, ε3, and
ε4, whose relative distribution varies among populations;
the frequency of the ε4 allele declines from northern to
southern Europe. These variant alleles encode proteins that
differ in their affinity both for lipoprotein particles and for
low-density lipoprotein receptors. The ε4 allele increases
the risk for late-onset Alzheimer’s disease and arterioscle-
rosis with low penetrance. Carriers of the ε2 allele tend to
display lower levels of total plasma cholesterol, whereas car-
riers of the ε4 allele, which may be ancestral, display higher
cholesterol levels. Therefore, serum cholesterol levels are
likely to be more responsive to low-fat and low-cholesterol
diets in carriers of the ε4 allele.69,70
One-Carbon metabolism
Folate-mediated one-carbon metabolism is required for
purine, thymidylate, and methionine biosynthesis, and it
affects genome synthesis, stability, and gene expression.71
Several polymorphic alleles have been identified as associ-
ated with metabolic perturbations that can confer both pro-
tection and risk for specific pathologies and developmental
anomalies.72 SNPs in MTHFR (A222V) and MTHFD1
(R653Q),73 which encode folate-dependent enzymes,
are associated with increased risk for neural tube defects;
N utrition a l G e no m ic s   •   1 8 7
the MTHFR (A222V) is protective against colon cancer
in folate replete subjects.74 The MTHFR A222V variant
protein has reduced affinity for riboflavin cofactors and is
thermolabile, resulting in reduced cellular MTHFR activ-
ity; its stability is increased when folate is bound.75 The
prevalence of the MTHFR allelic variant varies markedly
among human population and occurs with an allelic fre-
quency of nearly 40% in some Hispanic populations, but
it is mostly absent in African populations.76 However, it has
not been reported to display signatures of positive selection.
Although the biochemical role (if any) of these polymor-
phisms in the etiology of neural tube defects and cancer is
unknown, it is demonstrated that some carriers of MTHFR
variants require higher folate intakes than others to: 1) sta-
bilize the MTHFR protein, 2) lower the concentration of
the metabolic intermediate homocysteine, and 3) decrease
a women’s risk of bearing children with developmental
anomalies, including neural tube defects.76 The fortification
of the food supply with folic acid that occurs in many coun-
tries targets women of childbearing age for birth-defect
prevention, with genetically at-risk subgroups receiving the
most benefit.
Iron metabolism
Hereditary hemochromatosis is a recessive iron-storage dis-
ease that is prevalent in populations of European descent,
with an incidence of one in 300 persons. The HFE gene
is polymorphic and encodes a protein that regulates iron
homeostasis. A  common polymorphism, HFE C282Y,
emerged approximately 138 generations ago.19,77,78 This
SNP is associated with the disease phenotype in 60–100%
of Europeans. The HFE C282Y allele is not present in
Asian and African populations, despite the presence of
iron-storage diseases in those populations indicating that
other genes are associated with hereditary hemochroma-
tosis. Furthermore, the penetrance of the C282Y HFE
allele for the iron-overload phenotype varies widely among
homozygotes, with some individuals being asymptomatic,
indicating the presence of modifying alleles. The recent
expansion of this polymorphism may have conferred selec-
tive advantages in iron-poor environments77,78 or resistance
to microbial infection.79
Alcohol metabolism
Ethanol metabolism efficiency varies considerably among
human populations.80 Ethanol is oxidized to acetaldehyde
by the enzyme alcohol dehydrogenase, encoded by the ADH
genes. Acetaldehyde, a toxic metabolite, is subsequently
oxidized to acetic acid by the enzyme aldehyde dehydroge-
nase, which is encoded by ALDH2. Seven ADH genes that
are clustered on chromosome 4 encode proteins with dis-
tinct catalytic properties and tissue-specific expression pat-
terns. Two of the genes encoding class I enzymes (ADH1B
and ADH1C) are expressed in liver, function in systemic
ethanol clearance, and display functional polymorphism.
A variant ADH1B* 47His allele predominates in Japanese
and Chinese populations but is rare in European and
northern-African populations.81 The variant allele encodes
an enzyme with elevated enzyme activity, leading to more
rapid formation of acetaldehyde. The ADH1C*349Ile vari-
ant is found in Europeans while the ADH1B*369Arg vari-
ant is mostly restricted to individuals of African descent.
ALDH2 is also highly polymorphic, and Asian populations
carry a common dominant null allelic variant (E487K) and
develop a characteristic “flush” reaction when consuming
alcohol, resulting from acetaldehyde accumulation.82 ADH
and ALDH alleles that predominate in east Asian popula-
tions display signatures of positive selection, and the expres-
sion of these variant alleles results in elevated acetaldehyde
concentrations following alcohol consumption, which may
have conferred advantage by protecting against parasite
infection.83 The high frequency of ADH1B *47His in East
Asians may have resulted from adaptation to rice domesti-
cation and the consumption of fermented beverages.84
Glyoxylate metabolism and kidney stone disease
Kidney stone disease resulting from calcium oxalate
(CaOx) formation is common in Western populations
and results from multiple etiologies.85 Polymorphism in
the AGXT gene that encodes the enzyme alanine:  gly-
oxylate aminotransferase (AGT) results in an accumula-
tion of glyoxylate, a toxic intermediary metabolite that is
converted to oxalate. Oxalate does not undergo further
metabolism and accumulates as insoluble CaOx precipi-
tates that accumulate in the kidney and urinary tract. There
are two major precursors of glyoxylate: glycolate, which is
present in plant-based foods, and hydroxyproline, which is
present in meat collagen. Glycolate metabolism occurs in
the peroxisomes, while hydroxyproline is metabolized in
mitochondria. Among mammals, the intracellular localiza-
tion varies among carnivores and omnivores; it is primar-
ily peroxisomal in herbivores, mitochondrial in carnivores,
and both mitochondrial and peroxisomal in omnivores. In
humans, AGT is usually localized to peroxisomes. A com-
mon AGT variant, Pro11Leu, decreases enzyme activity by
about 70% but also results in the formation of a mitochon-
drial leader sequence and 5% of AGT protein localization
1 8 8   •   P rincip l e s o f G e no m ic M e dicin e
in mitochondria. This variant has been proposed to confer
advantage to meat-eating populations but is detrimental to
vegetarians. It displays evidence for positive selection.85 The
allelic frequency varies among human populations and cor-
relates with historical dietary patterns; it is present at a fre-
quency of 28% in Saami and 2.3% and 3% in Chinese and
Indian Hindus respectively.
Vitamin C transport
The individuality of vitamin C needs was first recog-
nized in 1967.86 Vitamin C is required for the function
of at least 8 mammalian enzymes and can scavenge reac-
tive oxygen species.87 There are two genes that encode
sodium-dependent vitamin C transporters, SLC23A1 and
SLC23A2. These genes resulted from an early gene dupli-
cation event but appear to have acquired distinct func-
tions. SLC23A1 is responsible for intestinal and renal
absorption of vitamin C and therefore has the potential
to affect whole-body vitamin C accumulation.87 The over-
all mutation rate of these genes is similar, as evidenced by
the similarity in their non-synonymous substitution rates.
However, four population-specific non-synonymous sub-
stitutions are seen in SLC23A1 that arose after the migra-
tions out of Africa, indicating a potential role for selection
in the expansion of these variant alleles.87 Only synony-
mous substitutions are observed in SLC23A2, and deletion
of the orthologous slc23a2 in mice is lethal, indicating the
critical, non-redundant function of this transporter that
seems to be under selective constraint. A common SNP in
SLC23A2, rs1279386, has been linked to plasma vitamin
C concentrations, with carriers of the GG genotype exhib-
iting lower plasma vitamin C concentrations than do car-
riers of other genotypes.88 The effects of non-synonymous
SLC23A1on vitamin C transport or physiology have not
been investigated.87
Energy metabolism
The “thrifty gene” hypothesis was first proposed over
40  years ago to account for the epidemic of type 2 dia-
betes observed in non-Western cultures that adopted
Western-style diets and lifestyles.89,90 The hypothesis states
that exposure to frequent famine selected for gene variants
that enabled the more efficient conversion of food into
energy and fat disposition during times of unpredictable
and sometimes scant food supplies. The putative adapta-
tions also may have resulted in more efficient adaptations
to fasting conditions (e.g., more rapid decreases in basal
metabolism) and/or physiological responses that facilitate
excessive intakes in times of plenty. Conclusive genomic
data have not yet supported this hypothesis.90,91
Starch Digestion
A well-known example for the adaptive role of CNV
is involved in starch digestion. The salivary amylase
gene (AMY1) shows extensive variation in copy num-
ber among individuals and between populations. It
was also demonstrated that the gene copy number is
positively correlated with the protein level. Populations
consuming a high-starch diet, such as agricultural popu-
lations of European Americans and Japanese, and Hadza
hunter-gatherers, who rely extensively on starch-rich
roots and tubers, have higher copy number of AMY1
than populations consuming a low-starch diet, such as
hunter-gatherers in the rainforests and near the Arctic
Circle. Comparison with other great apes, chimpanzees,
bonobos, New World monkeys, and Old World monkeys
indicate that the increased copy number of AMY1 origi-
nated in the human lineage. The low amount of nucleo-
tide divergence among different gene copies indicates a
recent origin that may be within the timeframe of mod-
ern human origins (~200,000 years ago). Taken together,
the copy number variations of the AMY1 gene among dif-
ferent populations might represent regional adaptation to
diets with varying starch content, an interesting example
demonstrating the role of diet in modulating the human
genome.30
Oxidative metabolism
Variations that have an impact on human nutrition and
metabolism may have arisen independently of direct
nutritional challenges. The enzyme glucose-6-phosphate
dehydrogenase is solely responsible for the generation of
reduced nicotinamide adenine dinucleotide phosphate
(NADPH) in red blood cells and therefore is required
to prevent oxidative damage. Variants with low activ-
ity resulting from amino acid substitutions, including
the G6PD-202A allele, are enriched in sub-Saharan
African populations and arose 2,500 to 6,500  years
ago.92 Presumably, this allelic variant became enriched in
populations as a result of balancing selection because it
conferred resistance to malarial disease in heterozygous
females and hemizygous males.93,94
These examples illustrate the role of environmental
exposures, including pathogens and dietary components, as
selective forces that facilitated the expansion of alleles that
alter the utilization and metabolism of dietary components.
N utrition a l G e no m ic s   •   1 8 9
Adaptive alleles may become recessive-disease alleles, or
disease alleles even in heterozygote individuals, when the
environmental conditions change profoundly, such as those
brought about by the advent of civilization and agriculture,
including alterations in the nature and abundance of the
food supply.41,42,45,50,95–99 Adaptive alleles may be respon-
sible for the generation of metabolic disease alleles both
within and across ethnically diverse human populations,
and therefore are strong, nonbiased candidate genes for
disease association studies:  the interacting and modifying
environmental factors can be inferred from the nutrients
and/or metabolites that are known to interact with the gene
product.14
N U T R I T I O N AL E P I G E N ET I C S
Traits can be inherited from one generation to the
next through both genetic and epigenetic mechanisms.
Classical genetic inheritance refers to the transmission of a
DNA primary sequence from one generation to the next.
Concordance among monozygotic twins illustrates the pre-
dominant yet non-exclusive contribution of DNA primary
sequence to human phenotypes; other modes of inheritance
must also be operative.100,101 Epigenetics refers to the inheri-
tance of traits through mechanisms that are independent
of DNA primary sequence. There are now many examples
demonstrating the inheritance of gene expression patterns
and/or levels independent of DNA primary sequence, and
such differences can elicit phenotypical differences among
individuals, including monozygotic twins through multiple
generations.100
Interest in the relationships among epigenetic events
and human nutrition was ignited by the “fetal origins of
adult disease” hypothesis, originally put forward by David
Barker and colleagues.102 Barker proposed that nutrition
acts very early in life to program risk for adverse outcomes
in adult life (Figure 12.2). The notion that phenotypical
plasticity was associated with in utero environmental expo-
sures had been validated in the toxicology literature, but
not well considered in the nutrition literature.103,104 Until
recently, Barker’s hypothesis was supported only by epide-
miological associations among early nutritional exposures
and increased risk in adulthood for obesity, hypertension,
and insulin resistance, which are the antecedents of adult
chronic disease, diseases that include cardiovascular disease
(CVD), diabetes, and metabolic syndrome.102,105 But now
there is an emerging, basic science literature that supports
the concept that fetal environment can, in fact, “program or
reprogram” the fetal genome with lifelong consequences.103
The human genome has evolved in the context of a nutrient
environment that was often scanty and always unpredict-
able. Hence, organisms developed the capacity to “sense”
and “adapt” to the food supply. These genomics adapta-
tions have been referred to as “metabolic imprinting” or
“metabolic programming.” Such adaptations occur within
critical windows in development, are seemingly irreversible,
and permit in utero survival in the context of a suboptimal
nutrient environment, but they may predispose the affected
individual to metabolic disease in adulthood.106–111
Waterland and Garza described specific criteria to dif-
ferentiate adaptive metabolic imprinting phenomena from
toxicological responses that resulted in permanent genomic
alterations during the affected individual’s lifetime. These
criteria include:  1)  a susceptibility window limited to a
critical ontogenic window in development, 2) a persistent
effect lasting through adulthood, 3) a specific and measur-
able outcome, and 4)  a dose–response or threshold rela-
tionship between a specific exposure and an outcome.111
Mechanisms for metabolic imprinting are beginning to be
understood and modeled in animal systems; some exam-
ples are illustrated below. Metabolic programming mecha-
nisms are more complex than those associated with toxic
or deficiency states. Whereas many teratogens are exog-
enous agents that disrupt biological processes, metabolic
imprinting is a conserved and adaptive response that opti-
mizes biological function in one life stage through genomic
mechanisms that result in permanent functional character-
istics. These imprints, however, may prevent or limit the
range of other adaptive mechanisms that are protective in
subsequent life stages when environments change, such as
in a transition from “dearth” to “surplus.” Thus, once the
program is established, a system’s buffering capacity is lim-
ited. The novelty of sustained surplus food present in most
Western cultures may explain some of the limited response
capability apparently at the core of the present obesity/type
2 diabetes epidemic.112
Genome programming results from chemical modifi-
cations of chromatin, either DNA or histone proteins, at a
specific locus that leads to programming of transcriptional
Risk phenotype
Early nutrition
obesity
experiences
hypertension
insulin resistance
“Program”
“Imprint”
Figure 12.2 
The fetal origins of disease hypothesis. Fetal environmental exposures, especially nutritional, act in early life to program risk for adult
health outcomes.
1 9 0   •   P rincip l e s o f G e no m ic M e dicin e
and metabolic networks.113 DNA is modified by meth-
ylation of cytosine deoxyribonucleotides present in the
sequence CpG. Cytosine methylation is usually associ-
ated with gene silencing; methylcytosine is bound by
methylcytosine-binding proteins that heterochromatize
DNA and hence silence the gene. Histone proteins are
essential components of chromatin and are subject to mod-
ification by methylation, phosphorylation, ubiquitination,
ribosylation, and acetylation, all of which modify gene
transcription efficiency and can influence DNA stability.114
Alterations in DNA and histone methylation constitute
the epigenetic signatures that enable genome programming
because of their potential connections to metabolic net-
works, chromatin structure, and transcriptional networks.
Furthermore, DNA and histone methylations are meta-
stable, heritable, and alter genome expression and stabil-
ity. Methylation is a higher order genomic signal that can
override transient metabolic or hormonal signals such as
the regulation of transcriptional networks through nuclear
receptors (e.g., vitamin A, vitamin D, steroid hormones).
The molecular mechanisms that describe the interac-
tions among nutrients, metabolism, and gene expression/
genome programming are mostly unknown. Following
are two illustrative examples of genome programming by
nutrition.
M AT E R NA L F O L AT E , O N E - C A R B O N
M ETA B O L I S M , FETA L G E N O M E
P RO G R A M M I N G, A N D IN U TE RO S U RVI VA L
Folate is a B-vitamin and a family of metabolic cofactors that
carry and chemically activate one-carbon units for the de novo
synthesis of purine nucleotides and thymidylate (dTMP),
and for the remethylation of homocysteine to methionine, a
metabolic network known as “folate-mediated one-carbon
metabolism” (Figure 12.3). Methionine can in turn be ade-
nosylated to form S-adenosylmethionine (AdoMet), which
is a cofactor for numerous cellular methylation reactions,
including histone and DNA methylation.72 Impairments in
this metabolic network by nutritional deficiencies or highly
Adult onset disease
CVD
diabetes
metabolic syndrome
 Free ebooks ==> www.Ebook777.com
penetrant SNPs increase risk for pathologies such as cancer
and cardiovascular disease, and developmental anomalies
such as spontaneous abortion (SA) and neural tube defects
(NTDs).72 Folate supplementation can reduce the risk for
these disorders; maximal benefit is achieved in genetically
susceptible individuals and populations.
Methionine and dTMP synthesis are the most vul-
nerable pathways within the network; their impairments
compromise the fidelity of DNA synthesis and cellular meth-
ylation reactions.72,115;31 AdoMet-dependent methyltrans-
ferases, including histone and DNA methyltransferases, are
subject to product inhibition by S-adenosylhomocysteine
(AdoHyc), which accumulates during folate deficiency
(Figure 12.3).116–118 Hence, methylation of chromatin is sen-
sitive to the efficiency of the one-carbon metabolic network;
methyltransferases sense the efficiency of the folate metabolic
network because their activity is determined by the cellular
AdoMet/AdoHyc ratio, otherwise known as the “methyla-
tion potential” of the cell.117,118 Global genomic methylcyto-
sine content is highly sensitive to the AdoMet/AdoHyc ratio,
which can affect both gene expression and DNA stability.119
The mechanisms underlying folate-associated patholo-
gies, including NTDs, SA, and cancers are assumed to be
the result of insufficient flux through the dTMP and/or
AdoMet synthesis pathways.120,121 Therefore, the etiology
involves either impairments in genome synthesis (mitotic)
rates, genome stability and/or methylation-sensitive
gene expression.72 Numerous studies have indicated that
10-formylTHF PURINES
TS
methyleneTHF dTMP
MTHFR
5-methylTHF
homocysteine
Methionine
AdoHyc
AdoMet
Methylation Reactions
DNA
Histones
Figure 12.3 
Folate-mediated one-carbon metabolism. Tetrahydrofolate
(THF)-mediated one-carbon metabolism is required for the
synthesis of purines, thymidylate, and methionine. MTHFR,
methylenetetrahydrofolate reductase; TS, thymidylate synthase;
AdoMet, S-adenosylmethionine; AdoHcy, S-adenosylhomocysteine.
N utrition a l G e no m ic s   •   1 9 1
www.Ebook777.com
folate-dependent dTMP synthesis, catalyzed by thymi-
dylate synthase (TS) and 5-methylTHF synthesis (leading
to Adomet synthesis) catalyzed by methylenetetrahydrofo-
late reductase (MTHFR), are competitive pathways within
the one-carbon network.122 They compete for a limiting
pool of the cofactor methylenetetrahydrofolate (methy-
leneTHF) (Figure 12.3).72,117,118 This metabolic competition
is unbalanced by the previously described MTHFR A222V
polymorphism. This functional SNP reduces MTHFR
activity and has two effects on the network. It impairs the
remethylation of homocysteine to methionine, thereby
reducing global DNA methylation and thus also influences
gene expression.123,124 It also increases the conversion of
deoxyuridine monophosphate (dUMP) to deoxythymidine
monophosphate (dTMP).125 These changes in the network
are associated with increased risk for spontaneous abortions
(SA) and NTDs, but decreased risk for adult colon cancer,76
illustrating that optimal network function or outputs differs
between the fetal and adult environments. Identifying the
precise mechanism for folate-related pathologies in experi-
mental systems is challenging because any factor, genetic or
environmental, that influences the metabolic competition
for methyleneTHF may simultaneously alter the efficiency
of both dTMP and AdoMet synthesis (Figure 12.3).
Alterations in one-carbon metabolism, and the AdoMet
cycle in particular, can have dramatic effects on genome
methylation. Both genome-wide and allele-specific DNA
methylation are influenced by alterations in folate metab-
olism.126 DNA hypomethylation induced by folate defi-
ciency alters transcription of genes regulated by promoter
methylation, including tumor suppressor genes,126,127 and
enables interchromosomal recombination events through
common retroviral repeat sequences whose activity nor-
mally is silenced by methylation.128 Patients with hyper-
homocysteinemia, a clinical state that results from the
inability to effectively metabolize homocysteine, accu-
mulate cellular AdoHyc and exhibit alterations in gene
expression. Patients exhibit DNA hypomethylation and a
homocysteine-dependent shift from monoallelic to biallelic
expression of genetically imprinted genes, including H19.
Folate supplementation in these patients restores homo-
cysteine levels to baseline, reverses global DNA hypometh-
ylation, and restores monoallelic expression of imprinted
genes.129 Interestingly, SNPs in the H19 gene are associated
with cord blood Isulin Growth Factor II (IGF-II) levels and
birth size.130
Folate-mediated alterations in genome methylation can
be set irreversibly or “imprinted” during early development.
In the viable yellow agouti (Avy) mouse model, maternal diet
determines the coat color of offspring.131 This mouse strain
contains an Intracisternal A-particle (IAP) element, which
is a retrotransposon, that integrated into a 5′ exon of the
agouti gene, resulting in cryptic and constitutive expression
of the agouti gene. This aberrant expression of the agouti
results in a “yellow” coat color and an obesity phenotype.132
The IAP retroviral element also attracts DNA methylation
to that locus, and the degree of methylation determines the
expression level of the agouti gene.
The Avy mouse is sensitive to maternal folate and
one-carbon status during gestation. Within a critical win-
dow in development, maternal diet determines the density
of cytosine methylation at the agouti locus and hence the
level of agouti gene transcription, coat color, and propen-
sity for obesity.133 The methylation patterns and subsequent
effects on coat color and, presumably, associated metabolic
characteristics are maintained throughout the lifetime of
experimental animals21 and are heritable.132 The identifi-
cation of other genes that are influenced by alterations in
the AdoMet/AdoHyc ratio through chromatin modifica-
tions, and the critical developmental windows that enable
genome programming, are essential to elucidate the mech-
anisms of folate-related pathologies and developmental
anomalies. Equally important, this mouse model illustrates
the concept that epigenetic modifications in the developing
embryo induced by maternal diet can “rescue” deleterious
genetic insults, such as retroviral insertions in gene promot-
ers, and restore the “normal” phenotype.
The ability of maternal folate and one-carbon sources
to compensate or “rescue” genetic deficiencies may not be
limited to the Avy mouse model, and therefore may have
implications for women taking nutritional supplements
during pregnancy at intake levels that exceed dietary recom-
mendations. Fetal genotypes that cannot support basic bio-
logical processes in the embryonic and fetal stages usually
are eliminated. In primates, this is achieved by spontaneous
abortion (miscarriage). Humans may be unique compared
to other mammalian species in their high rates of fetal
loss134; high SA rates may be a selective pressure that accel-
erates the expansion of polymorphic alleles within human
populations. Approximately 75% of human conceptions are
lost spontaneously before term; 80% of all SA occur within
the first trimester.135–137 It is estimated that half of SA occur
before the first three weeks of gestation and generally are
unnoticed; many embryos fail to implant in the uterus.138
Risk for SA increases and fertility decreases in women
over the age of 30  years.139,140 Although the etiologies of
SA are generally not established, the etiologies of most SA
are likely to be multifactorial. Many SA fetuses have struc-
tural and/or genetic anomalies.135,141 Potential inducers of
SA include maternal immune responses; fetal genotypes;
1 9 2   •   P rincip l e s o f G e no m ic M e dicin e
maternal and fetal endocrine, nutritional, or hormonal
imbalances; maternal and fetal infections; and endome-
triosis (Table 12.3).139 Few specific environmental risk fac-
tors for SA have been identified, but known factors include
low maternal folate status, diabetes (type 1), and elevated
homocysteine (which most often results secondarily to pri-
mary or conditioned folate deficiency).142–148 For example,
variants in the transcobalamin II (TC-II) gene are associ-
ated with sporadic and recurrent miscarriage.149 TC-II plays
a role in the delivery of vitamin B12 to peripheral tissues,
indicating a role for maternal vitamin B12 status, and the
genes responsible for its processing and utilization, in the
etiology of miscarriage.
Developmental anomalies (DA) and SA have both inde-
pendent and shared etiologies (Table 12.3).135 The under-
lying mechanisms for over 75% of DA are unknown and
assumed to be multifactorial; only 15% of those whose etiol-
ogies have been identified are solely genetic, including chro-
mosomal abnormalities or autosomal/sex-linked genetic
disease. Ten percent of DA are attributed to environmental
factors, 4% of which are attributed to disruptions in mater-
nal/fetal metabolism or suboptimal nutrition that includes
micronutrient under-nutrition, starvation, Phenylketonuria
(PKU), diabetes, alcoholism, etc. Infectious agents account
for another 4% of DA; mechanical disruption accounts
for 2%; and known chemical/prescription toxins account
for less than 1% of DA.135 Polymorphic variants of two
genes that encode folate-dependent enzymes, MTHFS
A222V and MTHFD1 R653Q, are associated with risk for
developmental anomalies, including neural tube defects.73
Interestingly, human alleles associated with developmen-
tal anomalies that encode folate-dependent metabolic
enzymes are not in Hardy-Weinberg equilibrium in some
studies (alleles are not inherited at the expected frequency),
consistent with evidence that elevated homocysteine is a
risk factor for spontaneous miscarriage and decreased fetal
viability.71,73,147,150,151
The concept that embryos can be rescued by maternal
nutritional status, or that “good diet hides genetic muta-
tions,”152 is suggested by numerous examples of nutritional
rescue or compensation (viability or phenotype) of gene dis-
ruptions through diet in mice and yeast.71,72,153–157 Individual
nutrients can rescue severe genetic lesions in mice when
administered in supra-physiological levels during critical
developmental windows. Maternal retinoic acid adminis-
tration between 7.5 and 9.5 days post-conception rescued
deafness and inner ear development in Hoxa1–/– mice,153
and folic acid can rescue skeletal defects associated with
deletion of a Hox gene, as well as neural tube defects in mice
that have no evidence of disrupted folate metabolism.152
 Table 12.3 MATERNAL RISK GENOTYPES AND REPRODUCTIVE OUTCOMES

GENE VARIANT PATHWAY FETAL RISK REFERENCE NO.

MTHFR V222A one-carbon metabolism SA 278

NTD
Down syndrome
Adult CVD
MTFD1 one-carbon metabolism NTD 73

TC II (transcobalamin) vitamin B12/one-carbon metabolism SA 143

NTD
IL6 (-174 G--> C) Cytokine SA 279

IFN-gamma 874 A--> T Cytokine SA 280, 281

IL1RN*1 Cytokine SA 282

IL1RN*2 Cytokine preterm birth 282

CYP17A2 steroid biosynthesis SA 283

CYP1A1*2A phase 1detox SA 284

PR*2 progesterone receptor SA 285

GSTM1 phase 2 detox SA 286

prothrombin G20210A Clotting SA 287

Factor V G1691A Clotting SA 287

Nos3B vascular function SA 288

PGM1*2 phosphoglucomutase SA 289

The most comprehensive studies have been performed
in yeast. Gene-deletion studies indicate that 80% of yeast
genes are nonessential for survival under laboratory condi-
tions. A  recent examination of yeast metabolic networks
using an in silico model revealed that culture conditions,
especially the use of nutrient-rich culture media, can com-
pensate for the disruption of 37% to 68% of the organism’s
genes. In microbial systems, the maintenance of enzymatic
flux under highly diverse environmental conditions appears
to be a primary selective pressure that maintains gene
sequence, starvation being among the most common envi-
ronmental stresses.158 The concept of nutritional rescue of
genetic mutations is exceptionally salient to human moder-
nity because of the unprecedented degree to which we
can manipulate our nutritional environments. The inborn
error of metabolism phenylketonuria provides the classical
example for the effectiveness of dietary manipulations in
modifying deleterious phenotypes resulting from genetic
mutations that alter metabolism. Restriction of phenylala-
nine from the diet can prevent severe cognitive deficits in
children with mutations in the phenylalanine hydroxylase
gene.159 Likewise, maternal folic acid supplementation and
fortification of the food supply with folic acid reduces the
occurrence and recurrence of neural tube defect–affected
pregnancies in genetically susceptible populations.160
Clearly, maternal folate and other methyl donor supple-
mentation alters the methylation status of targeted alleles
in the mouse embryo, and methylation patterns and subse-
quent effects on gene expression persist throughout adult-
hood.21 Epigenetic phenomena may provide mechanistic
insight into the many observational studies that associate
risk for adult chronic disease with maternal nutrition and
embryonic nutrient exposures (as proposed by Barker and
colleagues).111
G LU C O C O RT I C O I D S A N D M ETA B O L I C
D I S E A S E : MO D I FI C AT I O N O F T H E
P L AC E N TA L BA R R I E R BY M AT E R NA L
NU T R IT I O N
The consequences of fetal glucocorticoid (GC) exposure
on adult chronic disease provide some of the best support-
ing evidence for the fetal origins of disease hypothesis.161–164
Complementary human clinical and animal studies have
revealed the long-term consequences and associated mech-
anisms of fetal GC exposure (Figure 12.4).165–167 Fetal GC
levels are maintained at low concentrations relative to mater-
nal concentrations, primarily through the action of placen-
tal 11β-hydroxysteroid dehydrogenase type 2 (11β-HSD2),
which catalyzes the oxidative inactivation of cortisol and

N utrition a l G e no m ic s   •   1 9 3
 Initiating signal
Low maternal protein
Reduced 11β-HSD2 expression
Loss of placental GC barrier
Loss of 11β-HSD2 activity
–– inhibitors
–– variation in populations
–– gene disruptions(humans/mice)
Maternal/Fetal GCT herapy
GR knock-out in mice
PEPCK over expressing mice
Figure 12.4  Programming of the HPA axis by glucocorticoids.
corticosterone.168 Elevated fetal exposures to GC dur-
ing late gestation169 (which can result from 11β-HSD2
inhibitors, rare mutations in the human 11β-HSD2 gene,
or large existing variation in placental 11β-HSD2 activity
among humans) have lifelong consequences for the fetus,
including low birth weight (IUGR), elevated plasma GC,
hypertension, hyperglycemia, insulin resistance, hyperin-
sulinemia, and anxiety.164 IUGR and preeclampsia also are
associated with elevated cortisol and low birth weight.170,171
Furthermore, elevated GC and the metabolic syndrome
(i.e., the combination of type 2 diabetes/insulin resistance,
dyslipidemia, and hypertension) are also characteristic of
Cushing’s syndrome.162
Interestingly, low maternal dietary protein intake dur-
ing gestation causes a specific loss of placental 11β-HSD2
expression as well as fetal outcomes similar to those
observed from elevated fetal GC exposure.172 11β-HSD2
activity in placenta correlates with birth weight,173 and dis-
ruption of the murine gene encoding 11β-HSD2 reduces
birth weight.164 Similarly, obstetrical GC therapy to acceler-
ate lung development prior to anticipated preterm deliveries
increases risk for reduced fetal birth weight, and long-term
susceptibility to hypertension, hyperglycemia, cardiovascu-
lar disease, and increased hypothalamic-pituitary-adrenal
1 9 4   •   P rincip l e s o f G e no m ic M e dicin e
Programming
IncreasedFetalGC
GC Induced events
–– Small placenta
–– CNS defects
–– Attenuated HPA axis feedback sensitivity
–– Altered GR promoter methylation/expression declines
–– Altered dopaminergic programming
–– Increased PEPCK expression
(in adulthood/2nd generation)
Outcomes
–– inhibited insulin suppression of –– Low Birth weight (IUGR)
gluconeogenesis
–– increased insulin –– Elevated plasma GC in adulthood
–– glucose intolerance –– Hypertension
–– Hyperglycemia
–– Insulin resistence
–– Hyperinsulinaemia
–– Anxiety
axis (HPA) activity. These disorders persist not only into
adulthood, but also into the next generation.174
GCs are steroid hormones that serve as ligands for the
glucocorticoid receptor (GR), a member of the nuclear
receptor superfamily.175 GRs are complexed with heat
shock proteins and do not affect transcription in the
absence of bound ligand. With GC bound, GR remod-
els chromatin through transient interactions that recruit
remodeling proteins to a defined locus and “open” chro-
matin thereby enabling transcription factor binding.175
GR also may recruit transcription factors directly to the
transcriptional preinitiation complex,175 and may target
promoter demethylation.176 GCs are required for normal
central nervous system (CNS) development, apoptosis,
and synapse formation.177 Both GR and 11β-HSD2 are
present in the developing brain; 11β-HSD2 expression
begins to dissipate and brain GC accumulates from week
19 to 26 during end stages of neurogenesis.173,178 Lifelong
consequences associated with fetal GC exposure may
result from premature GR-mediated chromatin remodel-
ing in the hippocampus.164 Interestingly, GR programming
resulting from fetal GC exposure, and its deleterious reper-
cussions, can be erased in the adult animal by treatment
with histone deacetylase inhibitors.164
 GC homeostasis is maintained by the HPA axis, which
is imprinted or programmed by fetal GC exposures dur-
ing gestation. Plasma GC concentrations are normally
regulated by a feedback loop that involves GR in the hip-
pocampus. Persistent prenatal GC exposure decreases fetal
GR expression, and these expression levels can be “set” or
“memorized,” with lifelong consequences. Low hippocam-
pal GR levels increase adult corticosterone plasma levels
and thereby may reinforce the decreased GR expression lev-
els. Maternal undernutrition can elicit the same effect, pre-
sumably by decreasing placental 11β-HSD2, and thereby
program hypertension and hyperglycemia in the fetus. Fetal
GC exposure also affects the dopaminergic system179 and
alters amygdala function, including the regulation of fear
and anxiety.180 In addition, maternal GC exposure affects
glucose and insulin homeostasis181 by programming hepatic
Phosphoenolpyruvate carboxykinase (PEPCK) levels with
effects that persist into adulthood182,183 and are carried to
the next generation.174
Interestingly, GR levels also can be imprinted epigeneti-
cally and postnatally in rodents through maternal behavior,
including handling, licking, and grooming.161 At day 20
post-conception, the GR exon 17 promoter is unmethyl-
ated. One week after birth, offspring from low licking/
grooming-arched-back nursing mothers uniformly exhib-
ited methylated CpG islands in the NGFI-A cis-element
located within the exon 17 promoter and elevated GR
expression in the hippocampus, whereas offspring from
high licking/grooming-arched-back nursing mothers rarely
methylated this sequence.184,185 These elevations in GR
expression lower plasma GC levels and the stress response
through the HPA axis throughout the animal’s lifetime. In
humans, elevated maternal choline intake, which increases
the cellular methylation capacity, results in elevated placen-
tal promoter methylation of the cortisol-regulating genes
corticotropin-releasing hormone and glucocorticoid recep-
tor, and 33% lower concentration of cortisol cord plasma.186
N U T R I T I O N AL SYS T EMS B I O L O GY
All biological processes, including gene transcription,
protein synthesis, as well as signaling and metabolic path-
ways, are interrelated, interdependent, and function as a
dynamic and complex cellular network. Food and nutrient
intakes, as well as other environmental exposures, modu-
late life processes by interacting with biological networks.
The impact of such external inputs is rarely limited to their
primary targets, due to the interconnectedness of pathways
within the networks.187,188 Human health and disease are
N utrition a l G e no m ic s   •   1 9 5
manifestations of the functioning or malfunctioning of bio-
logical systems. Therefore, in contrast to traditional reduc-
tionist approaches, a systems framework is fundamental to
understanding the concept of life and disease, and to pro-
vide evidence-based guidance for effective disease preven-
tion and management. This systems perspective has been
recognized for almost one century,189 but it was not until the
beginning of the twenty-first century that systems biology
emerged as a new discipline.190,191 The remarkable growth
of systems biology during the past decade was enabled by
the rapid development of technologies that permit global
cell measurements and computational analyses. Global cell
measurement technologies at the levels of the genome, tran-
scriptome, proteome, metabolome, and interactome have
opened the possibility of enumerating all players within
the biological networks and all types of interactions among
them. The availability of comprehensive high-throughput
data sets, in the context of long-accumulated biochemical
knowledge and advanced mathematical and computational
modeling approaches, permits comprehensive construction
of biological networks. Network models serve as scaffolds
to assist the analysis and interpretation of high-throughput
data190,192,193 and elucidate the structure and dynamics of
biological systems and their roles in physiological and path-
ological states. Nutritional systems biology is an integral part
of systems biology, with a specific focus on the impact of
nutrients and other environmental exposures on biological
networks, the interrogation of perturbed networks underly-
ing metabolic diseases using nutrients and other bioactive
food components, the discovery of metabolic patterns or
metabolites as biomarkers, and the network-assisted devel-
opment of nutritional interventions to prevent or manage
metabolic diseases.188
T H E N ET WO R K BA S I S O F M ETA B O L I C
DISEASES
A systems approach is essential to understanding molecular
processes underlying metabolic diseases. From the point of
view of systems biology, metabolic diseases are manifesta-
tions of the dysregulation of biological networks, which
emphasizes the impairment of multiple processes, includ-
ing signaling, regulation, catalysis, and transport, or altered
interactions among them. The proper functioning of a
metabolic network is represented by the coordinated flows
of metabolites through the network. The flow of metabo-
lites is measured by metabolic flux, which is defined as the
rate at which every metabolite is produced or consumed by
each reaction.187,194 Although metabolic diseases are gener-
ally characterized as a series of impaired metabolic fluxes,
these can originate from a single, causal primary defect
as occurs in most inborn errors of metabolism (IEM), or
from the multiple subtly altered metabolic fluxes observed
in most complex diseases. Complex diseases illustrate the
importance and complexity of interacting genes and path-
ways in pathophysiological processes, whereas monogenic
IEM illustrate the principles by which impaired networks
cause diseases. Although most IEM are caused by single
gene defects, the effect of this impairment can spread
across the metabolic network, as accumulated metabo-
lites can be toxic and impair developmental or neurologi-
cal processes and/or be shunted to secondary pathways
and influence their metabolic fluxes. Moreover, upstream
and downstream alterations in metabolite concentrations
can inhibit or activate other pathways. For example, gly-
cogen storage disease type Ia (GSD-Ia), an IEM caused
by defects in glucose-6-phosphatase alpha (G6PC), leads
to the accumulation of glucose-6-phosphate and reduced
supply of glucose; hypoglycemia following a short fast is
a hallmark of GSD-Ia. G6PD catalyzes the hydrolysis of
glucose-6-phosphate (G6P) into glucose and inorganic
phosphate, the last step of glycogenolysis and gluconeo-
genesis. In this disorder, G6P is shunted to other pathways,
including glycogenesis, glycolysis, and lipogenesis, leading
to complications such as hepatomegaly, nephromegaly,
hyperlipidemia, and lactic academia.195,196 Furthermore, the
accumulated G6P may also play a regulatory role in the acti-
vation of transcription of lipogenic genes, further disturb-
ing lipid metabolism.197
Similar clinical presentations can result from muta-
tions in different genes when the encoded genes are closely
linked within the network architecture, such as serving as
components of a multiprotein complex, a pathway, or a
cellular organelle, and contribute to the same downstream
metabolic fluxes.198 For example, glycogen-storage disease
type I (GSD-I), caused by impaired hydrolysis of G6P into
glucose and inorganic phosphate that originates by several
mechanisms, including mutations in G6PC, as explained
above for GSD-Ia. Related phenotypes result from the
impairment of the G6P translocase (G6PT), which is
responsible for the transport of G6P from cytosol into the
lumen of the endoplasmic reticulum, where the hydrolysis
takes place. GSD-I caused by G6PT deficiency is referred
to as “GSD-Ib,” which has clinical presentation essentially
identical to that of GSD-Ia.195,197
Network architecture defines both the clinical pre-
sentation and the underlying molecular basis of diseases,
as well as variations in symptoms and clinical biomarkers
that define disease, in disorders that include obesity, dia-
betes, Gaucher disease, and Parkinson disease.199 It also
1 9 6   •   P rincip l e s o f G e no m ic M e dicin e
accounts for comorbidity among related diseases.187,199–201
Clustering diseases based on their shared phenotypes may
assist us in the identification of their common molecular
basis, but this approach is challenged by variable penetrance
and by a lack of consistent diagnoses of phenotypes.198
Interestingly, this concept has instead been supported
through a reverse approach: disease etiology with common
molecular bases tends to occur simultaneously in the same
individuals.199,202 By integrating available databases of meta-
bolic networks203,204 and compilations of disease–gene asso-
ciations,205 it is possible to cluster diseases based on their
shared molecular basis. Diseases can be clustered by their
common relatedness to genes, expression patterns, partici-
pation in multiprotein complexes, or adjacency within a
metabolic network.199–201 A metabolism-based disease net-
work (MDN) was constructed by clustering diseases that
are associated with specific genes that function in common
pathways.199 In this MDN, connected disease pairs are three
times more likely to occur together in the same individu-
als than the average of all disease pairs. The more a disease
is connected to other diseases in the MDN, the higher its
prevalence in the population and the higher its mortal-
ity rate.199 Similar results were found in disease networks
built based on protein–protein interactions, co-expression
and disease-gene sharing.201 These disease networks are the
phenotypical representations of the underlying cellular
networks.
G L O BA L R EC O N S T RU C T I O N O F
HUM A N M ETA B O L I C N ET WO R K S
The global reconstruction of biological networks is a pre-
requisite for the application of a systems approach to study
human physiology and pathology.193,204 It provides a scaf-
fold and a computable model for analyzing and interpreting
large-scale omics data to reveal perturbed pathways under-
lying different pathophysiological states. Moreover, the
reconstruction could be transformed into an in silico model
that is amenable to computational analysis and math-
ematical modeling. Computational analysis of the in silico
model may unravel human-specific network properties
or principles regarding network structure and dynamics.
Mathematical modeling enables simulations and predic-
tions of network responses to genetic and environmental
perturbations, including drug treatments and nutritional
interventions, and it can assist in the discovery of biomark-
ers and the development of disease management strategies.
The comprehensive reconstruction of biological net-
works is a daunting undertaking. The various types of cellu-
lar networks, usually classified as signaling, gene regulatory,
and metabolic, must be integrated. Each of these networks
exhibits different network properties and requires different
strategies of reconstruction and modeling.206 Each network
usually contains thousands of components and a myriad of
interactions among them. A complete reconstruction of cel-
lular network requires integrations of these different types
of networks, and must take cellular compartmentalization
into account. Cell-type specificity also calls for the recon-
struction of cell-specific models. Furthermore, models at
the level of tissues have to consider the integration of mul-
tiple cell-specific networks and the communication among
them.193,206
In spite of these challenges, considerable progress has
been made in network reconstructions. The release of the
global reconstruction of the human metabolic network,
named “Recon 1,” was a milestone event.204 This reconstruc-
tion was built based on the human genome sequence and
the accumulated knowledge of human metabolism going
back more than 50 years, encompassing 1,496 genes, 2,004
proteins, 2,712 metabolites, and 3,311 metabolic reactions.
Recon 1 is mass- and charge-balanced and accounts for
seven intracellular compartments (cytoplasm, mitochon-
dria, nucleus, endoplasmic reticulum, Golgi apparatus,
lysosome, and peroxisome).193,204 Although Recon 1 is still
being perfected and there are ongoing efforts to continu-
ously update and fill gaps and missing information, it has
enabled development and applications in multiple areas.
Firstly, the current reconstruction is fueling a wave of
hypothesis-driven studies that will advance our understand-
ing of human metabolism.207,208 Also, Recon 1 serves as the
foundation for generating cell-specific, tissue-specific, and
condition-specific models, and as the scaffold for contex-
tualizing high-throughput data to unravel the mechanistic
processes of disease and drug-treated states. Another attrac-
tive application of Recon 1 is mathematical modeling and
computational simulation to identify biomarkers and to
develop disease-management strategies.193,204
N ET WO R K-A S S I S T E D SYS T E M I C
I N T E R RO G AT I O N O F T H E MO L ECU L A R
BA S I S O F M ETA B O L I C D I S E A S E S
A N D N U T R IT I O NA L I N T E RVE N T I O N
Network-assisted systems approaches are fundamental to
elucidating the molecular basis of diseases and develop-
ing nutritional and/or pharmaceutical interventions using
high-throughput data,193 including transcriptomic,204,209,210
proteomic, and metabolomic data.211,212 The mapping of
genes, transcripts, and proteins onto the metabolic network is
conducted according to the gene-transcript-protein-reaction
N utrition a l G e no m ic s   •   1 9 7
associations.204 Transcriptomic data are currently the most
readily available high-throughput data for this type of anal-
ysis, due to the well-developed microarray and RNA-seq
technologies.213,214 Recon 1 can contextualize transcrip-
tomic data within the human metabolic network model.204
The application of the human metabolic network model in
contextualization of transcriptomic data was demonstrated
along with the release of Recon 1.  In this demonstration,
gene expression data derived from skeletal muscle isolated
from morbidly obese patients before and after gastric-bypass
surgery were reanalyzed and mapped onto Recon 1 in order
to identify metabolic changes following gastric-bypass
surgery. Signature patterns in anaerobic metabolism,
including downregulated oxidative phosphorylation and
mitochondrial bioenergetics, were observed following sur-
gery.204 Others have used Recon 1 to investigate the effect
of dietary interventions on transcriptome profiles at differ-
ent stages of the intervention. Studies of human subjects
challenged with an energy-restriction phase followed by a
weight-stabilization phase209 revealed interactions among
metabolic and inflammatory pathways in adipose tissue
and their impact on insulin sensitivity. Gene expression
patterns for both adipocytes and macrophages were char-
acterized within an unbiased context provided by Recon
1.  Interestingly, adipocyte genes involved in metabolism
and macrophage genes participating in immune pathways
exhibited an opposite pattern of responses to the dietary
intervention. Adipocytes metabolic genes were downreg-
ulated during energy restriction and upregulated during
weight stabilization, whereas macrophage immune-related
genes were not changed or upregulated during energy
restriction and down-regulated in weight-stabilization
phase.209 Comprehensive reconstructions of the whole-body
metabolic network are required to simultaneously analyze
high-throughput data from different cell types and body
fluids.215,216 For example, to determine differential meta-
bolic activity between obese and diabetic obese individuals,
a multi-tissue type genome-wide metabolic network was
built by integrating three cell-specific networks (hepato-
cyte, myocyte, and adipocyte) representing three tissues
(liver, skeletal muscle, and adipose tissue, respectively) and
one blood compartment connecting the three cell types.216
Various multicellular models of brain energy metabolism
have been reconstructed to study Alzheimer’s disease.215
Facilitated by the rapid technological advances in mass
spectrometry (MS) and nuclear magnetic resonance (NMR)
spectrometry, and high-resolution separation technologies,
such as high performance liquid chromatography (HPLC)
and gas chromatography (GC),217 network-assisted
approaches are starting to be applied in a systemic analysis
of metabolomic data.211,212 Pathways and affected enzymes
responsible for Leigh’s syndrome (LS) were identified using
fibroblasts obtained from normal and LS subjects grown in
media with 13C-labeled glucose. Time-course metabolomic
data revealed that fibroblasts from LS patients exhibited
slower metabolism and less adenosine triphosphate (ATP)
production. The model predicted mutations in succinate
cytochrome c reductase as the underlying cause of LS.212
Unbiased metabolic network modeling approaches can
also identify gaps in our understanding of the contributions
of genes to metabolic disease. A  reanalysis of the meta-
bolic responses to an oral glucose tolerance test (OGTT)
was conducted in 25 normal subjects and 25 subjects with
impaired glucose tolerance.211 The initial analysis of the
data that considered only the established insulin-related
pathways (glycolysis, lipolysis, ketogenesis, and proteolysis)
revealed that responses in all these pathways are blunted in
subjects with impaired glucose tolerance. Further investi-
gation identified 18 plasma metabolites that were respon-
sive to glucose ingestion in normal individuals only. These
metabolites could not be mapped to established pathways
and were not previously linked to glucose homeostasis.218
Reanalysis of these metabolic profiles, assisted with the
human Recon 1, revealed the unexpected involvement of
solute carriers in the metabolic pathways of OGTT.211 Such
studies give us confidence that metabolomic profiling com-
bined with network-assisted data analysis will significantly
enhance our further understanding of the molecular basis
of diseases and assist in the development of effective nutri-
tional interventions.
There are meaningful limitations to the use of
Recon 1 and its derivatives as scaffolds to contextualize
high-throughput data. Although these approaches consider
the interconnectedness of genes, enzymes, and metabolites
in the network, the scaffold design of these network recon-
structions ignores reaction stoichiometry, network topol-
ogy, the conservation of mass, and the balance of charge.204
Accounting for this unexploited information and consid-
ering these factors as constraints within a mathematically
representative system of reactions enables a higher-level
use of the reconstructions, referred to as “constraint-based
computational modeling,” which provides an intriguing
application for the reconstruction of the human metabolic
network.193 Constraint-based modeling and simulation
allows for the prediction of the metabolic effects of genetic
variants, enzyme deficiencies, and environmental expo-
sures.219–222 The successful application of this simulation
approach for biomarker identification will be further dis-
cussed in the next section. Computational simulation of
the human metabolic network reconstructions or their
1 9 8   •   P rincip l e s o f G e no m ic M e dicin e
derivatives alone can provide remarkable insights into the
mechanistic bases of metabolic disease and health manage-
ment, and generate a large number of hypotheses await-
ing experimental verifications.220–223 One study utilized
Recon 1–based computational simulation and considered
additional constraints, from enzyme solvent capacity, to
investigate the causes of the Warburg effect, which is the
preferred use of glycolysis over respiration even in the pres-
ence of oxygen in cancer cells.223 The model predicts that
the Warburg effect is the direct consequence of a metabolic
adaptation of cancer cells to fast proliferation. In addition,
the computational modeling also captures several experi-
mentally observed phenotypes during cancer development,
including the preferential uptake of glutamine over other
amino acids.223 There is little doubt that network-based
mathematical modeling and computational simulation will
greatly accelerate hypothesis-driven studies and enhance
our understanding of human diseases and health.
The reconstruction of the human metabolic network
also helps us make large-scale identifications of various
types of interacting relationships among different network
components. Based on these relationships, new candidate
disease genes or drug targets that closely interact with
well-established targets can be identified. For example,
“correlated sets of reactions” (Co-Sets) under a specific
condition can be identified from the condition-specific
reconstruction of the human metabolic network as reac-
tions whose fluxes are perfectly correlated, such that the
flux through one reaction indicates equivalent flux through
correlated reactions.204,224,225 Reactions in the same sets can
be either continuous in a linear pathway or be present in dif-
ferent pathways. Although the continuous cases are intui-
tive in understanding, the non-linear cases are less apparent
and require additional analysis based on comprehensive
networks.224 For example, correlated sets of reactions were
identified in cells conducting aerobic glucose metabolism.
One of the largest sets of correlated reactions is related to
cholesterol biosynthesis, and specifically to the reaction
catalyzed by 3-hydroxy-3-methylglutaryl-CoA reductase, a
primary metabolic target of the anti-lipidemic drug statin.
It is therefore anticipated that other enzymes in the same set
will be drug targets for treating the same disorders.224
S YS T E M I C I D E N T I FI C AT I O N O F
M ETA B O L I C PAT T E R N S O R M ETA B O L IT E S
A S B I O M A R K E R S
Biomarkers are biological characteristics that are objectively
measured and evaluated as indicators of normal biologi-
cal processes, pathological processes, or pharmacological
responses to a therapeutic intervention.226 Biomarkers have
been widely used in clinical practices to predict, diagnose,
and classify diseases, or to monitor outcomes of disease
management or nutritional intervention.227,228 For instance,
elevated blood-glucose levels are used to diagnose diabe-
tes, and increased plasma level of low-density lipoprotein
(LDL) cholesterol is indicative of higher risk for cardio-
vascular diseases.227,229 As indicated in the definition,226 a
biomarker is not necessarily a single molecule (metabolite,
protein, etc.). Rather, it could be a combination of mul-
tiple molecules or their related features.228,230 The currently
available omics data, especially metabolomic data, provide
opportunities for identifying metabolic patterns or metab-
olites as biomarkers.227,228 The analysis of large-scale omics
data for biomarker identification also benefits from the use
of network models. For example gene-expression data were
recently used to identify novel biomarker candidates for
type 2 diabetes.210 Transcriptomic data from skeletal muscle
of individuals with different phenotypes (insulin resistance,
type 2 diabetes) were compared to reveal differentially
expressed genes, which were further mapped onto the net-
work models. Reporter metabolites were identified based
on their association with enzyme-coding genes that were
enriched in the transcriptional response to diabetes. Some
of these reporter metabolites were from pathways known
to be associated with type 2 diabetes, including tricarbox-
ylic acid cycle (TCA cycle), oxidative phosphorylation, and
lipid metabolism, whereas others were novel discoveries,
such as NADH and ATP, representing candidate biomark-
ers awaiting thorough experimental verification.210
In addition to network-assisted mining of omics data,
computational simulation based on reconstructions of
human metabolic networks also enables the prediction
of candidate biomarkers. This approach is especially effi-
cient for monogenic hereditary diseases, such as IEMs,
because of their relatively simple etiologies. The power of
network-assisted biomarker identifications is exemplified
by the pioneering effort to systemically predict biomark-
ers for more than 300 metabolic disorders.223 This novel
computation approach applied a constraint-based mod-
eling method206 to predict the level of metabolites under
both normal and diseased states. For each metabolic disor-
der, the morbid state was simulated by deleting the known
causative gene. For 304 metabolic disorders documented in
the Online Mendelian Inheritance in Man (OMIM) data-
base231 whose causative genes are also present in the Recon
1,204 the computational approach made 3,912 predictions
involving 233 metabolites whose concentrations change in
at least one disease state, and 176 diseases that are associ-
ated with at least one candidate biomarker. Predictions were
N utrition a l G e no m ic s   •   1 9 9
systematically validated by comparing them with experi-
mentally confirmed biomarkers for each disease recorded
in the OMIM database. The biomarker predictions from
this method were 10 times higher than random chance.219
This study also illuminated the current limitations to
network-based computational simulation in biomarker pre-
diction. These limitations lie in both the network model
and the simulation methods. The network model used
in the study (Recon 1) is still incomplete and is not inte-
grated with the gene-expression regulatory network. The
prediction of extracellular biomarkers will benefit from
a whole-body network model that incorporates different
cells and tissues.219 Moreover, simulation techniques with
human metabolic networks have to be further improved.207
N U T R I T I O N AL G E N O M I C S
A P P L I C AT I O N A N D P U B L I C H EALT H
The use of genetic information in current nutrition practice
and policy is very limited, but such consideration has the
potential to “personalize” nutrition and thereby prevent
chronic disease and promote healthy aging through diet
by targeting the molecular antecedents of disease. Cellular
networks are sensitive to internal perturbations that result
from genetic variation and gene mutations; identification
of the gene variant and/or metabolites that induce network
dysfunction can be informative in the diagnoses or predic-
tions of various health and disease outcomes. For example,
as discussed in the section on nutritional systems biology,
classic monogenic disorders, including the in-born errors of
metabolism phenylketoneuria and galactosemia, illustrate
the severe consequences that can result from catastrophic
metabolic network disruptions. Perhaps more importantly,
these early clinical studies also demonstrated that single-gene
disorders can be managed and/or alleviated through dietary
interventions (e.g., the use of phenylalanine-restricted diets
to prevent or mitigate the cognitive impairments resulting
from mutations in the phenylalanine hydroxylase gene) and
thereby established the principles of nutritional interven-
tion. Nutrients, like pharmaceuticals, are powerful modi-
fiers of genome and network function and stability, and
gene–nutrient interactions can be optimized for disease
prevention and management (Figure 12.1).
P E R S O NA L I Z E D N U T R I T I O N
Salient examples of nutritional intervention are pro-
vided by past experiences that demonstrate that mater-
nal and perinatal nutrition can improve birth outcomes,
including cognitive development (e.g., ensuring iodine
sufficiency), lifelong chronic disease resistance (e.g., pre-
venting small-for-gestational-age births), and increased
longevity (e.g., optimizing dietary fat consumption and
immune function).232 It also is apparent that inappropri-
ate uses of nutrition to maximize reproductive outcomes
will present new potential risks.71,157 High-dose vitamin
therapy has been advocated to rescue impaired metabolic
reactions that result from mutations and polymorphisms
that decrease the affinity of substrates and cofactors for
the encoded enzyme.48 ω-3 fatty acids and tocopherols may
promote healthy aging and longevity by modulating the
inflammatory response by altering gene transcription.232
Genes and their allelic variants that influence longevity (a
trait that is nonadaptive) are being identified at accelerated
rates,233 and their penetrance can be modified by the ratio-
nal design of nutrition-based interventions and therapies.232
Repression of energy metabolism, through caloric restric-
tion or transcriptional regulation of metabolic enzymes,
reduces oxidative stress and promotes longevity in many
experimental model systems. Manipulation of these tran-
scriptional and/or metabolic networks by designer vitamin
supplements may promote healthy aging. However, caution
is warranted. Genes encoding virtually all physiological
process are not adapted to excessive nutrient intake expo-
sures that exceed what has been achieved in historical and
healthy food-based diets. Therefore, new risks and toxicities
should be anticipated in human populations or population
subgroups when nutrients are administered at pharma-
cological levels, as illustrated by the introduction of high
levels of fructose into the food supply.67 Some of the unin-
tended consequences may involve genome programming,
including permanent alterations in genome-wide methyla-
tion patterns in stem cell populations, as has been observed
in mouse embryos whose mothers received elevated doses
of folic acid and one-carbon donors during gestation.21
Methylation patterns that are established in utero, and per-
haps in adult stem cell populations, can be metastable and
influence gene expression and, potentially, mutation rates
throughout the lifespan.21 Furthermore, although antioxi-
dants can decrease mutation rates, they can also function
as pro-oxidants in vivo234 and may be cancer-promoting
when consumed at elevated intakes by inhibiting cellular
death programs in transformed cells.235 In conclusion, elu-
cidation of robust gene-by-nutrient interactions will inform
dietary approaches for individuals and for populations that
aim to prevent and/or manage complex metabolic disease,
as has been accomplished for rare inborn errors of metabo-
lism. Equally important, these and other examples indicate
that rigorous hazard-identification is essential prior to the
2 0 0   •   P rincip l e s o f G e no m ic M e dicin e
establishment of policies that result in pharmacological
intakes of nutrients and other dietary components.
A S S I S T E D R E P RO D U C T I O N
A N D O P T I M A L C U LT U R E M E D I A
Elevated SA rates have been observed in some, but not all,
studies of human in vitro fertilization (IVF) pregnancies
compared to natural conceptions. These findings may be
the result of early harvest and early manipulations of eggs
and embryos in culture media.139,236 Furthermore, other
studies have found that human IVF procedures result in
higher-than-expected incidences of IUGR.237 Numerous
studies have shown that the composition of the embryo
culture medium affects the expression and methylation
status of imprinted genes, including H19, Igf2 and Igf2r
in ovine and other mammalian embryos, resulting in large
offspring syndrome.237–241 There appear to be many critical
windows associated with the establishment of environmen-
tally sensitive methylation patterns from early embryogen-
esis through the suckling period, all of which are cell-type
and/or allele specific.184,185 Some of these networks are
sensitive to folate-mediated one-carbon metabolism, as
illustrated by the impact of maternal nutrition on genomic
methylation in the viable yellow agouti (Avy) mouse131;
other networks may respond to the allelic- or locus-specific
targeting of methylase/demethylase activity, as seen with
glucocorticoid-receptor programming, described earlier.176
The increasing evidence that early nutritional exposures can
increase the risk for late-onset metabolic diseases through
epigenetic mechanisms illuminates the major challenges
that are made more immediate by the increased demand for
assisted reproduction.
D I ETA RY R EC O M M E N DAT I O N S
F O R P O P U L AT I O N S
Food-based and nutrient-based dietary guidelines were
established to help individuals and populations achieve
adequate dietary patterns to maintain health. The deriva-
tion and goals of these guidelines evolve as new knowledge
becomes available.242,243 Guidelines for single nutrients and
other food components are scientifically and quantitatively
derived, and are usually based on the level of nutrient intake
that prevents a clinical and/or biochemical outcome that is
associated with a particular nutrient deficiency. Numerical
standards for nutrients are essential to validate the effi-
cacy of food-based guidelines.243 Nutrient requirements
vary among individuals within all human populations,
and can be modified by age, gender, and life stage, among
 Free ebooks ==> www.Ebook777.com
other variables. Therefore, recommended nutrient intakes
are often derived separately for population subgroups.
Although genetic variation can modify the efficacy, dos-
age, and safety of pharmaceutical agents244 and tolerance or
intolerance for certain foods,62 the contribution of genetics
to nutrient requirements within and among human popula-
tions remains to be evaluated. However, the characterization
of gene variants that modify optimal nutrient requirements
will enable the classification of genetic subgroups for whom
generalized nutritional requirements may be valid.
R EC O M M E N D E D DA I LY A L L OWA N C E
A N D U P P E R L I M ITS
The “recommended daily allowance” (RDA) for each nutri-
ent is defined as the level of dietary intake that is sufficient
to meet the requirements of 97% of healthy individuals
in a particular life stage and gender group (Figure 12.5).
When there are insufficient data to calculate an RDA for a
nutrient, an “adequate intake” (AI), which is an estimated
recommended intake value, is established. Some nutrients
demonstrate toxicities at elevated intake levels; therefore,
a “tolerable upper intake level” (UL), which represents the
highest level of nutrient intake that can be achieved without
incurring risk for adverse health effects for the vast majority
of individuals in the general population, is established.
Human genetic variation is not anticipated to confer
extreme variations in optimal nutrient requirements among
individuals and populations. Nutrition, unlike pharmaceu-
ticals, is an in utero and lifelong exposure that can serve as
a selective pressure to eliminate genomes that are not com-
patible with the nutrient environment. Therefore, human
genotypes that do not support basic physiological processes
EAR
AI UL
Risk of inadequacy
0.5
RDA
Increasing intake
Figure 12.5 
Dietary requirements. Estimated average requirement (EAR)
represents the intake at which the risk of inadequacy is 0.5 (50%) to an
individual. Recommended dietary allowance (RDA) represents a level
of nutrient intake at which the risk of inadequacy is <3%. Adequate
intake (AI) is not based on an EAR but is assumed to be near or above
the RDA if one could be calculated. The tolerable upper intake level
(UL) represents risk of excess that is set close to 0.
N utrition a l G e no m ic s   •   2 0 1
www.Ebook777.com
will be selected against in large part through fetal loss and
the failure to survive to reproduce. Allelic variants that con-
fer atypical nutrient requirements that cannot be met by the
mother or that result in severe metabolic disruptions are
expected to be embryonically lethal. Alleles that confer sub-
tler differences in nutrient requirements or food tolerances
are expected to be enriched in subgroups or populations,
and contribute to disease in certain environmental contexts
(Table 12.1 and Table 12.2).
The concept of generalized nutrient requirements
within and among human populations is nullified only when
a level of nutrient intake that represents minimal nutrient
adequacy for one genetic subgroup exceeds a safe intake
level for another group, assuming that nutrient deficiency
and harm/toxicity-avoidance are the primary criteria for
that requirement. For example, it is widely recognized that
optimal folate intakes may differ among identifiable genetic
subgroups.245 However, it is not at all certain that the mag-
nitude of the genetic contribution to variations in adequate
dietary folate intake warrants genotype-specific recommen-
dations, especially considering that folic acid intakes up to
1 mg/day are not associated with known toxicities.124 Iron
is another candidate nutrient for genotype-specific nutrient
requirement recommendations.79,246–248 For these and other
cases, the penetrance (contribution of the individual allele
to variation in nutrient requirements) and the prevalence
of these functional gene variants must be elucidated both
within and among human populations to validate the con-
cept of generalized nutrient requirements for all human
populations. Unlike the effects of gender and life cycle, no
common allelic variant has been shown to be sufficiently
penetrant and to warrant genotype-specific numerical stan-
dards for nutrient adequacy or upper levels of intake associ-
ated with harm or toxicity. At this time, genetic variation
is known only to influence nutrient and food intolerance
(Table 12.1 and Table 12.2). However, genetic variation has
not been characterized in many geographically and cultur-
Risk of excess
ally isolated populations that have existed in nutrient-poor
0.5 or otherwise unique nutritional environments for many
generations, therefore the presence of adaptive alleles
should be expected in such populations. Recent experiences
have demonstrated the severe adverse health consequences
that can result from rapid alterations in dietary patterns,
especially energy intake, among certain populations.249
Finally, to achieve dietary guidelines that optimize
health for all individuals and populations, the many func-
tions of individual genes, and their regulation within
metabolic and transcriptional networks, must be under-
stood comprehensively. For example, the LCT gene also
encodes the enzyme pyridoxine-5′-beta-D-glucoside as a
result of differential processing of the LCT transcript.250
Pyridoxine-5′-beta-D-glucoside activity is necessary for the
bioavailability of pyridoxine-5′-beta-D-glucoside, the major
form of vitamin B6 in plant derived foods.251 Therefore,
LCT variation predicts both lactose tolerance as well as pre-
ferred dietary sources of vitamin B6 in adulthood.
G E N O M I C C R IT E R I A F O R S ET T I N G
R EQ U I R E M E N TS A N D TOX I C IT I E S
Genomic technologies may provide new criteria for estab-
lishing numerical standards for adequate levels of nutrient
intake by targeting the molecular antecedents of disease.
DNA mutations increase risk for degenerative diseases and
cancers, and can be quantified in controlled experimental
settings, indicating that the effect of mineral and vitamin
intake levels on somatic cell DNA mutation rates should be
considered when establishing RDAs.33 Marginal deficien-
cies in folate, vitamin B12, niacin, and zinc can influence
genome stability, and antioxidants such as carotenoids, vita-
min C, and vitamin E may prevent damage resulting from
oxidative stress. Validation of these protective effects on
DNA mutation rates in controlled human trials may indi-
cate benefits and lead to increased recommended intake lev-
els, perhaps at levels not normally achievable from a natural
food-based diet. Similarly, the use of functional genomic
approaches, including expression-profiling and proteomics
to quantify gene expression and metabolomics to quantify
metabolic pathway flux, may provide a comprehensive set
of quantitative and physiologically relevant biomarkers to
model and assess nutrient efficacy in the context of optimal
network function.252
P U B L I C NU T R IT I O NA L
I N T E RVE N T I O NS
Public health interventions, including micronutrient food
fortifications, have often been successful in preventing
adverse outcomes associated with maternal nutrient defi-
ciencies or excesses. Successful initiatives include the fortifi-
cation of grain products with folate to prevent NTDs253 and
of salt with iodine to prevent cretinism,254 discouraging alco-
hol consumption during pregnancy to prevent fetal alcohol
syndrome,255 and the fortification of various vehicles to pre-
vent iron deficiency.247 These successes encourage further
improvements of maternal diets and infant formulas to opti-
mize short- and long-term outcomes, and predict the likely
accelerated development of dietary approaches that aim to
optimize functional abilities rather than the more tradi-
tional goals of preventing deficiency and deficiency-related
2 0 2   •   P rincip l e s o f G e no m ic M e dicin e
disease. For example, there is evidence that docosahexaenoic
acid (DHA)256 or choline257 supplementation during prena-
tal and/or early postnatal life improves CNS function and
cognitive performance throughout life. Continued progress
in these areas requires the identification and understanding
of nutritionally modifiable networks that enable successful
adaptations in successive life stages, their associated critical
windows, and validation of the efficacy, effectiveness, and
safety of the nutritional interventions.
C O N C LU S I O N
Our current knowledge of the role of vitamin–genome
interactions in the evolution of the human genome, and its
consequences for current human dietary requirements, is
still in its infancy, especially when compared to our under-
standing of the role of vitamins in regulating gene expres-
sion. As detailed in this chapter, only a very few controlled
dietary studies have directly quantified the effects of genetic
variation on individual vitamin requirements, and few
genetic subgroups have been identified for whom the cur-
rent dietary recommendations may be inadequate. Virtually
nothing is known about the role of genetic variation in alter-
ing tolerable upper levels of nutrient intake. Likewise, there
is much to be learned about the role of vitamins and vita-
min availability in the evolution of the human genome and
its associated variations, including how dramatic changes in
our food environment have altered the impact of individual
genetic variants that in one environmental context may have
been advantageous, but in others has resulted in the genera-
tion of disease-associated genes. A  more complete under-
standing of genome–vitamin interactions will lead to the
development of dietary recommendations for healthy indi-
viduals that account for an individual’s genetic background,
as well as dietary recommendations for the management of
nutrition-related chronic disease.
R EFE R E N C ES
1. Subbiah MT. Nutrigenetics and nutraceuticals: the next wave riding
on personalized medicine. Transl Res. Feb 2007;149(2):55–61.
2. Fenech M, El-Sohemy A, Cahill L, et al. Nutrigenetics and nutrig-
enomics:  viewpoints on the current status and applications in
nutrition research and practice. J Nutrigenet Nutrigenomics.
2011;4(2):69–89.
3. Ordovas JM, Mooser V. Nutrigenomics and nutrigenetics. Curr
Opin Lipidol. Apr 2004;15(2):101–108.
4. Guttmacher AE, Collins FS. Genomic medicine—a primer. N Engl
J Med. 7 Nov 2002;347(19):1512–1520.
5. Young VR. 2001 W.O. Atwater Memorial Lecture and the 2001
ASNS President’s Lecture: Human nutrient requirements: the chal-
lenge of the post-genome era. J Nutr. Apr 2002;132(4):621–629.
 6. Venter JC, Adams MD, Myers EW, et al. The sequence of the human
genome. Science. 16 Feb 2001;291(5507):1304–1351.
7. Tishkoff SA, Kidd KK. Implications of biogeography of human
populations for “race” and medicine. Nat Genet. Nov 2004;36(11
Suppl):S21–S27.
8. Venter JC. Multiple personal genomes await. Nature. 1 Apr
2010;464(7289):676–677.
9. Zhang F, Gu W, Hurles ME, Lupski JR. Copy number variation in
human health, disease, and evolution. Annu Rev Genomics Hum
Genet. 2009;10:451–481.
10. Redon R, Ishikawa S, Fitch KR, et  al. Global variation in copy
number in the human genome. Nature. 23 Nov 2006;444(7118):
444–454.
11. Sherry ST, Ward MH, Kholodov M, et  al. dbSNP:  the NCBI
database of genetic variation. Nucleic Acids Res. 1 Jan 2001;29(1):
308–311.
12. Nachman MW, Crowell SL. Estimate of the mutation rate per
nucleotide in humans. Genetics. Sep 2000;156(1):297–304.
13. Frazer KA, Murray SS, Schork NJ, Topol EJ. Human genetic varia-
tion and its contribution to complex traits. Nat Rev Genet. Apr
2009;10(4):241–251.
14. Tishkoff SA, Verrelli BC. Patterns of human genetic diver-
42. Wolfe KH, Li WH. Molecular evolution meets the genomics revo-
sity: implications for human evolutionary history and disease. Annu
Rev Genomics Hum Genet. 2003;4:293–340.
15. Van Blerkom LM. Role of viruses in human evolution. Am J Phys
Anthropol. 2003;Suppl 37:14–46.
16. Brookfield JF. Mobile DNAs: the poacher turned gamekeeper. Curr
Biol. 28 Oct 2003;13(21):R846–R847.
17. Batzer MA, Deininger PL. Alu repeats and human genomic diver-
sity. Nat Rev Genet. May 2002;3(5):370–379.
19. Bamshad M, Wooding SP. Signatures of natural selection in the
human genome. Nat Rev Genet. Feb 2003;4(2):99–111.
20. Gal-Mark N, Schwartz S, Ast G. Alternative splicing of Alu
exons—two arms are better than one. Nucleic Acids Res. Apr
2008;36(6):2012–2023.
21. Waterland RA, Jirtle RL. Transposable elements:  targets for early
nutritional effects on epigenetic gene regulation. Mol Cell Biol. Aug
2003;23(15):5293–5300.
22. Deininger PL, Batzer MA. Alu repeats and human disease. Mol
Genet Metab. Jul 1999;67(3):183–193.
23. Conrad DF, Pinto D, Redon R, et al. Origins and functional impact
of copy number variation in the human genome. Nature Apr 2010;
1;464(7289):704–712.
24. Freeman JL, Perry GH, Feuk L, et al. Copy number variation: new
insights in genome diversity. Genome Res. 2006;16(8):949–961.
25. Zhang F, Gu W, Hurles ME, Lupski JR. Copy number variation in
human health, disease, and evolution. Annu Rev Genomics Hum
Genet 2009;10:451–481.
26. Eichler EE, Nickerson DA, Altshuler D, et al. Completing the map
of human genetic variation. Nature. May 10, 2007;447(7141):
161–165.
27. Xue Y, Sun D, Daly A, et  al. Adaptive evolution of UGT2B17
copy-number variation. Am J Hum Genet. 2008;83(3):337–346.
28. Tuzun E, Sharp AJ, Bailey JA, et al. Fine-scale structural variation of
the human genome. Nat Genet. 2005;37(7):727–732.
29. Sebat J, Lakshmi B, Troge J, et al. Large-scale copy number polymor-
phism in the human genome. Science 2004;305(5683):525–528.
30. Perry GH, Dominy NJ, Claw KG, et  al. Diet and the evolution
of human amylase gene copy number variation. Nat Genet. Oct
2007;39(10):1256–1260.
31. Blount BC, Mack MM, Wehr CM, et  al. Folate deficiency causes
uracil misincorporation into human DNA and chromosome break-
age: implications for cancer and neuronal damage. Proc Natl Acad
Sci U S A. 1 Apr 1997;94(7):3290–3295.
32. Fenech M. Micronutrients and genomic stability: a new paradigm
for recommended dietary allowances (RDAs). Food Chem Toxicol.
Aug 2002;40(8):1113–1117.
N utrition a l G e no m ic s   •   2 0 3
33. Fenech M. Genome health nutrigenomics:  nutrition and the
science of optimal genome maintenance. Asia Pac J Clin Nutr.
2004;13(Suppl):S15.
34. Fenech M. Chromosomal biomarkers of genomic instability rele-
vant to cancer. Drug Discov Today. 15 Nov 2002;7(22):1128–1137.
35. Fenech M. Recommended dietary allowances (RDAs) for genomic
stability. Mutat Res. 1 Sep 2001;480:51–54.
36. Ames BN. DNA damage from micronutrient deficiencies is likely to
be a major cause of cancer. Mutat Res. 18 Apr 2001;475(1–2):7–20.
37. Goswami T, Rolfs A, Hediger MA. Iron transport: emerging roles in
health and disease. Biochem Cell Biol. 2002;80(5):679–689.
38. Hediger MA, Rolfs A, Goswami T. Iron transport and hemochro-
matosis. J Investig Med. Sep 2002;50(5):239S–246S.
39. Mishra HN, Das C. A review on biological control and metabolism
of aflatoxin. Crit Rev Food Sci Nutr. 2003;43(3):245–264.
40. Roach JC, Glusman G, Smit AF, et  al. Analysis of genetic inheri-
tance in a family quartet by whole-genome sequencing. Science. 30
Apr 2010;328(5978):636–639.
41. Clark AG, Glanowski S, Nielsen R, et al. Inferring nonneutral evo-
lution from human-chimp-mouse orthologous gene trios. Science.
12 Dec 2003;302(5652):1960–1963.
lution. Nat Genet. Mar 2003;33 Suppl:255–265.
43. Kimura M. Evolutionary rate at the molecular level. Nature. 17 Feb
1968;217(129):624–626.
44. King JL, Jukes TH. Non-Darwinian evolution. Science. 16 May
1969;164(881):788–798.
45. Akey JM, Zhang G, Zhang K, Jin L, Shriver MD. Interrogating a
high-density SNP map for signatures of natural selection. Genome
Res. Dec 2002;12(12):1805–1814.
46. Gabriel SB, Schaffner SF, Nguyen H, et  al. The structure
of haplotype blocks in the human genome. Science. 21 Jun
2002;296(5576):2225–2229.
47. Crawford DC, Nickerson DA. Definition and clinical importance
of haplotypes. Annu Rev Med. 2005;56:303–320.
48. Ames BN, Elson-Schwab I, Silver EA. High-dose vitamin therapy
stimulates variant enzymes with decreased coenzyme binding affin-
ity (increased K(m)):  relevance to genetic disease and polymor-
phisms. Am J Clin Nutr. Apr 2002;75(4):616–658.
49. Bracken MB. Genomic epidemiology of complex disease: the need
for an electronic evidence-based approach to research synthesis. Am
J Epidemiol. 15 Aug 2005;162(4):297–301.
50. Wooding S. Natural selection:  sign, sign, everywhere a sign. Curr
Biol. 7 Sep 2004;14(17):R700–R701.
51. Akey JM, Eberle MA, Rieder MJ, et al. Population history and natu-
ral selection shape patterns of genetic variation in 132 genes. PLoS
Biol. Oct 2004;2(10):e286.
52. Chakravarti A. To a future of genetic medicine. Nature. 15 Feb
2001;409(6822):822–823.
53. Reich DE, Lander ES. On the allelic spectrum of human disease.
Trends Genet. Sep 2001;17(9):502–510.
54. Pritchard JK. Are rare variants responsible for susceptibility to com-
plex diseases? Am J Hum Genet. Jul 2001;69(1):124–137.
55. Pritchard JK, Cox NJ. The allelic architecture of human disease
genes: common disease-common variant . . . or not? Hum Mol Genet.
1 Oct 2002;11(20):2417–2423.
56. Jorde LB, Watkins WS, Bamshad MJ. Population genomics: a bridge
from evolutionary history to genetic medicine. Hum Mol Genet.
1 Oct 2001;10(20):2199–2207.
57. Jorde LB, Watkins WS, Bamshad MJ, et  al. The distribution
of human genetic diversity:  a comparison of mitochondrial,
autosomal, and Y-chromosome data. Am J Hum Genet. Mar
2000;66(3):979–988.
58. Jorde LB, Wooding SP. Genetic variation, classification and “race.”
Nat Genet. Nov 2004;36(11 Suppl):S28–S33.
59. Charlesworth D. Balancing selection and its effects on sequences in
nearby genome regions. PLoS Genet. Apr 2006;2(4):e64.
 60. Collins FS. What we do and don’t know about “race,” “ethnicity,”
genetics and health at the dawn of the genome era. Nat Genet. Nov
2004;36(11 Suppl):S13–S15.
61. Bufe B, Hofmann T, Krautwurst D, Raguse JD, Meyerhof W. The
human TAS2R16 receptor mediates bitter taste in response to
beta-glucopyranosides. Nat Genet. Nov 2002;32(3):397–401.
62. Enattah NS, Sahi T, Savilahti E, Terwilliger JD, Peltonen L, Jarvela
I. Identification of a variant associated with adult-type hypolactasia.
Nat Genet. Feb 2002;30(2):233–237.
63. Bersaglieri T, Sabeti PC, Patterson N, et  al. Genetic signatures of
strong recent positive selection at the lactase gene. Am J Hum Genet.
Jun 2004;74(6):1111–1120.
64. Poulter M, Hollox E, Harvey CB, et al. The causal element for the
lactase persistence/non-persistence polymorphism is located in a 1
Mb region of linkage disequilibrium in Europeans. Ann Hum Genet.
Jul 2003;67(Pt 4):298–311.
65. Wooding S. PopHist: inferring population history from the spec-
trum of allele frequencies. Bioinformatics. 1 Mar 2003;19(4):
539–540.
66. Esposito G, Vitagliano L, Santamaria R, Viola A, Zagari A,
Salvatore F. Structural and functional analysis of aldolase B mutants
related to hereditary fructose intolerance. FEBS Lett. 6 Nov
2002;531(2):152–156.
67. Cox TM. The genetic consequences of our sweet tooth. Nat Rev
Genet. Jun 2002;3(6):481–487.
68. Fullerton SM, Clark AG, Weiss KM, et al. Apolipoprotein E varia-
tion at the sequence haplotype level: implications for the origin and
maintenance of a major human polymorphism. Am J Hum Genet.
Oct 2000;67(4):881–900.
69. Ordovas JM, Corella D. Nutritional genomics. Annu Rev Genomics
Hum Genet. 2004;5:71–118.
70. Ordovas JM. The quest for cardiovascular health in the genomic
era:  nutrigenetics and plasma lipoproteins. Proc Nutr Soc. Feb
2004;63(1):145–152.
71. Stover PJ. Nutritional genomics. Physiol Genomics. 15 Jan
2004;16(2):161–165.
72. Stover PJ. Physiology of folate and vitamin B12 in health and dis-
ease. Nutr Rev. Jun 2004;62(6 Pt 2):S3–S12; discussion S13.
73. Brody LC, Conley M, Cox C, et al. A polymorphism, R653Q, in
the trifunctional enzyme methylenetetrahydrofolate dehydroge-
nase/methenyltetrahydrofolate cyclohydrolase/formyltetrahydro-
folate synthetase is a maternal genetic risk factor for neural tube
defects:  Report of the Birth Defects Research Group. Am J Hum
Genet. Nov 2002;71(5):1207–1215.
74. Ma J, Stampfer MJ, Giovannucci E, et al. Methylenetetrahydrofolate
reductase polymorphism, dietary interactions, and risk of colorectal
cancer. Cancer Res. 15 Mar 1997;57(6):1098–1102.
75. Guenther BD, Sheppard CA, Tran P, Rozen R, Matthews RG,
Ludwig ML. The structure and properties of methylenetetra-
hydrofolate reductase from Escherichia coli suggest how folate
ameliorates human hyperhomocysteinemia. Nat Struct Biol. Apr
1999;6(4):359–365.
76. Bailey LB. Folate, methyl-related nutrients, alcohol, and the
MTHFR 677C-->T polymorphism affect cancer risk:  intake rec-
ommendations. J Nutr. Nov 2003;133(11 Suppl 1):3748S-3753S.
77. Toomajian C, Ajioka RS, Jorde LB, Kushner JP, Kreitman M. A
method for detecting recent selection in the human genome from
allele age estimates. Genetics. Sep 2003;165(1):287–297.
78. Toomajian C, Kreitman M. Sequence variation and haplotype
structure at the human HFE locus. Genetics. Aug 2002;161(4):
1609–1623.
79. Beutler E. Iron absorption in carriers of the C282Y hemochromato-
sis mutation. Am J Clin Nutr. Oct 2004;80(4):799–800.
80. Bosron WF, Li TK. Genetic polymorphism of human liver alco-
hol and aldehyde dehydrogenases, and their relationship to alco-
hol metabolism and alcoholism. Hepatology. May-Jun 1986;6(3):
502–510.
2 0 4   •   P rincip l e s o f G e no m ic M e dicin e
81. Osier MV, Pakstis AJ, Soodyall H, et  al. A global perspective
on genetic variation at the ADH genes reveals unusual patterns
of linkage disequilibrium and diversity. Am J Hum Genet. Jul
2002;71(1):84–99.
82. Loew M, Boeing H, Sturmer T, Brenner H. Relation among alcohol
dehydrogenase 2 polymorphism, alcohol consumption, and levels
of gamma-glutamyltransferase. Alcohol. Apr 2003;29(3):131–135.
83. Oota H, Pakstis AJ, Bonne-Tamir B, et  al. The evolution and
population genetics of the ALDH2 locus:  random genetic drift,
selection, and low levels of recombination. Ann Hum Genet. Mar
2004;68(Pt 2):93–109.
84. Peng Y, Shi H, Qi XB, et al. The ADH1B Arg47His polymorphism
in East Asian populations and expansion of rice domestication in
history. BMC Evol Biol. 2010;10:15.
85. Danpure CJ. Molecular etiology of primary hyperoxaluria
type 1:  new directions for treatment. Am J Nephrol. May-Jun
2005;25(3):303–310.
86. Williams RJ, Deason G. Individuality in vitamin C needs. Proc
Natl Acad Sci U S A. Jun 1967;57(6):1638–1641.
87. Eck P, Erichsen HC, Taylor JG, et al. Comparison of the genomic
structure and variation in the two human sodium-dependent vita-
min C transporters, SLC23A1 and SLC23A2. Hum Genet. Sep
2004;115(4):285–294.
88. Zanon-Moreno V, Ciancotti-Olivares L, Asencio J, et  al.
Association between a SLC23A2 gene variation, plasma vitamin
C levels, and risk of glaucoma in a Mediterranean population. Mol
Vis. 2011;17:2997–3004.
89. Neel JV. Diabetes mellitus:  a “thrifty” genotype rendered detri-
mental by “progress?” Am J Hum Genet. Dec 1962;14:353–362.
90. Diamond J. The double puzzle of diabetes. Nature. 5 Jun
2003;423(6940):599–602.
91. Rockman MV, Wray GA. Abundant raw material for
cis-regulatory evolution in humans. Mol Biol Evol. Nov 2002;19
(11):1991–2004.
92. Verrelli BC, McDonald JH, Argyropoulos G, et  al. Evidence for
balancing selection from nucleotide sequence analyses of human
G6PD. Am J Hum Genet. Nov 2002;71(5):1112–1128.
93. Watkins WS, Rogers AR, Ostler CT, et  al. Genetic variation
among world populations: inferences from 100 Alu insertion poly-
morphisms. Genome Res. Jul 2003;13(7):1607–1618.
94. Tishkoff SA, Varkonyi R, Cahinhinan N, et  al. Haplotype
diversity and linkage disequilibrium at human G6PD:  recent
origin of alleles that confer malarial resistance. Science. 20 Jul
2001;293(5529):455–462.
95. Baier LJ, Permana PA, Yang X, et al. A calpain-10 gene polymor-
phism is associated with reduced muscle mRNA levels and insulin
resistance. J Clin Invest. Oct 2000;106(7):R69–R73.
96. Inoue I, Nakajima T, Williams CS, et  al. A nucleotide substitu-
tion in the promoter of human angiotensinogen is associated with
essential hypertension and affects basal transcription in vitro. J
Clin Invest. 1 Apr 1997;99(7):1786–1797.
97. Wray GA, Hahn MW, Abouheif E, et al. The evolution of transcrip-
tional regulation in eukaryotes. Mol Biol Evol. Sep 2003;20(9):
1377–1419.
98. Wright AF, Carothers AD, Pirastu M. Population choice in map-
ping genes for complex diseases. Nat Genet. Dec 1999;23(4):
397–404.
99. Wooding S, Kim UK, Bamshad MJ, Larsen J, Jorde LB, Drayna D.
Natural selection and molecular evolution in PTC, a bitter-taste
receptor gene. Am J Hum Genet. Apr 2004;74(4):637–646.
100. Dennis C. Epigenetics and disease: altered states. Nature. 13 Feb
2003;421(6924):686–688.
101. Meda F, Folci M, Baccarelli A, Selmi C. The epigenetics of autoim-
munity. Cell Mol Immunol. May 2011;8(3):226–236.
102. Barker DJ. Intrauterine programming of coronary heart disease
and stroke. Acta Paediatr Suppl. Nov 1997;423:178–182; discus-
sion 183.
 103. Wu G, Imhoff-Kunsch B, Girard AW. Biological mechanisms for
nutritional regulation of maternal health and fetal development.
Paediatr Perinat Epidemiol. Jul 2012;26 Suppl 1:4–26.
104. Blake MJ, Castro L, Leeder JS, Kearns GL. Ontogeny of drug
metabolizing enzymes in the neonate. Semin Fetal Neonatal Med.
Apr 2005;10(2):123–138.
105. Rasmussen KM. The “fetal origins” hypothesis:  challenges and
opportunities for maternal and child nutrition. Annu Rev Nutr.
2001;21:73–95.
106. Waterland RA, Garza C. Potential mechanisms of metabolic
imprinting that lead to chronic disease. Am J Clin Nutr. Feb
1999;69(2):179–197.
107. Waterland RA. Is epigenetics an important link between early life
events and adult disease? Horm Res. Jan 2009;71 Suppl 1:13–16.
108. Waterland RA. Early environmental effects on epigenetic regula-
tion in humans. Epigenetics. 16 Nov 2009;4(8):523–525.
109. Hochberg Z, Feil R, Constancia M, et al. Child health, develop-
mental plasticity, and epigenetic programming. Endocr Rev. Apr
2011;32(2):159–224.
110. Waterland RA, Kellermayer R, Laritsky E, et al. Season of concep-
tion in rural gambia affects DNA methylation at putative human
metastable epialleles. PLoS Genet. 2010;6(12):e1001252.
111. Waterland RA, Garza C. Potential mechanisms of metabolic
imprinting that lead to chronic disease. Am J Clin Nutr. Feb
1999;69(2):179–197.
112. Waterland RA. Epigenetic epidemiology of obesity:  applica-
tion of epigenomic technology. Nutr Rev. Aug 2008;66 Suppl
1:S21–S23.
113. Cedar H, Bergman Y. Programming of DNA methylation pat-
terns. Annu Rev Biochem. 2012;81:97–117.
114. Rando OJ. Combinatorial complexity in chromatin structure and
function:  revisiting the histone code. Curr Opin Genet Dev. Apr
2012;22(2):148–155.
115. Suh JR, Herbig AK, Stover PJ. New perspectives on folate catabo-
lism. Annu Rev Nutr. 2001;21:255–282.
116. Clarke S, Banfield K. S-adenosylmethionine-dependent meth-
yltransferases. In:  Carmel R, Jacobson DW, eds. Homocysteine in
Health and Disease. Cambridge, UK: Cambridge University Press;
2001. 65–80.
117. Finkelstein JD. Homocysteine: a history in progress. Nutr Rev. Jul
2000;58(7):193–204.
118. Finkelstein JD. Pathways and regulation of homocysteine metabo-
lism in mammals. Semin Thromb Hemost. 2000;26(3):219–225.
119. Choi SW, Mason JB. Folate and carcinogenesis:  an integrated
scheme. J Nutr. Feb 2000;130(2):129–132.
120. Beaudin AE, Stover PJ. Insights into metabolic mechanisms under-
lying folate-responsive neural tube defects:  a minireview. Birth
Defects Res A Clin Mol Teratol. Apr 2009;85(4):274–284.
121. Beaudin AE, Abarinov EV, Malysheva O, Perry CA, Caudill
M, Stover PJ. Dietary folate, but not choline, modifies neural
tube defect risk in Shmt1 knockout mice. Am J Clin Nutr. Jan
2012;95(1):109–114.
122. Herbig K, Chiang EP, Lee LR, Hills J, Shane B, Stover PJ.
Cytoplasmic serine hydroxymethyltransferase mediates com-
petition between folate-dependent deoxyribonucleotide and
S-adenosylmethionine biosyntheses. J Biol Chem. 11 Oct
2002;277(41):38381–38389.
123. Oyama K, Kawakami K, Maeda K, Ishiguro K, Watanabe G. The
association between methylenetetrahydrofolate reductase poly-
morphism and promoter methylation in proximal colon cancer.
Anticancer Res. Mar-Apr 2004;24(2B):649–654.
124. Shelnutt KP, Kauwell GP, Gregory JF, 3rd, et  al.
Methylenetetrahydrofolate reductase 677C-->T polymorphism
affects DNA methylation in response to controlled folate intake in
young women. J Nutr Biochem. Sep 2004;15(9):554–560.
125. Quinlivan EP, Davis SR, Shelnutt KP, et al. Methylenetetrahydro­
folate reductase 677C-->T polymorphism and folate status affect
N utrition a l G e no m ic s   •   2 0 5
one-carbon incorporation into human DNA deoxynucleosides. J
Nutr. Mar 2005;135(3):389–396.
126. Zingg JM, Jones PA. Genetic and epigenetic aspects of DNA meth-
ylation on genome expression, evolution, mutation and carcino-
genesis. Carcinogenesis. May 1997;18(5):869–882.
127. Mason JB, Kim Y. Nutritional strategies in the prevention of
colorectal cancer. Curr Gastroenterol Rep. Aug 1999;1(4):341–353.
128. Kim M, Trinh BN, Long TI, Oghamian S, Laird PW. Dnmt1 defi-
ciency leads to enhanced microsatellite instability in mouse embry-
onic stem cells. Nucleic Acids Res. 2004;32(19):5742–5749.
129. Ingrosso D, Cimmino A, Perna AF, et  al. Folate treatment and
unbalanced methylation and changes of allelic expression induced
by hyperhomocysteinaemia in patients with uraemia. Lancet. 17
May 2003;361(9370):1693–1699.
130. Petry CJ, Ong KK, Barratt BJ, et al. Common polymorphism in
H19 associated with birthweight and cord blood IGF-II levels in
humans. BMC Genet. 10 May 2005;6(1):22.
131. Waterland RA, Jirtle RL. Early nutrition, epigenetic changes at
transposons and imprinted genes, and enhanced susceptibility to
adult chronic diseases. Nutrition. Jan 2004;20(1):63–68.
132. Morgan HD, Sutherland HG, Martin DI, Whitelaw E. Epigenetic
inheritance at the agouti locus in the mouse. Nat Genet. Nov
1999;23(3):314–318.
133. Waterland RA, Travisano M, Tahiliani KG. Diet-induced
hypermethylation at agouti viable yellow is not inherited trans-
generationally through the female. FASEB J. Oct 2007;21(12):
3380–3385.
134. Delhanty JD. Preimplantation genetics:  an explanation for poor
human fertility? Ann Hum Genet. Jul 2001;65(Pt 4):331–338.
135. Brent RL, Beckman DA. The contribution of environmental
teratogens to embryonic and fetal loss. Clin Obstet Gynecol. Sep
1994;37(3):646–670.
136. Edmonds DK, Lindsay KS, Miller JF, Williamson E, Wood
PJ. Early embryonic mortality in women. Fertil Steril. Oct
1982;38(4):447–453.
137. Edwards RG. Recent scientific and medical advances in assisted
human conception. Int J Dev Biol. Apr 1997;41(2):255–262.
138. Wilcox AJ, Weinberg CR, O’Connor JF, et al. Incidence of early
loss of pregnancy. N Engl J Med. 28 Jul 1988;319(4):189–194.
139. Bulletti C, Flamigni C, Giacomucci E. Reproductive failure due
to spontaneous abortion and recurrent miscarriage. Hum Reprod
Update. Mar-Apr 1996;2(2):118–136.
140. Brock DJ, Holloway S. Fertility of older women. Lancet. 16 Jun
1990;335(8703):1470.
141. Cowchock FS, Gibas Z, Jackson LG. Chromosome errors as a
cause of spontaneous abortion: the relative importance of mater-
nal age and obstetric history. Fertil Steril. May 1993;59(5):
1011–1014.
142. Gris JC, Perneger TV, Quere I, et al. Antiphospholipid/antiprotein
antibodies, hemostasis-related autoantibodies, and plasma homo-
cysteine as risk factors for a first early pregnancy loss: a matched
case-control study. Blood. 15 Nov 2003;102(10):3504–3513.
143. Zetterberg H. Methylenetetrahydrofolate reductase and trans-
cobalamin genetic polymorphisms in human spontaneous abor-
tion: biological and clinical implications. Reprod Biol Endocrinol.
17 Feb 2004;2(1):7.
144. Zetterberg H, Regland B, Palmer M, et  al. Increased frequency
of combined methylenetetrahydrofolate reductase C677T and
A1298C mutated alleles in spontaneously aborted embryos. Eur J
Hum Genet. Feb 2002;10(2):113–118.
145. Zetterberg H, Regland B, Palmer M, et  al. The transcobalamin
codon 259 polymorphism influences the risk of human spontane-
ous abortion. Hum Reprod. Dec 2002;17(12):3033–3036.
146. Zetterberg H, Zafiropoulos A, Spandidos DA, Rymo L, Blennow
K. Gene–gene interaction between fetal MTHFR 677C>T and
transcobalamin 776C>G polymorphisms in human spontaneous
abortion. Hum Reprod. Sep 2003;18(9):1948–1950.
 147. Nelen WL, Blom HJ, Steegers EA, den Heijer M, Eskes TK.
Hyperhomocysteinemia and recurrent early pregnancy loss:  a
meta-analysis. Fertil Steril. Dec 2000;74(6):1196–1199.
148. Nelen WL. Hyperhomocysteinaemia and human reproduction.
Clin Chem Lab Med. Aug 2001;39(8):758–763.
149. Yamada H, Sata F, Saijo Y, Kishi R, Minakami H. Genetic factors
in fetal growth restriction and miscarriage. Semin Thromb Hemost.
Jun 2005;31(3):334–345.
150. Nelen WL, Blom HJ, Steegers EA, den Heijer M, Thomas CM,
Eskes TK. Homocysteine and folate levels as risk factors for recur-
rent early pregnancy loss. Obstet Gynecol. Apr 2000;95(4):519–524.
151. Nelen WL, Blom HJ, Thomas CM, Steegers EA, Boers GH,
Eskes TK. Methylenetetrahydrofolate reductase polymor-
phism affects the change in homocysteine and folate concen-
trations resulting from low dose folic acid supplementation in
women with unexplained recurrent miscarriages. J Nutr. Aug
1998;128(8):1336–1341.
152. Pennisi E. Evolution of developmental diversity. Evo-devo
devotees eye ocular origins and more. Science. 10 May
2002;296(5570):1010–1011.
153. Pasqualetti M, Neun R, Davenne M, Rijli FM. Retinoic acid res-
cues inner ear defects in Hoxa1 deficient mice. Nat Genet. Sep
2001;29(1):34–39.
154. Zhao R, Russell RG, Wang Y, et  al. Rescue of embryonic lethal-
ity in reduced folate carrier-deficient mice by maternal folic acid
supplementation reveals early neonatal failure of hematopoietic
organs. J Biol Chem. 30 Mar 2001;276(13):10224–10228.
155. Finnell RH, Spiegelstein O, Wlodarczyk B, et al. DNA methyla-
tion in Folbp1 knockout mice supplemented with folic acid during
gestation. J Nutr. Aug 2002;132(8 Suppl):2457S-2461S.
156. Pal C, Papp B, Hurst LD. Genomic function:  Rate of evolution
and gene dispensability. Nature. 30 Jan 2003;421(6922):496–497;
discussion 497–498.
157. Stover PJ, Garza C. Bringing individuality to public health recom-
mendations. J Nutr. Aug 2002;132(8 Suppl):2476S-2480S.
158. Gasch AP, Werner-Washburne M. The genomics of yeast responses
to environmental stress and starvation. Funct Integr Genomics. Sep
2002;2(4–5):181–192.
159. van Spronsen FJ. Phenylketonuria management from an
European perspective:  a commentary. Mol Genet Metab. Jun
2010;100(2):107–110.
160. Beaudin AE, Stover PJ. Folate-mediated one-carbon metabo-
lism and neural tube defects:  balancing genome synthesis
and gene expression. Birth Defects Res C Embryo Today. Sep
2007;81(3):183–203.
161. Seckl JR, Meaney MJ. Glucocorticoid programming. Ann N Y
Acad Sci. Dec 2004;1032:63–84.
162. Seckl JR, Walker BR. Minireview:  11beta-hydroxysteroid dehy-
drogenase type 1- a tissue-specific amplifier of glucocorticoid
action. Endocrinology. Apr 2001;142(4):1371–1376.
163. McMillen IC, Robinson JS. Developmental origins of the meta-
bolic syndrome: prediction, plasticity, and programming. Physiol
Rev. Apr 2005;85(2):571–633.
164. Seckl JR. Prenatal glucocorticoids and long-term programming.
Eur J Endocrinol. Nov 2004;151 Suppl 3:U49–62.
165. Reinisch JM, Simon NG, Karow WG, Gandelman R. Prenatal
exposure to prednisone in humans and animals retards intrauterine
growth. Science. 27 Oct 1978;202(4366):436–438.
166. French NP, Hagan R, Evans SF, Godfrey M, Newnham JP.
Repeated antenatal corticosteroids:  size at birth and subsequent
development. Am J Obstet Gynecol. Jan 1999;180(1 Pt 1):114–121.
167. Newnham JP, Evans SF, Godfrey M, Huang W, Ikegami M, Jobe A.
Maternal, but not fetal, administration of corticosteroids restricts
fetal growth. J Matern Fetal Med. May–Jun 1999;8(3):81–87.
168. White PC, Mune T, Agarwal AK. 11 beta-Hydroxysteroid dehy-
drogenase and the syndrome of apparent mineralocorticoid excess.
Endocr Rev. Feb 1997;18(1):135–156.
2 0 6   •   P rincip l e s o f G e no m ic M e dicin e
169. Nyirenda MJ, Lindsay RS, Kenyon CJ, Burchell A, Seckl JR.
Glucocorticoid exposure in late gestation permanently programs
rat hepatic phosphoenolpyruvate carboxykinase and glucocorti-
coid receptor expression and causes glucose intolerance in adult
offspring. J Clin Invest. 15 May 1998;101(10):2174–2181.
170. Goland RS, Jozak S, Warren WB, Conwell IM, Stark RI, Tropper
PJ. Elevated levels of umbilical cord plasma corticotropin-releasing
hormone in growth-retarded fetuses. J Clin Endocrinol Metab. Nov
1993;77(5):1174–1179.
171. Goland RS, Tropper PJ, Warren WB, Stark RI, Jozak SM, Conwell
IM. Concentrations of corticotrophin-releasing hormone in the
umbilical-cord blood of pregnancies complicated by pre-eclampsia.
Reprod Fertil Dev. 1995;7(5):1227–1230.
172. Bertram C, Trowern AR, Copin N, Jackson AA, Whorwood CB.
The maternal diet during pregnancy programs altered expression
of the glucocorticoid receptor and type 2 11beta-hydroxysteroid
dehydrogenase:  potential molecular mechanisms underlying
the programming of hypertension in utero. Endocrinology. Jul
2001;142(7):2841–2853.
173. Stewart PM, Rogerson FM, Mason JI. Type 2 11 beta-hydroxysteroid
dehydrogenase messenger ribonucleic acid and activity in human
placenta and fetal membranes: its relationship to birth weight and
putative role in fetal adrenal steroidogenesis. J Clin Endocrinol
Metab. Mar 1995;80(3):885–890.
174. Drake AJ, Walker BR, Seckl JR. Intergenerational consequences of
fetal programming by in utero exposure to glucocorticoids in rats.
Am J Physiol Regul Integr Comp Physiol. Jan 2005;288(1):R34–R38.
175. Hebbar PB, Archer TK. Chromatin remodeling by nuclear recep-
tors. Chromosoma. May 2003;111(8):495–504.
176. Thomassin H, Flavin M, Espinas ML, Grange T.
Glucocorticoid-induced DNA demethylation and gene memory
during development. Embo J. 17 Apr 2001;20(8):1974–1983.
177. Weinstock M. Alterations induced by gestational stress in brain
morphology and behaviour of the offspring. Prog Neurobiol. Dec
2001;65(5):427–451.
178. Stewart PM, Murry BA, Mason JI. Type 2 11 beta-hydroxysteroid
dehydrogenase in human fetal tissues. J Clin Endocrinol Metab. Jun
1994;78(6):1529–1532.
179. Diaz R, Fuxe K, Ogren SO. Prenatal corticosterone treat-
ment induces long-term changes in spontaneous and
apomorphine-mediated motor activity in male and female rats.
Neuroscience. Nov 1997;81(1):129–140.
180. Dunn AJ, Berridge CW. Physiological and behavioral responses to
corticotropin-releasing factor administration:  is CRF a mediator
of anxiety or stress responses? Brain Res Brain Res Rev. May-Aug
1990;15(2):71–100.
181. Gatford KL, Wintour EM, De Blasio MJ, Owens JA, Dodic M.
Differential timing for programming of glucose homoeostasis, sen-
sitivity to insulin and blood pressure by in utero exposure to dexa-
methasone in sheep. Clin Sci (Lond). May 2000;98(5):553–560.
182. Valera A, Pujol A, Pelegrin M, Bosch F. Transgenic mice over-
expressing phosphoenolpyruvate carboxykinase develop
non-insulin-dependent diabetes mellitus. Proc Natl Acad Sci U S
A. 13 Sep 1994;91(19):9151–9154.
183. Rosella G, Zajac JD, Kaczmarczyk SJ, Andrikopoulos S, Proietto
J. Impaired suppression of gluconeogenesis induced by overexpres-
sion of a noninsulin-responsive phosphoenolpyruvate carboxyki-
nase gene. Mol Endocrinol. Nov 1993;7(11):1456–1462.
184. Weaver IC, Cervoni N, Champagne FA, et al. Epigenetic program-
ming by maternal behavior. Nat Neurosci. Aug 2004;7(8):847–854.
185. Weaver IC, Diorio J, Seckl JR, Szyf M, Meaney MJ. Early environ-
mental regulation of hippocampal glucocorticoid receptor gene
expression: characterization of intracellular mediators and potential
genomic target sites. Ann N Y Acad Sci. Jun 2004;1024:182–212.
186. Jiang X, Yan J, West AA, et  al. Maternal choline intake alters
the epigenetic state of fetal cortisol-regulating genes in humans.
FASEB J. Aug 2012;26(8):3563–3574.
 187. Lanpher B, Brunetti-Pierri N, Lee B. Inborn errors of metabo-
lism: the flux from Mendelian to complex diseases. Nat Rev Genet.
Jun 2006;7(6):449–460.
188. Panagiotou G, Nielsen J. Nutritional systems biology: definitions
and approaches. Annu Rev Nutr. 2009;29:329–339.
189. Arkin AP, Schaffer DV. Network news: innovations in 21st cen-
tury systems biology. Cell. 18 Mar 2011;144(6):844–849.
190. Ideker T, Galitski T, Hood L. A new approach to decoding life: sys-
tems biology. Annu Rev Genomics Hum Genet. 2001;2:343–372.
191. Kitano H. Systems biology:  a brief overview. Science. 1 Mar
2002;295(5560):1662–1664.
192. Chuang HY, Hofree M, Ideker T. A decade of systems biology.
Annu Rev Cell Dev Biol. 10 Nov 2010;26:721–744.
193. Bordbar A, Palsson BO. Using the reconstructed genome-scale
human metabolic network to study physiology and pathology. J
Intern Med. Feb 2012;271(2):131–141.
194. Orth JD, Thiele I, Palsson BO. What is flux balance analysis? Nat
Biotechnol. Mar 2010;28(3):245–248.
195. Chou JY, Mansfield BC. Mutations in the glucose-6-phosphatase-
alpha (G6PC) gene that cause type Ia glycogen storage disease.
Hum Mutat. Jul 2008;29(7):921–930.
196. Froissart R, Piraud M, Boudjemline AM, et  al. Glucose-6-
phosphatase deficiency. Orphanet J Rare Dis. 2011;6:27.
197. Bandsma RH, Smit GP, Kuipers F. Disturbed lipid metabolism in
glycogen storage disease type 1. Eur J Pediatr. Oct 2002;161 Suppl
1:S65–S69.
198. Oti M, Brunner HG. The modular nature of genetic diseases. Clin
Genet. Jan 2007;71(1):1–11.
199. Lee DS, Park J, Kay KA, Christakis NA, Oltvai ZN, Barabasi
AL. The implications of human metabolic network topol-
ogy for disease comorbidity. Proc Natl Acad Sci U S A. 22 Jul
2008;105(29):9880–9885.
200. Goh KI, Cusick ME, Valle D, Childs B, Vidal M, Barabasi AL.
The human disease network. Proc Natl Acad Sci U S A. 22 May
2007;104(21):8685–8690.
201. Park J, Lee DS, Christakis NA, Barabasi AL. The impact of cellular
networks on disease comorbidity. Mol Syst Biol. 2009;5:262.
202. An S, Kumar R, Sheets ED, Benkovic SJ. Reversible compartmen-
talization of de novo purine biosynthetic complexes in living cells.
Science. 4 Apr 2008;320(5872):103–106.
203. Kanehisa M, Goto S, Hattori M, et al. From genomics to chemical
genomics: new developments in KEGG. Nucleic Acids Res. 1 Jan
2006;34(Database issue):D354–D357.
204. Duarte NC, Becker SA, Jamshidi N, et  al. Global recon-
struction of the human metabolic network based on
genomic and bibliomic data. Proc Natl Acad Sci U S A. 6 Feb
2007;104(6):1777–1782.
205. Hamosh A, Scott AF, Amberger JS, Bocchini CA, McKusick VA.
Online Mendelian Inheritance in Man (OMIM), a knowledge
base of human genes and genetic disorders. Nucleic Acids Res. 1 Jan
2005;33(Database issue):D514–D517.
206. Machado D, Costa RS, Rocha M, Ferreira EC, Tidor B, Rocha I.
Modeling formalisms in Systems Biology. AMB Express. 2011;1:45.
207. Orth JD, Palsson BO. Systematizing the generation of missing meta-
bolic knowledge. Biotechnol Bioeng. 15 Oct 2010;107(3):403–412.
208. Heinemann M, Sauer U. Systems biology of microbial metabolism.
Curr Opin Microbiol. Jun 2010;13(3):337–343.
209. Capel F, Klimcakova E, Viguerie N, et al. Macrophages and adipo-
cytes in human obesity: adipose tissue gene expression and insu-
lin sensitivity during calorie restriction and weight stabilization.
Diabetes. Jul 2009;58(7):1558–1567.
210. Zelezniak A, Pers TH, Soares S, Patti ME, Patil KR. Metabolic
network topology reveals transcriptional regulatory signatures of
type 2 diabetes. PLoS Comput Biol. Apr 2010;6(4):e1000729.
211. Deo RC, Hunter L, Lewis GD, et  al. Interpreting metabolomic
profiles using unbiased pathway models. PLoS Comput Biol. Feb
2010;6(2):e1000692.
N utrition a l G e no m ic s   •   2 0 7
212. Vo TD, Paul Lee WN, Palsson BO. Systems analysis of energy
metabolism elucidates the affected respiratory chain complex in
Leigh’s syndrome. Mol Genet Metab. May 2007;91(1):15–22.
213. Ozsolak F, Milos PM. RNA sequencing: advances, challenges and
opportunities. Nat Rev Genet. Feb 2011;12(2):87–98.
214. Mortazavi A, Williams BA, McCue K, Schaeffer L, Wold
B. Mapping and quantifying mammalian transcriptomes by
RNA-Seq. Nat Methods. Jul 2008;5(7):621–628.
215. Lewis NE, Schramm G, Bordbar A, et al. Large-scale in silico mod-
eling of metabolic interactions between cell types in the human
brain. Nat Biotechnol. Dec 2010;28(12):1279–1285.
216. Bordbar A, Feist AM, Usaite-Black R, Woodcock J, Palsson BO,
Famili I. A multi-tissue type genome-scale metabolic network
for analysis of whole-body systems physiology. BMC Syst Biol.
2011;5:180.
217. Dunn WB, Bailey NJ, Johnson HE. Measuring the metab-
olome:  current analytical technologies. Analyst. May
2005;130(5):606–625.
218. Shaham O, Wei R, Wang TJ, et  al. Metabolic profiling of the
human response to a glucose challenge reveals distinct axes of insu-
lin sensitivity. Mol Syst Biol. 2008;4:214.
219. Shlomi T, Cabili MN, Ruppin E. Predicting metabolic biomarkers
of human inborn errors of metabolism. Mol Syst Biol. 2009;5:263.
220. Gille C, Bolling C, Hoppe A, et al. HepatoNet1: a comprehensive
metabolic reconstruction of the human hepatocyte for the analysis
of liver physiology. Mol Syst Biol. 7 Sep 2010;6:411.
221. Folger O, Jerby L, Frezza C, Gottlieb E, Ruppin E, Shlomi T.
Predicting selective drug targets in cancer through metabolic net-
works. Mol Syst Biol. 2011;7:501.
222. Frezza C, Zheng L, Folger O, et al. Haem oxygenase is synthetically
lethal with the tumour suppressor fumarate hydratase. Nature. 8
Sep 2011;477(7363):225–228.
223. Shlomi T, Benyamini T, Gottlieb E, Sharan R, Ruppin E.
Genome-scale metabolic modeling elucidates the role of prolifera-
tive adaptation in causing the Warburg effect. PLoS Comput Biol.
Mar 2011;7(3):e1002018.
224. Jamshidi N, Palsson BO. Systems biology of SNPs. Mol Syst Biol.
2006;2:38.
225. Burgard AP, Nikolaev EV, Schilling CH, Maranas CD. Flux cou-
pling analysis of genome-scale metabolic network reconstructions.
Genome Res. Feb 2004;14(2):301–312.
226. Biomarkers Definitions Working Group. Biomarkers and surrogate
endpoints: preferred definitions and conceptual framework. Clin
Pharmacol Ther. Mar 2001;69(3):89–95.
227. Mamas M, Dunn WB, Neyses L, Goodacre R. The role of metabo-
lites and metabolomics in clinically applicable biomarkers of dis-
ease. Arch Toxicol. Jan 2011;85(1):5–17.
228. Kouskoumvekaki I, Panagiotou G. Navigating the human metabo-
lome for biomarker identification and design of pharmaceutical
molecules. J Biomed Biotechnol. 2011; Jan-Feb;29(1):94–110.
229. Tobert JA. Lovastatin and beyond:  the history of the
HMG-CoA reductase inhibitors. Nat Rev Drug Discov. Jul
2003;2(7):517–526.
230. Weljie AM, Dowlatabadi R, Miller BJ, Vogel HJ, Jirik FR. An
inflammatory arthritis-associated metabolite biomarker pat-
tern revealed by 1H NMR spectroscopy. J Proteome Res. Sep
2007;6(9):3456–3464.
231. McKusick VA. Mendelian Inheritance in Man and its online ver-
sion, OMIM. Am J Hum Genet. Apr 2007;80(4):588–604.
232. Kornman KS, Martha PM, Duff GW. Genetic variations and
inflammation:  a practical nutrigenomics opportunity. Nutrition.
Jan 2004;20(1):44–49.
233. Vijg J, Suh Y. Genetics of longevity and aging. Annu Rev Med.
2005;56:193–212.
234. Seifried HE, Anderson DE, Sorkin BC, Costello RB. Free radi-
cals: the pros and cons of antioxidants. Executive summary report.
J Nutr. Nov 2004;134(11):3143S–3163S.
 235. Zeisel SH. Antioxidants suppress apoptosis. J Nutr. Nov
2004;134(11):3179S-3180S.
236. Stillman RJ, Rosenberg MJ, Sachs BP. Smoking and reproduction.
Fertil Steril. Oct 1986;46(4):545–566.
237. De Rycke M, Liebaers I, Van Steirteghem A. Epigenetic risks
related to assisted reproductive technologies: risk analysis and epi-
genetic inheritance. Hum Reprod. Oct 2002;17(10):2487–2494.
238. Sinclair KD, Dunne LD, Maxfield EK, et  al. Fetal growth and
development following temporary exposure of day 3 ovine
embryos to an advanced uterine environment. Reprod Fertil Dev.
1998;10(3):263–269.
239. Young LE, Fernandes K, McEvoy TG, et al. Epigenetic change in
IGF2R is associated with fetal overgrowth after sheep embryo cul-
ture. Nat Genet. Feb 2001;27(2):153–154.
240. Khosla S, Dean W, Reik W, Feil R. Culture of preimplantation
embryos and its long-term effects on gene expression and pheno-
type. Hum Reprod Update. Jul-Aug 2001;7(4):419–427.
241. Gao S, Latham KE. Maternal and environmental factors in
early cloned embryo development. Cytogenet Genome Res.
2004;105(2–4):279–284.
242. Schneeman BO, Mendelson R. Dietary guidelines:  past
experience and new approaches. J Am Diet Assoc. Oct
2002;102(10):1498–1500.
243. Schneeman BO. Evolution of dietary guidelines. J Am Diet Assoc.
Dec 2003;103(12 Suppl 2):S5–S9.
244. Weinshilboum R. Inheritance and drug response. N Engl J Med. 6
Feb 2003;348(6):529–537.
245. Solis C, Veenema K, Ivanov AA, et al. Folate intake at RDA lev-
els is inadequate for Mexican American men with the methy-
lenetetrahydrofolate reductase 677TT genotype. J Nutr. Jan
2008;138(1):67–72.
246. Hunt JR, Zeng H. Iron absorption by heterozygous carriers of the
HFE C282Y mutation associated with hemochromatosis. Am J
Clin Nutr. Oct 2004;80(4):924–931.
247. Swanson CA. Iron intake and regulation:  implications for iron
deficiency and iron overload. Alcohol. Jun 2003;30(2):99–102.
248. Moirand R, Guyader D, Mendler MH, et  al. HFE based
re-evaluation of heterozygous hemochromatosis. Am J Med Genet.
1 Sep 2002;111(4):356–361.
249. Baier LJ, Hanson RL. Genetic studies of the etiology of type 2 dia-
betes in Pima Indians: hunting for pieces to a complicated puzzle.
Diabetes. May 2004;53(5):1181–1186.
250. Tseung CW, McMahon LG, Vazquez J, Pohl J, Gregory JF, 3rd.
Partial amino acid sequence and mRNA analysis of cytosolic
pyridoxine-beta-D-glucoside hydrolase from porcine intestinal
mucosa: proposed derivation from the lactase-phlorizin hydrolase
gene. Biochem J. 15 May 2004;380(Pt 1):211–218.
251. Mackey AD, McMahon RJ, Townsend JH, Gregory JF, 3rd. Uptake,
hydrolysis, and metabolism of pyridoxine-5′-beta-D-glucoside in
Caco-2 cells. J Nutr. Apr 2004;134(4):842–846.
252. Csete M, Doyle J. Bow ties, metabolism and disease. Trends
Biotechnol. Sep 2004;22(9):446–450.
253. Honein MA, Paulozzi LJ, Mathews TJ, Erickson JD, Wong
LY. Impact of folic acid fortification of the US food sup-
ply on the occurrence of neural tube defects. JAMA. 20 Jun
2001;285(23):2981–2986.
254. Delange FM. Control of iodine deficiency in Western and Central
Europe. Cent Eur J Public Health. Sep 2003;11(3):120–123.
255. Center for Disease Control. Alcohol consumption among women
who are pregnant or who might become pregnant—United States,
2002. Morb Mortal Wkly Rep. 2004;53(50):1178–1181.
256. Mellott TJ, Williams CL, Meck WH, Blusztajn JK. Prenatal
choline supplementation advances hippocampal development
and enhances MAPK and CREB activation. FASEB J. Mar
2004;18(3):545–547.
257. Heird WC, Lapillonne A. The role of essential fatty acids in devel-
opment. Annu Rev Nutr. 2005;25:549–571.
2 0 8   •   P rincip l e s o f G e no m ic M e dicin e
258. Jacques PF, Bostom AG, Williams RR, et  al. Relation between
folate status, a common mutation in methylenetetrahydrofolate
reductase, and plasma homocysteine concentrations. Circulation.
1 Jan 1996;93(1):7–9.
259. Tsai MY, Yang F, Bignell M, Aras O, Hanson NQ. Relation between
plasma homocysteine concentration, the 844ins68 variant of the
cystathionine beta-synthase gene, and pyridoxal-5′-phosphate con-
centration. Mol Genet Metab. Aug 1999;67(4):352–356.
260. Afman LA, Trijbels FJ, Blom HJ. The H475Y polymorphism in
the glutamate carboxypeptidase II gene increases plasma folate
without affecting the risk for neural tube defects in humans. J Nutr.
Jan 2003;133(1):75–77.
261. Devlin AM, Ling EH, Peerson JM, et al. Glutamate carboxypep-
tidase II:  a polymorphism associated with lower levels of serum
folate and hyperhomocysteinemia. Hum Mol Genet. 22 Nov
2000;9(19):2837–2844.
262. Uitterlinden AG, Fang Y, Bergink AP, van Meurs JB, van
Leeuwen HP, Pols HA. The role of vitamin D receptor gene
polymorphisms in bone biology. Mol Cell Endocrinol. 29 Nov
2002;197(1–2):15–21.
263. Griffiths W, Cox T. Haemochromatosis: novel gene discovery and
the molecular pathophysiology of iron metabolism. Hum Mol
Genet. Oct 2000;9(16):2377–2382.
264. Geller DS. A genetic predisposition to hypertension? Hypertension.
Jul 2004;44(1):27–28.
265. Jeck N, Waldegger S, Lampert A, et al. Activating mutation of the
renal epithelial chloride channel ClC-Kb predisposing to hyper-
tension. Hypertension. Jun 2004;43(6):1175–1181.
266. Bentzen J, Jorgensen T, Fenger M. The effect of six polymorphisms
in the apolipoprotein B gene on parameters of lipid metabolism in
a Danish population. Clin Genet. Feb 2002;61(2):126–134.
267. Hubacek JA, Pistulkova H, Skodova Z, Berg K, Poledne R.
Association between apolipoprotein B promotor haplotypes and
cholesterol status. Ann Clin Biochem. Jul 2001;38(Pt 4):399–400.
268. Brown S, Ordovas JM, Campos H. Interaction between the
APOC3 gene promoter polymorphisms, saturated fat intake and
plasma lipoproteins. Atherosclerosis. Oct 2003;170(2):307–313.
269. Hein DW. Molecular genetics and function of NAT1 and
NAT2:  role in aromatic amine metabolism and carcinogenesis.
Mutat Res. 30 Sep 2002;506: 65–77.
270. Hein DW, Doll MA, Fretland AJ, et al. Molecular genetics and epi-
demiology of the NAT1 and NAT2 acetylation polymorphisms.
Cancer Epidemiol Biomarkers Prev. Jan 2000;9(1):29–42.
271. Ferre N, Camps J, Fernandez-Ballart J, et al. Regulation of serum
paraoxonase activity by genetic, nutritional, and lifestyle factors in
the general population. Clin Chem. Sep 2003;49(9):1491–1497.
272. Chistyakov DA, Savost’anov KV, Zotova EV, Nosikov VV.
Polymorphisms in the Mn-SOD and EC-SOD genes and their
relationship to diabetic neuropathy in type 1 diabetes mellitus.
BMC Med Genet. 2001;2(1):4.
273. Van Landeghem GF, Tabatabaie P, Kucinskas V, Saha N, Beckman
G. Ethnic variation in the mitochondrial targeting sequence poly-
morphism of MnSOD. Hum Hered. Jul 1999;49(4):190–193.
274. Messier W, Stewart CB. Episodic adaptive evolution of primate
lysozymes. Nature. 9 Jan 1997;385(6612):151–154.
275. Zhang J, Zhang YP, Rosenberg HF. Adaptive evolution of a dupli-
cated pancreatic ribonuclease gene in a leaf-eating monkey. Nat
Genet. Apr 2002;30(4):411–415.
276. Wu W, Goodman M, Lomax MI, Grossman LI. Molecular evo-
lution of cytochrome c oxidase subunit IV:  evidence for positive
selection in simian primates. J Mol Evol. May 1997;44(5):477–491.
277. Wooding SP, Watkins WS, Bamshad MJ, Dunn DM, Weiss RB,
Jorde LB. DNA sequence variation in a 3.7-kb noncoding sequence
5′ of the CYP1A2 gene: implications for human population history
and natural selection. Am J Hum Genet. Sep 2002;71(3):528–542.
278. Bailey LB, Gregory JF, 3rd. Folate metabolism and requirements.
J Nutr. Apr 1999;129(4):779–782.
 279. von Linsingen R, Bompeixe EP, Bicalho Mda G. A case-control
study in IL6 and TGFB1 gene polymorphisms and recurrent
spontaneous abortion in southern Brazilian patients. Am J Reprod
Immunol. Feb 2005;53(2):94–99.
280. Prigoshin N, Tambutti M, Larriba J, Gogorza S, Testa R. Cytokine
gene polymorphisms in recurrent pregnancy loss of unknown
cause. Am J Reprod Immunol. Jul 2004;52(1):36–41.
281. Daher S, Shulzhenko N, Morgun A, et  al. Associations between
cytokine gene polymorphisms and recurrent pregnancy loss.
J Reprod Immunol. Feb 2003;58(1):69–77.
282. Perni SC, Vardhana S, Tuttle SL, Kalish RB, Chasen ST, Witkin
SS. Fetal interleukin-1 receptor antagonist gene polymor-
phism, intra-amniotic interleukin-1beta levels, and history of
spontaneous abortion. Am J Obstet Gynecol. Oct 2004;191(4):
1318–1323.
283. Sata F, Yamada H, Yamada A, et al. A polymorphism in the CYP17
gene relates to the risk of recurrent pregnancy loss. Mol Hum
Reprod. Nov 2003;9(11):725–728.
284. Suryanarayana V, Deenadayal M, Singh L. Association of
CYP1A1 gene polymorphism with recurrent pregnancy loss in
N utrition a l G e no m ic s   •   2 0 9
the South Indian population. Hum Reprod. Nov 2004;19(11):
2648–2652.
285. Schweikert A, Rau T, Berkholz A, Allera A, Daufeldt S, Wildt
L. Association of progesterone receptor polymorphism with
recurrent abortions. Eur J Obstet Gynecol Reprod Biol. 15 Mar
2004;113(1):67–72.
286. Sata F, Yamada H, Kondo T, et al. Glutathione S-transferase M1
and T1 polymorphisms and the risk of recurrent pregnancy loss.
Mol Hum Reprod. Mar 2003;9(3):165–169.
287. Finan RR, Tamim H, Ameen G, Sharida HE, Rashid M, Almawi
WY. Prevalence of factor V G1691A (factor V-Leiden) and pro-
thrombin G20210A gene mutations in a recurrent miscarriage
population. Am J Hematol. Dec 2002;71(4):300–305.
288. Tempfer C, Unfried G, Zeillinger R, Hefler L, Nagele F, Huber
JC. Endothelial nitric oxide synthase gene polymorphism in
women with idiopathic recurrent miscarriage. Hum Reprod. Aug
2001;16(8):1644–1647.
289. Gloria-Bottini F, Lucarini N, Palmarino R, et al. Phosphoglucomu­
tase genetic polymorphism of newborns. Am J Hum Biol. Jan–Feb
2001;13(1):9–14.
 13.
GENOMICS IN PUBLIC AND POPULATION HEALTH
Anastasia L. Wise and Teri A. Manolio
INTRODUCTION
Public health seeks to improve health at a population level
through interventions that increase the net health benefit
to the population as a whole. Advances in genomics knowl-
edge and technologies can add to this endeavor, but they
also pose a challenge when faced with often conflicting pub-
lic health (population) and genomic medicine (individual)
perspectives. Combining the fields of genomic, popula-
tion, and social sciences, “population genomics” or “public
health genomics” looks at the promotion of health and pre-
vention of disease using genomic knowledge through the
lens of populations rather than individuals. In this chapter,
we will survey the three major disciplines contributing to
population genomics (genomic, population, and social sci-
ences) and explore two cross-cutting issues—global health
and population versus individual health—using specific
examples from diseases such as asthma, colon cancer, and
cystic fibrosis. Although the fields of population genom-
ics and genomic medicine look to prevent or treat disease
through different perspectives, they can act complementa-
rily to enhance overall health outcomes for both individuals
and populations at large.
M AJOR DISCIPLINES CONTRIBUTING
TO P O P U L AT I O N G E N O M I C S
Population genomics seeks to integrate knowledge from
genomic, population, and social sciences to improve popu-
lation health.
GENOMIC SCIENCES
The genomic sciences focus on studying whole genomes,
such as the entire DNA sequence making up the human
genome. Through studying genomics, researchers can iden-
tify genetic variants influencing human health. Studies of
210
particular genes can then further elucidate the function of
genetic variants.
Genomic information can be used clinically to inform
determinations of disease risk, diagnosis, drug selec-
tion, and drug dosing. Colorectal cancer provides a good
example of an area where population level screening along
with genomic medicine approaches are coming together to
improve overall population health.
Over 1 million individuals are diagnosed with colorec-
tal cancer each year worldwide, accounting for approxi-
mately 9–10% of cancer diagnoses in 2008.1 It is the third
leading cause of cancer-related death in the United States
and the fourth worldwide.1,2 In colorectal cancer, a patient’s
genomic information can be used to determine risk of
inherited colorectal cancer syndromes, whether certain bio-
logical agents will work in specific patients, and what start-
ing dose to use on specific chemotherapeutics.
As many as 20–25% of colorectal cancer cases have
a family history of colorectal cancer (two or more first-
degree relatives with colorectal cancer), yet only 5–6%
have an established familial genetic syndrome with a
known genetic variant.3,4 Of those with established famil-
ial genetic syndromes, approximately 3% will be diag-
nosed with Lynch syndrome (including variants in the
genes MLH1, MSH2, MSH6, PMS2, and EPCAM) and
1% with familial adenomatous polyposis (FAP; including
variants in APC and MUTYH).3 Individuals with a fam-
ily history of colorectal cancer have a two- to threefold
greater risk of developing colorectal cancer than the gen-
eral population; therefore, genetic testing for individuals
with a known family history conveys a substantial pub-
lic health benefit.3 Those with a known family history of
colorectal cancer are also recommended to be screened at
younger ages, typically 10 years younger than the onset of
the youngest case in their family.
Genetic testing is also used to determine treatment
options in colorectal cancer (Table 13.1).3,4 For example,
genetic variants that make KRAS constitutively active have
Table 13.1  PHARMACOGENOMIC VARIANTS IN COLORECTAL CANCER 3
GENE/VARIANT FUNCTION
BRAF Downstream pathways constitutively active
ERCC-1 DNA excision repair
Interleukin 8 Increased VEGF expression
KRAS Downstream pathways constitutively active
Microsatellite instability Reduced DNA repair
TSER Increased or decreased thymidylate synthase, depending
on variant
UTGA1 Responsible for metabolism of irinotecan
VEGF Increased VEGF expression
been shown to provide resistance to monoclonal antibod-
ies directed against the upstream epidermal growth factor
receptor (EGFR), as both are components of a cellular
pathway leading to abnormal cell growth and cancer. Thus,
cetuximab and panitumumab (anti-EGFR antibodies) are
given only to individuals with normally functioning KRAS,
where blocking EGFR can have an effect.3
Pharmacogenomics can also be useful in determining
drug dosage for colorectal cancer. For example, the Food
and Drug Administration (FDA) recommends testing
for UGT1A1 variants when administering irinotecan, as
individuals who are homozygous or heterozygous for the
UGT1A1*28 allele are at increased risk of developing neu-
tropenia and severe infections.5 Individuals with inactivat-
ing UGT1A1 variants are therefore recommended to be
started at a lower dosage of irinotecan to reduce the risk of
neutropenia.5
In addition to modifying drug dosing, pharmacoge-
nomic information can also be used in drug selection,
to choose agents more likely to give a beneficial response
based on a patient’s genetically driven ability to metabolize
them. For example, in patients of Asian ancestry who are
given carbamazepine (used to treat epilepsy and bipolar
disorder) the HLA-B*1502 allele has been associated with
Stevens-Johnson syndrome/toxic epidermal necrolysis, a
life-threatening skin condition. This allele can be found
in over 15% of the population in some regions in Asia,
including Hong Kong, Thailand, Malaysia, and parts of the
Philippines, and is very rare in other populations outside
Asia.6
Within populations of Asian ancestry, there can also
be great variation, such as is seen within China, where the
HLA-B*1502 allele frequency varies from 0–36% depend-
ing upon ethnicity (Table 13.2). Thus, the FDA recom-
mends genetic testing for the HLA-B*1502 variant before
prescribing carbamazepine for patients of Asian ancestry,
G enomics in P ublic and P opulation H ealth   •   2 1 1
CONSEQUENCES
Resistance to anti-EGFR monoclonal antibodies
Resistance to platinum-based chemotherapy drugs
Increased cancer recurrence
Resistance to anti-EGFR monoclonal antibodies
Improved prognosis
Response to fluorouracil reduced/increased (negative
relationship)
Dosing for irinotecan
Increased cancer recurrence
and for those with one or two copies of the variant, only
treating with carbamazepine when the benefits outweigh
the risks of the drug.6
P O P U L AT I O N S C I E N C E S
The population sciences, such as epidemiology, focus
on studying whole populations rather than individuals.
Through studying environmental, genomic, and social fac-
tors that affect human health, population-level interven-
tions can be identified.
Population variation is an important consideration
when studying common complex conditions that are influ-
enced by multiple genetic, environmental, and social risk
factors, such as asthma. Over 300 million individuals of
all ages worldwide have asthma.7 Prevalence estimates can
vary greatly by ethnicity, however, from 2–33%.8 In the
United States, prevalence ranges from approximately 8%
in European Americans to 12% in African Americans and
7% in Hispanic Americans.9 Within admixed populations,
such as Hispanic Americans, even greater variation can
be seen when populations are further sub-stratified, with
Mexican American populations around 6%, while Puerto
Rican populations are closer to 19%.9 Genetic studies have
shown that at least some of this variation is due to differ-
ences in genetic variants, with 35–80% of the variation in
asthma heritability explained by genetic factors.10,11 For
example, variants in ADAM33 have been seen in European,
African American, and some Hispanic populations, but not
in other European American, Mexican, Puerto Rican, and
Korean populations, all of which found different variants in
ADAM33 associated with asthma (Figure 13.1).12
Studying the interplay between environmental,
genetic, and social risk factors is also critical to under-
standing the etiology of this complex disease. For exam-
ple, the effect of air pollution on asthma case reports is
Table 13.2  HLA-B*1502 ALLELE FREQUENCY IN revealed differences in the bacterial populations present
WORLDWIDE POPULATIONS 46 in individuals with Crohn’s disease (a form of inflamma-
POPULATION HL A-B*1502 ALLELE FREQUENCY tory bowel disease).20 The genomic signatures of the gut
China 36%–0% microbiome in patients with Crohn’s disease show some
Indonesia 17%–11%
bacterial populations to be decreased, while others are
more abundant.20–23
Malaysia 16%–2%
Vietnam 14%
Thailand 9%–8% SOCIAL SCIENCES
India 6%–0%
The social sciences focus on studying society and human
Singapore 6% behavior through fields such as anthropology, economics,
Taiwan 6%–4% law, psychology, and sociology. The study of the ethical,
United States 4%–0% legal, and social implications (ELSI) of genomics plays
South Korea 2%–0.2% an important role in applying genomics to population
Australia 1%–0% health.
Japan 0.10% For example, though BRCA1 and BRCA2 variants
Germany 0% occur in 5–10% of breast cancer cases, these variants are
Brazil 0% found in less than 1% of the general population.24 Specific
Bulgaria 0% populations, such as those of Ashkenazi Jewish descent,
have an increased frequency of BRCA1 or BRCA2 variants.
Burkina Faso 0%
Two variants in BRCA1 and one in BRCA2 are found at
Cuba 0%
a frequency five times higher in Ashkenazi Jews than the
Ireland 0%
general population.25,26
Italy 0%
For these reasons, family members of those with known
Mexico 0% BRCA1/2 variants or those with a family history of breast
Morocco 0% cancer may be offered genetic testing. Men with BRCA1/2
Oman 0% variants are also at an increased risk of developing breast
South Africa 0% cancer.27,28 Thousands of variants have been discovered in
BRCA1 and BRCA2, yet only a minority are of known del-
eterious effect.29 Genetic testing therefore has the possibil-
modified by genetic factors as well, showing potential ity of finding a variant of unknown effect, whose functional
gene–environment interactions. A key measure of air pol- significance is unclear. One recent study found that 10% of
lution is the PM10 level, the concentration in parts per women undergoing BRCA1 and BRCA2 testing receive an
million of particulate matter that is 10 micrometers in ambiguous test result due to the detection of a variant of
diameter or less, which can penetrate and irritate small unknown significance.30 Thus, while deleterious variants
airways. PM10 has been shown in multiple epidemiologi- are known to increase the risk of developing breast cancer
cal studies to be an independent risk factor for increased approximately fivefold, deciding how to react to variants of
respiratory symptoms, including asthma.13–17 Similarly, unknown effect can be challenging for all involved:  both
variants in over 100 genes have been associated with clinicians and patients.31
asthma in genome-wide association (GWA) studies.12,18,19 Many genetic loci have also been associated with mul-
Looking at the two risk factors together, though, reveals tiple phenotypes, as evidenced in the NHGRI Catalog of
a potential gene–environment interaction where variants Published GWAS19,32 (Figure 13.2). Such pleiotropic genes
in GSTP1, SOD2, and NFE2L2, all related to oxidative (i.e., genes associated with multiple phenotypes) can pres-
stress pathways, were also associated with increased hos- ent additional challenges when we are considering the ELSI
pital admissions for asthma-related symptoms during days of returning genetic testing results. For example, variants
with high PM10 levels.13 in APOE are associated with multiple phenotypes, includ-
Genomics can also be used to help identify and bet- ing Alzheimer’s disease, cholesterol level, coronary disease,
ter define environmental risk factors in population C-reactive protein, hyperlipoproteinemia type III, low
studies. For example, genomic data profiling the bacte- density lipoprotein (LDL) level, macular degeneration,
ria inhabiting the human gut, or gut microbiome, has and response to statin therapy.33 The APOE*e4 variant

2 1 2   •   P rinciples of G enomic M edicine

 SNP: rs2280089
Ancestral Allele: G
Derived Allele: A
60°
30°
0°
−30°
0° 30°
Example of the variation in allele frequency by population for rs2280089 in ADAM33. The A allele has been previously associated with
Figure 13.1 
predisposition to asthma and bronchial hyper-responsiveness in populations from the United States, the United Kingdom, and China.48–51
in particular has been associated with increased risk for
developing both Alzheimer’s disease and atherosclerosis,
along with a protective effect against developing macular
degeneration.33
CROSS-CUTTING ISSUES
O F   P O P U L AT I O N G E N O M I C S
Though all three of the sciences contributing to popula-
tion genomics work together, there are also some issues that
more broadly span the field of population genomics and its
relationship to medicine and public health. As was touched
upon in many of the examples above, it is important to
consider the broader implications of population genomics
to global health and how population- and individual-level
views of health can work together to improve health
worldwide.
G L O BA L H E A LT H
Cardiovascular disease (CVD) is a leading cause of death
worldwide, with over 13.5  million deaths from ischemic
heart disease, stroke, or another form of cerebrovascu-
lar disease in 2008, and it is highly amenable to study
using population genomics techniques.34 For example,
G enomics in P ublic and P opulation H ealth   •   2 1 3
30°
0°
270° 300°
60° 90° 120° 150°
adding rs10757274 genotyping to the Framingham risk
score improved its ability to determine the individuals who
would suffer later cardiovascular events, independent of
their family history.35 Such models can be used to screen
populations to determine individuals at increased risk of
disease and recommend further testing and individualized
genomic medicine.
While chronic conditions such as cardiovascular disease
make up the majority of deaths in the developed world,
infections are still a major health concern within devel-
oping countries and are equally amenable to study using
population genomics. Genomics has made possible the
rapid identification of the organisms causing recent pan-
demic outbreaks, including H1N1 (swine flu) and severe
acute respiratory syndrome (SARs), as well as identifying
the source of foodborne illnesses. Genomic sequence infor-
mation on malaria parasites, mosquito vectors, and their
human hosts is being leveraged to produce more rapid diag-
nosis and better drugs, vaccines, and intervention strategies
to fight malaria.36,37
To maximize the benefit of population genomics
advances to global health, it is also important to include
multiple populations of diverse ages, ethnicities, and gen-
ders in disease research. As evidenced by the example of
asthma genomics above, the prevalence of disease can be
highly variable across ancestral groups, and genetic variants
 2011 3rd quarter

9 10 11 12
8
7
3 6
4 5
1
2

Y
19 21
20
22
16 18
15 17
14
13
X

APOE (19q13.32)
14 Alzheimer's disease GWAS
8 Cholesterol GWAS
19 3 C-reactive protein GWAS
1 Response to statin therapy GWAS
1 Brain iamging GWAS

The NHGRI GWAS Catalog showing that many genetic loci are associated with multiple phenotypes. APOE on 19q13.32 is highlighted,
Figure 13.2 
along with examples of the disease phenotypes associated with the gene.19

often vary in frequency as well. Thus, while a single pathway
may be implicated in disease across many populations, the
most frequent variant in each population may lie in differ-
ent genes or gene regions.
Local environmental and social factors that affect dis-
ease and population health should also be incorporated
into studies of population genomics to produce the most
complete picture of disease etiology. For example, the
prevalence of type 2 diabetes is increasing globally and
has been associated with multiple genetic (more than 60
genes to date), epigenetic (such as methylation or histone
modification), environmental (such as diet), and social fac-
tors (such as exercise), all of which contribute to this com-
plex disease.38 The prevalence of type 2 diabetes varies by
country, from approximately 5–29% with risk alleles such
as the C allele in rs11196205, and decreasing in frequency
from sub-Saharan Africa to Asia.39,40 (Figure 13.3). Effects
of other risk factors also vary across different populations,
with the relative risk of diabetes for each 5-kg/m2 increase
in body mass index (BMI), for example, being 2.4 in Asian
Americans, 2.2 in Hispanic Americans, 2.0 in European
Americans, and 1.6 in African Americans.41

2 1 4   •   P rinciples of G enomic M edicine

 SNP: rs11196205 30°
Ancestral Allele: C
Derived Allele: G
60° 0°

270° 300°

30°

0°

−30°

0° 30° 60° 90° 120° 150°

Figure 13.3  Example of the variation in allele frequency by population for rs11196205 in TCF7L2, a SNP previously associated with type 2 diabetes
risk.40,48

P O P U L AT I O N VE R S US I N D I VI D UA L the most common variant associated with cystic fibrosis

H E A LT H
In many ways, the family serves as an intermediary between
individual- and population-level views of health. It is an
important viewpoint that should be considered in popu-
lation genomics, as genomic information is inherently rel-
evant not only to the individual tested, but also to their
family members, with whom they share a large propor-
tion of their genetic variants. How and with whom such
family-related health information can or should be shared is
an important consideration for advancing both individual
and family health.
The availability of genomic information is also blurring
the line between population- and individual-level views of
health. For example, genetic testing for cystic fibrosis cur-
rently spans population screening–based carrier, prenatal,
and newborn tests to individualized genomic medicine–
based diagnostic and pharmacogenomic testing for treat-
ment selection. From the population screening perspective,
genetic testing is offered to prospective parents of European
descent and others who may be at increased risk of having
a child affected by cystic fibrosis, as the prevalence of cystic
fibrosis is highest in Northern Europe.42
Over 1500 variants have been found in the CFTR gene,
but the functional significance of many is unknown, with
being ∆F508.42 In 2012 the FDA approved ivacaftor, the
first drug to treat a specific cystic fibrosis variant, G551D in
CFTR (Table 13.3).43,44
The G551D variant impairs the ability of the CFTR
channel to open.42,43,45 Ivacaftor functions by increasing the
likelihood that the CFTR channel will be open, improving
chloride transport and restoring the function of the CFTR
gene.43–45 As the cost of genomic sequencing continues to
drop and electronic health records improve, the cost of col-
lecting and interpreting genomic data may fall below the
cost of conducting individual genetic tests, further blurring
the line between clinical and public health data.
S U M M A RY
In this chapter, we have explored how the integration of
genomic, population, and social sciences in population
genomics can improve health, through examples in phar-
macogenomics, population variation, and genetic pleiot-
ropy. We have also investigated cross-cutting issues in global
health and population versus individual health where pop-
ulation genomics can play a crucial role in the translation
of genomic health discoveries worldwide, and population

G enomics in P ublic and P opulation H ealth   •   2 1 5

Table 13.3  NHLBI EXOME SEQUENCING PROJECT
RESULTS FOR AFRICAN-AMERICAN AND
EUROPEAN-AMERICAN PARTICIPANTS FOR THE
15. Preutthipan A, Udomsubpayakul U, Chaisupamongkollarp T,
G551D VARIANT (RS75527207) ASSOCIATED WITH
CYSTIC FIBROSIS 47
16. Gordian ME, Choudhury AH. PM10 and asthma medication in
COHORT ALLELE COUNT A ALLELE COUNT G
African-American 0 4406
European-American 18 8582
screening can work with genomic medicine to provide the
greatest health benefit to both individuals and populations
at large. Thus multidisciplinary research in population
genomics can improve clinical care through improving our
understanding of the genetic variation in populations that
contributes to complex disease.
REFERENCES
1. Bray F, Ren JS, Masuyer E, Ferlay J. Global estimates of cancer
prevalence for 27 sites in the adult population in 2008. Int J Cancer.
2013;132(5):1133–1145.
2. American Cancer Society. Cancer Facts & Figures, 2012. http://
www.cancer.org/acs/groups/content/@epidemiologysurveilance/
documents/document/acspc-031941.pdf. Accessed June 27th, 2012.
3. Cunningham D, Atkin W, Lenz HJ, et al. Colorectal cancer. Lancet.
20 Mar 2010;375(9719):1030–1047.
4. Gala M, Chung DC. Hereditary colon cancer syndromes. Semin
Oncol. Aug 2011;38(4):490–499.
5. US Food and Drug Administration. Dugs@FDA:  Irinotecan
label and approval history. http://www.accessdata.fda.gov/scripts/
c d er / d r ug s atf da / in d e x .cf m ? f us e a c ti on = S e arc h . L a b e l _
ApprovalHistory#labelinfo.
6. US Food and Drug Administration. Drugs@FDA: Carbamazepine
label and approval history. http://www.accessdata.fda.gov/scripts/
cder/drugsatfda/.
7. Masoli M, Fabian D, Holt S, Beasley R. The global burden of
asthma: executive summary of the GINA Dissemination Committee
report. Allergy. May 2004;59(5):469–478.
8. Sembajwe G, Cifuentes M, Tak SW, Kriebel D, Gore R, Punnett L.
National income, self-reported wheezing and asthma diagnosis from
the World Health Survey. Eur Respir J. Feb 2010;35(2):279–286.
9. Current Asthma Prevalence Percents by Age, United States. 2010
National Health Interview Survey (NHIS) Data. http://www.cdc.
gov/asthma/nhis/2010/table4-1.htm.
10. Nieminen MM, Kaprio J, Koskenvuo M. A population-based study of
bronchial asthma in adult twin pairs. Chest. Jul 1991;100(1):70–75.
11. Duffy DL, Martin NG, Battistutta D, Hopper JL, Mathews JD.
Genetics of asthma and hay fever in Australian twins. Am Rev Respir
Dis. Dec 1990;142(6 Pt 1):1351–1358.
12. Drake KA, Galanter JM, Burchard EG. Race, ethnicity and social
class and the complex etiologies of asthma. Pharmacogenomics. Apr
2008;9(4):453–462.
13. Canova C, Dunster C, Kelly FJ, et al. PM10-induced hospital admis-
sions for asthma and chronic obstructive pulmonary disease:  the
modifying effect of individual characteristics. Epidemiology. Jul
2012;23(4):607–615.
14. Weinmayr G, Romeo E, De Sario M, Weiland SK, Forastiere
F. Short-term effects of PM10 and NO2 on respiratory health
2 1 6   •   P rinciples of G enomic M edicine
among children with asthma or asthma-like symptoms:  a sys-
tematic review and meta-analysis. Environ Health Perspect. Apr
2010;118(4):449–457.
Pentamwa P. Effect of PM10 pollution in Bangkok on children with
and without asthma. Pediatr Pulmonol. Mar 2004;37(3):187–192.
schoolchildren. Arch Environ Health. Jan 2003;58(1):42–47.
17. Donaldson K, Gilmour MI, MacNee W. Asthma and PM 10.
Respir Res. 2000;1(1):12–15.
18. Ober C, Hoffjan S. Asthma genetics 2006:  the long and winding
road to gene discovery. Genes Immun. Mar 2006;7(2):95–100.
19. Hindorff LA, Sethupathy P, Junkins HA, et  al. Potential etio-
logic and functional implications of genome-wide association
loci for human diseases and traits. Proc Natl Acad Sci U S A. Jun 9
2009;106(23):9362–9367.
20. Nagalingam NA, Lynch SV. Role of the microbiota in inflammatory
bowel diseases. Inflamm Bowel Dis. May 2012;18(5):968–984.
21. Frank DN, Robertson CE, Hamm CM, et  al. Disease phenotype
and genotype are associated with shifts in intestinal-associated
microbiota in inflammatory bowel diseases. Inflamm Bowel Dis. Jan
2011;17(1):179–184.
22. Walker AW, Sanderson JD, Churcher C, et  al. High-throughput
clone library analysis of the mucosa-associated microbiota reveals
dysbiosis and differences between inflamed and non-inflamed
regions of the intestine in inflammatory bowel disease. BMC
Microbiol. 2011;11:7.
23. Gophna U, Sommerfeld K, Gophna S, Doolittle WF, Veldhuyzen
van Zanten SJ. Differences between tissue-associated intestinal
microfloras of patients with Crohn’s disease and ulcerative colitis. J
Clin Microbiol. Nov 2006;44(11):4136–4141.
24. Schwartz GF, Hughes KS, Lynch HT, et  al. Proceedings of the
International Consensus Conference on Breast Cancer Risk,
Genetics, and Risk Management, April 2007. Cancer. 15 Nov
2008;113(10):2627–2637.
25. Struewing JP, Hartge P, Wacholder S, et  al. The risk of can-
cer associated with specific mutations of BRCA1 and BRCA2
among Ashkenazi Jews. N Engl J Med. 15 May 1997;336(20):
1401–1408.
26. Warner E, Foulkes W, Goodwin P, et al. Prevalence and penetrance
of BRCA1 and BRCA2 gene mutations in unselected Ashkenazi
Jewish women with breast cancer. J Natl Cancer Inst. 21 Jul
1999;91(14):1241–1247.
27. Thompson D, Easton DF. Cancer incidence in BRCA1 mutation
carriers. J Natl Cancer Inst. 18 Sep 2002;94(18):1358–1365.
28. Breast Cancer Linkage Consortium. Cancer risks in BRCA2 muta-
tion carriers. J Natl Cancer Inst. 4 Aug 1999;91(15):1310–1316.
29. An Open Access On-Line Breast Cancer Mutation Data Base.
http://research.nhgri.nih.gov/bic/?CFID=313172&CFTO
KEN=38988484.
30. Peshkin BN, DeMarco TA, Brogan BM, Lerman C, Isaacs C.
BRCA1/2 testing: complex themes in result interpretation. J Clin
Oncol. May 1 2001;19(9):2555–2565.
31. Howlader N, Noone A, Krapcho M, et  al., eds. SEER Cancer
Statistics Review, 1975–2009 (Vintage 2009 Populations). Bethesda,
MD: National Cancer Institute; 2012: available at http://seer.can-
cer.gov/csr/1975_2009_pops09/. Accessed June 27th, 2012.
32. Sivakumaran S, Agakov F, Theodoratou E, et al. Abundant pleiot-
ropy in human complex diseases and traits. Am J Hum Genet. 11
Nov 2011;89(5):607–618.
33. Online Mendelian Inheritance in Man, OMIM 2012;
Apolipoprotein E.  Available at http://omim.org/entry/107741?se
arch=APOE&highlight=apoe#contributors-shutter. Accessed June
24th, 2012.
34. World Health Organization. Death: top 10 causes. WHO Factsheet,
available at http://www.who.int/mediacentre/factsheets/fs310/
en/index.html.
 35. Talmud PJ, Cooper JA, Palmen J, et al. Chromosome 9p21.3 coro-
nary heart disease locus genotype and prospective risk of CHD in
healthy middle-aged men. Clin Chem. Mar 2008;54(3):467–474.
36. Volkman SK, Neafsey DE, Schaffner SF, Park DJ, Wirth DF.
Harnessing genomics and genome biology to understand malaria
biology. Nat Rev Genet. May 2012;13(5):315–328.
37. Agnandji ST, Lell B, Soulanoudjingar SS, et al. First results of Phase
3 trial of RTS,S/AS01 malaria vaccine in African children. N Engl J
Med. 17 Nov 2011;365(20):1863–1875.
38. McCarthy MI. Genomics, type 2 diabetes, and obesity. N Engl J
Med. 9 Dec 2010;363(24):2339–2350.
39. Danaei G, Finucane MM, Lu Y, et al. National, regional, and global
trends in fasting plasma glucose and diabetes prevalence since
1980: systematic analysis of health examination surveys and epide-
miological studies with 370 country-years and 2.7 million partici-
pants. Lancet. 2 Jul 2011;378(9785):31–40.
40. Chen R, Corona E, Sikora M, et al. Type 2 diabetes risk alleles demon-
strate extreme directional differentiation among human populations,
compared to other diseases. PLoS Genet. Apr 2012;8(4): e1002621.
41. Shai I, Jiang R, Manson JE, et al. Ethnicity, obesity, and risk of type
2 diabetes in women: a 20-year follow-up study. Diabetes Care. Jul
2006;29(7):1585–1590.
42. O’Sullivan BP, Freedman SD. Cystic fibrosis. Lancet. 30 May
2009;373(9678):1891–1904.
43. Davis PB, Yasothan U, Kirkpatrick P. Ivacaftor. Nat Rev Drug Discov.
May 2012;11(5):349–350.
G enomics in P ublic and P opulation H ealth   •   2 1 7
44. US Food and Drug Administration. Drugs@FDA:  Ivacaftor
label and approval history. Available at http://www.accessdata.
fda.gov/scripts/cder/drugsatfda/index.cfm?fuseaction=Search.
Label_ApprovalHistory#labelinfo.
45. Van Goor F, Hadida S, Grootenhuis PD, et al. Rescue of CF airway
epithelial cell function in vitro by a CFTR potentiator, VX-770.
Proc Natl Acad Sci U S A. 3 Nov 2009;106(44):18825–18830.
46. Gonzalez-Galarza FF, Christmas S, Middleton D, Jones AR.
Allele frequency net: a database and online repository for immune
gene frequencies in worldwide populations. Nucleic Acids Res. Jan
2011;39(Database issue):D913–D919.
47. Exome Variant Server. http://evs.gs.washington.edu/EVS/.
48. Li JZ, Absher DM, Tang H, et al. Worldwide human relationships
inferred from genome-wide patterns of variation. Science. 22 Feb
2008;319(5866):1100–1104.
49. Raby BA, Silverman EK, Kwiatkowski DJ, Lange C, Lazarus
R, Weiss ST. ADAM33 polymorphisms and phenotype asso-
ciations in childhood asthma. J Allergy Clin Immunol. Jun
2004;113(6):1071–1078.
50. Qu S, Sun D, Wang Y, Zhang C, Lv Y, Yao L. Association of
ADAM33 polymorphisms with childhood asthma in a northern
Chinese population. Exp Mol Pathol. Dec 2011;91(3):775–779.
51. Van Eerdewegh P, Little RD, Dupuis J, et  al. Association of the
ADAM33 gene with asthma and bronchial hyperresponsiveness.
Nature. 25 Jul 2002;418(6896):426–430.
 14.
GENETIC TESTING AND GENOMIC SCREENING
Angus John Clarke
INTRODUCTION
This chapter addresses the changes taking place in labora-
tory genetic investigations, the impact these are having
on clinical practice, and some of the difficult social, ethi-
cal, and communication issues that arise as a consequence.
In short, targeted genetic investigations are now being
replaced by genomic (“genome-wide”) investigations. This
change is a process rather than an abrupt transformation,
and a role will persist for the more traditional technologies,
but the change is nonetheless real and progressive; it marks
an important shift in clinical practice driven by changes in
technology. What impact is this having on the practice of
medicine and more broadly on society and the experience
of ill health?
The very first genetic laboratory investigation to enter
clinical practice was chromosome analysis, in the late
1960s—the entire karyotype—but since the 1980s, genetic
technologies have progressively focused their gaze, so that
the target of the more precise diagnostic investigations has
become correspondingly smaller. With the current develop-
ment of genomic diagnostics, however, investigations need
no longer remain focused but can instead—once more—
interrogate the whole genome.
The development of chromosome analysis as a clinical
investigation provided a tool both to explain many disor-
ders of physical and cognitive development and to catego-
rize such disorders, creating a taxonomy. Where parents
were concerned about a possible recurrence in their family,
cytogenetic testing also permitted risk estimation, prena-
tal diagnosis, and—when compatible with the law of the
land—the selective termination of affected pregnancies.
Advances in cytogenetic techniques—first Giemsa band-
ing and later in situ hybridization with cloned fragments of
DNA—greatly improved the resolution of testing so that it
became possible to detect specific submicroscopic deletions
when the associated disorder (usually known as a syndrome,
in the context of dysmorphology) was suspected on clinical
218
grounds. Array-based hybridization technologies now per-
form this same investigation across the whole chromosome
set as a first-line investigation (array-based comparative
genomic hybridization, or aCGH), and the karyotype has
been relegated to use in specific, restricted circumstances.
In the rather different circumstances of antenatal
screening, programs were established that enabled large
numbers of pregnancies to be screened for the fetal tri-
somies, usually with a preliminary filter such as maternal
age, biochemical assays of maternal serum, or ultrasound
examination of fetal nuchal thickness. Such chromosomal
studies have given rise to two categories of “problematic”
results: incidental findings (IFs) and variants of uncertain
significance (VUSs). Incidental findings, such as the sex
chromosome aneuploidies, are regular although infre-
quent findings. Apparently balanced de novo chromo-
somal translocations are also found with regularity but
will often be of uncertain significance, in that a modest
proportion will be associated with developmental prob-
lems for the child, but the majority will have no adverse
implications.
The course of molecular diagnostics first recapitulated
and then reversed the progressive focusing of cytogenetic
methods. In the 1980s, linkage analysis was applied to
predictive and prenatal testing for several important dis-
orders, and haplotype analysis was used in carrier testing.
This became progressively more accurate as closer markers
became available, flanking the various loci, and then intra-
genic markers, such as CA microsatellite repeats.
As the knowledge of gene structure and sequence
improved, diagnostics moved from inference (based on
linkage analysis) to the direct detection of disease-causing
point mutations, intragenic deletions, and triplet repeat
expansions. After a detour into genome-wide association
studies (GWAS), of substantial utility for research but with
little application to diagnostics, we have now entered the
stage of high-throughput molecular diagnostics. These are
array-based comparative genomic hybridization (CGH),
largely replacing chromosome analysis, and next-generation
sequencing (NGS).
One application of NGS is to analyze in parallel (i.e.,
simultaneously) a set of many genes, mutation in any one
of which may lead to a similar pattern of disease. Thus,
all the genes implicated in particular symptom clusters—
such as retinal degeneration, early-onset epilepsy, muscu-
lar dystrophy, or hypertrophic cardiomyopathy—may be
analyzed simultaneously; this can accelerate the diagnosis
of such conditions and highlight potential interactions
between variants at different loci, which might otherwise
have remained unrecognized. Problems of interpreta-
tion can arise but do so less frequently with these selective
approaches to sequencing.
As it becomes both simpler and cheaper to be less selec-
tive about the sequence information to be generated, labora-
tories are moving towards the use of exome sequencing (ES)
and whole-genome sequencing (WGS) for diagnostic as
well as research purposes. The limiting factor in laboratory
diagnostics is no longer the generation of sequence infor-
mation but is becoming the interpretation of the sequence
data generated.
N OT J U S T T E R M I N O L O GY:  G E N E
A N D G E N O ME , T ES T I N G A N D
S C R EE N I N G
Some comments on terminology will be helpful. The term
“genetic testing” can be used to refer to a wide range of
activities. We must be clear as to the scope intended by this
term, so we must distinguish between two pairs of words—
genetic and genomic, testing and screening.
For the purposes of clinical investigation, a “genetic”
test will refer to a test performed on one particular gene;
performing a test on several different genes would be a series
of genetic tests. Performing a genome-wide investigation—
WGS or ES, investigating all genes simultaneously, in par-
allel—would be a genomic investigation. A  genome-wide
association study (GWAS) is another example of a genomic
investigation but would be carried out for research rather
than in clinical practice. Sequencing a cluster of phenotypi-
cally related genes, such as all the genes known to be impli-
cated in causing a particular disorder, would still be a set of
genetic tests for that disorder but might now employ NGS
rather than Sanger sequencing.
The word “screening” also requires some clarification,
as it is slippery and is used with a range of different mean-
ings. It may be regarded as the testing of a person (or of
people) at more-or-less population risk of a disorder, to see
G e n etic T e s ting a nd G e no m ic Scr e e ning   •   2 1 9
if they are already affected or at increased risk. The point of
doing this is to reap the benefits of an early (presymptom-
atic) diagnosis, where this brings the advantages of a bet-
ter prognosis, or to gain access to screening for those in a
high-risk group. In contrast, “testing” in the narrower sense
refers to the examination or investigation of an individual
in the context of their personal or family history of disease;
that is, where there is a specific reason to believe that their
individual chance of carrying a disease-associated genetic
variant, or of developing a particular type of genetic illness,
may be greater than the population’s average.
An individual who seeks access to over-the-counter
genetic testing in a pharmacy or through the Internet may
therefore be arranging a genetic test on him- or herself if
s/he has a strong family history of a relevant disease, or they
may be arranging a genetic screening test if there are (as far as
they know) no particular features in their personal or fam-
ily history suggesting an increased risk. It is therefore the
context that distinguishes a specific test applied to one indi-
vidual from a population screening test applied to others;
there may also be differences in the technology employed,
but they will reflect differences in the scale of the enterprise
rather than the distinction between individual testing and
population screening.
This setting illustrates one of the serious problems
raised by direct-to-consumer (DTC) genetic screening:  a
company may claim that their single nucleotide polymor-
phism (SNP)-based genome-wide assessment of disease risk
indicates risk categories that people at population risk may
find helpful. However, those seeking such tests may well be
motivated by a family history of disease. They may actually
be in a high-risk group that only a thorough risk assessment
and the sequencing of relevant genes of major effect could
clarify, and which the SNP-based risk modifying screen
available DTC does not address. So they may buy a genetic
screen of no established utility under the illusion that it is,
for them, a highly specific genetic test relevant to their per-
sonal situation. When this confusion is placed alongside the
fact that most of these SNP-based tests use genetic variation
that accounts for only a small fraction of the genetic con-
tribution to risk of disease, it will be clear why such DTC
screening tests attract much professional hostility.
In addition to genetic testing of an individual in their
family context, and the population genetic screening of
individuals intended to identify those carrying a particu-
lar genetic variant predisposing them to disease, there is
another type of genetic test in clinical use—the testing of a
tissue sample from a patient with disease, where the genetic
changes may be present only in some tissues rather than
constitutionally (i.e., in all tissues). This usually entails the
testing of a malignant tumor to identify the genetic changes
that have either led to the tumor or that may be useful in
guiding treatment or predicting the likely response to treat-
ment. Testing the tumor may be a targeted genetic investi-
gation or an assessment of all potentially relevant variations
across the genome, when differences between the patient’s
constitutional genome and their tumor genome will be
the focus of interest. In the case of a genome-wide search
for variation, there is a high chance of stumbling across
incidental findings. One should note that there are other
occasions where somatic variation within an individual—
mosaicism—may have important implications for their
diagnosis.
There can be an additional confusion as to the scope of
the word “genetic” in the context of genetic testing. For our
purposes, it refers to the information being generated rather
than the mode of investigation. Hence, in the appropriate
family context, a renal ultrasound scan can generate essen-
tially predictive information about whether or not a young
adult is likely to develop his family’s polycystic kidney dis-
ease; a biochemical assay performed on a young male infant
at risk of Duchenne muscular dystrophy will usually give
very reliable predictive information as to whether he will
go on to develop that disease. It can be seen, therefore, that
genetic information can often be inferred from investiga-
tions that do not examine the genotype directly—the DNA
or chromosomes—but look instead at the phenotype or
some intermediary between the two. In this chapter, how-
ever, we focus primarily on molecular genetic investigations.
We use “genetic testing,” therefore, to refer to the exami-
nation of one or several genetic loci in an individual because
of their personal risk of a genetic disorder, usually because of
their personal diagnosis or their family history of disease. In
contrast, “genetic screening” now has two rather different
meanings. There is (i) population-based or population-risk
genetic testing in relation to specific diseases, and (ii)
genome-wide screening of an individual to identify vari-
ants of potential clinical significance. In population-risk
genetic screening, individuals are tested to see if they carry
one or more specific genetic variants, perhaps on the basis
of their ethnicity, but not on the basis of their personal or
family history. So those of Mediterranean extraction may
be offered screening to identify carriers of beta-thalassemia,
those from northern Europe may be tested for cystic fibro-
sis, and those of Ashkenazy Jewish origin may be tested for
Tay-Sachs disease. If someone has a close or strong family
history of such a disorder, then population screening might
not be appropriate, because it would be important to know
the specific genetic variant in their family to be certain
that the test performed was able to detect it. In contrast,
2 2 0   •   P rincipl e s o f G e no m ic M e dicin e
a genome-wide search for potentially important genetic
variation pays no attention to personal or family history
but attempts a “complete” assessment. The trigger for hav-
ing this investigation may be very specific and personal, or
perhaps relate to the person’s family history, but the inves-
tigation is conducted without reference to that. Genetic
screening programs aiming to identify carriers of autosomal
recessive disorders that vary in frequency between popula-
tion groups may then become irrelevant; when genomic
methods are applied that examine all recessive disease loci,
information about an individual’s ethnicity will no longer
be relevant to the laboratory procedures utilized.
It becomes apparent, then, that any genome-wide
search for clinically relevant genetic variation collapses the
distinction between targeted testing and population-risk
screening, because all variation is being assessed. While the
indication for investigation may be something very spe-
cific and personal, what is uncovered will depend upon the
interpretive lens through which any identified variants are
viewed. An investigator could have a patient’s WGS or ES
in their database but only seek to interrogate variants in a
finite number of specific genes, perhaps those known to have
been implicated in previous patients with the same clinical
disorder (e.g., hypertrophic cardiomyopathy, or HCM).
Then the investigation amounts to genetic testing for the
bundle of HCM loci. But if the investigator approaches the
patient’s sequence data without bias—without setting nar-
row criteria—then they will identify variants at many sites
that are unlikely to have any bearing on the personal or fam-
ily history of HCM. The specific operational decisions used
to manage the data-interpretation pipeline will therefore
permit fine gradations along a continuum between targeted
genetic testing and genome-wide screening.
What had been a clear distinction between testing and
screening no longer holds. This has major implications for
the public health impact of the high-throughput technolo-
gies, aCGH and NGS. Should the investigator deliberately
choose to wear blinkers and ignore all findings but those
specifically sought; that is, all but those pertinent to the indi-
cation for testing?(1) Or should all variants be considered?
G E N ET I C T ES T I N G
FA M I LY-BA S E D P R E D I C T I VE
G E N ET I C T E S T I N G
Genetic methods may, of course, be used to establish the
diagnosis when an individual presents with clinical fea-
tures likely to result from a genetic condition. Beyond that
 Free ebooks ==> www.Ebook777.com
circumstance, however, there may be several different family
contexts in which genetic testing of an individual may be
considered appropriate. In relation to their own health and
health care, genetic testing is most likely to be appropriate
when the person knows they are at risk of a disorder pres-
ent in other members of the family but does not yet know
if they will be affected; this amounts to predictive genetic
testing.
Such testing has been available for Huntington’s disease
for more than two decades, and for several familial cancer
syndromes for nearly as long. There have been numerous
studies of the impact of testing on individuals and families,
and of the circumstances in which testing can be clinically
helpful and emotionally tolerable or even beneficial. One
important factor is whether the genetic test result guides
the medical care of the individual, including surveillance
for treatable complications of the disease. Where the test
result does have implications for medical management,
the health professionals involved will need to ensure that
their client understands this. Indeed, professionals may
wish frankly to recommend testing if this is the best way
to safeguard their patient’s health and welfare. If the medi-
cal benefits are less clear, it may be more difficult for them
to help the client make the best decision for their personal
circumstances, when it is the family and emotional factors
that will be more relevant.
The usually “nondirective” ethos of genetic services will
not be so appropriate where the decision about testing has
clear implications for medical management(2). An example
of such a context might arise with an adolescent or young
adult at risk of the familial adenomatous polyposis coli
(FAP) present in other family members. Because the prog-
nosis in an affected individual is so poor in the absence of
tumor surveillance, and of colectomy in those with bowel
tumors at high risk of malignant transformation, it would
be good practice for any health professional involved to
recommend genetic testing. A similar situation arises with
infants at risk of inherited retinoblastoma, where the ben-
efits of genetic testing as an aid to the coordinated manage-
ment of the at-risk child are so substantial that the only
sensible course of action is to recommend it, so that frequent
surveillance for tumors can be provided to those who are at
high risk and can be avoided in those whose risk is low (that
of the general population). Because of the implications of
testing results for the use of other medical resources, a num-
ber of perspectives unfamiliar to clinical geneticists may
need to be introduced into decisions about which genetic
testing services it would be appropriate to fund(3).
In the context of an individual at risk of a usually late-
onset neurodegenerative disorder, such as Huntington’s
G e n etic T e s ting a nd G e no m ic Scr e e ning   •   2 2 1
www.Ebook777.com
disease (HD), the situation is very different and will remain
so until an effective treatment has been developed for symp-
tomatic disease, or effective, pre-symptomatic disease pre-
vention has become possible. In this context, the genetic
professional—whether clinical geneticist or genetics
counselor—will engage the at-risk individual who is seek-
ing predictive genetic testing in an extended conversation.
This process may entail challenging the individual’s state-
ments and attitudes, to help them make the best decision
for them(4,5); this is designed to help them decide whether
or not predictive testing would be helpful and, where the
client chooses to have the test, to prepare for the full range
of possible test results.
Even in such an apparently simple context, there are
complexities to consider before proceeding with the test.
Clients will be helped to think through how they, and oth-
ers, would react to a favorable or an unfavorable result,
or to a result of uncertain significance (an intermediate,
“gray-zone” allele), or to a change of heart—a decision not
to go ahead with testing (for the moment). Not everyone
wishes, or is able, to engage in discussions about such hypo-
thetical scenarios(6), but it is good practice to encourage
such reflection(7,8). Having clear reasons for testing, and the
willingness to engage in advance in reflection about the pos-
sible consequences of testing, may be associated with better
outcomes after testing(9,10).
There may be decisions to make about whom to tell
about the risk, the test, and the result, and there may be
important potential consequences of such testing for the
person’s insurance, employment, or career choice, as well
as personal relationships and emotional equilibrium. The
implications of testing for insurance and employment will
differ between countries, according to the legal context
and the system of health and social care. There will also be
consequences of testing for other individuals, including the
person’s partner, any surviving parents, and their siblings
and children; there will be an impact on the whole family
system(11). The most appropriate clinical approach to adopt
has been refined as experience has accumulated of predic-
tive testing in this context of high risk without effective
treatments(12–15). However, the early finding that, whatever
the result, there will be costs as well as benefits for those
who are tested has not been contradicted by subsequent
experience(16).
One type of difficulty can emerge, even with a favorable
result, if life-plans have been made and acted upon and the
result now undermines the basis of those decisions. Perhaps
a decision was made about marriage, children, or career
that seemed safe in the context of genetic risk, but with the
result known, years of life may now, in hindsight, be viewed
with regret. Difficulties may also arise if the results, whether
“good” or “bad,” conflict with a prior expectation of the per-
son tested or his or her family. Individuals and families can
have their own ideas about the pattern of inheritance that
applies to them: perhaps it is the oldest girl in each genera-
tion who is affected, or the youngest boy, or the one with red
hair. Such ideas, which can be referred to by the term “lay
beliefs about inheritance,” can lead to serious distortions
in medical care: for example, women at risk of breast can-
cer but related to affected relatives through their father are
less likely to seek risk assessment or additional surveillance
because many families imagine that susceptibility to breast
cancer has to be transmitted through the female line(17).
Particular individuals may also be picked out by others in
the family as destined to develop the family disease—they
are “preselected”—and this phenomenon can contribute to
complex processes of family psychodynamics(13).
Experience has shown that those given an unfavorable
predictive result for HD (and many other conditions) do
often experience distress, but, as with those given a favorable
result, there is a fall in their distress and anxiety over some
months, associated with the loss of uncertainty, that returns
to baseline levels by one year. Overall, the best predictor
of a person’s post-test emotional state is the pre-test state,
and those whose motivation for testing is unfocused and
non-specific tend to fare worse than others(10,18). Those who
show greater concern and distress during the counselling
before the testing may be those who can engage construc-
tively with their risk in advance, and they may cope better
with an unfavorable result(9). Those who proceed with test-
ing for HD are a self-selected group(16), and results of these
studies cannot be generalized to others—we could not draw
any conclusions from these studies about how those at low
risk of a serious disorder would respond to unfavorable test
results.
Using the framework of Burke and colleagues to evalu-
ate genetic tests(19–21), through examining both the clinical
validity of testing and the scope for useful medical inter-
ventions, we can turn to contexts where there may be some
health benefits from testing. Testing for familial predisposi-
tion to breast and ovarian cancer at the BRCA1 and BRCA2
loci is in this category, with the benefits of prophylactic sur-
gery in the prevention of malignancy now better defined.
Here, the potential benefits of testing need to be weighed
against the possible drawbacks jointly by the at-risk indi-
vidual and their health professionals, and the contextual
details of the individual case will often be crucial. One of
the factors to be taken into account is the ability or willing-
ness of the concerned individual to approach any surviving,
affected relatives to see if they would be willing to have a
2 2 2   •   P rincipl e s o f G e no m ic M e dicin e
genetic test, so issues of family communication (or lack of
it) are central. Equally, while many women at risk of breast
and ovarian cancer find testing helpful in both the practical
and emotional domains, there are often emotional barriers
to their discussing test results with relatives. Indeed, given
that the primary motivation of many is to provide informa-
tion for their children, the unanticipated emotional and
communication difficulties that arise for those who have
undergone testing constitute an important phenomenon(22).
Those already affected by cancer may have additional coun-
selling and support needs, in part because of the additional
implications for their own health(23), such as when a woman
with breast cancer finds that a test result shows that she is
now also at high risk of ovarian cancer.
Even apparently favorable results may have some dif-
ficult consequences, such as altered family relationships(13)
or an unanticipated reluctance to discontinue screening for
disease(24).
The value of a test—in the sense of its clinical utility—
can vary enormously with the details of the family history
and the availability for testing of other family members or of
stored pathological specimens. If a family mutation can be
identified in a sample from a definitely affected individual,
then the interpretation to be made of the results of testing
an at-risk relative can be defended with greater confidence
than if no family mutation is known and only the at-risk
individual can be tested. In the former case, a negative test
result (the failure to find the family’s mutation in one of
the two BRCA loci) will give very substantial reassurance,
although it will still leave the woman concerned with at least
the population risk of developing a breast cancer. In the lat-
ter setting, finding no mutation in either of the two loci will
provide much less reassurance. It is for this reason that some
services only perform mutation searches in samples from a
definitely affected individual and only when the chance of
finding a BRCA gene mutation exceeds a given (arbitrary)
threshold. Testing for the family’s particular mutation in an
individual at risk becomes possible only once it has been
identified by a mutation search in the affected individual.
This is one way of attempting to contain the costs of health
care, although other models of service provision could well
be defended and might permit at-risk women with fewer
surviving female relatives to be given access to testing.
One approach, which would maximize the efficiency of
cascade testing within families, would be to make mutation
searching available to those affected individuals found to
have a malignancy at a young age and who have a family his-
tory of relevant disease(25). This may be more cost-effective
in the long term than a genetics service operating solely
in response to concerns from unaffected, at-risk relatives.
Proactively suggesting to such affected individuals that
molecular testing might be appropriate would substantially
alter the emotional context for the patient. This greater
readiness to test the individual woman affected by breast
cancer needs to be assessed not only for its utility, but also
for the associated distress and emotional burden. There may
be several reasons for this burden, including both the direct
prognostic implications for themselves and the need for
them now to consider the implications for others in their
families, with the associated burden of having to raise this
topic with their loved ones.
Where testing requires access to information about, and
perhaps samples from, other family members, the decision
to seek testing will for many be difficult and, even if they
decide to go ahead, may be frustrated by a lack of coopera-
tion by their relatives. Furthermore, the possible adverse
consequences of unfavorable test results for insurance cov-
erage, and the possible distress or anger that may be antici-
pated in close family members and those whose risks will be
modified by the result, may deter an individual from pro-
ceeding. This also raises the question of “whose information
is it anyway?” and there are good grounds for weakening
the right of the affected individual to prevent the applica-
tion of their test result to the health care of their relatives.
Technology may come to the rescue, however, in that the
cost of testing the bundle of relevant genes in those at risk is
falling, so that it will soon make good sense to test anyone at
risk for mutations with the same panel of genes and to give
each individual their own risk level, without the interpreta-
tion depending so much upon their relatives’ results.
One context that has generated much debate among
professionals, and also within families, is that of testing an
individual at one in four (25%) prior risk of HD, when an
unfavorable result on the individual implies an unfavor-
able predictive result for the person’s parent, who was pre-
sumably at one in two (50%) prior risk but who may have
decided not to have predictive testing and may resent the
fact that they have, in effect, been “tested” without giving
consent(26).
As our knowledge of human genetic variation and our
ability to interpret it both improve, testing one individual
in a family will become progressively less dependent upon
information about other members of the family.
D I AG N O S T I C G E N ET I C T E S T I N G
When an individual presents with a disease that may have a
genetic basis, whether or not there is a relevant family his-
tory, genetic investigations may very reasonably be initiated.
While it will often be prudent and helpful to explain to the
G e n etic T e s ting a nd G e no m ic Scr e e ning   •   2 2 3
patient, and perhaps their relatives, that the investigations
may generate results with implications for others in the
family, the fact of these genetic implications—a potential
impact on the immediate or even the extended family—will
emerge once the diagnosis has been made, whatever tech-
nology is used to make the diagnosis. If a child has cystic
fibrosis (CF) or Duchenne muscular dystrophy (DMD),
for example, then there will be a risk of recurrence within
the immediate family, and there may be healthy carri-
ers of the disease in the extended family, too. This is true
whether the diagnostic process entails DNA technology or
more traditional methods (assay of sweat electrolytes in CF,
or the pathological and immunohistochemical examination
of a muscle biopsy in DMD). Similarly, if a healthy preg-
nant woman has renal cysts identified in herself during an
ultrasound scan carried out on her fetus, then the possible
implications for her own health, the fetus, and the family
will clearly emerge before any specifically genetic testing has
taken place.
It is the clinical imperative of arriving at the correct diag-
nosis for a sick individual that drives the process of investi-
gation, whether or not that entails genetic investigations.
Therefore, it is not only the patient but also other family
members who may need to come to terms simultaneously
with both the serious nature of the established diagnosis in
the patient and the fact of the genetic implications for oth-
ers. It is important to make sure that patients and families
caught up in such a situation have ready access to timely and
supportive clinical genetic assessment and genetic counsel-
ling, as well as the best clinical care for established disease.
There are rather different considerations in the context
of a child with developmental difficulties and dysmorphic
features. In this setting, diagnostic investigations will usu-
ally include tests for genetic conditions. The problems that
commonly arise in this context are threefold:
1. The failure to make a diagnosis that accounts for the
child’s difficulties, with resulting uncertainty for the
prognosis, the risk of recurrence, and the implications
for other family members.
2. The distress caused by the specific diagnostic label
that is attached, either because it confirms the
syndromic association of dysmorphic features with
the developmental problems, or because the parents
suddenly come to appreciate that the prognosis for their
child is substantially worse than they had previously
accepted: they may be forced to recognize that the
child is unlikely ever to talk, or to walk, or to lead an
independent life as an adult.
3. Distress and ambivalence about the risk of recurrence
of the condition, in siblings or in the extended family,
if this risk is significant. Many parents find that the
prospect of terminating a pregnancy causes great
distress, especially if it is associated in any way with the
perceived devaluation of a much-loved child. Again,
genetic counselling can be very helpful for families
in both understanding their situation and adjusting
constructively to it.
G E N ET I C S A N D R E P RO D U C T I O N: P R E NATA L
D I AG N O S I S A N D A N T E NATA L S C R E E N I N G
In the past, two very different types of screening were
applied to populations in the context of reproduction.
These are screening before or during a pregnancy to iden-
tify carriers of autosomal recessive disorders, or screening
during a pregnancy to identify chromosomal or structural
anomalies in the fetus. The first category to be considered
is screening to identify carriers—or carrier couples—at risk
of having a child affected by an autosomal recessive dis-
order. Such conditions are often specific to a population
group, such as Tay-Sachs disease, the hemoglobinopathies,
or cystic fibrosis. Identifying carriers before marriage allows
the marriage to be conditional upon the result of screen-
ing; screening before pregnancy allows the couple to decide
whether to embark upon a pregnancy; screening in an
established pregnancy only gives the choice of whether to
continue or terminate the pregnancy. What genomic test-
ing (NGS) provides is the ability to screen for virtually any
recessive disorder instead of just the one or the few most
prevalent in a population; it also makes it possible for ES or
WGS for a diagnostic purpose to generate carrier-screening
results as IFs.
Genomic screening to identify carriers of recessive dis-
ease could have an impact on several important features of
contemporary social life. First, it will identify many more
individuals as carrying a mutation in an important reces-
sive gene than will testing for just one or two disorders. It
is therefore more likely to affect a person’s “reproductive
self-esteem.” Second, and countering the first point, the
fact that anyone tested becomes aware of the important
disorders for which they are carriers makes it more diffi-
cult for stigmatization and discrimination to impose the
serious burdens that have been experienced in the past, as
most people are likely to carry some recessive disorder. If
a society or healthcare system decides to offer this type of
comprehensive genomic screening for recessive disease to
the general population—perhaps to couples planning to
marry—then it becomes a genomic screening test that is
2 2 4   •   P rincipl e s o f G e no m ic M e dicin e
made available as a population screening test. Given time,
the birth incidence of individuals with such diagnoses may
fall away as all recessive disease becomes potentially avoid-
able and may be seen as optional (by those who can access
good health care); one possible consequence of this may be
harsher blaming by others or stronger feelings of guilt when
parents have a child affected by such a “potentially avoid-
able” disorder. Might this reinforce discrimination against
those affected and their parents?
A related consideration is the question of the social
obligation to participate in antenatal screening. This sense
of obligation is experienced by some, and may be repack-
aged by “ethicists of the abstract” as the obligation to have
the healthiest child possible(27), but these debates become
most acute in middle- and low-income countries. In those
countries, effective health care for patients affected by
beta-thalassemia, for example, may depend upon “achiev-
ing” a large fall in the birth incidence of the disease. In
developed countries, an early finding—that uptake of car-
rier screening for recessive disease depends mostly upon the
manner and enthusiasm with which it is made available—is
likely to apply in this as in other contexts(28).
The distinctions between making a test available, pro-
moting its uptake, nudging to encourage uptake, and coer-
cion to maximize uptake are not easy to discern. It would
perhaps be appropriate to stimulate an international debate
on the circumstances under which a health care system may
persuade or push a population to accept a screening pro-
gram in this area of reproduction and genetics: is it coherent
to assert that the individual’s right to make such personal
decisions only applies in wealthy countries?
The question of whether to introduce carrier screen-
ing for a wide range of autosomal recessive disorders has
been mooted on several occasions. The United Kingdom’s
Human Genetics Commission and National Screening
Committeee(29) jointly proposed a framework within
which carrier screening could be made available. It has been
pointed out that it is a question not only of which diseases
to screen for, but also of which mutations to count as patho-
genic(30); it may be unhelpful to detect all variants in a locus
when many may not be disease-causing or may be associated
with only minor features of the condition. Thus, detecting
all variants in the CFTR locus as part of newborn screening
for cystic fibrosis would unhelpfully identify future cases of
male infertility as newborn infants.
One large study of an ethnically diverse population had
carrier screening carried out as a multiplex NGS applica-
tion examining roughly 400 loci; 24% of individuals were
carriers of at least one of these recessive diseases(31). There
is little to be gained by population-specific panels of genes
once so many can be examined, except perhaps in commu-
nities with many private recessive disorders if ES would be
substantially more costly.
One way of framing the concerns about social discrimi-
nation on genetic grounds is to consider how to reconcile
the different perspectives of the (affected) individual, the
family (at risk of having an affected child), and society
(which is challenged by the expectation to meet the medical
and social needs of all its members). How can the multiple
and conflicting goals of these parties be met, simultaneously
respecting the worth of each affected individual, promoting
the reproductive autonomy of prospective parents, all while
containing the cost of care of children and adults with all
forms of special needs, at a time when families’ expectations
of government support are rising(32)? These goals are framed
at different levels of social organization—the individual,
the family, and society at large—and there are powerful ten-
sions between them.
The other application of NGS to reproduction is in
the context of antenatal screening. While screening for
structural anomalies will continue to use ultrasound scan-
ning, the current methods of screening for chromosomal
anomalies are likely to give way to methods based on deep
sequencing of free DNA in maternal plasma, a proportion
of which will be fetal in origin. The formerly crucial distinc-
tion between prenatal diagnosis and antenatal screening
collapses as the screening method becomes the diagnostic
method, and the level of resolution (the whole chromosome
or the nucleotide base pair) depends largely upon the cover-
age and depth of sequencing; in effect, the cost.
Diagnosing fetal chromosome anomalies is now largely
based on invasive prenatal diagnosis, but it is beginning
to employ high-throughput methods, such as array CGH
analysis, in place of conventional cytogenetics(33). By setting
diagnostic criteria that minimize the number of VUSs, this
may be appropriate practice, although our experience is too
limited to regard aCGH as a complete substitute for cyto-
genetics in prenatal diagnosis. Avoiding too much informa-
tion—too many VUSs or IFs—remains a higher priority
in this setting than after birth, and assessing the effects of
a CNV will vary greatly with the precise diagnostic route
(the sequence of events) that led to the aCGH being
performed(34).
The other technology to consider here is noninvasive
prenatal diagnosis through the sequencing of free DNA
in maternal plasma, as this derives from both the mother
and the fetus. Sequencing of chromosome 21 elements
from maternal plasma enables the reliable identification
of fetal trisomy 21 (Down syndrome)(35,36). This changes
the domain of prenatal screening and diagnosis in two
G e n etic T e s ting a nd G e no m ic Scr e e ning   •   2 2 5
fundamental ways, for both the timing and the risks of the
test. Fetal DNA is present in maternal plasma in adequate
quantities from seven weeks of gestation onward, and
there is no risk of miscarriage from the procedure. These
are major changes to the context of prenatal screening or
diagnosis, and such changes will substantially alter the
ethical considerations surrounding prenatal testing and the
termination of pregnancies. Thoughtful responses to these
new challenges will be important to prevent the excessive
commercialization and/or the trivialization of antenatal
screening.
If a program of comprehensive carrier-screening is intro-
duced alongside antenatal screening by maternal blood
sample for fetal aneuploidy and chromosomal copy number
variants (CNVs), then the incidence of both chromosomal
and recessive disorders at birth may fall dramatically, with
unpredictable consequences for social attitudes to these
conditions and perhaps other causes of disability. Is this, or
to what extent, or under what circumstances, would this be
something to be welcomed?
Pre-implantation genetic testing (PGT) or screening
may be employed by those using IVF methods to achieve
a conception, whether because of fertility problems or to
avoid transmitting a serious genetic disorder. The applica-
tion of NGS methods in single-cell PGT raises concerns
that too much knowledge may be generated about a child,
whose full genome sequence may become known from
before birth. This will give rise to questions considered
below concerning the genetic testing of children.
G E N ET I C I N F O R M AT I O N A B O U T C H I L D R E N
There has been a broad professional consensus to avoid
generating predictive genetic information about children
unless it is to their direct medical benefit. This consensus
has been in place in the United Kingdom and Europe for
about two decades and is still in place(37–39), although recent
recommendations of the American Academy of Pediatrics
and the American College of Genetics and Genomics(40,41)
have somewhat weakened the previously similar U.S. policy.
These policies also apply to information of importance in
reproduction rather than personal health care, such as car-
rier status for recessive disorders, but there may be rather
less at stake in these contexts, especially when the condition
under consideration is autosomal recessive. There may be
more at stake for the child’s future when the condition is
sex-linked or results from a chromosomal rearrangement,
when all the child’s offspring will be at risk, whereas if the
condition is autosomal recessive, then the risk only materi-
alizes if the child’s future partner is also a carrier.
 The development of these professional policies reflects
the context within which such genetic testing evolved; that
is, family-based genetic counselling and testing. In that con-
text, the high prior risk of the child’s inheriting a specific
condition shapes the ethical concerns. The family usually
knows that the child is at high risk, and the strategies for
dealing with that risk have usefully guided professional
practice. However, in the new context of high-throughput
genetic technologies, predictive or carrier status informa-
tion is likely to emerge as IFs when the child is tested for
some other reason and without the family’s having appre-
ciated or addressed the question of the child being at risk.
The predictive and carrier information to emerge will usu-
ally arise without the family’s having been aware that either
the child or even the parents were at risk. This is a crucial
difference in the clinical context, and we will have to adjust
our thoughts, our intuitions, and our professional approach
to take this shift into account.
It is for this reason—the possible implications for
the health of the parents, as well as the child as a future
adult—that the American College of Medical Genetics has
recommended that important IFs of possible relevance to
the future health of the child should be made known to
the parents(42,43). While this may seem to run counter to the
previous professional guidance, that was in the different
setting, in which there was prior family awareness of risk
to both the parents and the child, while in the new con-
text this prior awareness is absent, so the family—including
this child—stands to gain a lot by disclosure of the child’s
result.
Another important reason why this recommendation is
attractive—that a defined set of IFs about children or adults
should be disclosed to the family—is that it sets limits to the
information (i.e., which IFs) should generally be disclosed.
The list of 56 genes in which definite or likely pathogenic
mutations should be disclosed is a starting point. This list of
genes can be modified over time as evidence and experience
accumulate, but it will be at least defensible for practitio-
ners and researchers not to disclose other results of much
less probable clinical applicability, whose disclosure is much
more likely to cause confusion and distress.
The primary concern of professionals to restrict the
release of information about children is to preserve their
“open future”(44,45). The different perspectives on the genetic
testing of children may be approached by framing the dis-
cussion in terms of the best interests of the child, which
must surely trump all other considerations. However, it
may be difficult in practice to distinguish the interests of
the child from those of her family as a whole; this is an area
where further research into communication about genetic
2 2 6   •   P rincipl e s o f G e no m ic M e dicin e
information (especially NGS information) within families
will be very important.
Parents can find it especially difficult to adjust to two
particular types of genetic information about their child,
whether the information arises in an already known fam-
ily context, or as an IF. These difficulties may particularly
apply in the case of cardiac disease with a risk of sudden
death (such as the inherited rhythm disorders and hyper-
trophic cardiomyopathy)(46) and potentially serious psychi-
atric disease(47). What both these areas have in common,
in addition to the parental “biological” guilt at having
imposed the risk on the child, is the paralysis likely to be
felt by parents not knowing what to do for the best. In rela-
tion to cardiac disease, how does one give the child any-
thing like a “normal,” psychologically healthy childhood if
his activities are always being restricted and if the child and
others around him are aware of the risk of sudden death? If
cardiac treatment can normalize the risks of sport, dancing,
and sex, then a normal life may be feasible; but if not, then
the benefits of knowledge may appear thin. Equally for the
risk of psychiatric disease, how should parents behave so as
to minimize the risk of such disease occurring? After all,
parents will be told that psychosis is not inevitable, even
with a strong inherent predisposition, so that opens the
door to guilt if psychosis does indeed develop:  “it must
have been something we did.” In daily life, should parents
aim to minimize stress and conflict—or to impose clear,
firm boundaries and insist upon high standards of disci-
pline? There is a real danger of self-fulfilling prophecies,
with parental anxiety in their management of the child’s
behavior actually becoming a stress that increases the
chance of the problem’s arising.
Another specific question to consider in relation to chil-
dren is that of ES or WGS as a screening program for new-
born infants. This practice has shifted into focus with the
decision of the U.S. National Institutes of Health (NIH)
to introduce a project examining the outcomes and impli-
cations of such sequencing in several thousand infants.
Whereas the shift in technology of newborn screening to
tandem mass spectrometry (TMS) has challenged the gen-
erally accepted criteria for population screening—the origi-
nal Wilson and Jungner criteria from 1968(48), as revised by
the United Kingdom’s National Screening Committee(49)—
this program of research appears to trample over them.
Wilson and Jungner proposed criteria for making a public,
transparent decision about screening disease-by-disease,
according to the evidence. There have been discussions
about broadening the criteria; for example, in relation to
newborn screening for essentially untreatable disorders for
which an early diagnosis may have benefits for the family
unit, although no strictly medical benefits for the child
(e.g., screening for Duchenne muscular dystrophy)(50). The
introduction of TMS as the method of biochemical mea-
surement has led practitioners to begin to assess “screening
by TMS” as a potential alternative to earlier assay methods
that diagnosed many fewer conditions. Instead of making
a decision “disease-by-disease,” the question of introducing
TMS could be approached by judging between methods of
screening, taking what is revealed by TMS as a single entity.
That approach has led to the packaging of a long list of met-
abolic disorders diagnosable through TMS as the new stan-
dard. The process of deciding which diseases to report may
have included a step in which the potential health benefits
of an early diagnosis for each disease were considered, but,
beneath the disease-by-disease rhetoric, the decision being
made was in fact about whether TMS should be used or
whether laboratories should close their eyes to the potential
of the new technology. A decision to use TMS, but practice
as if it had not been developed, would not have been sus-
tainable. The Wilson and Jungner approach has been super-
seded by a technology-led decision to make the diagnoses
that the new technology permits.
Now that the technology of NGS has arrived, will this
simply follow the precedent of TMS to take over labora-
tory practice and to drive the clinical practices of explana-
tion, taking consent, and reporting results? If so, will it be
applied solely in private health care, or in state-sponsored
programs, too? These questions will have particular reso-
nance in United States, not only because that is where this
will be happening first on a large scale, but also because
of the pattern of healthcare funding in the United States.
Will health risks identified through newborn screening by
WGS be managed as an entitlement through state health-
care, as would often occur for the continuing dietary needs
of children with phenylketonuria (PKU)? Or will a state
program of newborn early diagnosis leave a generation of
children with knowledge of their state of risk but without
the resources to act upon this information?
In addition, of course, there are some additional ques-
tions about the management of the information generated
by WGS of healthy newborn infants. Will this be used by
insurance companies to restrict access to health care? Will
the program store the sequence data indefinitely, as a life-
long resource, as suggested by Biesecker(51)? Will periodic
reanalyses be expected, as the ability to interpret genomic
data improves? How will information be passed to the
children as they mature? How will the liminal category of
“patient in waiting”(52) be managed by professionals, parents,
and then by the child? Will such results lead to unhelpful
distress and anxiety(53)?
G e n etic T e s ting a nd G e no m ic Scr e e ning   •   2 2 7
The costs of storage will be substantial, as active man-
agement of the data will be required because both soft-
ware and information technology hardware will change.
Furthermore, repeat or additional analyses on fresh samples
may be required (e.g., to capture markers of epigenetic influ-
ences and gene expression), and not simply data storage.
These would be real challenges for such a program, which
may render carefully stored data relatively useless. It will be
instructive to follow the emerging experience of this pro-
gram; one can only hope that the children do not become
its casualties.
G E N ET I C T E S T I N G A N D T H E R A P EU T I C
B E N E FIT
Moving from prediction through to therapy, one can be
optimistic that genome-based knowledge is set to transform
medicine. The first area in which major benefits can already
be seen in the clinic is that of oncology. Genomic analysis of
a patient’s malignancy is already guiding the choice of treat-
ments. However, it is also becoming clear in the area of rare
diseases that a gene-based understanding of disease is open-
ing up immense therapeutic possibilities.
In malignancies, rational treatments can be developed
following the model of inhibitors of the Philadelphia fusion
protein, the BCR-ABL tyrosine kinase(54), and the generation
of monoclonal antibodies against the disease-modifying
protein products of tumor-amplified genes. Understanding
the cellular role of DNA mismatch repair systems has led to
the use of Poly Adenosine diphosphate Ribose Polymerase
(PARP) inhibitors in treating breast cancers in those with
constitutional BRCA gene mutations(55). The post-genome
move into proteomics also creates opportunities to classify
tumors more sensitively (e.g., with certain lung cancers and
leukemias), which can then lead to the improved targeting
of therapies, although this chapter is not the place to review
these developments.
Among the rare (“orphan”) diseases, gene product
replacement can sometimes be helpful, as in the inherited
forms of leptin deficiency(56) and growth hormone deficiency.
The administration of developmental signaling proteins
at the appropriate stage has effectively treated the mouse
with hypohidrotic ectodermal dysplasia(57), and human tri-
als of this approach are proceeding. The gene-based under-
standing of disease mechanisms is also proving beneficial in
tuberous sclerosis (TS); thus, the tumors that cause so many
of the problems of TS can be stabilized or sometimes made
to regress with inhibitors of the mTOR pathway(58).
Understanding the precise mutation in a disease-
associated gene may also open up therapeutic avenues. One
approach of potentially broad application is the suppression
of nonsense mutations(59), where further development work
is active. In Duchenne muscular dystrophy, the frequency
of frame-shifting exonic deletions or duplications has led to
the development of oligo-induced exon skipping as a useful
approach(60). The detailed understanding of the cell biology
of disease also opens up avenues for the treatment of cystic
fibrosis associated with specific mutations in the relevant
gene, CFTR(61).
Within neurology, several strategies are being devised
for the treatment of Huntington’s disease with the implan-
tation of fetus-derived cells or stem cells within the brain,
or the use of allele-specific oligos to suppress expression
of the polyglutamine expansion-encoding allele. Gene
therapy appears promising in spinal muscular atrophy(62).
Even within the previously intractable area of neurodevel-
opmental disorders, there is optimism about the treatment
of Rett syndrome; from the effective reversal of the disease
in the mouse(63), it seems that any method of increasing
the availability of MeCP2 within those neurons lacking it
may be highly effective in ameliorating the disorder; using
a separate approach, the modulation of neurotransmitters
may be highly effective in stabilizing the autonomic dis-
turbances that are so common and so problematic in this
condition(64).
T H E C O MM O N C O M P L E X
D I SEASES
What we refer to as “the common complex diseases” are not
all distinct conditions but can be seen rather as symptom
clusters with a variety of contributing causal factors (includ-
ing environmental and life-history factors) and with con-
testable and somewhat arbitrary diagnostic criteria. While
this may be less true for the common cancers, this general-
ization applies to type 2 diabetes (T2D), coronary artery
disease, hypertension, hypercholesterolemia, cerebrovascu-
lar disease, Alzheimer’s disease, and perhaps schizophrenia.
In this group of conditions, there is a health-disease contin-
uum extending from clear normality through susceptibility
and then minor clinical features, to an unambiguous disease
state. Furthermore, many of these complex disorders share
at least some of the same causal factors, often grouped as the
“metabolic syndrome,” and the various associated degenera-
tive pathologies often coexist in the same patient.
Interestingly, epidemiological data now strongly suggest
that intrauterine experiences can influence the future health
of the fetus as an adult individual—with poor fetal growth
being associated with a higher incidence of these conditions
2 2 8   •   P rincipl e s o f G e no m ic M e dicin e
in adult life as well as certain neurodevelopmental disor-
ders in childhood. While the causal processes involved are
still being elucidated, the association between intrauterine
growth retardation and susceptibility to these conditions is
well established, and there are exciting indications that epi-
genetic modification of DNA is involved (65–67).
The mechanisms through which environmental fac-
tors contribute to disease are also being elucidated through
gene-based research. The way smoking interacts with
genetic variants to produce its damage is being revealed.
Smoking interacts with apoE4 in increasing the risk of
coronary artery disease CAD(68), it interacts with folate and
folate metabolism (MTHFR genotype) in causing colorec-
tal polyps(69), and it interacts with at least two maternal gene
loci in its effect on fetal growth(70).
IDENTIFYING THE MENDELIAN
S U B S ETS : TA K I N G A FA M I LY H I S TO RY
The application of genetic testing to the common com-
plex disorders is still (as with the first edition of this book)
largely restricted in evidence-based clinical practice to iden-
tifying those families where the disease has arisen in asso-
ciation with a monogenic (Mendelian), usually autosomal
dominant, disease susceptibility. The age of disease onset
is often somewhat earlier than average in these families.
Important Mendelian disease subsets are found in T2D
(maturity-onset diabetes of the young—MODY), coronary
artery disease associated with hypercholesterolemia (famil-
ial hypercholesterolemia—FH), the common cancers of
breast and bowel, and Alzheimer’s disease. The risk of an
affected individual’s transmitting the susceptibility to their
children is then 50%, so that the risk that a child will go on
to develop the condition at some stage is given as 50% of the
lifetime penetrance. In the families showing predisposition
to cancer, penetrance varies with the particular disease gene
mutation but may be virtually 100% in familial adenoma-
tous polyposis families, and perhaps 70–80% for breast can-
cer in women carrying many of the mutations at one of the
BRCA gene loci, compared to the average lifetime risk for
women of 8–10%. In MODY, the prognosis and response
to treatment vary with the particular gene involved, and
the risk to offspring is high—very much greater than the
approximately 10% risk to the offspring of other parents
with T2D. Mendelian subsets exist for the other complex
disorders, too, often with a relatively early age of onset, as is
typical with familial Alzheimer’s disease.
Inquiring about a family history of CAD or cancer has
been more generally accepted by primary healthcare pro-
fessionals as a guide to clinical decision making than has
screening for carriers of recessive Mendelian disease such as
cystic fibrosis(71), at least in the United Kingdom. This may
be because: (i) a family history of CAD has immediate rele-
vance to the well-being of the practitioner’s patients instead
of only a long-term relevance to potential future patients,
and (ii) it does not raise the difficult and emotional topics
of prenatal diagnosis and the termination of wanted (but
affected) pregnancies, as inevitably occurs in the context
of screening for carriers of recessively inherited disorders.
There are several single-gene disorders that nevertheless
raise important issues for decisions about screening for the
complex, multifactorial, diseases.
FA M I L I A L H Y P E RC H O L E S T E RO L E M I A ( FH )
This autosomal dominant condition affects about one in
500 of the population in the United Kingdom and many
developed countries; it predisposes to CAD at an early age.
The prevention of such excess and early CAD is feasible in
many of those at risk by treatment with diet and medica-
tion (the statin drugs, inhibitors of HMG CoA reductase)
to achieve a reduction of serum cholesterol levels to within
the normal range. Before such safe and effective treatment
of hypercholesterolemia became available, measurement of
serum cholesterol was often advocated to assess the risk of
CAD, and that context—the awareness of risk without any
highly effective remedy—raised a number of issues.
It has long been recognized that families have their
own understanding of how they come to be at risk of heart
disease—acknowledging the effects of smoking, diet, and
genetic factors in an intuitive fashion and without necessar-
ily formulating a clear mechanism through which such fac-
tors could operate(72). While identifying those at increased
risk of disease can motivate some to comply with medical
advice, it can also lead to paradoxical consequences such as
inappropriate feelings of fatalism or of invulnerability, per-
haps encouraging indulgence in harmful behaviors(73) and
therefore leading on to unhelpful health consequences(74,75).
Interestingly, CAD is often seen as a disease predominantly
affecting men, so there can be a readiness to accept that men
are at increased risk but a reluctance to acknowledge that
women can also be affected(76). This could, of course, be
highly relevant to decisions about who seeks screening for
disease risk, and who complies with behavioral recommen-
dations or takes a prescribed risk-reducing medication(77).
Now that effective treatments are available for elevated
serum cholesterol, will the introduction of genetic testing
improve the management of at-risk individuals and fami-
lies? There are many mutations recognized—predominantly
in the LDLR gene but also in the apoB gene and some other
G e n etic T e s ting a nd G e no m ic Scr e e ning   •   2 2 9
loci—that are associated with hypercholesterolemia, so that
the molecular diagnostic work involved in cascade testing
within families is substantial. Measurement of serum cho-
lesterol remains the primary screening test, but molecular
diagnostics nevertheless improves the sensitivity of test-
ing within families once the family’s FH-associated muta-
tion has been identified(78,79). The question then arises as to
whether establishing the diagnosis of FH in an individual
through “traditional” biochemical measures of serum cho-
lesterol or with molecular methods of mutation detection
will have different consequences for patients’ perceptions of
the disease and for their motivation to comply with medi-
cal recommendations. There is some evidence from experi-
ence with newborn screening that a DNA-based diagnosis
of FH can lead to a sense of fatalism(80). This area of research
is currently being pursued vigorously, given its potential
public health importance, but whether these behavioral
considerations will influence the choice of screening or
cascade-testing laboratory methods, however, is rather
doubtful, as the test’s sensitivity, specificity, and cost are
likely to be decisive.
The general question of whether individuals with an
increased but readily modifiable risk of disease should
be identified through population screening or through
family-based cascade testing is important, and will have to
be kept under review as laboratory methods develop over
time for each relevant disease. Family-based cascade test-
ing for FH is certainly more cost-effective than population
screening(81,82), but there is debate about the “best” (most
effective and most ethical) way to achieve effective and
efficient cascading through a family(83). There are good rea-
sons for primary care practitioners to remain alert to those
at risk of FH—active, opportunistic case ascertainment—
and to consider serum cholesterol screening in healthy
young and middle-aged adults. Once one case in a family
has been identified biochemically, cascade testing may then
use molecular or biochemical methods, or both. Clearly,
it is important that cases of FH are identified so that they
can be offered treatment for their susceptibility to CAD;
this remains a Mendelian predisposition to cardiac disease,
however, rather than a common, complex disorder.
There are small but important Mendelian subsets con-
cealed within many of the common, complex disorders. We
have already referred to some of the familial cancer predis-
positions, including the BRCA1 and BRCA2 loci, in which
mutations predispose to breast, ovarian, and other cancers,
the apc locus associated with familial adenomatous polypo-
sis coli (FAP), and the mismatch repair loci associated with
hereditary nonpolypotic colon cancer (HNPCC). In cur-
rent medical practice, the role of genetic testing for cancer
susceptibility consists primarily of attempts to distinguish
the ~5% of affected cases where there is a strong familial
predisposition, with clear implications for the gene carrier
and other members of the family, from the other ~95%
where any inherited element in the predisposition—and
the corresponding implications for others in the family—is
much weaker. The small Mendelian subset among patients
with diabetes mellitus consists largely of families in which
the susceptibility is transmitted as an autosomal dominant
trait, and there is in addition a group whose predisposition
is mitochondrial. Clinical recognition of these Mendelian
groups can be helpful in disease management. MODY usu-
ally presents as the non-insulin-dependent or maturity-onset
type of diabetes but often occurring at rather a young age.
Knowledge of the diagnosis may be useful in ensuring that
other affected cases in the family are recognized promptly,
and families may seek genetic testing to resolve their uncer-
tainty about who will go on to develop the condition(84).
Attempts at the ascertainment of families affected
by Mendelian disorders through newborn screening has
been instructive, if our memories allow us to recall them.
This has been tried in the contexts of FH(80) and also
alpha1-antitrypsin deficiency(85), where the rationale for
neonatal screening is somewhat stronger, as the infants
could potentially be spared exposure to parental smoking,
so that the interval between testing and intervention in the
management of the individual child is much less—but nei-
ther of those programs gave encouraging results, and both
were discontinued.
More generally, it has become clear that health-related
lifestyle behaviors are not generally influenced by genetic
information. Highly specific decisions may be affected—
like the decision to have prophylactic surgery to mitigate
the risk of colorectal or breast/ovarian cancers—but “life-
style improvement” decisions are not so open to useful
change(86,87).
How can we reach agreement about when a test for
genetic susceptibility to one of the common, complex dis-
orders is ready for general clinical application? The mere
absence of continuing, overt psychological distress after
testing for such a susceptibility(88) is not at all sufficient to
justify such testing(21,89,90). The fraction of heritability that
can be accounted for by identifiable loci remains small
for most of these conditions, typically ~20%, so that the
case for population screening remains weak. Some of the
“missing heritability” will be the result of gene–gene and
gene–environment interactions that cannot yet be assessed,
and some (especially in psychiatric disease) will be the
result of new mutations that inflate the estimates of heri-
tability(91). And there are additional issues—political and
2 3 0   •   P rincipl e s o f G e no m ic M e dicin e
economic—to consider, such as how the test can be funded,
how individuals can be protected from the effects of adverse
discrimination in the face of such genetic tests(92), and who
should pay for the additional healthcare measures that are
then triggered for the index case and their family, by these
test results.
The expectation that genetic research will lead to a use-
ful ability to predict who will develop which of the com-
mon complex disorders appears, however, to be founded on
some fundamental misapprehensions. Furthermore, as the
interventions usually recommended to avoid such condi-
tions are all rather similar, with a focus on a healthy diet
(including some restriction of calorie intake) and an exer-
cise program, knowledge of the conditions for which one is
at somewhat increased risk may be unimportant, especially
if that knowledge is unlikely to be effective in influencing
one’s lifestyle.
The relevant misapprehensions about the ability to pre-
dict common complex disorders include:
1. The failure to recognize that the history of specific
populations will have exposed them to different
selective forces and generated different responses, so
that the usual claim that even larger studies will allow
researchers to reach statistical significance will not give
all the answers.
2. Discussions of human genetic polymorphism often
refer to “heterozygote advantage,” but in general
they pay much less attention to other important
categories of selection that act to maintain variation,
such as disruptive selection, density-dependent and
frequency-dependent selection, and the difference in
direction of selection acting on a variant at different
stages of the lifecycle, or in the two different sexes, or
in different (e.g., fluctuating or otherwise changing)
environments (as are found in the biologically simpler
Drosophila species,[93]).
3. The very fact that environments do alter, through
migration, climate change, population density, etc., will
give an advantage to organisms whose phenotype can
be fine-tuned to the particular circumstances likely to
be encountered—promoting the “predictive adaptive
responses” mediated by epigenetic mechanisms, as
discussed by Gluckman and Hanson(66).
The operation of predictive adaptive responses is con-
ceptually akin to the notion that, when circumstances
change to give an advantage to new mutations, selection
will also and inevitably be operating in favor of increased
 Free ebooks ==> www.Ebook777.com
mutation rates—the latter will hitchhike along with the
success of the former. These are reasons why the analysis of
clinical and genetic data to give insights into the causation
of the common complex diseases (CCDs) will be very com-
plex and difficult. This is not a reason for not setting out to
perform such research, but it is a reason, in conjunction with
the limited effect of genetic information on health-related
lifestyles, for caution in making claims about the ready clini-
cal applicability of such research.
Indeed, the attempt to focus on genetic risk factors and
individualized, behavioral responses serves to distract our
attention from the risk factors that are the result of collec-
tive, societal practices and that may be far more amenable
to collective solutions (such as controls on industrial pollu-
tion, transport policy, vitamin supplementation of essential
foods, etc.). To individualize the problems, when the only
effective solutions are likely to be collective, is profoundly
unhelpful and likely to be driven by a specific political
agenda.
Challenges that remain, and that have to be tackled
afresh in each study, include the possibility of population
stratification (which might obscure real effects or generate
confounding and misleading associations)(94), the appro-
priate way to make allowance for the performance of mul-
tiple tests when assessing the significance of findings, and
the difficulty of determining the biological basis of a true
SNP–disease association (identifying the causal factor in
disequilibrium with the SNP, lying in the same haplotype
block, if the SNP is not it). While methods are being devel-
oped to recognize SNPs that influence local and remote
gene expression, the interactions between polymorphic
structural arrangements, gene expression, and disease asso-
ciation remain challenging.
Another temptation that some investigators succumb to
is the use of social categories as if they were discrete, bio-
logical entities. In particular, finding differences (e.g., in
SNP allele frequencies) between racial or national groups
does not establish the two populations as discrete; clines in
allele frequencies across continents are common, so allele
frequencies will differ between arbitrarily defined groups in
the absence of biological boundaries. Indeed, the relevance
of population group to the interpretation of genetic infor-
mation should diminish as the genome-wide causal factors
involved are understood, so that the inadequate proxy of
“race” will become irrelevant.
Finally, we need to learn to bring together our under-
standing of genetics, epigenetics, and selection, and the role
of these factors in shaping the variation in disease within
and between populations. This can be seen as the founda-
tion stone of the grand synthesis required to understand
G e n etic T e s ting a nd G e no m ic Scr e e ning   •   2 3 1
www.Ebook777.com
the functional consequences of genetic diversity for disease.
Other necessary elements of this are the Gluckman and
Hanson model of predictive adaptive responses, derived
from Barker’s “thrifty phenotype” hypothesis of physiologi-
cal adaptation and an understanding of history’s contribu-
tion to genetic variation through drift (e.g., fall in population
size during pogroms); selection for physical endurance, such
as during hazardous migrations, relating to Neel’s “thrifty
gene” hypothesis(95); and Wilkinson and Pickett’s “Spirit
Level” study on the biological consequences of the distri-
bution of wealth and power(96). The political consequences
of the biological understanding of these population effects
must not be forgotten(97,98). The physical, nutritional envi-
ronment and the social setting work together to influence
the development of disease through genetic, epigenetic, and
physiological mechanisms.
C O N C LU S I O N
Developments in the technologies of genomics—the
high-throughput, genome-wide technologies, especially
array CGH and NGS—are bringing rapid change to the
clinical practice of genetic medicine. The proportion of rare
diseases for which a definitive diagnosis can be achieved
is rising, and the treatment of many malignancies, as well
as some rare genetic disorders, is being greatly improved
by these laboratory developments in genetic laboratory
diagnostics.
Some difficulties arise, however, from the very same
features of these technologies that make them so attractive.
The sheer volume of sequence generated leads to difficul-
ties in interpreting the findings, especially when novel vari-
ants are found whose pathological significance is unclear.
Furthermore, because these are genome-wide investigations,
they will often generate findings incidental to the original
indication for genetic testing, whether or not there are also
helpful findings pertinent to that original diagnostic ques-
tion. These changes mean that the formerly clear distinction
between a focused genetic test and genome-wide screening
collapses; the laboratory process and interpretive pipeline
may be the same, but the context determines what counts as
a pertinent result from a focused genetic test or as an inci-
dental product of genomic screening.
The occurrence of these VUSs and IFs requires changes
in the conversation between patient and professional, espe-
cially the talk in which a proposed genetic investigation is
explained, and in the process of documenting appropriate
consent for the investigation. There must be an appropri-
ate process for deciding what incidental results—especially
about a child—should be disclosed to the patient (or their
parents). If samples are stored long-term, or their test results
are maintained on file for future reference, then specific
plans should be established about when a reinterpretation
would be performed and how to contact the patient in the
future if additional findings emerge from our developing
collective ability to interpret genome data.
When a family is unaware of the potential finding
of important, family-specific IFs when their child has a
genome assessment, then the considerations that usually
caution against revealing predictive genetic information
about a child in relation to a future risk of an adult-onset
disease may no longer apply. The alternative to generating
and disclosing such predictive information to an unpre-
pared family, who may then have the opportunity to pre-
pare in whatever way is possible, is for all parties to remain
ignorant until someone in the family comes to be affected,
with even less opportunity to prepare. How to manage
these human, interactive consequences of the technological
developments in genomics is being debated among profes-
sionals with a passionate intensity. These challenges will
become simpler over the next decade as laboratory scientists
and clinical practitioners jointly develop their approaches
to handling genomic uncertainty and negotiate strategies
for an appropriate narrowing or widening of the focus of
investigation. As our collective experience of interpreting
genomic investigations accumulates over the years, the cor-
responding level of justified confidence in these interpreta-
tions will improve.
R EFE R E N C ES
1. PHG Foundation. 2013. Managing incidental and pertinent find-
ings from WGS in the 100,000 Genome Project (www.phgfounda-
tion.org.uk).
2. Elwyn G, Gray J, Clarke A. Shared decision making and
non-directiveness in genetic counselling. J Med Genet.
2000;37:135–138.
3. Fulda KG, Lykens K. Ethical issues in predictive genetic testing: a
public health perspective. J Med Ethics. 2006;32:143–147.
4. Wolff G, Jung C. Nondirectiveness and genetic counseling. J Gen
Counsel. 1995;4:3–25.
5. Clarke A. The process of genetic counselling: beyond nondirective-
ness. In: Harper P, Clarke AJ. Genetics, Society, and Clinical Practice.
Oxford, UK: Bios Scientific Publishers; 1997:179–200.
6. McAllister M. Predictive genetic testing and beyond:  a theory of
engagement. J Health Psychol. 2002;7(5):491–508.
7. Sarangi S, Bennert K, Howell L, Clarke A, Harper P, Gray J.
Initiation of reflective frames in counselling for Huntington’s dis-
ease predictive testing. J Genet Counsel. 2004;13:135–155.
8. Sarangi S, Bennert K, Howell L, Clarke A, Harper P, Gray J. (Mis)
alignments in counselling for Huntington’s disease predictive
testing:  clients’ responses to reflective frames. J Genet Counsel.
2005;14:29–42.
2 3 2   •   P rincipl e s o f G e no m ic M e dicin e
9. DudokdeWit AC, Tibben A, Duivenvoorden HJ, Niermeijer MF,
Passchier J, Trijsburg RW, and the Rotterdam/Leiden Genetics
Workgroup. Distress in individuals facing predictive DNA test-
ing for autosomal dominant late-onset disorders:  comparing
questionnaire results with in-depth interviews. Am J Med Genet.
1998;75:62–74.
10. Decruyenaere M, Evers-Kiebooms G, Cloostermans T, et  al.
Psychological distress in the 5-year period after predictive testing for
Huntington’s disease. Eur J Hum Genet. 2003;11:30–38.
11. Sobel S, Cowan DB. Impact of genetic testing for Huntington dis-
ease on the family system. Am J Med Genet. 2000;90:49–59.
12. Harper PS. Presymptomatic testing for late-onset genetic disor-
ders. Lessons from Huntington’s disease. In:  Harper P, Clarke AJ.
Genetics, Society and Clinical Practice. Oxford, UK: Bios Scientific
Publishers; 1997:31–48.
13. Kessler S, ed. Resta RC. Psyche and Helix. Psychological Aspects of
Genetic Counselling. New York and Chichester, England: Wiley-Liss;
2000.
14. Soldan J, Street E, Gray J, Binedell J, Harper PS. Psychological
model for presymptomatic test interviews:  lessons learned from
Huntington disease. J Genet Counsel. 2000;9(1):15–31.
15. Tibben A. Genetic counselling and presymptomatic test-
ing. In:  Bates G, Harper PS, Jones L, eds. Huntington’s Disease.
New York, USA: Oxford University Press; 2002:198–248.
16. Codori AM, Brandt J. Psychological costs and benefits of predictive
testing for Huntington’s disease. Am J Med Genet (Neuropsychiatric
Genet). 1994;54:174–184.
17. Richards MPM. Families, kinship and genetics. In:  Marteau T,
Richards M, eds. The Troubled Helix. Cambridge, UK: Cambridge
University Press; 1996:249–273.
18. Broadstock M, Michie S, Marteau T. Psychological consequences
of predictive genetic testing: a systematic review. Eur J Hum Genet.
2000;8:731–738.
19. Burke W, Pinsky LE, Press NA. Categorizing genetic tests to iden-
tify their ethical, legal and social implications. Am J Med Genet.
2001;106:233–240.
20. Burke W, Zimmern RL. Ensuring the appropriate use of genetic
tests. Nature Rev Genet. 2004;5:955–959.
21. PHG Foundation. 2011. Next steps in the sequence: the implica-
tions of whole genome sequencing for health in the UK.
22. Lim J, Macluran M, Price M, Bennett B, Butow P and the kCon-
Fab Psychosocial Group. Short- and long-term impact of receiving
genetic mutation results in women at increased risk for hereditary
breast cancer. J Genet Counsel. 2004;13(2):115–133.
23. Hallowell N, Foster C, Eeles R, Ardern-Jones A, Murday V, Watson
M. Balancing autonomy and responsibility: the ethics of generating
and disclosing genetic information. J Med Ethics. 2003;29:74–83.
24. Michie S, Smith JA, Senior V, Marteau TM. Understanding why
negative genetic test results sometimes fail to reassure. Am J Med
Genet. 2003;119A:340–347.
25. Wevers MR, Ausems MG, Verhoef S, Bleiker EM, Hahn DE,
Hogervorst FB, et  al. Behavioral and psychosocial effects of rapid
genetic counseling and testing in newly diagnosed breast cancer
patients:  design of a multicenter randomized clinical trial. BMC
Cancer. 10 Jan 2011;11:6.
26. Lindblad AN. To test or not to test: an ethical conflict with pres-
ymptomatic testing of individuals at 25% risk for Huntington’s dis-
order. Clin Genet. 2001;60:442–446.
27. Savulescu J, Kahne G. The moral obligation to create children with
the best chance of the best life. Bioethics. 2009;23:274–290.
28. Bekker H, Modell M, Denniss G, et  al. Uptake of cystic fibro-
sis testing in primary care:  supply push or demand pull? BMJ.
1993;306:1584–1586.
29. HGC (Human Genetics Commission). Increasing Options,
Informing Choice:  A  Report on Preconception Genetic Testing and
Screening. London: Department of Health; 2011.
 30. Grody WW. Expanded carrier screening and the law of unin-
tended consequences: from cystic fibrosis to fragile X. Genet Med.
2011;13(12):996–997.
31. Lazarin GA, Haque IS, Nazareth S, Iori K, Patterson AS, Jacobson
JL, et al. An empirical estimate of carrier frequencies for 400+ causal
Mendelian variants: results from an ethnically diverse clinical sam-
ple of 23,453 individuals. Genet Med. 2013;15(3):178–186.
32. de Wert GM, Dondorp WJ, Knoppers BM. Preconception care and
genetic risk: ethical issues. J Community Genet. 2012;3(3):221–228.
33. Faas BH, Feenstra I, Eggink AJ, et al. Non-targeted whole genome
250K SNP array analysis as replacement for karyotyping in fetuses
with structural ultrasound anomalies: evaluation of a one-year expe-
rience. Prenat Diagn. 2012;32(4):362–370.
34. Benn PA. Prenatal counseling and the detection of copy-number
variants. Genet Med. 2013;15(4):316–317.
35. Ehrich M, Deciu C, Zwiefelhofer T, Tynan JA, Cagasan L, Tim R,
et  al. Noninvasive detection of fetal trisomy 21 by sequencing of
DNA in maternal blood: a study in a clinical setting. Am J Obstet
Gynecol. 2011;204(3):205.e1–11.
36. Palomaki GE, Kloza EM, Lambert-Messerlian GM, et  al. DNA
sequencing of maternal plasma to detect Down syndrome:  an
international clinical validation study. Genet Med. 2011;13(11):
913–920.
37. Borry P, Evers-Kiebooms G, Cornel MC, Clarke A, Dierickx K,
and the Public and Professional Policy Committee (PPPC) of
the European Society of Human Genetics (ESHG). Genetic test-
ing in asymptomatic minors:  background considerations towards
ESHG Recommendations. Eur J Hum Genet. 2009;17(6):711–719
(Recommendations, 720–721).
38. European Society of Human Genetics. Genetic testing in asymp-
tomatic minors:  recommendations of the European Society of
Human Genetics. Eur J Hum Genet. 2009;17(6):720–721.
39. British Society for Human Genetics. 2010. Genetic Testing of
Children. Report of a working party of the British Society for
Human Genetics. http://www.bsgm.org.uk/media/678741/gtoc_
booklet_final_new.pdf
40. American Academy of Pediatrics and the American College of
Medical Genetics and Genomics. Ethical and policy issues in genetic
testing and screening of children. Pediatrics. 2013;131:620–622.
41. Ross LF, Saal HM, David KL, Anderson RR; American Academy of
Pediatrics; American College of Medical Genetics and Genomics.
Technical report:  ethical and policy issues in genetic testing and
screening of children. Genet Med. 2013;15(3):234–245.
42. Green RC, Berg JS, Grody WW, et  al. ACMG recommendations
for reporting of incidental findings in clinical exome and genome
sequencing. Genet Med. 2013;15(7):565–574.
43. American College of Medical Genetics and Genomics. Incidental
Findings in Clinical Genomics: A Clarification. A policy statement
of the American College of Medical Genetics and Genomics.
Bethesda, MD: ACMGG; 2013.
44. Feinberg J. The child’s right to an open future. In:  Aiken W, La
Fallette H, eds. Whose Child? Children’s Rights, Parental Authority
and State Power. Totowa, NJ: Littlefield, Adams; 1980:124–153.
45. Davis D. Genetic Dilemmas:  Reproductive Technology, Parental
Choices, and Children’s Futures. 2nd ed. Oxford, UK:  Oxford
University Press; 2010.
46. Hendriks KSWH, Grosfeld FJM, van Tintelen JP, et al. Can parents
adjust to the idea that their child is at risk of sudden death? Am J
Med Genet. 2005;138A:107–112.
47. Hercher L, Bruenner G. Living with a child at risk for psychotic ill-
ness. Am J Med Genet. 2008;146A; 2355–2360.
48. Wilson JMG, Junger G. Principles and Practices of Screening for
Disease. Geneva: World Health Organisation; 1968.
49. National Screening Committee. Second Report of the National
Screening Committee. London: Health Departments of the United
Kingdom; 2000.
G e n etic T e s ting a nd G e no m ic Scr e e ning   •   2 3 3
50. Parsons EP, Clarke AJ, Hood K, Lycett E, Bradley DM. Newborn
screening for Duchenne muscular dystrophy: a psychosocial study.
Arch Dis Child Fetal Neonatal Ed. 2002;86:F91–F95.
51. Biesecker LG. Opportunities and challenges for the integration of
massively parallel genomic sequencing into clinical practice: lessons
from the ClinSeq Project. Genet Med. 2012;14(4):393–398.
52. Timmermans S, Buchbinder M. Patients-in-waiting: living between
sickness and health in the genomics era. J Health Soc Behav.
2010;51(4):408–423.
53. Goldenberg AJ, Sharp RR. The ethical hazards and program-
matic challenges of genomic newborn screening. JAMA. 1 Feb
2012;307(5):461–462.
54. Savage DG, Antman KH. Imatinib mesylate—a new oral targeted
therapy. N Engl J Med. 2002;346(9):683–693.
55. Lee JM, Ledermann JA, Kohn EC. PARP inhibitors for BRCA1/2
mutation-associated and BRCA-like malignancies. Ann Oncol. Nov
12, 2013. Online Access, https://www.acmg.net/docs/Incidental_
Findings_in_Clinical_Genomics_A_Clarification_081413.pdf.
56. Oral EA, Simha V, Ruiz E, et  al. Leptin-replacement therapy for
lipodystrophy. N Engl J Med. 2002;346(8), 5700578.
57. Gaide O, Schneider P. Permanent correction of an inherited
ectodermal dysplasia with recombinant EDA. Nature Med.
2003;9:614–618.
58. Davies DM, de Vries PJ, Johnson SR, et al. Sirolimus therapy for angi-
omyolipoma in tuberous sclerosis and sporadic lymphangioleiomyo-
matosis: a phase 2 trial. Clin Cancer Res. 2011;17(12):4071–4081.
59. Du M, Liu X, Welch EM, Hirawat S, Peltz SW, Bedwell DM.
PTC124 is an orally bioavailable compound that promotes suppres-
sion of the human CFTR-G542X nonsense allele in a CF mouse
model. Proc Natl Acad Sci U S A. 2008;105(6):2064–2069.
60. Aartsma-Rus A, Fokkema I, Verschuuren J, et al. Theoretic applica-
bility of antisense-mediated exon skipping for Duchenne muscular
dystrophy mutation. Hum Mutat. 2009;30(3):293–299.
61. Jih KY, Hwang TC. Vx-770 potentiates CFTR function by promot-
ing decoupling between the gating cycle and ATP hydrolysis cycle.
Proc Natl Acad Sci U S A. 2013;110(11):4404–4409.
62. Benkhelifa-Ziyyat S, Besse A, Roda M, et  al. Intramuscular
scAAV9-SMN injection mediates widespread gene delivery to the
spinal cord and decreases disease severity in SMA mice. Mol Ther.
Feb 2013;21(2):282–290.
63. Guy J, Gan J, Selfridge J, Cobb S, Bird A. Reversal of neuro-
logical defects in a mouse model of Rett syndrome. Science
2007;315:1143–1147.
64. Abdala AP, Dutschmann M, Bissonnette JM, Paton JF. Correction
of respiratory disorders in a mouse model of Rett syndrome. Proc
Natl Acad Sci U S A. 2010;107:18208–18213.
65. Barker DJP, ed. Fetal and Infant Origins of Adult Disease.
London: British Medical Journal; 1992.
66. Gluckman P, Hanson M. The Fetal Matrix. Evolution, Development,
and Disease. Cambridge, UK: Cambridge University Press; 2005.
67. Pembrey M, Bygren LO, Kaati G, et  al., and the ALSPAC Study
Team. Sex-specific, male-line transgenerational responses in humans.
Eur J Hum Genet. 2006;14:159–166.
68. Humphries SE, Talmund PJ, Hawe E, Bolla M, Day INM, Miller GJ.
Apolipoprotein E4 and coronary heart disease in middle-aged men
who smoke: a prospective study. Lancet. 2001;358:115–119.
69. Ulvik A, Evensen ET, Lien EA, et al. Smoking, folate and methy-
lenetetrahydrofolate reductase status as interactive determinants
of adenomatous and hyperplastic polyps of colorectum. Am J Med
Genet. 2001;101:246–254.
70. Wang X, Zuckerman B, Pearson C, et al. Maternal cigarette smok-
ing, metabolic gene polymorphism, and infant birth weight. JAMA.
2002;287:195–202.
71. Payne Y, Williams M, Cheadle J, et al. Carrier screening for cystic
fibrosis in primary care: evaluation of a project in South Wales. Clin
Genet. 1997;51:153–163.
 72. Davison C, Frankel S, Smith GD. The limits of lifestyle: re-assessing
“fatalism” in the popular culture of illness prevention. Soc Sci Med.
1992;34(6):675–685.
73. Davison C, Frankel S, Smith GD. Inheriting heart trouble:  the
relevance of common-sense ideas to preventive measures. Health
Education Research. 1989;4:329–340.
74. Kinlay S, Heller RF. Effectiveness and hazards of case finding for a
high cholesterol concentration. BMJ. 1990;300:1545–1547.
75. Clarke A. Population screening for genetic susceptibility to disease.
BMJ. 1995;311:35–38.
76. Emslie C, Hunt K, Watt G. Invisible women? The importance
of gender in lay beliefs about heart problems. Sociol Health Ill.
2001;23(2):203–233.
77. Senior V, Smith JA, Michie S, Marteau TM. Making sense of risk: an
interpretative phenomenological analysis of vulnerability to heart
disease. J Health Psychol. 2002;7(2):157–168.
78. Heath KE, Humphries SE, Middleton-Price, Boxer M. A molecu-
lar genetic service for diagnosing individuals with familial hyper-
cholesterolaemia (FH) in the United Kingdom. Eur J Hum Genet.
2001;9:244–252.
79. Umans-Eckenhausen MAW, Defesche JC, Sijbrands EJG,
Scheerder RLJM, Kastelein JJP. Review of first 5  years of screen-
ing for familial hypercholesterolaemia in the Netherlands. Lancet.
2001;357:165–168.
80. Senior V, Marteau TM, Peters TJ. Will genetic testing for predispo-
sition for disease result in fatalism? A qualitative study of parents’
responses to neonatal screening for familial hypercholesterolaemia.
Soc Sci Med. 1999;48:1857–1860.
81. Marks D, Wonderling D, Thorogood M, Lambert H, Humphries
SE, Neil, HAW. Cost effectiveness analysis of different
approaches of screening for familial hypercholesterolaemia. BMJ.
2002;324:1303–1308.
82. Leren TP. Cascade genetic screening for familial hypercholesterol-
emia. Clin Genet. 2004;66:483–487.
83. Newson AJ, Humphries SE. Cascade testing in familial hypercholes-
terolaemia: how should family members be contacted? European J
Hum Genet. 2005;13:401–408.
84. Shepherd M, Hattersley AT, Sparkes AC. Predictive genetic testing
in diabetes: a case study of multiple perspectives. Qualitative Health
Research. 2000;10(2):242–259.
2 3 4   •   P rincipl e s o f G e no m ic M e dicin e
85. McNeil TF, Sveger T, Thelin T. Psychosocial effects of screening
for somatic risk:  the Swedish α1-antitrypsin experience. Thorax.
1988;43:505–507.
86. Marteau TM, French DP, Griffin SJ, et al. Effects of communicat-
ing DNA-based disease risk estimates on risk-reducing behaviours
(review). The Cochrane Library. 2010;10:1–74.
87. McBride CM, Koehly LM, Sanderson SC, Kaphingst KA. The
behavioural response to personalised genetic risk information: will
genetic risk profiles motivate individuals and families to choose more
healthful behaviours? Annu Rev Public Health. 2010;31:89–103.
88. Romero LJ, Garry PJ, Schuyler M, et  al. Emotional responses to
APO E genotype disclosure for Alzheimer disease. J Genet Counsel.
2005;12(2):141–150.
89. Wang C, Gonzalez R, Merajver SD. Assessment of genetic testing
and related counselling services: current research and future direc-
tions. Soc Sci Med. 2004;58:1427–1442.
90. Sanderson S, Zimmern R, Kroese M, Higgins J, Patch C, Emery
J. How can the evaluation of genetic tests be enhanced? Lessons
learned from the ACCE framework and evaluating genetic tests in
the United Kingdom. Genet Med. 2005;7(7):495–500.
91. Clarke A, Cooper DN. GWAS: heritability missing in action. Eur J
Hum Genet. 2010;18:859–861.
92. Clayton EW. Ethical, legal and social implications of genomic medi-
cine. N Engl J Med. 2003;349(6):562–569.
93. Vieira C, Pasyukova EG, Zeng Zh-B, Hackett JB, Lyman RF,
Mackay TFC. Genotype–environment interaction for quantitative
trait loci affecting life span in Drosophila melanogaster. Genetics.
2000;154:213–227.
94. Berger M, Stassen HH, Kohler K, et  al. Hidden population sub-
structures in an apparently homogeneous population bias associa-
tion studies. Eur J Hum Genet. 2006;14:236–244.
95. Neel JV. Diabetes mellitus: a “thrifty” genotype rendered detrimen-
tal by “progress?” Am J Hum Genet. 1962;14:353–362.
96. Wilkinson R, Pickett K. The Spirit Level: Why Equality Is Better for
Everyone. London: Penguin Books; 2010.
97. McDermott R, Ethics, epidemiology and the thrifty gene: biological
determinism as a health hazard. Soc Sci Med. 1998;47:1189–1195.
98. Räisänen U, Bekkers M-J, Boddington P, Sarangi S, Clarke A. The
causation of disease: the practical and ethical consequences of com-
peting explanations. Med Health Care Philos. 2006;9:293–306.
 15.
BIOBANKING FOR GENOMICS-BASED
TRANSLATIONAL MEDICINE
Steven J. Madore
INTRODUCTION
The term “translational research” can be defined as the pro-
cess of converting basic scientific discoveries into new clini-
cal diagnostics and treatments. The process is a two-way
street in that basic research discoveries can drive patient
treatment, and, conversely, medical information obtained
at the patient bedside can be used to drive basic labora-
tory research(1–4). Genomics is the study of the complete
genetic material, including genes and their functions, of an
organism and has its origins with the successful completion
in 2001 of generating a complete sequence of the human
genome(5–7). The information obtained by genomic analyses
will enable the identification of all human genes and their
regulatory regions, thus facilitating the elucidation of the
genetic mechanisms of many human diseases. Genomics
holds great promise in advancing translational research,
with specific benefits in three areas—a better understand-
ing of pathophysiology, development of new and better
diagnostic tools, and discovery and validation of novel
therapies(8,9).
The engines driving the development of new and bet-
ter approaches and technologies resulting in advancing the
“bench-to-bedside” paradigm of translational medicine are
basic and clinical research(2,4). Key discoveries rely on sev-
eral factors, including solid experimental design, relevant
in vitro and in vivo models, and new technological advance-
ments, as well as the ready availability of high quality, well
annotated biospecimens. Sources for such valuable biologi-
cal samples are biobanks or biorepositories. The term “bio-
bank” appeared for the first time in PubMed in 1996 and
is the general term for a repository of biological samples
or biospecimens(10). Biobanks add significant value and
integrity to samples, given that sample collection, anno-
tation, processing, and storage are executed according to
well-established and standardized procedures(11–13). Thus,
biobanks offer researchers a “one-stop shopping” resource
235
for useful biological samples. The importance of biospeci-
men resources and their potential in driving basic and
translational research is demonstrated by the inclusion of
biobanks in a special March 2009 edition of Time magazine
among the “Ten Ideas Changing the World Right Now”(14).
In this review, I will introduce the concept of biobanking
science and discuss the importance of these repositories as
contributors to translational research, as well as highlight
some of the challenges associated with biospecimens as
material for analysis by genomic technologies.
BIOBANKING
The basic functions of biobanks are the collection, acces-
sioning, processing, quality assessment, safe storage, and
distribution of biospecimens(15,16). Biobanks range in size
from single-freezer collections in a research laboratory or
pathology suite, to much larger collections of thousands
to millions of samples that are linked to relevant personal
and health information, such as medical and family history
and lifestyle. Biobanks can operate as a centralized facility
that accepts samples from multiple locations (like national
biobanks) or in a “federated” mode where separate institu-
tions maintain distinct collections but agree to list them
in a central shared database(17,18). Collections may consist
of “project-driven” specimens collected and distributed to
answer specific research questions, or “general” reference
collections, which may not be collected to meet a particu-
lar research goal but are made available for an assortment
of research uses(17). Large, national, population-based bio-
banks contain very large collections of well-annotated
human DNA collected from volunteers with and without
disease, and now exist or are being developed in a number
of countries, including Iceland, Japan, Canada, Sweden,
and the United Kingdom(19–21). Many of these biobanks
not only receive, process, and store samples, but also collect
and compile medical histories and lifestyle and genealogi-
cal information from the original sample donor. Samples
stored in large population biobanks have proven to be
extremely useful as input material for genome-wide associa-
tion studies (GWAS) to identify genes that contribute to
human disease(22–25).
Critical to the downstream usefulness of biospecimens
is that they be obtained under appropriate ethical and legal
criteria(26). Proper consent for biospecimen acquisition
is least problematic when donors are healthy and compe-
tent adults. In contrast, consent considerations can arise
when potential donors are sick and/or vulnerable(27–30).
While the consent of participants is usually required before
samples can be used in research, the nature of this consent
and how it is obtained can vary widely. Given the impor-
tance of human biospecimens, their collection and pro-
cessing must adhere to standards that maintain biological
integrity as well as ensure that relevant patient-related and
biospecimen-specific information remains linked to each
specimen(26–30). While a full discussion of the issue of sample
consent is extremely important, it is beyond the scope of
this review; however, researchers requesting biospecimens
from established biobanks need to be aware of legal and
ethical implications of obtaining properly consented sam-
ples for their research programs.
Human biospecimen sample types found in biobanks
include normal and diseased tissue, whole blood, plasma,
serum, urine, cerebral spinal fluid, and saliva, while tumor
banks consist of well-annotated tumor tissue (either
stored frozen or as formalin-fixed paraffin-embedded
[FFPE] blocks or slides) obtained from cancer patients.
Immortalized cell lines derived by transformation of periph-
eral blood premature B-cells with Epstein-Barr virus (EBV)
remain a rich source for genomic-based research. Other
types of cell lines found in biobanked collections include
lines derived from tumor biopsies, as well as those that have
been genetically manipulated to express mutant or altered
versions of certain genes. Biobanks may also contain large
collections of animal and plant specimens, as well as various
bacterial and viral strains. Finally, reagents useful for basic
research, including genomic DNA, RNA, plasmid con-
structs, purified proteins, tissue extracts, and polyclonal and
monoclonal antibodies, can also be important components
of the biorepository.
So what are the origins of biospecimens stored within
biobanks? Probably the most valuable samples for transla-
tional research originate from well-planned, large clinical
studies designed with specific aims and outcomes. These
studies are often longitudinal and powered to recruit a
sufficient patient population that also includes healthy
2 3 6   •   P rincip l e s o f G e no m ic M e dicin e
donors or unaffected controls. Many epidemiological
studies are focused on or include the goal of biomarker
identification(31). Collection of the various sample types
and accompanying relevant medical records from study
participants is often coordinated by the biobank. The
degree and accuracy of sample-linked data adds tremen-
dous value to the sample when linking genetic findings
with donor phenotype.
Biospecimens are collected at sites away from the biore-
pository; thus the mode of packaging and transportation is
a critical parameter in assuring biospecimen stability during
transit(32–35). This may include the addition of wet or dry ice
in sufficient quantity and in insulated containers to account
for the anticipated period of transit from collection site
to biobanking facility. Blood packages shipped overnight
by commercial carrier may encounter extreme seasonal
temperatures, and careful precautions need to be taken to
ensure that the shipping containers include adequate pro-
tection against extreme ambient temperature deviations.
Prolonged storage of blood at room temperature prior to its
processing can affect the stability of the macromolecules like
DNA and proteins, as well as the efficiency of lymphocyte
separation by Ficoll density centrifugation(36–40). Biofluids
such as plasma, serum, urine, or cerebrospinal fluid (CSF)
are shipped frozen on dry ice. Established cell lines can be
shipped as live cultures in growth media or cryopreserved
under liquid nitrogen vapor.
Human tissue biobanking—the procurement of tissue
samples from live donors or cadavers—requires special con-
siderations related to minimizing pre-analytical variation.
Pre-analytical variables, such as the time the tissue remains
at room temperature, can cause significant variability and
bias in downstream molecular analysis, including reactive
changes that begin with oxidative, hypoxic, and meta-
bolic stress, and culminate in apoptosis(41). Standardized
methods for tissue procurement to minimize sources of
pre-analytical variation as well as the development of bio-
markers indicative of the biological state of the tissue prior
to freezing remain active areas of research in biobanking
science(41–43).
B I O B A N K E D S A M P LE T Y P ES
A N D   G E N O M I C T E C H N O L O G I ES
Upon arrival at the biobank, samples are usually assigned
a unique reference ID to facilitate tracking during labora-
tory processing and inventory during storage and distribu-
tion. While some submitted samples remain unprocessed
and are placed in safe storage after accessioning, biobanks
may extract or isolate important components or “prod-
ucts” from submitted samples like whole blood or saliva.
Large epidemiological studies focused on genetics include
whole-blood specimens that yield sufficient quantities of
high-quality DNA amendable to several genomic analysis
platforms. Human blood is collected into polypropylene
tubes containing the anticoagulants ethylenediaminetet-
raacetic acid (EDTA), heparin, or acid citrate dextrose.
If shipped properly, DNA can be extracted directly from
fresh blood or from blood that has been stored frozen
at –80 C. Most large biobanks rely on robotic instrumen-
tation for DNA extraction, using a number of different
extraction methodologies that are scalable and appropri-
ate for automation(44). These include salt precipitation
(“salting out”), and direct capture on silica membranes or
magnetic beads, with typical DNA yields ranging from
100 µg to 400 µg per 10 ml of whole blood(45–47). Purified
genomic DNA must undergo rigorous quality assessment.
For example, DNA purity can be determined using spec-
trophotometry, and DNA size by agarose gel electropho-
resis or lab-on-a-chip techniques using a set of molecular
size standards for comparison(48). Due to shearing forces
introduced during most extraction methods, average DNA
size with typical extraction methods described above aver-
ages around 100,000 base pairs, which is sufficient for most
existing genomic technologies(44–47).
Whole blood collected in PAXgene tubes containing
a stabilization reagent to neutralize intrinsic RNase activ-
ity can be used for the extraction of high-quality total
RNA that can serve as input material for high-throughput
expression-profiling analysis. Whole blood can also serve as
raw material for the isolation of live cells such as peripheral
blood mononuclear cells (PBMCs). Primary fibroblasts can
be isolated from skin-punch biopsies by outgrowth in tissue
culture dishes using appropriate growth media. Blood can
be further processed into plasma or serum for proteomics
studies. Biofluids such as urine and cerebral spinal fluid
can serve as substrate for both proteomic and metabolomic
analyses.
DNA samples of sufficient quantity and quality for
GWAS can be isolated from saliva, which can be easily
self-collected in prefabricated vessels that contain a stabili-
zation solution(49–52). Once collected, the saliva can be stored
for months at room temperature with minimal effects on the
yield or quality of extracted genomic DNA(53). The analysis
of saliva-derived genomic DNA by ultra-high-throughput
genotyping, in which up to one million distinct single-base
positions can be interrogated in parallel, has greatly facili-
tated the ability to assess common variation across the
genome(54). Along with the ease of collection and advances
B iobanking f or G e no m ic s -B a s e d T ran s l ationa l M e dicin e   •   2 3 7
in GWAS technologies, human genetic association studies
have moved from looking at a single SNP or a small set of
SNPs to very large studies involving international consor-
tium analyzing genome-wide analysis of tens of thousands
of individuals(55–57). The increase in cohort size and breadth
of coverage across the genome lends authenticity and reli-
ability to GWAS studies(54).
The availability of large collections of archived saliva
from GWA studies represents a valuable resource for
advancing the study of the microbiota in the mouth as
well(58–60). The salivary microbiota are a potential diagnos-
tic indicator of several diseases, and because the oral cavity
is often the entry point for bacteria, there is the likelihood
that interactions between the saliva microbiome and other
microbiomes in the human body (particularly within the
intestinal tract) play a role in human disease(58–60). Genomic
DNA purified from human saliva can be analyzed by
high-throughput DNA sequencing to characterize the
enormous diversity of microbial organisms present in the
human salivary microbiome. This characterization should
provide insight into what role the oral microbial commu-
nity has in human health and disease(58).
Another rich source of genomic DNA found in bio-
banks is that isolated from lymphoblastoid cell lines, or
LCLs, which are EBV-immortalized B-cells derived from
donor blood. LCLs, which can be grown easily in the labo-
ratory and represent a renewable source of genomic DNA
from the original donor, have been used in a wide variety
of genomic studies, including GWAS and gene-expression
profiling, and next-generation sequencing data are now
publicly available for hundreds of established LCLs. Large,
well-characterized collections of LCLs that have been used
in pharmacogenomic studies include the HapMap and
Human Variation Panel lymphoblastoid cell lines(61,62)—
both of which are publicly available from the Coriell Institute
for Medical Research (www.coriell.org). Researchers have
also used cultured LCLs as a cost-effective, unlimited cell
resource to identify genes and variants capable of predicting
both drug response and drug toxicity. The utility of LCLs in
cancer pharmacogenomic studies has recently been demon-
strated by the identification of polymorphisms associated
with clinical phenotypes such as survival and response to
chemotherapy(63).
The recent development of “next-generation sequenc-
ing” (NGS) technologies holds great promise in accel-
erating genomic studies by increasing efficiency and
significantly reducing the cost of generating whole
genome sequences. Most new techniques require exten-
sive biochemical labeling and sample preparation, and
do not allow long, single-molecule read lengths to be
achieved(64–66). The push for higher resolution, lower cost,
and the ability to obtain longer read lengths and DNA
strand-phasing continues, with even more advanced
technologies such as nanopore-based DNA sequencing
looming on the horizon(66,67). While NGS platforms tend
to require smaller input DNA fragment sizes, the next
wave of DNA sequencers, with the ability to read very
large single DNA molecules, will require modifications
of existing DNA  extraction methods such that much
larger, megabase-size DNA is obtained(68,69). Having made
significant financial investments in automated nucleic
acid extraction instrumentation, biobanks will be chal-
lenged to develop and validate new methodologies for
DNA extraction to meet the demand of newer sequenc-
ing technologies. Extraction methods for megabase-sized
DNA isolation were pioneered in the 1980s, and these
methods, in which whole cells are embedded in agarose
gels or sequestered onto microbeads, are cumbersome
and technically challenging (for review, see reference
70). Most importantly, each isolation method is not
amenable to scaling up—an important issue in regard
to high-throughput processing of the large number of
samples processed at biobanks. Clearly there is an unmet
need for a scalable, low cost, and efficient method for
obtaining genomic DNA of sufficient size and quality
that can be used as a template for newer DNA sequencing
methodologies.
Given that malignant neoplasms are the most common
cause of death in humans, the application of genomic tech-
nologies to the study of cancer is a leading force in devel-
oping new diagnostic and therapeutic strategies(71). Over
the last few years, advances in NGS technologies have
brought us the ability to detect unique patterns of muta-
tions and genomic rearrangements within individual tumor
samples(72). Given the heterogeneity of tumors, the increase
in the sensitivity of genomic technologies like NGS and
RNA-seq (see below) in detecting rare sequence alterations
allows clinicians to obtain distinct genetic profiles of malig-
nant cellular subtypes composing each tumor biopsy(73–78).
While most of these large-scale sequencing efforts have
relied on DNA isolated from fresh tissues(79), which can be
difficult to obtain, biobanks often contain large collections
of archived FFPE samples derived from surgical tumor
resections and histopathological examinations(80). DNA
extracted from FFPE samples can be sequenced by existing
methods, and the ability to obtain high-quality reads from
small amounts of archived tissue will allow access to a vari-
ety of clinically relevant samples(80,81).
In addition to the genetic information encoded within
genomic DNA, biobanked specimens can also be valuable
2 3 8   •   P rincip l e s o f G e no m ic M e dicin e
as a source for assessing chromatin modifications such as
DNA methylation and post-translational modifications of
histones. This epigenetic information can be transmitted
from mother to daughter cells during development and
determines cell fate by regulating gene expression. Because
the epigenetic state of the cell can be significantly altered
by developmental and environmental cues, the “epigenome”
is a critical interface between the human genome and the
environment(82). Dysregulation or disruption in chromatin
modifications can lead to human diseases such as cancer(83).
Along with new DNA sequencing methods, advances in
mass spectrometry-based proteomics technologies provide
a powerful toolbox for epigenetic analysis, thereby increas-
ing our understanding of chromatin structure and func-
tion(82). Immunoprecipitation of cross-linked chromatin
(Chromatin IP, or ChIP) in combination with microarrays
(ChIP-chip) or DNA sequencing (ChIP-seq) can now be
used to comprehensively map transcription factor-binding
sites in vivo(84).
In addition to advances in DNA sequencing, the abil-
ity to determine the primary sequence of the entire RNA
population in a biological sample is now available. This
method—called RNA-seq, has revolutionized the field of
transcriptomics by eliminating technological limitations
of microarray technologies and enabling the quantita-
tive and qualitative analysis of all types of RNA mole-
cules—including messenger RNA (mRNA), microRNAs
(miRNAs), and other non-coding RNAs. In comparison
to a “closed” analysis system like solid-phase microar-
rays, which are limited by fixed sequence capture probes
and limited dynamic range(85,86), RNA-seq technology
enables the detection of novel RNA molecules through
massive-scale complementary DNA (cDNA) sequencing.
Thus, RNA-seq captures the extent of gene-expression
variation and resulting protein diversity generated by
alternative mRNA start site usage, mRNA alternative
splicing, polyadenylation site selection, and RNA edit-
ing. RNA-seq technology has proven to be especially
beneficial in cancer research through the identification
of gene fusion transcripts and miRNAs. Direct RNA-seq
technology that bypasses the need for cDNA synthesis
from RNA template enables the identification of anti-
sense RNA transcripts that are functional molecules
with roles in both normal and disease states(87). Several
examples exist of the identification of non-coding RNAs
(ncRNAs) whose dysregulation has been shown to be
involved in tumorigenesis as well as other diseases (for
review, see reference 88).
RNA-seq has been especially useful in identifying novel
miRNAs, key regulators of gene expression that have been
shown to be involved in a variety of biological processes.
Some miRNAs exhibit differential expression levels in can-
cer and have been shown to affect cellular transformation,
carcinogenesis, and metastasis, acting either as oncogenes or
as tumor suppressors(89,90). Careful attention must be paid
to RNA extraction methods so that small RNAs, including
miRNAs, are retained in the final RNA preparation. For
blood collected in PAXgene tubes, there are several com-
mercially available kits amenable to automation that use
silica membranes embedded in disposable spin columns
or magnetic bead technology to capture pure RNA from a
biological sample. Binding and elution conditions must be
optimized so that the final purified RNA sample includes
the miRNA fraction.
Fluid biospecimens such as blood, urine, saliva, CSF,
seminal fluid, and bronchial lavage contain cell-free pro-
teins, nucleic acids, lipids, and metabolites, all of which
can be utilized as biomarkers of disease. The availability of
high-quality biospecimens is a requirement for novel bio-
marker discovery that can promote disease risk-assessment,
diagnosis, management, and treatment(91). Biospecimen
quality depends upon a number of pre-analytical variables
such as collection and processing techniques, as well as
freezing/thawing stability and long-term storage stabil-
ity(91–93). Millions of fluid biospecimens are currently stored
in biobanks across the world, and with appropriate annota-
tion and consent, and if properly collected, processed, and
stored, archived fluid biospecimens are valuable raw mate-
rial for translational research.
C O N C LU S I O N S A N D S U MM A RY
The availability of high-quality, well-annotated bio-
specimens to both basic and clinical research efforts is a
critical component for advancing biomedical research.
Because biospecimen quality is affected by a number of
pre-analytical factors that are introduced through the
variety of biospecimen collection, processing, and storage
procedures, there is a need to establish best practices and
recognize and further advance the field of “biospecimen
science”—the emerging field of study that is attempting
to quantify and control such variability(94). Various efforts
are now under way around the world to establish research
programs, evidence-based biospecimen protocols, and
standards to improve the overall quality of biospecimens
for research(95). Thus, biobanks will continue to be a valu-
able resource for biological raw material for both basic
and clinical researchers, and hold great promise as provid-
ers of the raw material necessary for fueling the genomic
B iobanking f or G e no m ic s -B a s e d T ran s l ationa l M e dicin e   •   2 3 9
technology engine towards key discoveries and advance-
ments in translational research.
AC K N OW LE D G ME N T S
The author wishes to apologize to all whose published works
were not referenced in the manuscript due to space limitations.
Special thanks to Drs. Joseph Jarvis, Christine Beiswanger,
and Victoria Kelly for their helpful review of this manuscript.
R EFE R E N C ES
1. Goldblatt EM, Lee WH: From bench to bedside: the growing use
of translational research in cancer medicine. Am J Transl Res 2:1–18,
2000.
2. Woolf SH: The meaning of translational research and why it mat-
ters. JAMA 299:211–213, 2008.
3. Zerhouni EA:  Translational and clinical science—time for a new
vision. NEJM 353:1621–1623, 2005.
4. Marincola FM: Translational medicine—a two-way road. J Transl
Med 1:1, 2003.
5. International Human Genome Sequencing Consortium: Finishing
the euchromatic sequence of the human genome. Nature 431:931–
945, 2004.
6. Venter JC, Adams MD, Myers EW, et  al.:  The sequence of the
human genome. Science 291:1304–1351, 2001.
7. NCI Dictionary of Cancer Terms, at http://www.cancer.gov/
dictionary?cdrid=446543.
8. Hawkins N, de Vries J, Boddington P, Kaye J, Heeney C: Planning
for translational research in genomics. Genome Med 1:87–94, 2009.
9. Kumar, D:  Genomic medicine:  a new frontier of medicine in the
twenty-first century. Genomic Med 1:3–7, 2007.
10. Loft S, Poulsen HE:  Cancer risk and oxidative DNA damage in
man. J Mol Med 74:297–312, 1996.
11. Cambon-Thomsen A, Ducoumau P, Gourraud P-A, Pontille
D: Biobanks for genomics and genomics for biobanks. Comp Funct
Genom 4:628–634, 2003.
12. Department of Health and Human Services (Internet). Personalized
Healthcare, at http://www.hhs.gov/myhealthcare/news/presonalized-
healthcare-9-2007.html.
13. Compton C: Getting to personalized cancer medicine—taking out
the garbage. Cancer 110:1641–1643, 2007.
14. Park A: Ten ideas changing the world right now. Time, 173(11):17–
22, 2009.
15. Baker, M:  Biorepositories:  Building better biobanks. Nature
486:141–146, 2012.
16. Riegman PHJ, Morente MM, Betsou F, de Blasio P, Geary P, the
Marble Arch International Working Group on Biobanking for
Biomedical Research: Biobanking for better healthcare. Mol Oncol
2:213–222, 2008.
17. Steinberg K, Beck J, Nickerson D, et al.: DNA banking for epidemi-
ologic studies: a review of current practices. Epidemiology 13:246–
254, 2002.
18. Murtagh MJ, Demir I, Harris JR, Burton PR: Realizing the promise
of population biobanks: a new model for translation. Hum Genet
130:333–345, 2011.
19. Swede H, Stone CL, Norwood AR: National population-based bio-
banks for genetics research. Genet Med 9:141–149, 2007.
20. Austin MA, Harding S, McElroy C:  A comparison of eight pro-
posed international genetic databases. Comm Genet 6:37–45, 2003.
 21. Maschke KJ: Navigating an ethical patchwork—human gene banks.
Nat Biotechnol 23:539–545, 2005.
22. Norrgard K:  Genetic variation and disease:  GWAS. Nature Educ
1:1, 2008.
23. Steinthorsdottir V et al.: A variant in CDKAL1 influences insulin
response and risk of type 2 diabetes. Nat Genet 39:770–775, 2007.
24. Wellcome Trust Case Control Consortium:  Genome-wide asso-
ciation study of 14,000 cases of seven common diseases and 3,000
shared controls. Nature 447:661–682, 2007.
25. Visscher PM, Montgomery GW: Genome-wide association studies
and human disease: from trickle to flood. JAMA 302:2028–2029,
2009.
26. Elger BS, Caplan AL:  Consent and anonymization in research
involving biobanks. EMBO Rep 7:661–666, 2006.
27. Kegley JA:  Challenges to informed consent. EMBO Rep 5:832–
836, 2004.
28. Lin Z, Owen AB, Altman RB: Genomic research and human sub-
ject privacy. Science 305:183, 2004.
29. Knoppers BM: Biobanking: international norms. J Law Med Ethics
33:7–14, 2005.
30. Spencer B, Koutaissoff D, Lehr H-A:  Informed consent:  Biobank
donors should have a say. Nature 481:443, 2012.
31. Hewitt RE: Biobanking: the foundation of personalized medicine.
Curr Opin Oncol 23:112–119, 2011.
32. Olson WC, Smolkin ME, Farris EM, et  al.:  Shipping blood to a
central laboratory in multicenter clinical trials:  effect of ambient
temperature on specimen temperature, and effects of temperature
on mononuclear cell yield, viability and immunologic function.
J Transl Med 8:26–38, 2011.
33. Vaught JB:  Blood collection, shipment, processing, and storage.
Cancer Epidemiol Biomarkers Prev 15:1582–1584, 2006.
34. Vaught JB, Caboux E, Hainaut P: International efforts to develop
biospecimen best practices. Cancer Epidemiol Biomarkers Prev
19:912–915, 2010.
35. Hallmans G, Vaught JB:  Best practices for establishing a bio-
bank. In:  Methods in Biobanking. Edited by:  Dillner J. Springer
Science+Business Media New York, NY; 241–260, 2011.
36. Leyland-Jones BR, Ambrosone CB, Bartlett J, et  al.:
Recommendations for collection and handling of specimens from
group breast cancer clinical trials. J Clin Oncol 26:5638–5644,
2008.
37. Ashmore LM, Shopp GM, Edwards BS: Lymphocyte subset analy-
sis by flow cytometry. Comparison of three different staining tech-
niques and effects of blood storage. J Immunol Meth 118:209–215,
1989.
38. Garraud O, Moreau T: Effect of blood storage on lymphocyte sub-
populations. J Immunol Meth 75:95–98, 1984.
39. Weiblen BJ, Debell K, Giorgio A, Valeri CR: Monoclonal antibody
testing of lymphocytes after overnight storage. J Immunol Meth
70:179–183, 1984.
40. Kaplan J, Nolan D, Reed A: Altered lymphocyte markers and blas-
togenic responses associated with 24-hour delay in processing of
blood samples. J Immunol Meth 50:187–191, 1982.
41. Espina V, Mueller C, Edmiston K, Sciro M, Petricoin EF, Liotta
LA: Tissue is alive: new technologies are needed to address the prob-
lems of protein biomarker pre-analytical variability. Proteomics—
Clin Appl 3:874–882, 2009.
42. Jackson DH, Banks RE: Banking of clinical samples for proteomic
biomarker studies: a consideration of logistical issues with a focus on
pre-analytical variation. Proteomics—Clin Appl 4:250–270, 2010.
43. Bevilacqua G, Bosman F, Dassesse T, et al.: The role of the patholo-
gist in tissue banking: European Consensus Expert Group report.
Virchows Arch 456:449–454, 2010.
44. Riemann K, Adamzik M, Frauenrath S, et al.: Comparison of man-
ual and automated nucleic acid extraction from whole-blood sam-
ples. J Clin Lab Anal 21:244–248, 2007.
2 4 0   •   P rincip l e s o f G e no m ic M e dicin e
45. Miller SA, Dykes DD, Polesky HF: A simple salting out procedure
for extracting DNA from nucleated cells. Nuc Acid Res16:1215,
1988.
46. Baker MP, Mitchell A, Bridge C, et al.: Isolation of genomic DNA
from blood using a novel filter-based DNA purification technology.
BioTechniques 31:142–145, 2001.
47. Rudi K, Kroken M, Dahlberg OJ, Deggerdal A, Jakobsen KS, Larsen
F: Rapid, universal method to isolate PCR-ready DNA using mag-
netic beads. BioTechniques 22:506–511, 1997.
48. Gallagher SR, Desjardins PR: Quantitation of DNA and RNA with
absorption and fluorescence spectroscopy. Curr Protoc Mol Biol.
76:A.3D.1–A.3.D.21 2006.
49. Quinque D, Kittler R, Kayser M, Stoneking M, Nasidze I: Evaluation
of saliva as a source of human DNA for population and association
studies. Anal Biochem 353:272–277, 2006.
50. Schipper RG, Silletti E, Vingerhoeds MH: Saliva as research mate-
rial: biochemical, physicochemical and practical aspects. Arch Oral
Biol 52:1114–1135, 2007.
51. Ng DP, Koh D, Choo S, Chia KS: Saliva as a viable alternative source
of human genomic DNA in genetic epidemiology. Clin Chim Acta
367:81–85, 2006.
52. Rylander-Rudqvist T, Hakansson N, Tybring G, Wolk A: Quality
and quantity of saliva DNA obtained from the self-administrated
Oragene DNA method—a pilot study on the cohort of Swedish
men. Cancer Epidemiol Biomarkers Prev 15:1742–1745, 2006.
53. Nunes AP, Oliveira IO, Santos BR, et  al.:  Quality of DNA
extracted from saliva samples collected with the Oragene™ DNA
self-collection kit. BMC Med Res Methodol 12:65–69, 2012.
54. Topol EJ, Murray SS, Frazer KA: The genomics gold rush. JAMA
298:218–221, 2007.
55. Easton DF, Pooley KA, Dunning AM, et al.: Genome-wide associa-
tion study identifies novel breast cancer susceptibility loci. Nature
447:1087–1093, 2007.
56. Hunter DJ, Kraft P, Jacobs KB, et al.: A genome-wide association
study identifies alleles in FGFR2 associated sporadic postmeno-
pausal breast cancer. Nat Genet 39:870–874, 2007.
57. Saxena R, Vioght BF, Lyssenko V, et al.: Genome-wide association
analysis identifies loci for type 2 diabetes and triglyceride level.
Science 316:1331–1336, 2007.
58. Lazarevic V, Whiteson K, Gaïa N, et  al.:  Analysis of the sali-
vary microbiome using culture-independent techniques, J Clin
Bioinforma 2:4, 2012.
59. Li J, Quinque D, Tang K, Stoneking M:  Global diversity in the
human salivary microbiome. Genome Res 19:636–643, 2009.
60. Turnbaugh PJ, Ley RE, Hamady M, Fraser-Liggett CM, Knight R,
Gordon JI: The human microbiome project. Nature 449:804–810,
2007.
61. International HapMap Consortium:  A haplotype of the human
genome. Nature 437:1299–320, 2005.
62. The International HapMap Consortium:  The International
HapMap Project. Nature 426:789–796, 2003.
63. Wheeler HE, Dolan ME:  Lymphoblastoid cell lines in pharma-
cogenomic discovery and clinical translation. Pharmacogenomics
13:55–70, 2012.
64. Kircher M, Kelso J: High-throughput DNA sequencing—concepts
and limitations. Bioessays 32:524–536, 2010.
65. Ansorge WJ:  Next-generation DNA sequencing techniques.
New Biotechnol 25:195–203, 2009.
66. Schneider GF, Dekker C:  DNA sequencing with nanopores.
Nat Biotechnol 30:293–370, 2012.
67. Metzker M:  Sequencing technologies—the next generation.
Nat Rev Genet 11:31–45, 2010.
68. Cherf GM, Lieberman KR, Rashid H, Lam CE, Karplus K,
Akeson, M: Automated forward and reverse ratcheting of DNA
in a nanopore at 5-angstrom precision. Nat Biotechnol 30:344–
348, 2012.
 69. Manrao EA, Derrington IM, Laszlo AH, et  al.:  Reading DNA at
single-nucleotide resolution with a mutant MspA nanopore and
Phi29 DNA polymerase. Nat Biotechnol 30:349–353, 2012.
70. Zhang M, Zhang Y, Scheuring CF, Wu C-C, Dong JJ, Zhang
H-B: Preparation of megabase-sized DNA from a variety of organ-
isms using the nuclei method for advanced genomics research. Nat
Protocols 7:467–478, 2012.
71. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D: Global
cancer statistics. CA Cancer J Clin 61:69–90, 2011.
72. Cheng KWA, Tanyi J, Mills GB: Genomic technologies in cancer
research, drug discovery and development. CME J Gynecol Oncol
7:16–24, 2002.
74. Bell DW: Our changing view of the genomic landscape of cancer.
J Pathol 220:231–243, 2010.
75. Greenman C, Stephens P, Smith R, et al.: Patterns of somatic muta-
tion in human cancer genomes. Nature 446:153–158, 2007.
76. Jones S, Zhang X, Parsons DW et al.: Core signaling pathways in
human pancreatic cancers revealed by global genomic analyses.
Science 321:1801–1806, 2008.
77. Parsons DW, Jones S, Zhang X, et al.: An integrated genomic analy-
sis of human glioblastoma multiforme. Science 321:1807–1812,
2008.
79. Taylor BS, Schultz N, Hieronymus H, et  al.:  Integrative genomic
profiling of human prostate cancer. Cancer Cell 18:11–122, 2010.
78. Kan Z, Jaiswal BS, Stinson J, et al.: Diverse somatic mutation pat-
terns and pathway alterations in human cancers. Nature 466:869–
873, 2010.
79. Gilbert MT, Haselkorn T, Bunce M, et al.: The isolation of nucleic
acids from fixed, paraffin-embedded tissues—which methods are
useful when? PLoS ONE 2:e537, 2007.
80. Schweiger MR, Kerick M, Timmermann B, et  al.:  Genome-wide
massively parallel sequencing of formaldehyde fixed-paraffin embed-
ded (FFPE) tumor tissues for copy number and mutational analysis.
PLoS ONE 4:e5548, 2009.
81. Kerick M, Isau M, Timmermann B, et al.: Targeted high throughput
sequencing in clinical settings: formaldehyde fixed-paraffin embed-
ded (FFPE) tumor tissues, input amount and tumor heterogeneity.
BMC Med Genomics 4:68–72, 2011.
B iobanking f or G e no m ic s -B a s e d T ran s l ationa l M e dicin e   •   2 4 1
82. Stunnenberg HG, Verneulen M:  Towards cracking the epigen-
etic code using a combination of high-throughput epigenomics
and quantitative mass spectrometry-based proteomics. Bioessays
33:547–551, 2011.
83. Chi P, Allis CD, Wang GG: Covalent histone modifications—mis-
written, misinterpreted and mis-erased in human cancers. Nat Rev
Cancer 10:457–469, 2010.
84. Li XY, Biggin MD:  Genome-wide in vivo cross-linking of
sequence-specific transcription factors. Methods Mol Biol. 809:
3–26, 2012.
85. Wang Z, Gerstein M, Snyder M: RNA-Seq: a revolutionary tool for
transcriptomics. Nat Rev Genet 10:57–63, 2009.
86. Ozolak F, Milos PM:  RNA sequencing:  advances, challenges and
opportunities. Nat Rev Genet 12:87–98, 2011.
87. Faghihi MA, Wahlstedt C:  Regulatory roles of natural antisense
transcripts. Nat Rev Mol Cell Biol 10:637–643, 2009.
88. Esteller M:  Non-coding RNAs in human disease. Nat Rev Genet
12:861–874, 2011.
89. Ryu S, Joshi N, McDonnell K, et  al.:  Discovery of novel human
breast cancer microRNAs from deep sequencing data by analysis of
pri-microRNA secondary structures. PLoS ONE 6: e16403, 2011.
90. Medina PP, Slack FJ:  MicroRNAs and cancer:  an overview. Cell
Cycle 7:2485–2492, 2008.
91. Hubel A, Aksan A, Skubitz APN, Wendt C, Zhong X:  State of
the art in preservation of fluid biospecimens. Biopreserv Biobank,
9:237–244, 2011.
92. Landi MT, Caporaso N: Sample collection, processing and storage.
IARC Sci Publ 142:223–236, 1997.
93. Elliot P, Peakman TC, UK Biobank:  The UK Biobank sample
handling and storage protocol for the collection, processing and
archiving of human blood and urine. Int J Epidemiol 37:234–244,
2008.
94. Vaught JB, Henderson MK, Compton CC:  Biospecimens and
biorepositories:  from afterthought to science. Cancer Epidemiol
Biomarkers Prev 21:253–5, 2012.
95. Moore HM, Compton CC, Alper J, Vaught JB:  International
approaches to advancing biospecimen science. Cancer Epidemiol
Biomarkers Prev 20:729–32, 2011.
 16.
GENETICS AND GENOMICS EDUCATION: THE PATH
FROM HELIX TO HEALTH
Reed E. Pyeritz
INTRODUCTION
It has become cliché to note the rapid pace of discovery
in the biomedical sciences. This is nowhere more evident
than in human and medical genetics, now expanded to
include genomics. The sheer volume of information that is
published in the literature each year continues to expand
rapidly. This trend is accelerating due to a surge in the num-
ber of journals that are open-access, only online, marginally
peer-reviewed, or a combination thereof. The purpose of
this chapter is to review some of the problems associated
with the explosion of information and knowledge (not the
same!) and measures that are being and could be taken to
address them. Finding a specific nugget of information, the
old needle-in-a-haystack problem in continually enlarging
haystacks, is facilitated by the worldwide web and by better
search engines. However, health professionals find them-
selves overwhelmed with information, and scant time to
turn information into knowledge, if only for the moment
they encounter a patient. The result is frequent failures of
appropriate diagnosis, management, and counseling.
The education of health professionals about genetics
and genomics is gradually improving, but whether the pace
is sufficient to keep up with knowledge and application is
open to question. Given all the enthusiasm for “personal-
ized medicine,” it is incumbent on the medical community
to become facile with genomics. However, the responsibil-
ity to become educated and to stay updated should not fall
to physicians and other health professionals alone. There
are many stakeholders, and the most important of these are
identified in Table 16.1.
Education can and should occur at many levels, begin-
ning in primary and secondary education. Sensitizing young
people to the excitement and potential of human genetics
and its application to their own lives and the lives of their
relatives serves two purposes. First, they will be better pre-
pared when later faced with situations pertinent to their
242
own health, especially prevention. Second, more young
people will be stimulated to pursue a career in the health
sciences that involve genetics and genomics. To accomplish
this requires resources, enlightened educational administra-
tors, and most important, well-prepared teachers. However,
few studies address the effectiveness of current and innova-
tive approaches to education.
In 2011, the Secretary’s Advisory Committee on
Genetics, Health and Society, which reported to the U.S.
Secretary of Health and Human Services, conducted a
detailed study of genetics education and training. The final
document addressed the needs of health professionals,
patients, and consumers and proposed further work in a
number of areas, including improved communication strat-
egies, diversity in the workforce, and use of family history
tools[1]‌.
E D U C AT I N G T H E G E N E R A L
PUBLIC
Dating at least from the mid–twentieth century, geneti-
cists recognized that the presentation of their discipline in
primary and secondary (K–12) education was deficient. In
large part, this was the fault of inadequate textbooks[2–4].
The teaching of biology, while highly prevalent—even
more than chemistry and physics, suffered from a focus on
facts rather than concepts and a lack of discussion of the
personal and societal relevance of topics. Often the launch
of Sputnik on October 4, 1957, is identified as the stimulus
for a heightened emphasis in the United States on educa-
tion in science and mathematics. The National Defense
Education Act of 1958 provided funding to develop mod-
ern textbooks. One response that same year was a grant from
the National Science Foundation (NSF) to establish the
Biological Sciences Curriculum Study (BSCS; www.bscs.
org). The initial BSCS efforts were controversial for their
Table 16.1  INDIVIDUALS WHO SHOULD BECOME
EDUCATED ABOUT MEDICAL GENETICS
GENERAL PUBLIC
Non-Health Professionals:
Educational administrators
Teachers
Legislators and their aides
Policymakers
Health Professionals:
Physicians generally, especially
  Primary care physicians
 Oncologists
 Cardiologists
 Neurologists
 Medical geneticists
Genetic counselors
Nurses and nurse practitioners
Physician assistants
focus on concepts and investigation (not “what to think,”
but “how to think”), but the fact that the program persists
to this day is a testimony to its many successes[5]‌. More than
three decades ago, a survey in Canada indicated that high
school students had a strong preference for learning more
about genetics, including genetic screening[2]. However,
deficiencies in the education of our youth are not limited to
biology and genetics.
Students in the United States continue to fall behind
their peers in other developed nations in the quality and
the quantity of STEM (science, technology, engineering
and mathematics) subjects from primary through second-
ary school. One result is that fewer American youth major
in these subjects when they do have an opportunity to
pursue higher education. Another result is that they are
less likely to be interested in or understand reports of new
scientific advances that appear in all of the various popular
media. A number of organizations are working to improve
STEM learning, especially the American Association for
the Advancement of Science (AAAS) and the National
Science Foundation (NSF)[6–8]. The federal Department
of Education also identifies improving STEM learning as
a priority. Too few states and localities, which are the true
arbiters of curricula, have prioritized STEM subjects.
With regard to genetics and genomics, a 2011 survey by
the American Society of Human Genetics (ASHG) found
that few states have standards for core concepts in basic
genetics[9]‌. Of the 19 core concepts assessed, 85% of states
were judged to have inadequate standards.
G enetic s and G eno m ic s E ducation :   T he Path f ro m H eli x to H ealth   •   2 4 3
The U.S. Department of Health and Human Services
identified public education in genetics as one of a number
of critical areas that need to be reviewed and addressed as
the healthcare system prepares for genomic medicine and
direct-to-consumer genetic testing. For example, when
patients were surveyed in 2007, most reported receiving no
genetic information from their health care provider about a
condition that was known to be familial[10]. In response, the
National Human Genome Research Institute (NHGRI)
established an Education and Community Involvement
Branch that hosts a website with a rich variety of informa-
tion for the public (http://www.genome.gov/Education)
and also provides a compendium of information from other
sources[11]. A recent initiative is GeneEd, designed for high
school students, teachers, and the general public[12].
Notable recent initiatives include “DNA Day” on April
25 (in honor of the anniversary of the publication of Watson
and Crick’s seminal paper on the double helix) sponsored
by the NHGRI and the ASHG[13]. One component is a
national competition among high school students to select
the most outstanding essays submitted on a topic that var-
ies each year[14]. Additionally, the ASHG sponsors a one-day
workshop for high school science teachers and students
immediately preceding their annual national meeting[15].
Colleges and universities may partner with local commu-
nity schools to enhance teaching and learning in STEM. For
example, Johns Hopkins, supported by an NSF grant, has
developed the STEM Achievement in Baltimore Elementary
Schools, which initially will benefit 1,600 students in grades
three through five in nine elementary schools[16]. A project
at Harvard, the Personal Genetics Education Project (pged.
org) has focused on high school students and teachers
through workshops and curriculum development[17].
Planned for 2013 is a major exhibition at the
Smithsonian Institution in Washington developed in col-
laboration with the NHGRI[18]. The project is funded
by the National Institutes of Health (NIH) Foundation
(non-governmental funds), the Smithsonian, and private
sources. After a stay of a year in Washington, the exhibit will
travel internationally.
Providers of genetic services, mainly commercial ones
thus far, have also developed educational materials for
the  general public. While typically slanted toward use
of  the specific genetic and genomic testing being offered,
the material is nonetheless typically sound and approach-
able. Some companies also provide mechanisms for poten-
tial customers to pose questions and to have them answered.
Thus far these companies reach only a small segment of
the population. For example, several groups have shown
that people who avail themselves of personal genetic and
genomic testing through direct-to-consumer portals, even
if the services are free, are typically highly educated whites
and Asians from upper socioeconomic strata[19].
There should be ways to capitalize on the growing inter-
est in tracing ancestry. While some people will be primar-
ily interested in genealogy, more need to be encouraged to
record their family histories for medical reasons. A variety
of web-based tools is now available, including some that
can be transmitted in electronic or hard-copy formats to
personal health care providers[20]. November, because the
American Thanksgiving is traditionally a time for fam-
ily gatherings, has been declared Family Health History
Month. The NHGRI[21], the U.S. Surgeon General[22], and
the Centers for Disease Control and Prevention all support
family history tools for the general public[23].
E D U C AT I N G N O N –H E A LT H
P R O F E SS I O N A L S
A number of goals might inform who receives focused atten-
tion in genetics and genomics education, and how they do.
Preparing those who educate others should generate little
controversy as a major goal, so teachers at all levels of primary,
secondary, and post-secondary schooling should be targeted.
Others who deserve special attention are politicians and
other involved in formulating federal, state, and local policies
dealing with the provision and financing of medical care and
public health. This is important, since well-informed politi-
cians and their staff members are more likely to maintain and
enhance support for basic, translational, and clinical research
in genetics and genomics. The AAAS Science and Technology
Policy Fellows has been a major effort since 1973[24]. Through
2012, 2,616 scientists and engineers have served as Fellows.
Administrators of federal, state, and private health insur-
ance who are sensitized to the relevance of genetics and
genomics in health care are more likely to establish and pri-
oritize reimbursement for medical genetic services. It will be
interesting to follow how provision of genetic services evolves
under the Affordable Care Act through health exchanges and
accountable care organizations, but improved knowledge
about genetics and genomics can only assist the cause.
E D U C AT I N G H E A LT H P R O F E SS I O N A L S
P H Y S I C I A NS
Medical School Curricula
The amount of information that has accumulated in the
past decade exceeds the sum total of information available
2 4 4   •   P rinciple s o f G eno m ic Medicine
in all preceding human history. This is especially perti-
nent in medical education, and genetics and genomics as
a combined discipline certainly is at the head of the pack
in this regard. In terms of education in medical school, the
demands on both curriculum planners and students are
daunting. New information must be relayed in the confines
of a fixed number of hours in the four-year course of study.
A  recent study found that the average medical student is
assigned so much material outside of classroom hours, that
28 to 41 hours per week would be required to complete the
readings only once[25]. If there is any good news, it is that
far less material needs to be memorized since the digital age
renders details readily, even instantly, available. However,
having information, or access to it, is not the same as having
knowledge.
Given that human and medical genetics are relative
newcomers to the traditional curriculum, the efforts to
expand course hours have met with mixed success. And
even when time is available, inefficiencies arise. Some
members of genetics faculties still delight in teaching how
technologies work. Twenty years ago, many of us taught
the intricacies of restriction enzymes and Southern blots,
only to see them largely disappear from use. This experi-
ence should inspire caution about what is taught about
next-generation sequencing and cytogenomic arrays today.
The vast majority of health professionals need to under-
stand when to order a test and what to do with the results,
not what is done in the laboratory or the steps taken to pro-
duce the clinical report. Often the fundamentals of prob-
ability and statistics are more relevant than the methods
used to produce a result. Genome-wide association studies
(GWAS) are a good example. Understanding risk (relative
and absolute), clinical validity and clinical utility, and eth-
nicity are fundamental to interpreting any GWAS result,
yet physicians are notably deficient in these skills. As Feero
and Green emphasized, “ensuring that high-quality soft-
ware tools are available to clinicians will be more impor-
tant than forcing them to understand the intricacies of how
those tools work”[26].
Medical schools regularly revise their curricula sub-
stantially. There is no consistent pattern or rationale to
these revisions. Given the century-old conflicts produced,
it might seem ironic that serious discussions are occurring
about reducing the length of undergraduate medical educa-
tion to three years[27]. From the perspective of this chapter,
one example bears mentioning. Beginning with a strategic
planning initiative in 2002, the Johns Hopkins School
of Medicine posed the question, “How will medicine be
practiced in 10 years?”[28] Once the decision was made that
comprehensive curricular reform was necessary, the process
was informed by the six-step approach recently defined[29].
The underlying theme was “genetic, environmental, and
societal influences are subject to variation. These variations
lead to the enormous heterogeneity of health phenotypes.”
The choice of genetics as a focus was heavily influenced by
the writings of one of the founders of the field of medi-
cal genetics and a senior pediatrician at Hopkins, Barton
Childs[30]. The new Hopkins curriculum, entitled “Genes
to Society,” was introduced in the fall of 2009. What is
different? First, there is a focus on individuality, based on
examples. Second, evolution plays a central role[31]. Third,
there is a heightened appreciation of the complexity and
importance of the relationships among health, risk, pre-
vention, disease, and therapy. There is analogy with the “P4
medicine” concept of Leroy Hood and colleagues:  pre-
dictive, preventive, personalized, and participatory[32].
A structural nuance is the intersession in between clinical
clerkships, which is led by basic scientists and clinicians
and updates and integrates basic science into the expand-
ing clinical experiences of the students. Even before this
new curriculum can be evaluated, it is being exported to
Malaysia, where Hopkins is building a new medical school
from scratch in Kuala Lumpur, and the curriculum will be
based on Genes to Society.
Licensure of physicians in the United States is con-
trolled by the medical board in each state, and each such
board requires that the candidate pass a three-part U.S.
Medical Licensing Examination (USMLE®) administered
jointly by the Federation of State Medical Boards (www.
fsmb.org) and the National Board of Medical Examiners
(NBME; www.nbme.org). Since the first two parts of the
USMLE are taken by medical students, the content of
the examinations and students’ performance influence in
important ways the curriculum of each medical (and osteo-
pathic) school. The Association of Professors of Human
and Medical Genetics (APHMG; www.aphmg.org) is the
professional organization representing the genetics leader-
ship of medical and graduate schools in the United States
and Canada. From the mid-1990s, every few years for more
than a decade, the APHMG collaborated with the NBME
to assess the genetics content assessed on the USMLE.
Over this period of time, both the quantity and quality
of material covered on the examination improved. While
further assessment and improvement are clearly war-
ranted, the overall infusion of genetics into medical school
curricula has improved markedly from the 1970s, when
Barton Childs observed that the likelihood of a medical
school’s having a course, division, or department of genet-
ics was inversely proportional to its having a department of
­community or family medicine[33].
G enetic s and G eno m ic s E ducation :   T he Path f ro m H eli x to H ealth   •   2 4 5
A number of schools in the United States (Tufts,
Stanford, Vanderbilt, Penn) have utilized direct-to-
consumer genotyping of willing students to introduce top-
ics such as risk (absolute and relative), genetic variation,
genetic factors in common disease, and ethical, legal, and
economic issues[34,35].
Non-Genetics Residency Training
After medical school, the curricula of residency train-
ing programs in primary specialties are largely specified
by the various organizations that accredit training pro-
grams (residency review committees; RRC) and certify
specialists (e.g., the American Board of Pediatrics). Here
the requirements for formal or practical introduction to
genetics and genomics are decidedly mixed. Pediatrics,
family medicine, and obstetrics-gynecology have long
been at the forefront, while internal medicine, among
many, has lagged.
Graduate medical education (GME), the training of
physicians between medical school and independent prac-
tice, has been criticized in the United States for not ade-
quately preparing physicians for their future practices and
for not being sufficiently responsive to the needs of soci-
ety. A recent report of the Josiah Macy Foundation report
entitled “Ensuring an Effective Physician Workforce for the
United States: Recommendations for Reforming Graduate
Medical Education to Meet the Needs of the Public” is sure
to exert wide influence[36]:
Advances in medical diagnostics, therapeutics, and
information technology can significantly improve
health outcomes. However, we have fallen short in
consistently using technology optimally to improve
the quality and efficiency of health care. We need to
train the next generation of physicians to optimally
use medical and information technology, to follow
the principles of quality improvement and patient
safety, and to practice medicine based on the best
evidence.
One major recommendation states, “The content of
training should expand to include topics essential for cur-
rent and future practice, particularly those related to pro-
fessionalism, population medicine, and working effectively
in the health care system.” However, nowhere in the entire
document do the words “genetics” or “genomics” appear. To
give them the benefit of the doubt, perhaps the absence of
proven clinical utility for many emerging genetic technolo-
gies led to the Foundation’s failure to mention genetics.
 The Physician Workforce genetics and in a number of subspecialties (cytogenetics,
biochemical genetics, clinical molecular genetics).
Education of physicians after residency, known as post-
There are multiple paths for education in clinical genet-
graduate medical education or continuing medical edu-
ics. A  handful of programs accept students right out of
cation (CME), lasts the entire professional lifespan. The
medical school and provide four years of training in pediat-
practitioner needs to be updated about familiar topics
rics, internal medicine, and medical genetics. At the conclu-
and introduced to new concepts and applications. Due to
sion, the resident is eligible to take the ABMG examination
licensing and maintenance of certification requirements,
in clinical genetics. The majority of programs, some 51 in
new technologies, and decreased funding by industry,
number, accept residents after they have completed at least
approaches to CME are in rapid evolution[37]. A  number
two years of residency in another primary specialty. Two
of studies have documented the obvious: that the practic-
additional years are required for board-eligibility in clinical
ing physicians who know the most about genetics are those
genetics, although many programs encourage a third year
who most recently graduated from medical school. Thus,
of research. Alternatively, a medical student can apply to a
the challenge of CME is to bring older physicians up to
combined residency that involves five years of training in
speed while at the same time keeping all informed about
clinical genetics and either pediatrics (16 accredited pro-
new developments[38]. This is an issue throughout the
grams) or internal medicine (6 programs). A separate com-
world, and has been documented in the United States and
bined track exists for obstetricians training in maternal-fetal
Europe (reviewed in reference 39).
medicine and medical genetics (6 programs).
In the United States, family physicians have taken the
The Residency Review Committee for Medical
lead in assessing the needs and providing material through
Genetics, which accredits training programs, currently
a program called Genetics in Primary Care[40]. More
has approved programs that provide about 90 slots for
recently, the American Academy of Pediatrics established
first-year residents. On average, 50–60% of these slots fill,
the Genetics in Primary Care Institute, with the support of
so there is considerable unused training capacity for phy-
the federal Health Resources and Service Administration
sician clinical geneticists[51]. There was a broad consensus
Maternal and Child Health Bureau[41]. One recurring
that too few medical geneticists were being trained, even
theme is that physicians still believe that genetics is periph-
before the emergence of genomic medicine, but few stud-
eral to their everyday practices. Hence, education needs to
ies have examined workforce issues[52]. Currently 1,419
be placed in a clinical context[39,42–44]. Some specialty societ-
individuals in the United States are certified as clinical
ies have established policy statements about specific genetic
geneticists. Since most work in academic medical centers,
issues, such as carrier screening[45] and genetic testing for
most are engaged in teaching, research, and administra-
cancer susceptibility[46].
tive activities, so the number of full-time clinical equiva-
Some practicing physicians are aware of clinical testing
lents available is unknown. In an effort to increase the
based on genetics and genomics, either through their per-
medical genetics workforce, during its annual meeting
sonal use or results that their patients bring to them[47,48].
the American College of Medical Genetics and Genomics
Many primary care physicians recognize that to stay cur-
(ACMG) sponsors a career day for undergraduate, gradu-
rent in their abilities they need to learn more genetics[49].
ate, and medical students[53].
Personal genomic testing, such as is available to consumers
In keeping with the requirements for all specialties, the
through a number of companies, has been used as a teaching
APHMG and the ACMG have developed competencies
tool at one major academic health center[50].
for medical geneticists[54], and these have been incorporated
into the programs at each accredited medical genetics resi-
dency. More recently, a task force of medical geneticists has
E D U C AT I N G M E D I C A L G E N ET I C I S TS
been working with the ACGME to develop “milestones” to
The American Board of Medical Specialties (www.abms. assess a resident’s progress through training. Additionally,
org/) oversees the certification of physicians in 24 primary the ABMG established maintenance-of-certification
specialties and 121 subspecialties. Medical genetics is a program more than a decade ago. All clinical geneticists
primary specialty, and the last officially so designated, in with time-limited certificates must maintain certification
1991. The certifying body for the specialty is the American through self-study and examination every ten years. The
Board of Medical Genetics (www.abmg.org), which judges professional society for medical geneticists, the ACMG,
whether an applicant is eligible to sit for the board exami- provides continuing education at its annual meeting and at
nation and also administers the examination in clinical a biennial board review course.

2 4 6   •   P rinciple s o f G eno m ic Medicine

 H E A LT H P RO FE S S I O NA L S OT H E R
T H A N P H Y S I C I A NS
Non–Genetics Professionals
Included in this broad group are dentists, physician assis-
tants, nurse practitioners, advanced-practice nurses, mid-
wives, physical and occupational therapists, and so on.
There is little debate now about the relevance of genetics
and genomics in the curricula that educate these individu-
als[55], and many groups have recommended competency
objectives[56–59].
The National Coalition for Health Professional
Education in Genetics (NCHPEG; www.nchpeg.org)
describes itself as an “organization of organizations,” the lat-
ter now numbering over 80. The AMA, American Nurses
Association, and the NHGRI established NCHPEG in
1996 with several goals relevant to genetics education:
Integrate genetics content into the knowledge base
of health professionals and students of the health
professions;
Develop educational tools and information resources to
facilitate the integration of genetics into health profes-
sional practice; and,
Strengthen and expand the Coalition’s interdisciplinary
community of organizations and individuals committed
to coordinated national genetics education for health
professionals.
NCHPEG is now partnered with the Jackson Laboratory.
Non-Physician Medical Geneticists
In addition to the Ph.D.-level medical genetics laboratory
specialist, another vital component of the workforce is the
genetic counselor, and to a lesser extent numerically, the
genetics nurse. Genetic counselors train in two-year mas-
ter’s-level graduate programs accredited by the American
Board of Genetic Counseling (ABGC; www.abgc.net).
There are 31 such programs in the United States and three
in Canada. The ABGC also provides certification for coun-
selors through an examination. A growing number of states
license genetic counselors. Their professional organization,
the National Society of Genetic Counselors (NSGC), con-
ducts an annual meeting and arranges credit for continuing
education.
There are currently over 2,400 certified genetic coun-
selors in the United States. Graduate training programs are
routinely filled. There is a broad consensus that the number
G enetic s and G eno m ic s E ducation :   T he Path f ro m H eli x to H ealth   •   2 4 7
of genetic counselors is markedly inadequate to serve the
needs of genetic—let  alone genomic—medicine. One
important impediment is the ability of counselors to gener-
ate clinical revenue independent of a physician.
Nurses have also considered what the baseline com-
petencies in genetics and genomics should be in their
curricula[60].
N UA N C E S C O N C E R N I N G E D U C AT I O N
A B O U T G E N ET I C S A N D G E N O M I C S
Having reviewed in a somewhat hierarchical format those in
need of enhanced education, let us now examine the nuances
of the genomic era that affect education (Table 16.2).
The explosion in information has already been noted
above. Utilizing this information clinically is one of the
most active areas of translational research. The technologies
available for studying the genome continue to evolve. We
must hold genetic technologies to the same rigorous stan-
dards applied to all medical innovations[61]. Moreover, the
focus cannot be just on the nucleotide sequence, even if dele-
tions, duplications, and copy number variants are included.
Epigenetic modifications are increasingly recognized as
important trans-generational effects that are strongly influ-
enced by the environment. As many have emphasized,
genetic (and epigenetic) variations do not operated in a
vacuum. A systems approach to understanding—and edu-
cation—is required[30,32,62]. Academic departments and pro-
grams in systems biology and its cousin disciplines are still
relatively novel, and as a result, this approach to integrating
knowledge and teaching it has not permeated curricula at
many institutions.
Table 16.2 FACTORS THAT IMPACT EDUCATION IN
GENETICS AND GENOMICS
Explosion of information
Rapid evolution of technologies
(family history remains an underuti-
lized, inexpensive test)
Direct-to-consumer genetic testing
Economics
Legal issues:
Case law
Statutory law
Suits and settlements: NCAA
and HbS testing
Clinical studies or the lack thereof
(meta-analyses, decision analysis,
“crowd sourcing”)
 Feero and Green identified a number of impediments
to physicans’ learning new information and applying it to
their practices[26]:
MDs are practical; most genomic advances have been rel-
evant to only a few clinical situations;
New ideas, even well-tested ones, get neglected unless they
are relatively easy to understand;
Few recommendations for the use of genomic information
have evidence-based support (either because they apply
to rare diseases, or because research about their appli-
cation to common disease is lacking), so few practice
guidelines adopt them;
There are lots of other educational reforms and demands.
Genomics needs to be integrated, but this is difficult
without wholesale alteration of curricula (see above);
and
All of this has been done in the context of the “push
model”; what we need is a “pull model.”
This will require support for research that documents
the clinical utility of genetic technologies. Evidence for
cost-effectiveness/benefit would also help stimulate inter-
est in education[63]. The Centers for Disease Control and
Prevention (CDC) Office of Public Health Genomics
launched the Evaluation of Genomic Applications in
Practice and Prevention (EGAPP) program in 2004[64].
Most of the analyses performed thus far found a lack of
firm evidentiary basis for clinical applications of genetic
tests that assess risk of common diseases or affect drug
choice or dosing. As important as this activity could be for
health professionals, insurers, and policymakers, EGAPP
is under-resourced and potentially overwhelmed. Given
that EGAPP and others have identified a lack of rigor-
ous research on clinical utility of genomics, and that the
NHGRI strategic plan clearly prioritizes establishing major
clinical applications in the next decade, it is ironic that little
funding exists for the requisite research. Ultimately research
on the clinical utility of genomics, coupled with due atten-
tion to ethical, economic, legal, and social implications, will
determine both the approach to, and the success of, educa-
tional efforts for all stakeholders.
C O DA
This brief review of education in genetics and genom-
ics has, in the interests of conserving space, been highly
selective and focused primarily on efforts in the United
2 4 8   •   P rinciple s o f G eno m ic Medicine
States. I apologize to those whose work I have not cited or
important projects occurring in other parts of the world.
REFERENCES
1. National Institutes of Health Office of Science Policy. http://
oba.od.nih.gov/oba/SACGHS/reports/SACGHS_education_
report_2011.pdf. Accessed 14 April 2014.
2. Scriver CR, et  al. The education of citizens:  human genetics. Am
Biol Teacher 40:280–284, 1978.
3. Fitzsimmons JS. The teaching of human genetics in schools. J Med
Genet 20:244–248, 1983.
4. McInerney JD. DNA in medicine:  school-based education. Am
J Hum Genet 42:635–636, 1988.
5. McInerney JD. The Human Genome Project relevant to genetics
education in high school. Am J Hum Genet 52:235–238, 1993.
6. Alberts B. Trivializing science education. Science 335:263, 2012.
7. Alberts B. Teaching real science. Science 335:380, 2012.
8. Fedoroff NV. The global knowledge society. Science 335:503, 2012.
9. Dougherty M, et  al. A comprehensive analysis of high school
genetics standards:  are states keeping pace with modern genetics?
CBE-Life Sciences 10:318–327, 2011.
10. Harvey EK, et al. Providers’ knowledge of genetics: a survey of 5915
individuals and families with genetic conditions. Genet Med 9:
259–267, 2007.
11. NHGRI Online Genetics Education Resources. http://www.
genome.gov/10000464. Accessed 14 April 2014.
12. GeneEd. http://geneed.nlm.nih.gov/. Accessed 14 April 2014.
13. NHGRI DNA Day. http://www.genome.gov/10506367. Accessed
14 April 2014.
14. ASHG DNA Day. http://www.ashg.org/education/dnaday_­
winners_2012.shtml. Accessed 14 April 2014.
15. ASHG High School Workshop 2012. http://www.ashg.org/
2012meeting/pages/workshops.shtml#2. Accessed 14 April 2014.
16. Johns Hopkins receives $7.4  million grant to boost STEM edu-
cation in Baltimore city. http://releases.jhu.edu/2012/09/25/
johns-hopkins-receives-7-4-million-grant-to-boost-stem-education-
in-baltimore-city/. Accessed 14 April 2014.
17. Kung JT, Gelbart ME. Getting a head start: the importance of per-
sonal genetics education in high schools. Yale J Biol Med 85:87–92,
2012.
18. NHGRI: Genome: unlocking life’s code. http://www.genome.gov/
Smithsonian/. Accessed 14 April 2014.
19. Gollust S, et  al. Motivations and perceptions of early adopters of
personalized genomics:  perspectives from research participants.
Publ Health Genomics 15:22–23, 2012.
20. Pyeritz RE. The family history: the first genetic test, and still useful
after all those years? Genet Med 14:3–9, 2012.
21. NHGRI:  The U.S. Surgeon General’s family history initiative.
http://www.genome.gov/17516481. Accessed 14 April 2014.
22. U.S. Department of Health & Human Services: Surgeon General’s
family health history initiative. http://www.hhs.gov/familyhis-
tory/. Accessed 14 April 2014.
23. CDC Public Health Genomics. http://cdc.gov/genomics. Accessed
14 April 2014.
24. American Association for the Advancement of Science: Fellowships.
http://fellowships.aaas.org/. Accessed 14 April 2014.
25. Klatt EC, Klatt CA: How much is too much reading for medical
students? Assigned reading and reading rates at one medical school.
Acad Med 86:1079–1083, 2011.
26. Feero WG, Green RD. Genomics education for health care profes-
sionals in the 21st century. JAMA 306:989–990, 2011.
27. Emanuel EJ, Fuchs VR. Shortening medical training by 30%. JAMA
307:1143–1144, 2012.
 28. Wiener CM, et  al. “Genes to Society”—the logic and process of
the new curriculum for the Johns Hopkins University School of
Medicine. Acad Med 85:498–506, 2010.
29. Kern D, et  al., eds. Curriculum Development for Medical
Education:  A  Six-step Approach. Baltimore, MD:  Johns Hopkins
University Press; 1998.
30. Childs B. Genetic Medicine:  A  Logic of Disease. Baltimore,
MD: Johns Hopkins University Press; 1999.
31. Nesse RM, et  al. Making evolutionary biology a basic science for
medicine. Proc Natl Acad Sci U S A 107:1800–1807, 2010.
32. Hood L, Fores M. A personal view on systems medicine and the
emergence of proactive P4 medicine: predictive, preventive, person-
alized and participatory. New Biotech 29:613–624, 2012.
33. Childs B. Garrod, Galton, and clinical medicine. Yale J Biol Med
46:297–313, 1973.
34. Walt DR, et  al. Lessons learned from the introduction of person-
alized genotyping into a medical school curriculum. Genet Med
13:63–66, 2011.
35. Salari K, et  al. Commentary:  to genotype or not to genotype?
Addressing the debate through the development of a genomics and
personalized medicine curriculum. Acad Med 86:925–927, 2011.
36. Josiah Macy Jr. Foundation:  Conference Summary. http://josia-
hmacyfoundation.org/docs/macy_pubs/Macy_GME_Report,_
Aug_2011.pdf. Accessed 14 April 2014.
37. Wentz DK. Continuing Medical Education: Looking Back, Planning
Ahead. Lebanon, NH: Dartmouth College Press; 2011.
38. Childs B, et al. Human genetics teaching in U.S. medical schools.
Am J Hum Genet 33:1–10, 1981.
39. Guttmacher AE, et  al. Educating health-care professionals about
genetics and genomics. Nat Rev Genet 8:151–157, 2007.
40. Laberge AM et al: Long-term outcomes of the “Genetics in Primary
Care” faculty development initiative. Fam Med 41:266–270, 2009.
41. Genetics in Primary Care. http://www.geneticsinprimarycare.org.
Accessed 14 April 2014.
42. Andermann A, Blancquaert I. Genetic screening: a primer for pri-
mary care. Can Fam Phys 56:333–339, 2010.
43. Galinkin JL, et  al. Genetics for the pediatric anesthesiologist:  a
primer on congenital malformations, pharmacogenetics, and pro-
teomics. Anesth Analg 111:1264–1274, 2010.
44. Houwink EJF, et al. Prioritization of future genetics education for
general practitioners: a Delphi study. Genet Med 14:323–329, 2012.
45. American College of Obstetricians and Gynecologists Committee
on Genetics:  ACOG Committee Opinion No. 486:  update on
G enetic s and G eno m ic s E ducation :   T he Path f ro m H eli x to H ealth   •   2 4 9
carrier screening for cystic fibrosis. Obstet Gynecol 117(4):1028–
1031, 2011.
46. Robson ME, et al. American Society of Clinical Oncology policy
statement update: genetic and genomic testing for cancer suscepti-
bility. J Clin Oncol 28:893–901, 2010.
47. Bernhardt BA, et  al. Incorporating direct-to-consumer genomic
information into patient care: attitudes and experiences of primary
care physicians. Personal Med 9:683–692, 2012.
48. Pyeritz RE. The coming explosion in genetic testing: is there a duty
to recontact? N Engl J Med 365:1367–1369, 2011.
49. Scheuner MT, et  al. Delivery of genomic medicine for common
chronic adult diseases: a systematic review. JAMA 299:1320–1324,
2008.
50. Sharp RR, et al. Addressing gaps in physician education using per-
sonal genomic testing. Genet Med 13:750–751, 2011.
51. Brotherton SE, Etzel SI. Graduate medical education, 2011–2012.
JAMA 308:2264–2279, 2012.
52. Cooksey JA, et  al. The medical genetics workforce:  an analysis of
clinical geneticist subgroups. Genet Med 8:603–614, 2006.
53. American College of Medical Genetics http://www.acmg.net.
Accessed 14 April 2014.
54. Korf BR, et al: Competencies for the physician medical geneticist in
the 21st century. Genet Med 13(11):911–912, 2011.
55. Gaff CL, et  al. Genetics in health practice and education:  special
issue. J Genet Counsel 17:143–144, 2008.
56. McInerney JD. Genetics education for health professionals: a con-
text. J Genet Counsel 17:145–151, 2008.
57. Skirton H, et al. Genetic competence of midwives in the UK and
Japan. Nursing Health Sci 12:292–303, 2010.
58. Jenkins J, Calzone KA. Establishing the essential nursing competen-
cies or genetics and genomics. J Nurs Scholarsh 39:10–16, 2007.
59. Belt CM, Weinar MZ. Genetic educational, health policy, research,
and networking resources. Sem Oncol Nursing 27:72–81, 2011.
60. Jenkins J. Essential genetic and genomic nursing competencies for
the oncology nurse. Sem Oncol Nursing 27:64–71, 2011.
61. Burke W, Evans JP. Teaching with single nucleotide polymor-
phisms: learning the right lessons. Genet Med 13:17–18, 2011.
62. Nadeau JH, Dudley AM. Systems genetics. Science 331:1015–1016,
2011.
63. Ioannidis JPA. Personalized genetic prediction:  too limited, too
expensive, too soon? Ann Intern Med 150(2):139–141, 2009.
64. Evaluation of Genomic Applications in Practice and Prevention.
http://www.egappreviews.org. Accessed 14 April 2014.
 17.
ETHICAL, LEGAL, AND SOCIAL ISSUES
IN CLINICAL GENOMICS
Caroline F. Wright, Anna Middleton, and Michael Parker
INTRODUCTION
The astonishing development of massively parallel,
high-throughput DNA sequencing technologies over the
last decade means that sequencing multiple genes or even
whole genomes is now becoming a clinical reality with
enormous diagnostic potential.1,2 This has far-reaching
consequences for the practice of clinical genetics as well as
mainstream medicine and public health.
Sequencing a genome should not only be regarded as a
clinical test, but also as an assay that creates a data resource
that has the potential to be repeatedly interrogated with
specific analytical questions. Under a model wherein indi-
vidual genome sequences are stored and linked to personal
medical records, each new analysis is essentially free of cost.
The clinician will no longer need to decide what labora-
tory test to order based on a set of clinical phenotypes, but
which bioinformatics analyses to perform and when. The
challenge will therefore become one of data interpretation
rather than data acquisition. Ultimately, both the scope
and breadth of testing are likely to expand, from the niche
specialty of clinical genetics focused primarily on targeted
diagnostic testing of families with inherited disorders and
birth defects, to genome sequencing of individuals through-
out mainstream medicine to allow increasingly stratified
diagnosis and treatment.
Does the shift from genetics to genomics raise any new
ethical, legal, or social issues? Although at first sight there
might appear to be nothing new beyond the scale and flex-
ibility of genomic testing, the creation of unprecedented
amounts of personal, identifiable data with a multiplic-
ity of medical (and other) applications has novel ethical
implications,3 particularly for responsible data stewardship.
Genome sequencing is not only likely to be the first medical
test that could potentially offer everyone a positive result of
some clinical value, but is also likely to be one where the vast
majority of results will be of little or no value whatsoever.
250
This change in scale therefore creates enormous challenges
in itself, from accurately interpreting variants in individu-
als, families, and populations, to protecting individual pri-
vacy and managing public expectations, to the delineation
of the responsibilities and duties of care of clinicians and
researchers.
In the first section of this chapter, we review the ethical
values and norms at the heart of traditional clinical genet-
ics (often termed “genethics”)4; in the second section, we
outline the key ethical, legal, and social challenges in an era
of whole-genome sequencing (which we term “genometh-
ics”). Finally, we discuss the implications for the boundary
between clinical practice and research.
G E N ET H I C S
Clinical genetics has traditionally focused on diagnostic
and predictive testing for rare, highly penetrant germline
genetic variants. These variants are either inherited and
are uncovered through family history, or occur spontane-
ously (de novo) and are generally diagnosed in childhood,
in relation to reproduction or linked to the inheritance of
adult-onset cancer. Therefore, unlike most other areas of
medicine, clinical management is generally centered around
the family rather than individual patients. Of the thousands
of rare disease-causing variants known, many have cata-
strophic biological and phenotypical effects, and determin-
ing the presence (or absence) of a particular genetic variant
in an individual is highly predictive of current and future
disease both in that individual and their relatives.
Many of the ethical principles and guidelines that have
evolved in the practice of clinical genetics stem directly
from these properties of rare Mendelian diseases—that
variants are extremely predictive, and they may have pro-
found implications that reach beyond the individual being
tested. Similarly, many of society’s concerns about genetics
 Free ebooks ==> www.Ebook777.com
can be traced to the same origin. The perceived power and
inescapably deterministic feature of Mendelian genetics
has led to a fear of stigmatization and unfair discrimina-
tion, which in turn has led to the introduction of genetic
non-discrimination legislation and insurance moratoria in
many countries. The treatment of genetic information in
this way, as needing special protection above and beyond
other biomedical data—a practice known as “genetic excep-
tionalism”—also derives to some extent from the wide-
spread misunderstanding that genetic tests deliver certainty.
Although genetic exceptionalism has been widely criticised,5
and is based on the false belief that most genetic informa-
tion is deterministic, clinicians must nonetheless address
and respond to these preconceptions and worries when
working with patients.
The emotion attached to a diagnosis of many Mendelian
diseases may be very significant for both the individual and
family. The discipline of genetic counselling has developed
from the (patient-led) necessity for psychosocial and infor-
mational support to help individuals and families cope with
the impact of a genetic condition. Genetic counselling for
rare, highly penetrant, serious—often life-threatening—
conditions is available from specially trained clinical
geneticists and genetic counsellors. These professionals use
established, evidence-based communication models that
offer time and space to individuals and families to consider
the emotional and psychological implications of being
tested for a family condition. Many recognized ethical,
legal, and social issues have emerged from genetic counsel-
ling practice over the last 50 years, and any discussion about
genethics must involve genetic counselling practice.
C O NS E N T A N D AU TO N O MY
A key ethical commitment in clinical genetics is a respect
for individual autonomy, which manifests itself in a widely
agreed recognition of the importance of providing genetic
counselling and ensuring informed consent prior to
undertaking testing. This often involves gathering consent
for testing from potentially affected relatives, particularly
where the individual referred for testing is not himself
affected. The obtaining of valid consent (or refusal) is,
however, not always a straightforward matter. Individuals
may struggle to fully comprehend the future implica-
tions of a test result, and obtaining informed consent
from family members can sometimes be extremely chal-
lenging; for example, due to difficulty in knowing how to
communicate genetics to relatives, possible differences in
opinion about testing, or simply problems in even making
contact due to family breakdown. Moral dilemmas faced
Ethic a l , Leg a l , a nd S oci a l I ssues in C l inic a l G eno m ics   •   2 5 1
www.Ebook777.com
by genetic health professionals occur when individual
autonomy conflicts with familial solidarity. Should an
individual be able to consent alone to a test that will reveal
information about a family member who does not want to
know their result?
Maintaining patient confidentiality and an individu-
al’s right to privacy is important in clinical genetics and, as
such, genetic diagnoses are generally treated no differently
from other potentially sensitive personal medical informa-
tion. However, unlike the case with most other medical
data, respecting individual privacy and or choice can be
problematic in the context of “at risk” families in which
it is possible that individual family members will have dif-
ferent values and conflicting opinions. Does an individual
have a right not to know their own genetic makeup,6 or
to withhold access to it when a family member is in need
of the same information7? In such cases, the value of pri-
vacy needs to be balanced against the rights and freedoms
of others, and in certain circumstances it may be justified
to break confidence in order to avoid serious harm to a
relative.8
Particular difficulties arise around testing those who
cannot give consent (minors and those lacking capacity).
Testing for preexisting conditions or those where early pre-
ventative actions may be taken is usually advised, and offered
with consent from a parent or legal guardian. However,
most professional guidelines in Europe9 recommend that
minors should not be tested for adult-onset conditions
for which no immediate preventative action exists, which
reflects a general consensus that this would infringe on the
autonomy of the future adult to make their own decision
about genetic testing.10
R E P RO D U C T I VE AU TO N O MY
A N D ITS A P P RO P R I AT E L I M ITS
The use of preconception, pre-implantation, and prenatal
genetic testing to facilitate reproductive autonomy is a criti-
cal part of modern clinical genetics, and for many couples
who choose this option, it has substantially reduced the
burden of serious inherited diseases.11 Individual autonomy,
non-directive counselling, and patient empowerment12 are
central to supporting decisions that may include opting for
assisted reproduction, destruction of unwanted embryos,
or termination of affected pregnancies. Respecting the
individual’s reproductive autonomy also means supporting
and providing appropriate care to women and couples who
choose not to opt for these routes—for example, women
who, on discovering that their pregnancy is “at risk” or
“affected,” opt to continue the pregnancy to term.
 Ongoing political debate surrounding embryo research
and the ethics of abortion, amplified by the unpleasant
historical specter of eugenics, means that developments in
this area continue to be somewhat contentious. The scope
of individual reproductive choice remains unclear, and the
majority of genetic tests available remain within the con-
fines of highly penetrant clinical diseases. Controversies
have arisen relating to what constitutes a “disease” and to
what extent the autonomous choices of parents—to choose
to have a deaf child using preimplantation genetic diagnosis,
for example13—should be respected. Sex selection for social
and cultural reasons or family balancing is generally viewed
as unacceptable in most countries and is only permitted to
prevent X-linked diseases.
I N C I D E N TA L FI N D I N G S
The issue of obtaining informed consent for genetic testing
is further complicated by the potential for uncovering inci-
dental findings (IFs)—unexpected results that do not relate
to the original clinical inquiry but that may nonetheless
have equivalent or greater clinical or personal significance.
This is a familiar problem within the medical imaging com-
munity, where scans may often reveal unexpected findings of
unknown significance, many of which turn out—after fur-
ther investigation—to be benign.14 Genetic examples range
from discovering an adult-onset cancer-predisposition gene
in a child being investigated for developmental disorders,15
or uncovering misattributed paternity (or maternity)16 in
the course of a routine test. Although the use of targeted
molecular testing for specific variants largely mitigates this
problem for many conditions, use of genome-wide tech-
nologies such as karyotyping and DNA microarrays have
made IFs a more frequent clinical occurrence. To date, there
is very little consensus on how to handle these findings, and
practice tends to vary between services and clinicians, often
based on perceived clinical utility.
G E N O M ET H I C S
The move from genetics to genomics will bring about a
profound change in the practice of clinical genetics, pri-
marily due to the dramatic fall in costs and the impending
data tsunami. Since every individual has around 3 to 4 mil-
lion genomic variants (versus the reference sequence),17
data management and interpretation will be an enormous
challenge. High-resolution DNA microarrays have already
given laboratories and clinicians a glimpse of the problem: a
plethora of genetic variants present in every individual,
2 5 2   •   P rincip l es o f G eno m ic Medicine
most of which have unknown clinical significance and are
unrelated to the reasons for which these tests were ordered.
In twentieth-century genetics, the majority of variants
seen clinically were rare and assumed to be pathogenic;
however, twenty-first–century genomics has shown
that non-pathogenic genetic variation is far more com-
mon. Knowledge of normal population genetic varia-
tion is therefore crucial to interpreting genomic data.
Whole-genome sequencing has the potential to reveal,
not only rare highly predictive variants with heritable
consequences, but also novel and common variants with
unknown or no clinical or phenotypical consequence.
Some variants will be risk factors for common complex
diseases; others may play a role in drug metabolism and
toxicity; a number will relate to behavioral phenotypes;
but many will have no discernable effect. In an era of
multi-gene panel testing and clinical whole-genome
sequencing, most variants are likely to be assumed to be
benign until proven otherwise.
Whilst genetic counselling and the ethical prac-
tices developed in the rare-disease-genetics field offer a
solid foundation upon which to build, their relevance
is weakened when we consider whole-genome testing.
The paradigm of genetics as deterministic and familial is
unconvincing in the context of common variation, minor
genetic risk factors, and somatic mutations. In reality, most
human traits and diseases are complex and multifactorial,
most variants have variable penetrance due to environmen-
tal interactions or other genetic modifiers, and the major-
ity of germline genetic variation has little or no predictive
power for individuals or their families. All genomes con-
tain some loss of function variants and recessive alleles,18
and a whole-genome analysis could yield reams of infor-
mation pertaining to a multitude of traits, providing risk
figures that are either small, weakly predictive, or uninter-
pretable. One might expect that such benign information
will have minimal emotional or psychosocial value for the
individual or their family. This contrasts enormously with
a single test for a highly predictive, serious, life-threatening
condition. Therefore, although the ethical, legal, and social
issues around the rare-disease component of genomics are
well established, the framework required for genomethics
necessarily has a broader scope, due to the unprecedented
scale and range of genomic data as well as the seemingly less
evocative nature of it.
R E S P O N S I B L E DATA S T EWA R D S H I P
The first and most obvious principle in genomethics,
stemming from respect for autonomy and the importance
of avoiding, or minimizing, harms, is the need to ensure
that individuals’ data are handled in an ethical man-
ner. In particular, difficulties arise where respect for an
individual’s privacy conflicts with public beneficence
and the need for data sharing. There is no question that
individual medical records, which include genomic data,
should be stored securely and protected effectively (like
any other sensitive medical data), with access limited
to the patient themselves, medical professionals who
need access to deliver high-quality clinical care, and the
researchers involved in studies to which the patient/par-
ticipant has consented.
However, there is also no doubt that data-sharing across
jurisdictions is crucial for both clinical interpretation of
genomic test results, as well as future scientific research and
development. Discriminating between classes of variants
for different diseases in any individual’s genome will rely
entirely upon large genotype–phenotype databases of pre-
viously sequenced genomes, against which each variant can
be compared. These databases will inevitably be the result of
international collaboration in many, if not most, cases. How
can this be achieved without infringing on an individual’s
right to medical privacy? What is the just and fair way to
treat an individual’s genomic data shared across multiple
jurisdictions?
Although ethical practice in this area is still evolv-
ing,19 the principle of responsible data stewardship has
already been established, and models of good practice are
being developed by numerous biobanks and data reposi-
tories globally.20 This includes strong protections for indi-
viduals, such as explicit consent for inclusion in genomic
databases at the point of testing and appropriately
de-identifying or anonymizing publicly accessible data,
whilst promoting managed data access to those who have
a legitimate need for it. Because genome sequences (and
some rare phenotype combinations) are uniquely identi-
fiable, it may never be possible to completely remove the
chance of re-identifying an individual from within a full
dataset; hence limited, aggregate, or pooled datasets may
be more suitable for wider release. However, the likeli-
hood of, and harms associated with, this outcome must
be appropriately weighed against the certain benefits of
data sharing. Genomic databases frequently have differ-
ent levels of access with alternative security provisions
based on professional and institutional responsibility and
accountability. However, while managed data sharing
amongst academic and medical centers is now common-
place, granting access to commercial organizations—
ranging from biotech, diagnostics, and pharmaceutical
companies, to insurance, advertising, and employment
Ethic a l , Leg a l , a nd S oci a l I ssues in C l inic a l G eno m ics   •   2 5 3
agencies—is varied and likely to remain controversial for
the foreseeable future.
VA L I D C O N S E N T
Although obtaining informed consent remains the corner-
stone of good practice, many have pointed out the difficulty
of obtaining truly informed consent for whole-genome
sequencing.21,22 The potential scope and use of the data,
both now and in the future, is enormous and unpredictable;
hence, the potential benefits and risks of genome sequenc-
ing cannot be accurately or comprehensively assessed. This
has led to proposals for “open” or “broad” models of con-
sent,23 which do not attempt to restrict the data to specific
uses but keep the dynamic nature of scientific research in
mind. For example, using the data from specific individuals
or cohorts as control datasets in unrelated research studies is
an invaluable method for interpretation and discovery, but
it is clearly impossible to predict what future studies will
either investigate or uncover. Regardless of the context for
testing, depositing data in global databases to facilitate the
interpretation of individual variants and for use in future
research is absolutely critical to reaping the benefits of
genomics for healthcare.
GENOME SCREENING
The issue of consent for multigene testing or whole-genome
sequencing is further complicated by the occurrence of
IFs. The magnitude of this issue is so vastly increased in
whole-genome sequencing versus traditional genetic test-
ing that there are suggestions that IFs should no longer be
regarded as incidental or unexpected, but as anticipated
secondary findings that will occur regardless of the primary
purpose of testing. Everyone carries a number of recessive
variants of relevance to reproductive choice, as well as vari-
ants relating to drug metabolism, disease susceptibility, eth-
nicity, and family background. There remains much debate
over how to deal with the spectrum of information contained
in a genome sequence, ranging from whose responsibility it
is to look for and interpret IFs, to what types of IFs should
be shared with patients and research participants, and who
should have access to the information. Genomic analysis
will necessarily be targeted using computational methods,
but these analyses could be limited to diagnostically rele-
vant variants or used to facilitate wider genomic screening.
In 2013 the American College of Medical Genetics and
Genomics recommended that all clinical genomes should
be screened for clinically actionable variants in a short list
of genes relating to serious genetic conditions.23a Should
individuals be able to consent to receive specific genomic
findings, but not others? Do healthcare professionals have
a duty to reanalyze individual genomes and re-contact
patients in light of new scientific discoveries relating to any
clinically important findings?
Although most of the work on dealing with IFs has
focused thus far on research participants24—primarily
because most genome-sequencing data to date have been
generated in a research context—the same conceptual
frameworks for thinking about IFs apply equally in the
clinic. The main difference, if indeed there is one, is in
the responsibility of a clinician to act in the best interests
of their patient (although the clinician may actually wear
two hats—one as the main clinical caregiver, and one as
a clinical researcher). When clinically actionable variants
are uncovered, it would be usual for a clinician (wear-
ing her or his clinical hat) to share these with patients,
regardless of whether the result relates to the primary
purpose for testing or not. Difficulties nonetheless arise
around variants with unknown or minor clinical sig-
nificance, variants associated with diseases for which no
therapeutic or preventative actions can be taken, recessive
variants that may be relevant to reproductive choice, and
so on. Although numerous proposals have been made to
group variants into different categories according to their
clinical validity and utility, and potentially offer a choice
of which variants to analyze and disclose to the patients,25
no consensus has yet been reached. From a public health
perspective, trawling through an individual’s genome
looking for potentially pathogenic variants in the absence
of any associated symptoms, phenotypes, or family his-
tory is perhaps more akin to screening than diagnostic
testing and hence is likely to be prone to false positives
and over-diagnosis. Even relatively well-characterized
known pathogenic variants have often been studied only
in symptomatic individuals and families, so little is known
about their frequency and penetrance in the asymptom-
atic general population.25a
There are specific concerns about IFs and genome
screening in relation to prenatal genome sequencing.26
Currently, there are clear guidelines for tests that are
offered prenatally, and where possible targeted tests are
generally preferred to open-ended genome-wide assays.
However, were genome sequencing to be used, and poten-
tially offered non-invasively by testing cell-free fetal DNA
circulating in the maternal blood,27 there is a potential to
creep outside of the purpose of testing. In addition to their
potential for use in reproductive choice beyond those
envisaged at the time of testing, such as decisions to termi-
nate pregnancies at very low or uncertain risks, secondary
2 5 4   •   P rincip l es o f G eno m ic Medicine
information from prenatal genome sequencing might be
used to inform parents about traits of future interest in
the child. This could change the norms and expectations
of pregnancy, and undermine the child’s future autonomy
to choose not to know about their genome, while perpetu-
ating an inappropriately deterministic view of the role of
genetics in child-rearing.
A concern often raised over IFs and variants of unknown
significance is the potentially unmanageable workload that
dealing with large numbers of variants in every patient might
bring to the health profession. Although this is unlikely to
be problematic if IFs are limited to known clinically action-
able variants,28 the informatics infrastructure required to
develop and maintain a clinical-grade analysis system to
alert clinicians to important genomic findings is currently
nonexistent within healthcare services. The possibility that
patients might be able to choose what results they wish to
receive from a menu of options is highly speculative at this
point, and would require substantial investment in infor-
matics, medical training, and public education. Moreover,
the economic and legal implications of providing a detailed
interpretation of every individual’s entire genome sequence
have not yet been established, and doing so will be cru-
cial for determining how best to invest limited healthcare
resources.
M A I N S T R E A M I N G G E N ET I C S
As the science of genomics develops, and more single-gene
subsets of common diseases are uncovered, it is likely
that genetic or genomic tests will increasingly be ordered
and interpreted by medical specialists outside of clinical
genetics. The move towards mainstream medicine is likely
to be accompanied by a shift away from managing fami-
lies in favor of testing individual patients—even if it is
likely to continue to be the case that at-risk family mem-
bers will be identified as a consequence of unexpected
clinical manifestations of potentially inherited disor-
ders. Consent for testing may become more laissez-faire,
as a genomic analysis comes to be seen as just another
test along with a battery of other standard tests used to
diagnose an individual’s condition. Indeed, for the vast
majority of people who do not have a highly penetrant,
heritable genetic condition—where the genetic informa-
tion will be used primarily to stratify the disease subgroup
or choose the most suitable treatment regime—enforcing
a model of genetic counselling and informed consent that
requires individuals to consider the future implications
for family members prior to testing may be unwarranted
and possibly unwelcome.
 G E N O M I C S I N P U B L I C H E A LT H
There are also potentially far-reaching applications of genom-
ics in public health.29 Existing genetic-screening programs—
such as preconception carrier-screening, antenatal screening,
and the newborn bloodspot screening test—could poten-
tially be expanded to include more conditions using genome
sequencing, without the need for a major reassessment of
the overarching ethical context in which these programs
are offered. However, it has been suggested in the media
that newborn babies will or even should have their genome
sequenced at birth and stored for future use, replacing the
existing newborn bloodspot and any future requirement for
genetic data. Such an enormous change to medical and pub-
lic health practice is unlikely to be considered seriously until
data security and public acceptance can be guaranteed, and
clinical utility and cost-effectiveness proven.30
A population database of individuals’ genomes would
allow a plethora of screening tests to be systematically per-
formed, both for heritable single-gene conditions and for
genetic risk factors associated with common diseases. This
might allow existing screening programs to be better tar-
geted at populations with the highest risk.31 Whether the
systematic collection and storage of individual genome
sequence would improve population health to an extent
that justifies the resources required is currently unclear.
Public acceptability of the storage and use of genomic data
for such purposes remains largely unexplored, and the clini-
cal impact of specific genomic variants in healthy individu-
als is largely unknown.
R E S E A R C H VE R S U S C L I N I C AL C A R E
One peculiarity of genetics and genomics is the particu-
larly close relationship between clinicians and researchers,
and hence between patients and research participants. Both
genomic technologies and scientific understanding have
advanced at such a pace over the last decade that the best
way to access state-of-the-art technology and knowledge
is through research studies. Many patients have become
research participants in the hope of finding a genetic diag-
nosis for their condition, and many clinicians have turned
to research for the same reason. This has led to a substantial
blurring of traditional boundaries between providing indi-
vidual clinical care and doing scientific research.
As a consequence, in the genomic era, many of the ethi-
cal principles discussed in the previous section are going
to become equally applicable within the research and the
clinical context. However, unlike clinicians, whose pri-
mary responsibility is to their patient, the researchers’
Ethic a l , Leg a l , a nd S oci a l I ssues in C l inic a l G eno m ics   •   2 5 5
main responsibility of has until now been primarily seen
as being to society (or to their funders). The relationship
between patients and clinicians, and their respective rights
and responsibilities, are well established and enshrined in
best-practice guidance, medical regulation, and gover-
nance; in contrast, the relationship between research par-
ticipants and researchers is essentially unregulated (aside
from the input of local research ethics committees). What
responsibilities does a genomic researcher have towards an
individual research participant?
I N F O R M E D A N D VO LU N TA RY
C O NS E N T
Informed and freely given consent is a vital part of research
on human subjects, enshrined in the Declaration of
Helsinki,32 which states that “in medical research involving
human subjects, the well-being of the individual research
subject must take precedence over all other interests.”
Participants should have a good understanding of the
research and its implications, and should feel able to refuse
to take part or withdraw at any time without penalty. Aside
from issues concerning obtaining meaningful informed
consent for genome sequencing, discussed earlier, the other
key element of informed consent is that it should be given
voluntarily. Ensuring true voluntariness in the face of ever
blurring boundaries between clinical practice and research
is a challenge, and the therapeutic or diagnostic misconcep-
tions that accompany that blurring are potentially impor-
tant. Individuals may feel they have no choice but to join
a study if they want to get a diagnosis, particularly where
the clinician and the researcher are one and the same per-
son. Cultural and linguistic barriers, as well as a general
ignorance about genetics, may also play a role in whether
individuals or families fully understand what they are con-
senting to and whether they feel able to decline to partici-
pate. With respect to samples, consent must detail whether
samples will be stored for future use; though again, broad
consent may be preferable to allow for future avenues of
research. Importantly, there is a developing consensus that
unauthorized (unconsented) genome testing should be pro-
hibited,19 and performing a genetic test on any sample origi-
nating from an individual who has not consented (“DNA
theft”) is illegal in many countries.
R E S E A RC H FI N D I N G S
Individuals may participate in medical research for largely
altruistic purposes, to contribute to human knowledge, or
they may be motivated to enter a research study primarily
for personal reasons, such as access to a new treatment or C A S E S T U DY: T H E D EC I P H E R I N G
diagnostic technology. Regardless of their motivations, D EVE L O PM E N TA L D I S O R D E R S P RO J EC T
it has been argued, the ethical principles of autonomy,
One large genomics study that is pioneering the system-
beneficence, and reciprocity are directly relevant to the
atic return of individual research results is the Deciphering
relationship between scientific researchers and research
Developmental Disorders (DDD) project,35 which aims to
participants.33 This has led to the suggestion that, in addi-
improve understanding of the genetic architecture of severe
tion to publishing the aggregate results of their research
developmental disorders while facilitating the translation
in the public interest, researchers should offer to return
of high-throughput genomic technologies into the United
individual-level research findings to individual partici-
Kingdom’s National Health Service (NHS).35 Families have
pants. Importantly, the policy regarding the return of
been recruited into the study by regional genetic services
individual results should be made clear at the consent
across the United Kingdom (starting in 2011 and continu-
stage, so that individuals can choose whether they wish to
ing until 2015), and clinical and phenotype data are entered
participate.
online by local clinicians. Samples are sent to the Wellcome
Once again, the issue of IFs is highly topical here.
Trust Sanger Institute, where various high-resolution
Unlike clinical testing, there is no requirement or expecta-
genomic assays (DNA microarrays and exome sequenc-
tion in most research studies to return any individual-level
ing) are performed to attempt to identify the cause of the
data, even results relating to the specific research purpose.
child’s developmental disorder. When a diagnosis has been
However, the academic debate over returning individual
made, it is fed back to the family’s referring clinician via a
research findings to individual research participants is mov-
secure-log-in website15 using a semi-automated system, and
ing in favor of offering to return a variety of different find-
the local clinician can then decide whether to contact the
ings, possibly with an option at the point of consent for an
family to confirm the result and provide genetic counselling
individual to decide what types of results (if any) they would
as required.
wish to receive.34 Options might range from raw genome
Because the study is focused on children with severe,
data, to data about a wide variety of traits and diseases, to
undiagnosed developmental disorders, it was felt that return-
pathogenic variants that cause a specific predefined condi-
ing carrier information or results relating to adult-onset dis-
tion. Non-clinical data, such as consanguinity or misattrib-
orders would be inappropriate, so the policy of the project is
uted paternity, should be also be explicitly considered and
not to return IFs at all (except where it is unavoidable; e.g., a
discussed.
large deletion removing both a developmental disorder and
Some proponents of this model go further, and sug-
a cancer-predisposition gene). Crucially, since the study is
gest there is a moral imperative to return life-saving clini-
returning pathogenic changes likely to be pertinent for indi-
cally actionable findings to individuals (or their healthcare
viduals, the practical requirements for returning genomic
providers), arguing that not to do so is tantamount to
variants of any kind have been developed and put in place,
disregarding the “Rule of Rescue—the perceived duty to
including sample tracking, variant filtering algorithms, and
save endangered life where possible”33 This implies that
informatics pipelines, as well as linked-anonymized patient
researchers have a duty not only to society, to perform the
records. The experience of shifting resources in this project
research that they have been funded for, but also to indi-
from pure genomics research into providing a translational
vidual research participants—to provide genomic analysis
service should be invaluable in assessing the viability of such
across a wide variety of clinical conditions and re-contact
an approach in future studies.
individuals or their clinicians where it is deemed appropri-
ate. In practice, placing this additional burden on research
teams or biobanks has enormous resource implications and C O N C LU S I O N S
may ultimately be deemed inappropriate and unnecessary
in many cases, particularly where the cohort is simply too High-throughput multi-gene or whole-genome sequencing
large or geographically dispersed to maintain high enough is now reaching clinical application. This will bring substan-
standards of sample-tracking and data quality. In addition tial new challenges to clinical genetics and mainstream med-
to concerns over feasibility, feedback might exacerbate the icine if we are to maximize its utility and reap the benefits
therapeutic misconception by further blurring the distinc- in terms of healthcare whilst providing appropriate protec-
tion between research and clinical care, and may cause harm tions for the interests of patients and research participants.
through incorrect interpretation of a result by either the Although the principles of genethics will still be relevant
researcher, clinician, or research participant themselves.33 in the genomethics era, particularly for the management

2 5 6   •   P rincip l es o f G eno m ic Medicine

of disorders with a strong heritable component, the poten-
tial for much wider applicability of whole-genome data
may require the development of a new ethical paradigm.
Together with this, considerable thought needs to be given
to what principles of genetic counselling can also be use-
fully applied to whole-genome data-sharing. It is unlikely
that genetic counselling in its current form—established
for serious, often life-threatening genetic conditions—will
translate directly to dealing with the data gained from a
whole genome. Thus, genomethics challenges us to reevalu-
ate the relevance of genetic counselling in its current form,
and it is likely that a new model of communication about
genomics will emerge.
The defining feature of the genomic world is the gen-
eration of data on an unprecedented scale, making ethical
data-management crucial. Storage, access, and interpreta-
tion of individual genomes will be vastly facilitated by global
genotype-phenotype databases, which need governance
frameworks that promote responsible data-stewardship and
use suitable consent procedures. This will require an appro-
priate balance between respecting individual autonomy
and the right to privacy on one hand, and the benefits of
data sharing and the duty to care for family members on the
other.
Individual clinical and research teams will need to
decide on a policy for the return of incidental or second-
ary findings and ensure that patients and research partici-
pants understand and consent to this policy. Once again,
this will require an appropriate balance between individual
autonomy and beneficence, versus public beneficence and
fair allocation of resources. Against this background, it is
our view that there is an urgent need in this area for empiri-
cal social science research, critical ethical analysis, and the
creation of new conceptual frameworks, to identify and
analyze the key ethical issues and to work towards the devel-
opment of models of good practice.
An interesting and somewhat unexpected result of the
decreasing cost of genomic technologies, coupled with
scientific and medical uncertainty around their interpreta-
tion and implementation, is the rise of consumer genetics.
Although many have been quick to criticize this nascent
industry and the validity of the some of the information
provided,36 these companies have provided fertile ground
for exploring what sort of information individuals might
wish to receive, how best to store and communicate com-
plex probabilistic information, how individuals use the
information and what levels of uncertainty consumers are
willing to accept in the analysis of genome data. These are
the very questions with which the emerging discipline of
genomethics must concern itself.
Ethic a l , Leg a l , a nd S oci a l I ssues in C l inic a l G eno m ics   •   2 5 7
REFERENCES
1. O’Sullivan, J., et al. A paradigm shift in the delivery of services for
diagnosis of inherited retinal disease. Journal of Medical Genetics 49,
322–326 (2012).
2. Worthey, E.A., et al. Making a definitive diagnosis: Successful clini-
cal application of whole exome sequencing in a child with intrac-
table inflammatory bowel disease. Genetics in Medicine 13, 255–262
(2011).
3. Wright, C.F., et al. Next steps in the sequencing: The implications
of whole genome sequencing for health in the UK. Public Health
Genomics Foundation (2011).
4. Parker, M. Ethical Problems and Genetics Practice. Cambridge,
UK: Cambridge University Press (2012).
5. Green, M.J., & Botkin, J.R. “Genetic exceptionalism” in medi-
cine:  Clarifying the differences between genetic and nongenetic
tests. Annals of Internal Medicine 138, 571–575 (2003).
6. Andorno, R. The right not to know: An autonomy based approach.
Journal of Medical Ethics 30, 435–439 (2004).
7. Lucassen, A., & Parker, M. Confidentiality and sharing genetic
information with relatives. The Lancet 375, 1507–1509 (2010).
8. Royal College of Pathologists and British Society of Genetic
Medicine. ‘Consent and confidentiality in clinical genetic prac-
tice:  Guidance on genetic testing and sharing genetic informa-
tion’. In: Report of the Joint Committee on Medical Genetics, 2nd ed.
London: RCP, RCPath (2011).
9. European Society of Human Genetics. Genetic testing in asymptom-
atic minors: Recommendations of the European Society of Human
Genetics. European Journal of Human Genetics 17, 720–721 (2009).
10. Mand, C., Gillam, L., Delatycki, M.B., & Duncan, R.E. Predictive
genetic testing in minors for late-onset conditions: A chronological
and analytical review of the ethical arguments. Journal of Medical
Ethics 38(9), 519–524 (2012). doi:10.1136/medethics-2011-100055
11. Clarke, A. Response to: What counts as success in genetic counsel-
ling? Journal of Medical Ethics 19, 47–49 (1993).
12. McAllister, M., et al. Patient empowerment in clinical genetics ser-
vices. Journal of Health Psychology 13, 895–905 (2008).
13. Emery, S.D., Middleton, A., & Turner, G.H. Whose deaf genes
are they anyway? The deaf community’s challenge to legislation on
embryo selection. Sign Language Studies 10, 144–169 (2010).
14. Booth, T.C., Jackson, A., Wardlaw, J.M., Taylor, S.A., & Waldman,
A.D. Incidental findings found in “healthy” volunteers during imag-
ing performed for research: Current legal and ethical implications.
British Journal of Radiology 83, 456–465 (2010).
15. Pichert, G., Mohammed, S.N., Ahn, J.W., Ogilvie, C.M., & Izatt,
L. Unexpected findings in cancer predisposition genes detected
by array comparative genomic hybridisation: What are the issues?
Journal of Medical Genetics 48(8), 535–539 (2011).
16. Lucassen, A., & Parker, M. Revealing false paternity: Some ethical
considerations. The Lancet 357, 1033–1035 (2001).
17. The 1000 Genomes Project Consortium. An integrated map of
genetic variation from 1,092 human genomes. Nature 491, 56–65
(2012).
18. MacArthur, D.G., et  al. A systematic survey of loss-of-function
variants in human protein-coding genes. Science 335, 823–828
(2012).
19. Presidential Commission for the Study of Bioethical Issues. Privacy
and Progress in Whole Genome Sequencing. Washington, DC (2012).
http://bioethics.gov/node/764
20. Firth, H.V., et  al. DECIPHER:  Database of Chromosomal
Imbalance and Phenotype in Humans using Ensembl Resources.
American Journal of Human Genetics 84, 524–533 (2009).
21. Mascalzoni, D., Hicks, A., Pramstaller, P., & Wjst, M. Informed con-
sent in the genomics era. PLoS Medicine 5, e192 (2008).
22. McGuire, A.L., & Beskow, L.M. Informed consent in genomics and
genetic research. Annual Review of Genomics and Human Genetics
11, 361–381 (2010).
 23. Lunshof, J.E., Chadwick, R., Vorhaus, D.B., & Church, G.M. From
genetic privacy to open consent. Nature Reviews Genetics 9, 406–
411 (2008).
23a. Green, R. C., et al. ACMG recommendations for reporting of inci-
dental findings in clinical exome and genome sequencing. Genetics
in Medicine 15, 565–574 (2013).
24. Wolf, S.M., et al. Managing incidental findings in human subjects
research:  Analysis and recommendations. The Journal of Law,
Medicine & Ethics 36, 219–248 (2008).
25. Berg, J.S., Khoury, M.J., & Evans, J.P. Deploying whole genome
sequencing in clinical practice and public health: Meeting the chal-
lenge one bin at a time. Genetics in Medicine 13, 499–504 (2011).
25a. Wright, C.F., et al. Policy challenges of clinical genome sequencing.
BMJ 347, f6845 (2013).
26. Donley, G., Hull, S.C., & Berkman, B.E. Prenatal whole genome
sequencing. Hastings Center Report 42, 28–40 (2012).
27. Lo, Y.M.D., et  al. Maternal plasma DNA sequencing reveals the
genome-wide genetic and mutational profile of the fetus. Science
Translational Medicine 2, 61ra91 (2010).
28. Johnson, G., Lawrenz, F., & Thao, M. An empirical examination of
the management of return of individual research results and inciden-
tal findings in genomic biobanks. Genetics in Medicine 14, 444–450
(2012).
29. Burke, W., et  al. Extending the reach of public health genom-
ics:  What should be the agenda for public health in an era of
2 5 8   •   P rincip l es o f G eno m ic Medicine
genome-based and “personalized” medicine? Genetics in Medicine
12, 785–791 (2010).
30. Human Genetics Commission. Profiling the newborn: A prospec-
tive gene technology? A report from a Joint Working Group of the
Human Genetics Commission and the UK National Screening
Committee. London, Department of Health (2005).
31. Pharoah, P.D.P., Antoniou, A.C., Easton, D.F., & Ponder, B.A.J.
Polygenes, risk prediction, and targeted prevention of breast cancer.
New England Journal of Medicine 358, 2796–2803 (2008).
32. World Medical Association. Declaration of Helsinki—Ethical
Principles for Medical Research Involving Human Subjects. (2008).
See http://www.wma.net/en/30publications/10policies/b3/
33. Bredenoord, A.L., Kroes, H.Y., Cuppen, E., Parker, M., & van Delden,
J.J.M. Disclosure of individual genetic data to research partici-
pants: The debate reconsidered. Trends in Genetics 27, 41–47 (2011).
34. Wolf, S.M., et al. Managing incidental findings and research results
in genomic research involving biobanks and archived data sets.
Genetics in Medicine 14, 361–384 (2012).
35. Firth, H.V., Wright, C.F., & the DDD Study. The Deciphering
Developmental Disorders (DDD) study. Developmental Medicine
& Child Neurology 53, 702–703 (2011).
36. Janssens, A.C.J.W., et al. A critical appraisal of the scientific basis of
commercial genomic profiles used to assess health risks and person-
alize health interventions. American Journal of Medical Genetics 82
593–599 (2008).
 18.
THE REGULATION OF HUMAN GENOMICS RESEARCH
Jane Kaye
INTRODUCTION
A key event in the regulation of medical research was the
Nuremberg Trials that were held after World War II, which
found Nazi physicians and researchers guilty of atroci-
ties carried out in the name of medical research. Since
then, research ethics committees and oversight mecha-
nisms for medical research on human beings have become
the accepted norm in most countries. The regulation of
genomic research, like other subsets of medical research, is
subject to the same regulatory controls that apply to medi-
cal research in general. Therefore, the focus of this chapter
will be on the regulation of medical research on human
beings (excluding the approval of medicines and medical
devices) in order to discuss the ways in which these more
general requirements apply to genomic research. This
chapter will also make reference to some of the specialist
bodies that have been established as expert advisory bod-
ies for genomics.
I will compare the regulatory systems in the United
Kingdom and the United States in order to provide a gen-
eral overview, while at the same time illustrating some of
the significant differences that exist between them. These
countries have quite different ways of regulating medi-
cal research on human beings. For example, the United
Kingdom is increasingly influenced by European law and is
subject to increasing harmonization across Europe, whereas
the United States has its own unique regulatory regime. In
order to elucidate these differences, this chapter will com-
pare the legal framework between the two jurisdictions, the
institutions that have been established to oversee medical
research on human beings, and the powers that they have. By
highlighting the differences between two well-established
legal systems in the Western world, it is hoped that this
review may assist readers in other countries where the issues
of the regulation of human genomics research are also
actively debated and considered.
259
T H E L E G A L F R A M EWO R K F O R
MEDICAL RESEARCH
T H E C O N S T I T U T I O NA L S T RU C T U R E
In order to understand the legal framework for medical
research, it is important to understand the role of law and
its origins in each jurisdiction. The legal frameworks in the
United Kingdom and the United States have been developed
through different processes. In the United Kingdom, there
is a national Parliament that makes law for the whole coun-
try. As a member of the European Community, the United
Kingdom is also subject to European law. Regulations are
directly applicable in all European Union member states,
but EU directives need to be implemented by the member
state in order to have effect. The Westminster Parliament
is obliged to implement directives that are passed by the
European Parliament into U.K.  law. Failure to do so will
result in legal proceedings against the United Kingdom
in the Court of Justice of the European Union. However,
each of the member states has its own “margin of apprecia-
tion” in the way that it implements and interprets European
Directives, and this has led to discrepancies in the legislation
of each member state. The United Kingdom can also sign
Conventions of the European Council, but this is entirely
voluntary. The most significant legal instrument regarding
medical research in terms of the European Community
is the Convention on Human Rights and Biomedicine
(1997). However, the United Kingdom has yet to sign this,
due to the provisions that relate to stem cell research.
In contrast, the U.S. Congress (consisting of the House
of Representatives and the Senate) is the highest law-making
body in the United States. There is no obligation to sign
or ratify the law of another authority, such as the United
Nations or an international treaty, unless this has been a
decision of the American Congress. The United States is a
federal system, which means that the Congress and the 50
state legislatures have different Constitutional powers. For
medical research there is federal legislation, but the states
also have considerable powers to legislate. The lack of a
national health service has also resulted in differences and
anomalies in the provision of healthcare in different parts
of the United States.
T H E RO L E O F S TAT U T E
The United States has specially drafted federal legisla-
tion that covers the regulation of medical research, which
applies across the United States. In 1991, the Department
of Health and Human Services issued the “Common
Rule.”1 The philosophical underpinnings of the Common
Rule come from the Belmont Report—Ethical Principles
and Guidelines for the Protection of Human Subjects. The
Common Rule has a very narrow ambit, as it only applies to
medical research that is conducted by a federal department
or agency or is federally funded. Research that comes under
this scope must be reviewed and approved by an institu-
tional review board (IRB) that operates in accordance with
the requirements of this policy.
However, under the Common Rule, the following
research does not require IRB approval:
1. Research conducted in established or commonly
accepted educational settings, involving normal
educational practices;
2. Research involving the use of educational tests
(cognitive, diagnostic, aptitude, achievement), survey
procedures, interview procedures, or observation of
public behavior, unless an individual can be identified
or the information provided by the individual will
incriminate or be detrimental to him or her;
3. Research involving the use of educational tests as
long as the human subjects are elected or appointed
public officials or candidates for public office; or
federal statute(s) require(s) without exception that
the confidentiality of the personally identifiable
information will be maintained throughout the research
and thereafter;
4. Research involving the collection or study of existing
data, documents, records, pathological specimens,
or diagnostic specimens, if these sources are publicly
available or if the information is recorded by the
investigator in such a manner that subjects cannot be
identified, directly or through identifiers linked to the
subjects;
2 6 0   •   P rinciples of G enomic M edicine
5. Research and demonstration projects which are
conducted by or subject to the approval of department
or agency heads;
6. Taste and food quality evaluation and consumer
acceptance studies.2
There have been significant advances in scientific
research since the Common Rule was developed, and it
is no longer seen to be as effective and comprehensive as
it could be. Therefore, in July 2011, the U.S. Department
of Health and Human Services announced a proposal to
improve the rules protecting human research subjects.3 The
suggested amendments would make considerable changes
to the Common Rule. They propose to amend seven aspects
of the current framework, as follows:4
1. Refining the existing risk-based regulatory framework;
2. Establishing a single IRB review of record for domestic
sites in multiple locations;
3. Improving consent forms and the consent process;
4. Establishing mandatory data security and information
protection standards for all studies involving
identifiable or potentially identifiable data;
5. Establishing an improved system for managing data on
unanticipated problems and adverse events;
6. Extending all federal regulatory protections to include
all research at a U.S. institution that is wholly or even
partly federally funded for research on human subjects;
and
7. Harmonization of regulations and related agency
guidance.
One of the consequences of these changes would be the
amendment of the exemption categories listed above. For
instance, category 4 would be expanded to include “all sec-
ondary research use of identifiable data and biospecimens
that have been collected for purposes other than the cur-
rently proposed research, provided that specified new con-
sent requirements are satisfied.”5 The aim of the proposals is
to place less emphasis on the need to de-identify data, and
to counteract the related loss of privacy by establishing a
general rule that the participant must consent in writing to
the use of their biospecimens in research. Whilst the initial
public consultation period for these proposals has closed,
it is likely to undergo further amendment before becoming
federal law.
 There are numerous pieces of legislation that give
power to supervisory bodies to act, or that have direct or
indirect implications for research. Some examples are the
Bayh-Doyle Act 1980,6 which has had an effect on research,
as it imposes a legal obligation on researchers who receive
government funding to protect their intellectual property
in order to allow the commercialization of their results.7
The Health Information Patient Privacy Act (HIPPA) was
introduced across the United States in 2003 to provide
uniform privacy protection for patients across the country,
and it has also had a significant impact on research practice.
The HIPAA Regulations require that a patient must autho-
rize the use of their records for medical research purposes.
This can be waived, however, by an IRB or a privacy board.
Alternatively, researchers can obtain access to medical
records from a covered entity (such as a hospital) without
IRB or patient authorization, in two situations:8
1. In preparing a research protocol (provided access to
medical records is required in order to do so, and no
protected medical information is removed from the
site); and
2. When performing research relating solely to those who
have already died.
However, there are various other pieces of legislation
that have relevance for medical research.9
In contrast, the U.K.  regulatory system for medical
research on human beings does not have one piece of leg-
islation that provides a framework for the governance of all
medical research.10 Animal research, on the other hand, is
governed by the Animal (Scientific Procedures) Act 1986,
which has its origins in a nineteenth-century piece of leg-
islation.11 Providing a statutory framework for an activity
not only gives a mandate from a democratically elected
body, it also provides the opportunity to allocate powers
of enforcement and accountability. Instead of a compre-
hensive piece of legislation, the law that applies to medi-
cal research on human beings in the United Kingdom is a
complex patchwork of legislation, regulations, common law
principles, guidelines, and codes of practice.12 At the level of
statute, there are many general pieces of legislation that have
clauses that refer to medical research. There are other stat-
utes that relate directly to either medical research, the use
of personal information, or the use of human tissue. These
are the Data Protection Act (1998), the Human Rights
Act (1998), the Mental Capacity Act (2005), the Human
Tissue Act (2004), the National Health Service Act (2006),
the Freedom of Information Act (2000), the Human
Fertilisation and Embryology Act (1990 and 2008), the
T he R egulation of H uman G enomics R esearch   •   2 6 1
Access to Health Records Act (1990), the Health and
Social Care Act (2001), the Computer Misuse Act (1990),
and the Electronic Communications Act (2000).
Alongside these are regulations that also have a bear-
ing on medical research, and are approved by the Secretary
of State for Health. The most significant of these is the
Medicines for Human Use (Clinical Trials) Regulations
2004. These regulations apply to clinical trials, and are the
only instrument in the United Kingdom that details the
requirements for informed consent of medical research.
These requirements are derived from the Declaration of
Helsinki. The other important features of these regulations
are the sections that relate to research ethics procedures
and standards. These regulations are a direct implementa-
tion of the European Community Clinical Trials Directive
2001/20/EC.13 This is an example of the increasing influ-
ence that the law of the European Community has on its
member states.
This European influence on medical research will
greatly increase if the proposed European Data Protection
Regulation is passed in its current form. The Regulation
specifically includes “genetic data” (Article 4[10]), and
attempts to amend the current restrictions on scientific
research that arose from the current Data Protection
Directive. It proposes implementing strict requirements as
to consent, in that it must be freely given, informed, specific
and explicit (Article 4[8]‌). The Regulation could result in
a greater degree of harmonization between member states,
but there will still be a considerable margin for apprecia-
tion. It is unlikely that it will be in force before 2014.
T H E RO L E O F GU I D E L I N E S
There are also several key guidelines that apply to all
employees in the National Health Service (NHS).
These are written by the Department of Health and the
National Institute for Health and Clinical Excellence
(NICE); an independent organization that is respon-
sible for providing national guidance relating to
health care. As there is a comprehensive health system
(though this is constantly being reorganized), most
medical professionals in the United Kingdom are cur-
rently employed by the NHS. However, the passing of
the Health and Social Care Act in 2012 means that
the commissioning of healthcare provision will be car-
ried out by general medical practitioners, which has
opened up the possibility for greater provision of NHS
services by private companies, with fewer medical pro-
fessionals being employed by the NHS. This will also
have an effect on medical research. Whilst the NHS is
currently one of the key access points for research par-
ticipants and patients, not all medical research is carried
out by this body. A great deal of research is also carried
out in universities and through the commercial sec-
tor. There are guidelines written by professional bod-
ies such as the General Medical Council (the statutory
body responsible for the registration and accreditation
of doctors), the British Medical Association, and the
Royal Colleges. Also, major funding bodies such as the
Medical Research Council, Cancer Research U.K., and
the British Heart Foundation write guidelines or notes
for practice. If a case is litigated, the courts will take
into account professional guidelines, particularly those
issued by the General Medical Council, as a basis for
determining whether the appropriate standard of care
has been followed. In the United States, there are simi-
lar bodies that write guidelines for medical research, but
these are not binding on researchers. This differs from
guidance issued by the National Institute of Health and
other nationally funded bodies, which must be followed.
T H E RO L E O F T H E C O U RT S
Whilst the courts in England and Wales are prepared to
consider guidelines when assessing the evidence relat-
ing to the standard of care, the court will make the final
decision on what is the appropriate standard. In the vast
majority of cases, the court will accept the medical stan-
dard of a body such as the General Medical Council,
unless “it is not capable of logical analysis.”14 The deci-
sions of the courts have an effect on the way that medi-
cal research is conducted in the United Kingdom, even
though many of the decisions are based on cases relat-
ing to medical treatment involving negligence. There
has only been one case concerning negligence in medi-
cal research, and that was the Creutzfeldt-Jakob Disease
Litigation, Plaintiffs v. United Kingdom Medical Research
Council and another case.15 It is worth quoting from the
judgement, as, while the courts acknowledge that they
should err on the side of caution in finding negligence,
they are also prepared to impose a high standard of care
on researchers, using the same legal standards that are
applied in all negligence cases.
The courts must be very cautious in condemning a
clinical trial or therapeutic programme. Too ready
a labeling of an act or omission as negligent by the
courts could stultify progress in medical and scien-
tific research and render eminent experts reluctant
to serve on committees voluntarily. However, during
2 6 2   •   P rinciples of G enomic M edicine
the clinical trial of a new drug or form of treatment,
and especially when the clinical trial is becoming a
general therapeutic programme, all reasonably prac-
ticable steps should be taken to minimise dangers
and side-effects. To discharge this duty, constant
alert and inquiring evaluation of the trial or pro-
gramme is required. I do not accept that a govern-
ment department or a quasi-governmental agency
such as the MRC [Medical Research Council] can
discharge this duty by a lower standard of care than
a commercial pharmaceutical company.  .  .  . In my
judgement, the same duty with the same standard
of care is owed to all patients who are the subjects
of clinical trials or new therapeutic programmes,
whether the responsibility of a pharmaceutical com-
pany, government department or other agency.16
The courts in England and Wales also have a key role in
determining the requirements for valid consent. The cho-
sen criteria primarily stem from the common law doctrine
of negligence and the medical professional’s duty to warn.
In contrast to the United States, Canada, and Australia,
the test in the United Kingdom has been based on what
a doctor would consider necessary to tell a patient,
rather than what a reasonable patient would expect to
know.17 However the House of Lords’ decision in Chester
v.  Afshar18 suggests that the courts may be moving more
to a position where the information that must be given to
the patient should be based on what a reasonable patient
would want to know.
The courts in the United States have also been very
influential in directing the standards that should be applied
to medical research. As in the United Kingdom, the courts
pass judgement on the gray areas in the law, and have estab-
lished the requirements on issues such as consent, duty to
warn, privacy, and anti-discrimination. Examples of impor-
tant U.S.  decisions are Washington University v.  William
J.  Catalona19on the ownership of biological samples col-
lected for research purposes, and the cases on genetic test-
ing, such as Pate v.  Threlkel,20 Safer v.  Pack,21 and Molloy
v.  Meier.22 But the courts are not responsible for estab-
lishing a legal framework for research, as their rulings are
dependent upon the cases that come before them, and are
an incremental extension of existing principles rather than a
complete overhaul of a particular area. Therefore, the courts
can explain and develop certain areas of law for guidance
for researchers, but it is not their role to develop completely
new legal frameworks, which is the responsibility of the
state and federal governments that have the democratic
mandate to do so.
 R E GU L ATO RY B O D I E S A N D
T H E I R P OW E R S
T H E S IT UAT I O N I N T H E U N I T E D S TAT E S
The U.S. Department of Health and Human Services funds
two key organizations for the regulation of medical research
on human beings. These are the Office for Human Research
Protections (OHRP) and the National Institutes of Health
(NIH). The NIH is responsible for research in general,
whereas the OHRP has more responsibility for compliance
and accountability in research. Both of these institutions
focus on federally funded research, though their activities
and guidelines have an effect on other bodies that carry out
medical research in the United States.
The NIH is the primary federal agency in the United
States for conducting and supporting medical research. The
organization has its own research centers, which are lead-
ers in their respective fields, but it also funds other research
and provides guidance and information to researchers. The
NIH Office of Biotechnology Activities (OBA) is respon-
sible for writing guidance, providing advice, monitoring
scientific progress, and maintaining a register of research
activities. There are many expert advisory committees
that develop guidance and feed back findings to the OBA
and the NIH director, including the Recombinant DNA
Advisory Committee (RAC). Researchers at institutions
receiving NIH funding for research involving recombinant
DNA must comply with the NIH guidelines. These have
become a “universal standard for safe scientific practice in
this area of research and are followed voluntarily by many
companies and other institutions not otherwise subject to
their requirements.”23
The Office for Human Research Protections (OHRP)
provides leadership in the protection of the rights,
welfare, and well-being of subjects involved in
research conducted or supported by the U.S.
Department of Health and Human Services (HHS).
OHRP helps ensure this by providing clarification
and guidance, developing educational programs and
materials, maintaining regulatory oversight, and
providing advice on ethical and regulatory issues in
biomedical and social-behavioral research.24
In order to fulfill all of these functions, they orga-
nize conferences, develop resource materials, and hold
quality-improvement consultations. As a part of its
accountability strategy, OHRP has a system of assur-
ances where research institutions both in the United
T he R egulation of H uman G enomics R esearch   •   2 6 3
States and abroad have formal agreements to comply with
the OHRP regulations relating to research on humans.
It also runs an accreditation scheme for researchers in
order to improve the ethos of research practice within an
organization.
The OHRP is also responsible for maintaining a regis-
ter of institutional review boards (IRBs) or independent
ethics committees (IECs). These must approve feder-
ally funded research on human subjects or research that
comes under the Common Rule before it begins. The
Federal Policy for the Protection of Human Subjects or
the “Common Rule” was published in 1991. It was then
codified in separate regulations by 15 Federal depart-
ments and agencies and so applies to all publicly-funded
research carried out in the USA.25 The OHRP is respon-
sible for investigating any allegations of non-compliance
with the Common Rule and has the power to take action
to protect human research subjects. The initial approach
is to ask the institution to investigate; then OHRP may
conduct further investigations through on-site evalua-
tions. The OHRP has the power to make recommenda-
tions about further training, education, and procedures.
In extreme circumstances, it has the power to suspend
research programs if they are proving to be unsafe, such as
in the case of the death of Jesse Gelsinger in a gene-therapy
trial.26 These powers of inspection and investigation
ensure that human subjects are protected throughout the
research process.
As well as these key institutions, there are also numer-
ous organizations that have responsibility for the regula-
tion of medical research on human beings: for example, the
Environmental Protection Agency, the Centers for Disease
Control and Prevention (CDC), the Food and Drug
Administration, the Federal Trade Commission, and the
Securities and Exchange Commission.
As in the United Kingdom, it is the IRBs that are
responsible for the approval of research projects and dealing
with the applications. The IRBs evaluate proposed research
activities, making sure that the design of each study is con-
sistent with sound scientific principles and ethical norms,
such as obtaining informed consent. Applications can be
rejected if they do not meet the criteria, which will prevent
the research from taking place. The IRBs also have author-
ity to seek a yearly review of the research. If the research
involves vulnerable participants, reviews may be carried out
more frequently. In addition, IRBs have the authority to
“suspend or terminate approval of research that is not being
conducted in accordance with the IRB’s requirements or
that has been associated with unexpected serious harm to
subjects.”27
 T H E S IT UAT I O N I N T H E U N IT E D
KINGDOM
Before 2011, there was no one body that was responsible
for oversight of the whole research process in the United
Kingdom. This has since changed, with the establishment
of the Health Research Authority (HRA), whose purpose
is to protect and promote the interests of patients and the
public in health research.28 The National Research Ethics
Service (NRES) now comes under the aegis of the HRA
and is responsible for providing guidance and the system
of accreditation for research ethics committees (RECs).
NRES does not have any formal compliance or inspection
powers, though it is required to audit RECs in the United
Kingdom on a regular basis. In the United Kingdom, there
are in excess of 30 bodies that write guidelines that have
relevance for medical research. The result is that there a
lot of activity at the front end of research, in the form of
guidance and the approval of medical research by research
ethics committees, but little supervision after it has been
approved and is underway. The HRA may go some way
towards changing this situation and providing a more
cohesive approach to the regulation of medical research in
the United Kingdom.
There are five other bodies in the United Kingdom that
have powers of enforcement and supervision. These are the
General Medical Council, the Information Commissioner,
the Human Tissue Authority, the Human Fertilisation
and Embryology Authority, and the National Health
Service. The Human Tissue Authority is also responsible
for licensing collections of biological samples, although
there are numerous exemptions for research collections.
The Information Commissioner’s Office (ICO) has a light
touch when it comes to implementing the regulation, as
it does not have a system of regular inspection or over-
sight. However, the ICO has recently been given powers
to impose fines of up to £500,000 for breaches of the Data
Protection Act, which has greatly increased the effective-
ness of the ICO to enforce compliance with the Act. The
Human Fertilisation and Embryology Authority has sev-
eral powers of inspection and enforcement for ensuring
compliance, but its scope is restricted to gametic materials
and embryos.29 The National Health Service as an employer
also has professional supervisory procedures in place, such
as requiring Caldicott Guardians to be appointed to ensure
that ethical practices are followed. “A Caldicott Guardian is
a senior person responsible for protecting the confidential-
ity of a patient and service-user information and enabling
appropriate information-sharing.”30 Under Health Service
Circular:  HSC 1999/012, each NHS organisation that
2 6 4   •   P rinciples of G enomic M edicine
has access to patient records is required to have a Caldicott
Guardian.
Research Ethics Committees are the main
decision-makers in determining if, and how, medical
research should proceed in the United Kingdom. All bio-
medical research projects must undergo both ethical review
and technical scientific appraisal before any research begins.
Only research covered by the U.K.  clinical trials regula-
tions31 and the Human Tissue Act of 2004 must be submit-
ted to research ethics committees for approval; researchers
in other areas are under no such obligation. Research Ethics
Committees have no power to inspect research projects
once they have commenced, so that approval has become a
hurdle that just needs to be passed. This is in contrast to the
powers of the IRBs in the United States.
Research Ethics Committees in the United Kingdom
are concerned primarily with research that involves NHS
patients, premises, staff, or tissue samples. If research does
not fall within this ambit, “this does not mean the proj-
ect team can pursue their work unrestrained. It would be
highly unusual if the study did not fall within the remit of at
least one (and possibly several) other regulatory bodies.”32
Whether research falls within the remit of the Research
Ethics Committees or not, there will inevitably be further
documents to be completed by the researcher in order to
carry out their research. These are likely to include univer-
sity registration documents and hospital approval forms,
amongst others.
C O N C LU S I O N
New developments in the regulation of medical research
have come about largely in response to public scandals. For
instance, in the United Kingdom, the unlawful collection
and storage of human tissue and organs at Alder Hey and
Bristol resulted in the Human Tissue Act of 2004. These
changes have led to a “layering” effect of oversight mecha-
nisms and an increase in the number of requirements for
research to be approved, which in turn has led to frustration
for researchers. As Shaw and her colleagues state:
In the UK the rapid growth of systems and proce-
dures for research management and governance
has generated confusion and resentment in the
research community. They bemoan the rising bur-
den of paperwork, the curtailment of research free-
dom, expensive delays caused by lengthy application
procedures, inconsistent decisions, and in some
cases, the halting of entire research programmes
 by allegedly heavy handed but misinformed ethics
committees.33
The establishment of the HRA is intended to address this
situation by having one body that is responsible for medical
research regulation.
In the United States, scandals have also led to changes
in the regulatory system for medical research. For exam-
ple, the Tuskegee Syphilis Study led to the enactment of
the National Research Act of 1974 and the formation of
the National Commission for the Protection of Human
Subjects of Biomedical and Behavioral Research. More
recently, the financial conflicts of interest in the early
1990s have led to monographs providing guidance by
the Association of American Medical Colleges and the
Association of Academic Health Centers.34
However, the essential difference between the United
States and the United Kingdom is that there is a clear legisla-
tive basis for the regulation of research in the United States.
Federal legislation provides a framework that gives clarity
and certainty as to how, and on what basis, medical research
should proceed. The changes to the Common Rule pro-
posed by the Department of Health and Human Services
would retain this basic framework, but in a modernized and
more streamlined form. The bodies that are involved in the
supervision of research, such as the IRB and the OHRP,
have considerable powers and legal authority to enforce
their decisions. They are therefore able to take action when
research projects are failing to meet the required standards.
But the weakness of the American system is that it does not
have universal application to both public and private sectors
where research is carried out.
The purpose of this chapter has been to give an over-
view of the legal framework for medical research in the
United Kingdom and the United States, the bodies that
are responsible for overseeing research in each nation,
and to illustrate the different powers that they exercise. In
the space available, I  have only been able to give a board
overview of the current regulatory frameworks for medi-
cal research in the two jurisdictions, which are in a con-
stant state of change. It is evident that there are differences
between the two systems, which reflect the societal con-
cerns and the historical contexts of medical research in
which they have been developed. Both systems have to
be understood and complied with if researchers in each
jurisdiction wish to collaborate. There is scope for further
comparison and review of the two jurisdictions in order
to make an assessment of the best procedures for medical
research on human beings, as well as the implications that
this may have for global genomic research.
T he R egulation of H uman G enomics R esearch   •   2 6 5
AC K N OW L E D G E M E N T S
My thanks go to Heather Griffin for her assistance in
updating this chapter in light of considerable changes to
the regulation of research in the United Kingdom and
the United States since the previous edition of this vol-
ume. This research was funded by the Wellcome Trust
(096599/2/11/Z; WT091310).
REFERENCES
1. U.S. Department of Health and Human Services. 45 CFR 46. Fed
Regist 1991;56: 28012.
2. U.S. Department of Health and Human Services. 45 CFR 46. Fed
Regist 1991;56: 28012.
3. Office for Human Research Protections website: http://www.hhs.
gov/ohrp/ Accessed Dec. 12, 2012.
4. Federal Register Vol. 76, No. 143, July 26, 2011, p44512.
5. Federal Register Vol. 76, No. 143, July 26, 2011, p44512.
6. Pub L No. 96-517, 35 USC (1980).
7. Although the Supreme Court’s decision in Stanford University
v. Roche Molecular Systems et al. (09-1159) Feb. 28, 2011, may have
curtailed the effects of this legislation.
8. George, A.J. 2002. Medical privacy and medical research—judging
the new federal regulations. New England Journal of Medicine, Vol.
346, No. 3: 216–220.
9. Library of Congress website, http://thomas.loc.gov/. Accessed Jan.
23, 2013.
10. The devolution of powers to the Scottish and Northern Ireland
Parliaments means that legally the concept of the United Kingdom
is becoming increasingly complex. Over time, the law in these coun-
tries will start to deviate from that of England and Wales.
11. The Cruelty to Animals Act of 1876.
12. Kaye, J., et. al. Governing Biobanks—Understanding the Interplay
Between Law and Practice (Hart Publishing:Oxford 2012).
13. L 121/34, Official Journal of the European Communities, 1.5.2001.
14. Bolitho v. City and Hackney Health Authority [1998] AC 232.
15. (2000) 54 BMLR 8 (QB).
16. Creutzfeldt-Jakob Disease Litigation, Plaintiffs v. United Kingdom
Medical Research Council and another QB 54 BMLR 8.
17. Sidaway v. Bethlem Royal Hospital Governors [1985] AC 871 (HL).
18. (2004) UKHL 41.
19. Case No. 4:03 cv 01065 SNL. http://prostatecure.wustl.edu/.
Accessed April 28, 2006. Appealed in 2007; 490 F.3d 667.
20. 661 So 2d. 278 (Florida 1995).
21. 677 A. 2d 1188, 683 A 2d 1163 (New Jersey 1996).
22. Nos. C9-02-1821, C9-02-1837 (Minnesota 2004).
23. Recombinant DNA Advisory Committee (RAC) website, http://
oba.od.nih.gov/rdna_rac/rac_about.html. Accessed Dec. 11,
2012.
24. Office of Human Research Protections website, http://www.hhs.
gov/ohrp/about/facts/index.html Accessed Dec. 11, 2012.
25. http://www.hhs.gov/ohrp/humansubjects/commonrule/
26. George, A.J.T., et al., 2002. Research governance at the crossroads.
Nature Medicine Vol. 8, No. 2: 99.
27. U.S. Department of Health and Human Services. 45 CFR 46. Fed
Regist 1991;56: 28012. Para 46.113.
28. HRA website: http://www.hra.nhs.uk/about-us/. Accessed Jan. 23,
2013.
29. Human Fertilisation and Embryology Act, 1990 (HFEA).
30. http://systems.hscic.gov.uk/data/ods/searchtools/caldicott/
index_html
31. SI 2004/1031 (Medicines for Human Use [Clinical Trials]
33. Shaw, S. et al. 2005. Research governance: where did it come from,
Regulations). what does it mean? Journal of the Royal Society of Medicine, Vol.
32. Shaw, S. et al. 2005. Research governance: where did it come from, 98: 496.
what does it mean? Journal of the Royal Society of Medicine, Vol. 34. Korn, D. 2000. Conflicts of interest in biomedical research. Journal
98: 496, 498. of the American Medical Association. Vol. 284, no. 17, 2234.

2 6 6   •   P rinciples of G enomic M edicine

 PA RT I I

G E N O M I C S I N C L I N I C A L P R AC T I C E
 19.
GENETIC AND GENOMIC APPROACHES
TO CLINICAL MEDICINE
Dhavendra Kumar
A
ncient civilizationsin India, China, Greece,and
Egypt recognized that heredity played an impor-
tant part in human health. Writings, paintings,
and sculptures belonging to people from these regions,
and dating back several thousand years, provide evidence
that they knew of a close association between heredity and
health. Hippocrates in his medical texts noted that “hered-
ity affects health.”It is likely that this concept and its prac-
tical applications were recognized by other societies but
lacked documentary evidence. But there are some concrete
examples of this awareness, such as the reference in ancient
Jewish texts to a young boy given exemption from circumci-
sion because he had a male relative who died from bleeding
following ritual circumcision. This reflects the practice of
heredity medicine. However, scientific support for hered-
ity and its biological relevance was not clearly understood
until the mid–nineteenth century, when the “evolution of
species” theory proposed by Charles Darwin, and Gregor
Mendel’s seminal observations on cross-breeding various
forms of green pea plants, laid the foundations of the sci-
ence of genetics. Darwin’s theory elucidated the biological
importance of genetic variation, and Mendel put in place
the fundamental principles of heredity based on individual
hereditary factors that were eventually to be identified as
genes. Unfortunately, the medical applications of these
major discoveries remained unrecognized for almost half a
century.
Archibald Garrod, one of the pioneers of “genetics in
medicine” first began applying this knowledge to human
health at the start of the twentieth century. In his Harveian
Oration of 1924, he quoted from a letter written by William
Harvey, the Founder of Modern Medicine:
Nature is nowhere accustomed more openly to dis-
play her secret mysteries than in cases where she
shows traces of her workings apart from the beaten
path; nor is there any better way to advance the
269
proper practice of medicine than to give our minds
to the discovery of the usual law of Nature, by care-
ful investigation of cases of rarer forms of disease.
For it has been found, in almost all things, that
when they contain of useful or applicable is hardly
perceived unless we are deprived of them, or they
become deranged in some way.
Garrod, in his seminal paper on the patient with
“alkaptonuria,” provided the necessary evidence to sup-
port the “one gene—one enzyme” theory. This laid the
foundation for molecular medicine, and showed the way
forward for genetic medicine. Surprisingly, all this hap-
pened well before the elucidation of the chromosomal and
nucleic acid basis of inheritance. However, it was widely
agreed that the individual genetic complement was locked
in the nucleus.
The science of human genetics probably began with the
discovery of the full chromosomal complement of man by
Tjio and Evans in 1956 (see Chapters 1and 2).[1]‌ This led
to the rapid expansion of the whole field of human genet-
ics, and its divergence into various specialist fields such as
medical and clinical genetics. Developments in medical
genetics and subsequent applications in clinical practice
provided a strong base for wider applications of genetics in
medicine. Genetic medicine gained wide recognition over a
period of 40 years and led to the historical decision of 1996
to sequence the whole human genome.[2] This mammoth
task, called “The Human Genome Project,” was imme-
diately acknowledged as similar to, or even larger than,
the Manhattan Project to build the atomic bomb, or the
Apollo Mission for the trip to the Moon and back.[3] Some
enthusiasts regard this as the “genomic period,” followed
by the “post-genomic” phase. This is probably incorrect.
The genomic era has only just begun. The first glimpse of
the human genome was made possible with the publication
of the full sequence of the human mitochondrial genome.[4]
The technology to sequence the human genome was pro-
vided by the publication of the genome of the bacterium
Haemophilus influenzaeand other microbes.[5] This and
many subsequent achievements were acknowledged as “a
point of entry into genomics.”
Perhaps it is reasonable to identify the period prior
to the completion of the human genome project as the
pre-genome era. However, Dr.  Francis Collins, direc-
tor of the Human Genome Project and a keen enthusi-
ast for genomic medicine, dismissed the use of the term
“post-genomic era” while discussing the scope of pro-
teomics following the completion of sequencing the
human genome.[6]‌ He queries whether this means that
from the beginning of the universe until 2001 we were
in the “pre-genome era,” and then suddenly, “Bang!” we
moved into the post-genome era, leading one to won-
der:  what happened to the genome era? He suggested
that it was presumptuous to say that the Human Genome
Project is already behind us. He pointed out that pro-
teomics is a subset of genomics, and genomics is more
than sequencing genomes, which will be continuing for
decades to come. The most appropriate term would be the
post–Human Genome Project era, which can be referred to
as the post-HGP era. Whatever the argument, this is truly
the “genomic era.” Perhaps “post-Mendelian” seems more
appropriate as we move from an era in which genetics has
been rooted in monogenic diseases with high penetrance,
to a greater awareness (but limited understanding) of
polygenic diseases (and traits), often with relatively low
penetrance.
The “genomic era” holds phenomenal promise for
identifying the mechanistic basis of anatomical develop-
ment, metabolic processes, and disease. This is supported
by bioinformatics research, which will have a dramatic
impact on improving our understanding of such diverse
areas as the regulation of gene expression, protein struc-
ture determination, comparative evolution, and drug dis-
covery. The availability of virtually complete data sets also
makes negative data informative: when we can map entire
pathways, for example, it becomes interesting to ask, not
only what is present, but also what is absent.[7]‌ This meta-
phor can also refer to the increasing emphasis on func-
tional genomics. With an increasing number of organisms
for which we have (more or less) complete genomes, we
are just beginning to glimpse the potential power of fully
mapped sequences.
Perhaps we could delineate the various phases of genom-
ics more easily by examining the evolution of the scientific
methods that led to the development of different analytical
methods. In the past (pre-genomic era), both experiments
2 7 0   •   G enomics in C linical P ractice
and analytical methods were based on hypotheses. As
we enter into the genomic phase (present), the majority
of experiments are still hypothesis-based, but analysis is
now more systematic. It is envisaged that in the future
(post-genomic era), experiments will be more systematic,
leading to automatic analytical methods.
A N EW TAXO N O M Y F O R
HUM AN DISE ASE
Realizing the full beneficial potential of the human genome
sequence will ultimately depend on its applications in
clinical medicine.[8]‌ Many aspects of modern-day clinical
practice will change with technological advances and the
understanding of disease mechanisms, developments in
diagnosis, and new drugs and therapeutic procedures. These
advances have largely influenced our approach towards
human disease and have contributed to developing a new
taxonomy for human disease (see also Chapter 20).[9]
Genetic etiology in human disease—for example, con-
genital heart disease—is recognized and includes diseases
resulting from chromosomal abnormalities, mutations in
nuclear and mitochondrial DNA gene sequences, and inter-
action of environmental factors with several low-penetrance
genes carrying a small additive effect.[10] Advances in genomic
technology have now made it possible to unravel pathogenic
mechanisms in certain genetic disorders that result from
unusual genetic factors that do not comply with the tradi-
tional concept of genetic etiology. The pathogenic mecha-
nisms include a number of complex alterations distributed
across various genomic regions. These are now being termed
“genomic disorders.”[11] Distinct categories of genomic disor-
ders include epigenomic diseases, disorders of genome archi-
tecture, trinucleotide repeat disorders, and diseases associated
with complex genomic polymorphisms.[12] It is likely that in
the future, more groups of genomic diseases will come to
light. It is essential that the underlying pathogenic mecha-
nisms in these diseases be understood in order to facilitate
diagnosis and development of targeted specific therapy.
TOX I C O G E N O M I C S
Toxicogenomics is a scientific field that aims to study the com-
plex interaction between the whole genome, chemicals in
the environment, and disease. When the cells are exposed to
a stress, drug, or toxin, they respond by altering the pattern
of expression of genes within their chromosomes. Genes are
transcribed into messenger RNA (mRNA), the chemical mes-
sage by which information encoded in genes is translated into
 Free ebooks ==> www.Ebook777.com
proteins that serve a variety of cellular functions in response
to the exposure.[13] The production of protein encoded by a
given gene may be increased, decreased, or remain unchanged,
depending upon the type of toxic exposure and the cellular
requirements. In a general way, this could be directly attrib-
uted to post-transcriptional or post-translational effect, a
fundamental step in the peptide chain assembly (see also
Chapter 3).[14] Another important effect of a toxic exposure
could be abnormal chromatin structure. Abnormal chromatin
structure spread over successive generations could lead to epi-
genetic or epigenomic changes manifesting as developmental
or systemic disorders (see Chapter 4).[15]
A technology that is central to the field of toxicogenom-
ics is known as micro-arrays(http://www.ohsu.edu/croet/
research/centers/toxicogenomics/whatis.html), which
enable scientists to simultaneously monitor interactions
among thousands of genes within the genome. This tech-
nology (see also Chapter 10) will help define the complex
regulatory circuitry within a cell, tissue, or organ and give
scientists a global perspective on how an organism responds
to a stress, drug, or toxin. The data generated will provide
information about cellular networks of responding genes,
define important target molecules associated with the toxic-
ity mechanism, and provide biomarkers for epidemiological
studies. Ultimately, this information may allow us to iden-
tify ways to reduce or prevent disease by pinpointing bio-
chemical and molecular functions that have been perturbed
by environmental chemicals. DNA micro-array technology
will undoubtedly become a major tool in environmental
medicine, because it will improve our diagnostic and prog-
nostic capabilities for specific diseases, as well as our ability
to examine drug interactions, sensitivities, and effective-
ness. This technology will also aid research on alternative
model-testing procedures and support the development of
new toxicity screening processes.
It is envisioned that DNA micro-array technologies will
permit the design of experiments in the occupational and
environmental sciences[16] that will clarify whether:
•
Specific toxicants have unique gene-expression profiling
signatures;
• Different cells in different tissues have profoundly differ-
ent response signatures for a given toxicant;
•
Different species show similar, overlapping, or distinct
patterns of gene responses to a toxicant;
• A specific toxicant signature is altered depending upon
the stage in the developmental process or defined health
condition;
G enetic and G enomic A pproaches to C linical M edicine   •   2 7 1
www.Ebook777.com
• Responses to complex chemical mixtures can
be more easily elucidated and defined by their
gene-expression-profiling signatures;
• Responses to chronic low doses of toxicants
or environmental pollutants can be defined by
gene-expression-profiling;
• Specific gene polymorphisms can be defined that are
characteristic of an increased susceptibility to the pathol-
ogy of environmental health diseases.
Toxicologists and environmental health scientists have
studied the effects of the environment on human health for
several decades. Many adverse environmental effects have
been identified, and important progress has been made in
preventing exposure to harmful agents such as γ-radiation,
ultraviolet light, lead, pesticides, and dioxins. Toxicological
research has attempted to develop an efficient, cost-effective,
comprehensive strategy for predicting and preventing toxic
responses in humans. However, progress towards this goal
has been proportionate to the existing technologies and
level of scientific knowledge. The field of toxicology could
not have risen to this challenge if it had only had access to
the less efficient technologies of the past several decades.
One challenge is to use the human genome sequence
as a first step toward understanding the genetic and bio-
logical basis of complex biological traits and diseases such
as cancer, diabetes, Alzheimer’s disease, and Parkinson’s dis-
ease. Another challenge is to utilize the increased volume
of toxicological data to construct genetic and biochemical
pathways that will explain the mechanism of toxic responses.
Advances in combinatorial chemistry and molecular biology
have dramatically accelerated the rate of drug discovery and
availability, and the rate at which populations are exposed to
new drugs. Such advances intensify the burden of exposure
in the population, making it critical that we rapidly increase
our understanding of the consequences of such exposure.
The National Center for Toxicogenomics (NCT) of the
National Institutes of Health (NIH) in the United States
is leading the development of a unified strategy for toxi-
cogenomics studies and a public knowledge-base. This will
have an informatics infrastructure that will allow all part-
ners in this unprecedented enterprise to share equally in
its benefits and products. By providing a focus for techno-
logical coordination and basic research, a centralized public
knowledge-base, and a center for coordination for all the
partners in the pharmaceutical and chemical industries, the
NCT will facilitate this diverse national effort. The NCT
will not only achieve economies of time, cost, and effort,
but will help ensure the successful development of a broad
scientific consensus on the application of toxicogenomics to
the improvement of human health.
In brief, toxicogenomics combines the conventional
tools of toxicology (such as enzyme assays, clinical chem-
istry, pathology, and histopathology) with the new
approaches of transcriptomics, proteomics, metabolomics,
and bioinformatics.[17] This marriage of toxicology and
genomics has created not only opportunities, but also new
informatics challenges. This field is likely to be of major
importance in genomic medicine.
M ETA B O N O M I C S A N D
M ETA B O L O M I C S
Genomics measures the entire genetic makeup of an organ-
ism, while proteomics measures all the proteins expressed
under given conditions. Metabonomics, as the name implies,
is defined as measurement of the complete metabolic
response of an organism to an environmental stimulus or
genetic modification. Some people prefer the term metabo-
lomics, which refers to a holistic metabolic profile, to meta-
bonomics, which focuses at single-cell level.[18] Essentially,
there is no biological difference between these two concepts.
The -omics can provide information for basic biological
research and for pharmaceutical and clinical applications.
One of the challenges is integrating the information from
the various omics: in the process, yet another term is coined,
systeomics, which refers to the integration of genomics, pro-
teomics, metabolomics, and metabonomics.[19]
Metabolomics may be one of the most recently included
members of the omics family; however, it is probably the old-
est. In fact, it dates back to old-fashioned biochemistry, with
its emphasis on metabolism, the sum of the processes that
operate to acquire and use energy in an organism, to biosyn-
thesize cellular components, and to catabolize waste. Many
toxicological and disease diagnostics are based on metabolic
profiling. This methodology has been in existence for around
50 years, well before the advent of genomics or proteomics.
It is probably true that metabolomicsis “more closely related
to things in the clinical world” than either genomics or pro-
teomics, owing to the fact that metabolic signatures reflect
both genetic information and environmental influences.[3]‌
NU T R I G E N O M I C S
In the past decade, nutrition research has undergone an
important shift in focus from epidemiology and physiol-
ogy to molecular biology and genetics. Nutrigenomics is the
application of transcriptomics, proteomics, metabolomics/
metabonomics, and bioinformatics in nutrition research
2 7 2   •   G enomics in C linical P ractice
(see also Chapter  12).[20] The European Nutrigenomics
Organisation (NuGO) has recently taken over the ambi-
tious challenge to translate the nutrigenomics data into
an accurate prediction of the beneficial, or adverse, health
effects of dietary components. This organization and asso-
ciated agencies have set out to address important issues,
including nutrigenomics technology standardization and
innovation, bioinformatics environment harmonization,
andintegrated information-system development.
The integration of genomics and nutritional sciences
has led to the field of nutritional genomics. This provides
a very important base from which we can study the com-
plexity of genome responses to nutritional exposure while
offering opportunities to enhance our understanding of the
effectiveness of dietary interventions, at both individual
and population levels. Nutrients influence multiple physi-
ological responses that affect genome stability, imprinting,
expression, and viability. Nutritional genomics challenges us
to understand the complex interactions between the human
genome and dietary components in normal physiology and
pathophysiology.[21] An understanding of these interactions
will enable us to assess the benefits and risks of various dietary
recommendations, minimizing the risk of unintended con-
sequences. Furthermore, nutritional genomics will enable
the design of effective dietary regimens for the prevention
and management of complex chronic diseases. New perspec-
tives in the nutritional sciences in the light of advances in
genomics are reviewed elsewhere (see Chapter 12).
P H A R M AC O G E N O M I C S
The study of the role of genetic inheritance in individual
variation in drug response and toxicity is referred to as
pharmacogenetics. Convergence during the past decade of
advances in pharmacogenetics and human genomics has
led to the emergence of the field of pharmacogenomics.[22]
Pharmacogenomics is the study of the relationship between
the specific drug, DNA-sequence variation, and drug
response (see Chapter 7).
Pharmacogenomics holds great promise for the future of
medicine and is one of the major principles for the practice
of personalized medicine.[23] This relies on genomic biomark-
ers for disease susceptibility, including both Mendelian and
complex diseases. Applications of human genomics will
result in improved understanding of the pathophysiology
of disease, identification of new therapeutic targets, and
improved molecular classification of disease. The promise
of individualized therapeutic interventions largely depends
on the identification of drug toxicity genomic biomarkers
that will enable differentiation of individuals likely to show
both positive and negative therapeutic response.[24] This
would also help in assessing the efficacy of new drugs, and
their side effects and toxicity. A  comprehensive review of
the principles and applications of pharmacogenomics is dis-
cussed separately in this book (see also Chapter 7).
THE MOLECUL AR BASIS OF MEDICINE
Understanding the molecular basis of human disease
provides insight into pathogenesis and helps in design-
ing therapeutic interventions. The rapid developments in
genomics have strengthened the field of molecular medi-
cine. High-throughput genome sequencing and systematic
experimental approaches have helped in developing strate-
gic programs to investigate gene function at the cellular, bio-
chemical, and organism levels. Comprehensive functional
analysis of all genes and genome sequences falls within the
remit of functional genomics. This field holds great promise,
and heralds the beginning of the genomic era.[25]
Understanding the functional significance of genes and
genomic variants would require full knowledge of the exist-
ing human gene mutations and as well as all variants, par-
ticularly single-nucleotide polymorphisms [SNPs] (see also
Chapter  2). Efforts are being made to catalogue all known
human gene mutations (www.hgmd.cf.ac.uk) and conduct
haplotype analysis of all known SNPs. The recent comple-
tion of the human haplotype map (HapMap) has provided
a valuable resource in the study design for disease associa-
tion studies in common complex diseases, such as cancer,
coronary heart disease, diabetes, schizophrenia, and oth-
ers.[26] A  natural successor to the HapMap project is likely
to be the “functional-variant database,” which will probably
include all SNPs that alter amino acids in proteins, and possi-
bly gene-splicing or transcription.[27] This functional-variant
database would be made available to all researchers and would
be an essential resource in all future disease-gene studies.
Functional genomics is a systematic effort to understand
the function of genes and gene products by high-throughput
analysis of gene products (transcripts, proteins) and bio-
logical systems (cell, tissue, or organism), using automated
procedures that allow scaling up of experiments classically
performed for single genes, such as generation of mutants,
analysis of transcript, protein expression, protein structure,
and protein–protein interactions on a genome-wide basis.[28]
Functional genomics is based on two approaches: gene-driven
and phenotype-driven (Figure 19.1).
The gene-driven approach uses genomic techniques
for identifying, cloning, expressing, and characteriz-
ing genes at the molecular level. On the other hand, the
G enetic and G enomic A pproaches to C linical M edicine   •   2 7 3
phenotype-driven approach depends on analyzing pheno-
types from random mutation screens or naturally occurring
variants, such as mouse mutants or human disease, to iden-
tify and clone the gene(s) responsible for the phenotype,
without having prior knowledge of the underlying molecu-
lar mechanisms. Both approaches are highly complementary
at virtually all levels of analysis and assist in understanding
genotype–phenotype correlations. An important compo-
nent of functional genomics is comparative genomics, which
allows in vivo understanding of the molecular mechanisms
of various cellular processes.
The chapter on proteomics (Chapter  3) reviews basic
tools and routes of investigations commonly employed
by researchers in this major field. Collectively, functional
genomics and proteomics approaches provide a matrix of
information on gene products and their functional attri-
butes.[29] Research in this field will lead to novel findings
that will remove the current bias towards novel genes and
proteins and thus will be of particular importance in novel
genomics-based therapeutic approaches.
B I O I N F O R M AT I C S A N D G E N O M I C
MEDICINE
Bioinformatics is a rapidly emerging field of biomedi-
cal research.[30] This relatively new discipline develops and
applies informatics to the field of molecular biology. The
field is broad and includes scientific tools and methods for
sequence analysis (nucleotide and protein sequences), ren-
dering of secondary and tertiary structures for these mol-
ecules, and protein fold prediction that is crucial to targeted
drug design and development.[31] Bioinformatics opens the
way for a new approach in molecular medicine, referred to
as “phenomics.” Clinical (medical) informatics has long
been recognized as an important methodology in biomedi-
cal research and clinical care, integrating experimental and
clinical information systems. Both clinical/medical infor-
matics and bioinformatics will eventually change the current
practice of medicine, including diagnostics, therapeutics, and
prognostics.[32] In some ways, this process is similar to the
clinical applications of biochemistry that happened about
half a century ago. Post-genome informatics, equipped with
high-throughput technologies and genomic-based databases,
is likely to transform biomedical understanding. Some of the
key applications of genome-based bioinformatics are multi-
variate data projection, gene-metabolic pathway mapping,
automated biomolecular annotation, text mining of factual
and literature databases, and the integrated management of
biomolecular databases (see also Chapter 6).
 Gene-driven analysis: Comparative genomics: Identification of genes with
from genes to function homologous functions

Identifying geneticists variants

TT Genome
TT associated with a phenotype
Gene cataloguing

Transcript populations expressed in each tissue

Transcriptome
Spatial and temporal gene expression patterns Genotyping

2D gels Analyzing molecular phenotypes

Proteomics/structural genomics
Global RNA/protein profilling
(ICAT)
Identifying/localizing proteins
Resolving protein structures
Protein-protein interaction maps
Classifying phenotypes
Protein ligands
Anatomic, metabolic, physiological, behavioral features

Modulating gene or protein activity in vitro/in vivo

Undestanding gene function
Human
Pathways identification - networks simulation
pathologies
Links to metabolome
Producing large mutant collections in model organisms Monogenic and
complex diseases

Phenotype-driven analysis: from organism trats to genes

TRENDS in Molecular Medicine

Figure 19.1 
Two complementary approaches in functional genomics: the gene-driven and the phenotype-driven, artificially separated along the
diagonal axis. Different levels of information are interconnected with main tools of analysis and routes of investigations. Adapted, with permission, from Trends
in Molecular Medicine.[28]

S T R AT I F I E D M E D I C I N E Several programs and incentives are now operational for

Stratified medicine is the grouping of patients based on
their risk of disease or response to therapy by using diag-
nostic tests or techniques.[33] Patients and healthcare
providers both benefit from more targeted and effective
treatments, whereas industry benefits from the potential
for more efficient therapeutic development as well as the
market expansion for these new treatments. The develop-
ment of stratified medicine is being pursued globally as
its benefits are increasingly recognized. The concept and
philosophy behind stratified medicine are not unfamiliar.
However, this approach is now remarkably strengthened
with the increasing accuracy and sophistication of the
genomic and molecular medicine.[34] Stratified approaches
to therapy are expected to become the standard for the
management of a whole range of diseases (for example,
chronic heart failure),provided that these match certain
criteria as recommended by leading clinicians and scien-
tists.[35] The Academy of Medical Sciences in the United
Kingdom (www.acmedsci.ac.uk) has recommended crite-
ria for Stratified Medicine (see Table 19.1).[36]
tratified medicine to enable partnership across academia,
industry, healthcare systems, regulatory/pricing authorities,
research funders, and patient groups. The progress towards
Table 19.1  CRITERIA FOR STRATIFIED MEDICINE
(ADOPTED WITH PERMISSION FROM THE ACADEMY
OF MEDICAL SCIENCES, UK [36])
1. C
 ontinued research to understand the genetic and molecular
bases of diseases.
2. D
 evelopment and use of increasingly sophisticated and powerful
informatics technology.
3. I mprovement and standardization of clinical data collection and
linkage with genomic and other databases.
4. I ncreased collection of tissues for biomarker research and evalua-
tion, and its organization in national and international biobanks.
5. G
 reater efficiency and productivity in the development of thera-
peutics and diagnostics.
6. Th
 e introduction of flexible and novel approaches for the regula-
tory assessments of innovative stratified medicine products.
7. I mproved flexibility in pricing for stratified medicine products—
both for the diagnostic and for the associated therapy—to ensure
cost-effectiveness for payers while encouraging innovation.

2 7 4   •   G enomics in C linical P ractice

stratified medicine, increasingly confused with “personalized whole-genome-sequence data, which will require that
medicine”(see also the section on personalized medicine in privacy and data protection concerns be addressed.
this chapter), relies fundamentally upon data, which are cen-
• Because of the complexity, capital expense of equipment,
tral to the applied and translational research to understand
and size of datasets, progress in molecular medicine is
the molecular basis of disease; the development of targeted
increasingly requiring collaboration between many aca-
interventions; effective regulation, health technology assess-
demic groups, public institutions, and industry, often
ment, and valuation of stratified medicine products; and the
internationally.
stratification of treatment by physicians.[37]
Among many examples of stratified approaches in plan- • Genomic information on its own, although useful, is
ning and executing treatment for common cancers, the case only part of the story. Greater knowledge is gained when
for non–small cell lung cancer (NSCLC) is noteworthy, and such genetic information is linked to clinical outcomes.
probably the best paradigm in the context of stratified medi- Thus there remains a major hurdle of linking genome
cine (Table 19.2; Figure 19.2). It has been known for some databases to healthcare records, which need to be elec-
time that mutations in KRAS were associated with squamous tronic for this to be done efficiently.
cell lung cancer. Further research indicated that mutations in
• Research is still required so that genetic variations are
the epidermal growth factor receptor gene (EGFR) could be
not only correlated to diseases, but causal links are
used in targeting the treatment, notably that of exons 19 and
established, if the underlying molecular mechanisms of
20.[39] Similarly, the EML4–ALK mutation can be used as an
disease are to be understood.
example of how molecular understanding accompanied by
targeted medicines has transformed the treatment of patients • Correlation of genetic variation and disease may
with NSCLC.[40] In 2007, research demonstrated that approx- sometimes not transcend ethnic groups. The
imately 5% of NSCLC cases involved this mutation. Within Pharmacogenetics for Every Nation initiative has been
three years, targeted therapies were developed, and demon- set up to address this issue.
strated dramatic efficacy; now patients with lung cancer can
• The effect of epigenetic variations on drug response,
have biopsy tissue sent for genetic analysis to ascertain their
pharmacoepigenomics, needs further research.[43]
suitability for this treatment, and receive an accurate, geneti-
Epigenetic variations are inheritable, and affect gene
cally derived diagnosis in 7 to 10  days.[41] These develop-
expression levels and therefore phenotype, yet they do
ments (Table 19.3) have transformed therapy for the 5% with
not result from changes in the DNA sequence.[44]
NSCLC driven by the EML4—ALK mutation, meaning
simply taking two capsules per day causes the cancer to shrink
or disappear for more than half of all people treated, rather T H E F U T U R E O F S T R AT I FI E D M E D I C I N E
than for one in every ten as was the case with traditional che-
There are multiple factors that will determine the develop-
motherapy.[42] Although this dramatic response is not always
ment and adoption of stratified approaches to medicine.
sustained over time, it is highly beneficial to patients.
There are “pull” factors, in that the healthcare system needs to
become increasingly effective and sustainable, in particular in
its economic policies for investment and cost-reimbursement.
C H A L L E N G E S F O R S T R AT I FI E D M E D I C I N E
There are also “push” factors, from recent advances in medi-
There are several challenges and obstacles that must be cal science and informatics, and the pharmaceutical industry’s
surmounted to realize the full potential of benefits of the requiring substantial improvements in research and devel-
substantial progress in genomic and molecular research in opment productivity to remain a viable sector in the long
pursuit of stratified approaches to clinical medicine. The term.[45] These factors accelerate the momentum of stratified
Academy of Medical Sciences recommendations (www.​ medicine and will be transformative in the provision of care.
acmedsci.ac.uk) include the following goals: Detailed discussion of this aspect of stratified medicine is
beyond the scope and remit of this chapter. However, the fol-
• Standardization of genome-sequencing platforms to lowing major areas are important[36] to consider for planners
avoid laboratory-to-laboratory variability that compli- and developers of stratified medicine:
cates the analysis of combined datasets.
• Effective and sustainable healthcare systems
• High levels of enrollment for sequencing are

required to benefit from the accumulation of • Scientific and technological advances

G enetic and G enomic A pproaches to C linical M edicine   •   2 7 5

Table 19.2  EXAMPLES OF CASE STUDIES FOR MODELING STRATIFIED MEDICINE (ADOPTED WITH PERMISSION FROM THE ACADEMY OF
MEDICAL SCIENCES, UK) [38]

DISEASE AREA DRUG (FTX) AND COMPANIONDIAGNOSTIC (CDX) US APPROVAL EU APPROVAL

RX CDX RX CDX RX CDX
Breastcancer Herceptin (trastuzumab) Hercep Test Dako Sep 1998 Sep 1998 Aug 2000 Yes
Roche/Genentech
Herceptin targets the HER2 protein, present on cell surfaces. In some cancers, HER2 overproduction causes the uncontrollable cell growth, driving the disease.
HercepTest identifies if an individual’s breast cancer involves HER2 overproduction: if so, they will respond to Herceptin. The HER2 marker was found during drug
development. This was the first simultaneous approval of Rx and CDx. The product received subsequent approval for use in HER2-positive gastric cancer.
HIV Ziagen (abacavir) HLA-B*57:01 screening Dec 1998 N/A:unbranded test Jul 1999 N/A:unbranded
GSK/ViiV Healthcare assay test
Action of HIV’s reverse transcriptase enzyme is critical to the replication of the virus. Abacavir is a nucleoside reverse transcriptase inhibitor (NRTI) with activityagainst
Human Immunodeficiency Virus Type 1 (HIV-1). Serious and sometimes fatal hypersensitivity reactions have been associated with abacavir and abacavir-containing
products. Extensive research established that patients who carry the HLA-B*5701 allele are at a high risk for experiencing a hypersensitivity reaction to abacavir.
Breastcancer N/A: Dx only Oncotype DX N/A:Dx only Not FDA-approved: N/A:Dx only 2007
Genomic Health usesupported by literature
Developed through retrospective studies on tissue archives, Oncotype Dx is a diagnostic tool that predicts the likelihood of breast cancer recurrence and the benefit of
chemotherapy in about 60% of breast cancer cases. The test is now included in major treatment guidelines for breast cancer in the United States, and receives a value-based
reimbursement, which is based on clinical data demonstrating the test’s ability to restrict healthcare costs.
Disease area Drug (Rx) and Companiondiagnostic (CDx) US approval EU approval
Rx CDx Rx CDx Rx CDx
Colorectal cancer Vectibix panitumumab) EGFR pharmDx kit- Sep 2006 Sep 2006*Jul 2012 Sep 2007 Yes, Yes
Amgen Dako therascreen®:
KRAS RGQ PCR kit
Qiagen
Vectibix was designed to treat colorectal cancers overproducing a protein called EGFR. After going to market, it was found that EGFR overproduction does not indicate
response to the Rx, and that individuals with this marker would not respond to therapy if they also carried a mutation in another protein, KFtAS. KRAS is now estab-
lished as a stratifying marker, and a marker for the safety of using Vectibix in combination with a certain type of chemotherapy.
Melanoma Zelboraf (vemurafenib) cobas® 4800 BRAF V600 Aug 2011 Aug 2011 Feb 2012 Yes
Roche/Plexxikon Mutation Test Roche
This drug was selected by Roche for development owing to knowledge of the biomarker: the drug showed effects in melanomas containing a particular mutation, V600E,
in a protein called BRAF. The Rx and CDx were developed in parallel, and co-approved in one of the fastest FDA approvals in history (four months). Zelboraf was
approved by NICE (National Institute of Clinical Excellence) in November 2012.
Non–small cell Xalkori (crizotinib) Pfizer Vysis ALK Break Apart Aug 2011 Aug 2011 Jul 2012** Sep 2011
lung cancer FISH probe kitAbbott
(NSCLC) Molecular Diagnostics
A 2007 study linked a subset of NSCLC to the ALK fusion gene. This prompted a partnership between Rx and CDx manufacturers, and patient stratification using this
CDx resulted in dramatic improvement in response rates. Approval was rapid both in the US and in the EU.
Cystic fibrosis Kalydeco (ivacaftor) GSSI.D Jan 2012 N/A:unbranded test Jul 2012 N/A:unbranded test
Vertex/Cystic Fibrosis mutation test
Foundation Therapeutics
Inc.
One of the first treatments to target the underlying cause of cystic fibrosis, Kalydeco was developed based on gene and protein data from sufferers of the disease.The abil-
ity to test for specific cystic fibrosis mutations was critical both during development and for post-approval use, yet a specific brand of test is not specified on the label.
Melanoma BRAF/MEK inhibitor BRAFTm mutation kit In development
(dabrafeniband trametinib) (v600E & K) bioMerieux
GSK
This Ftx-CDx combination is currently under development. The BRAF V600 mutations are present in approximately 50% of melanomas. Separately, the Rx showed posi-
tive results up to phase 3 trials. As a combination, they have shown promising resultsat phase 2, and are now at phase 3. GSK (Glaxo-Smith-Kline) has been collaborating
with bioMerieux to develop the CDx, which is being used to identify patients BRAF V600 status in the current phase 3 trials.
 Traditional view 1987

Adenocarcinoma Squamous KRAS

Unknown

Large-cell

2004 2009

KRAS KRAS
Unknown
Unknown EGFR EGFR

EML4ALK
HER2
BRAF
MET
AKT1
MAP2K1
PI3KCA
Mutations associated with drug sensitivity
EGR Gly719X, exon 19 deletion, Leu858Arg, Leu861Gln
Mutations associated with primary drug resistance
EGR exon 20 insertions
Mutations associated with acquired drug resistance
EGR Thr790met, Asp761Tyr, Leu747Ser, Thr854Ala

Figure 19.2  Stratified approaches in planning and executing treatment for common cancers: the case for non–small cell lung cancer (NSCLC).[36]

• Diagnostic applications to accommodate new disease
categories
• Challenges facing the pharmaceutical industry
• Role of the regulatory and statutory agencies
EXAMPLES OF GENOMIC AND
M O L E C U L A R A P P R OAC H E S
I N FEC T I O US D I S E A S E S
Genetic factors have been recognized to influence infec-
tious disease susceptibility, resistance, and response to
antimicrobial therapy. Numerous studies, including
twin studies, support the importance of host–genotype
interactions in wide-ranging infectious-disease clinical
phenotypes, population differences in susceptibility or
resistance, and favorable or adverse reactions to antimi-
crobial therapy. Population studies have also contributed
evidence supporting the selective advantage of human
evolution and population genetic structure. For instance,
a high frequency of heterozygotes for sickle-cell anemia
and thalassemias in some populations confers a selective
advantage for malaria in the face of deleterious effects in
homozygotes.[46] Interplay between the Duffy locus muta-
tions and sickle-cell anemia in certain populations is rec-
ognized to confer malaria resistance. The molecular basis
of this phenomenon lies in the erythrocyte chemokine

2 7 8   •   G enomics in C linical P ractice

Table 19.3  STRATIFIED MEDICINE IS CURRENT MEDICAL PRACTICE (ADOPTED WITH PERMISSION FROM THE ACADEMY OF MEDICAL SCIENCES,
UK) [36]

CASE STUDY DRUG (FTX) COMPANION DIAGNOSTIC US APPROVAL EU APPROVAL

(CDX) FTX CDX FTX CDX
1 Breast cancer Herceptin (trastuzumab) Hercep Test Dako Sep 1998 Sep 1998 Aug 2000 Yes
Roche/Genentech
2 HIV Ziagen (abacavir)GSK/ViiV HLA-B*57:01 screening assay Dec 1998 N/A:unbranded test Jul 1999 N/A:unbranded test
Healthcare
3 Breast cancer N/A:Dx only Oncotype DXGenomic Health N/A:Dx only Not N/A:Dx only 2007
FDAapproved:usesupported
byliterature
4 Colorectal cancer Vectibix(panitumumab)Amgen EGFR pharmDx Sep 2006 Sep 2006 Sep 2007 Yes,Yes
kitDakotherascreen®:KFtAS July 2012
RGQ PCR Kit
Qiagen
5 Melanoma Zelboraf(vemurafenib) cobas® 4800 BRAF Aug 2011 Aug 2011 Feb 2012 Yes
Roche/Plexxikon V600Mutation
Test Roche
6 Non–small cell- Xalkori(crizotinib) Vysis ALK Break Apart Aug 2011 Aug 2011 Jul 2012 Sep 2011
lung cancer Pfizer FISHprobe kitAbbott Molecular (Conditional
Diagnostics Marketing
Authorization)
7 Cystic fibrosis Kalydeco(ivacaftor)VertexPh G551D mutation test Jan 2012 N/A:unbrandedtest Jul 2012 N/A:unbrandedtest
armaceuticals/CysticFibrosis
FoundationTherapeutics Inc.
8 Melanoma BFtAF/MEK BRAFTM mutation kit In development
inhibitor(trametinib (v600E & K)bioMerieux
anddabrafenib)GSK
receptor, which also binds the malarial parasite.[47] Similar
observations have been made for another chemokine
receptor, CCR5 (CKR5),which offers resistance to HIV/
AIDS.[48] This has led to an interesting speculation that
this mutation emerged as a selective force from the plague
(Y. pestis), which affected the northeastern European pop-
ulations.[49] Similar observations have been made by scien-
tists studying host–pathogen interactions in determining
infectious disease susceptibility.
Many genome projects of both human and pathogen
genomes have provided an insight into microbial-ecogenetic
relationships. Several prominent examples of host–patho-
gen interactions include malaria (Plasmodium falci-
parum, Plasmodium vivax),tuberculosis (Mycobacterium
tuberculosis),AIDS (human auto-immune deficiency virus),
cholera (Vibrio cholerae),and meningitis-otitis (Haemophilus
influenzae).These studies have helped in designing targeted
antimicrobial therapy.[50]
Genomics has accelerated insights in microorgan-
isms, including genome architecture, sequence simi-
larities, mobile genetic elements, and large numbers of
genes of previously unknown function. This information
is vital in developing antimicrobial therapy and a new
class of DNA-based vaccines. An illustrative example is
H.  influenzae, which was the first organism to be fully
sequenced.[51] H. influenzaehas been immensely useful in
research in microbial genomics due to its small size (1830
kb), its importance as a major human pathogen, its capac-
ity for DNA transformation as seen in the mouse model,
and the rapid advance in knowledge of its genome. There
are 1703 proposed genes, of which 736 lack proposed
functions; of these, 347 are conserved across species, while
389 are unique to H. influenzae. These unique genes are
now the target for developing selective therapeutic agents
and vaccines (see also Chapter 11).
Similar approaches are currently being applied to many
pathogens, including the flu virus and HIV. Other notable
organisms targeted for this work include Mycobacterium
tuberculosis, E. coli strain O157:H7, V. cholerae, Helicobacter
pylori, and Yersinia pestis. The whole field of microbial
genomics is promising and lies in the core of the sphere of
genomic medicine (see Chapter 37).
C R IT I C A L C A R E M E D I C I N E
Advances in genomics increasingly affect all areas of clini-
cal medicine, including critical care medicine. Survival
after acute trauma and sepsis is now common, thanks to the
development of improved trauma systems, advanced resus-
citation methods, and organ support systems. However,
2 8 0   •   G enomics in C linical P ractice
survival is often complicated by nosocomial infection and
organ failure. Technological advances in genomics and pro-
teomics, together with techniques of bioinformatics, pro-
vide an opportunity to characterize the determinants of,
and the responses to, injury and sepsis on a genome-wide
scale. This includes large-scale collaborative efforts aimed
to investigate genomic variation (polymorphisms), and
characterize multiple levels of biological response (tran-
scriptome and proteome) to injury and infection, and relate
these to clinical situations.
Applications of in-depth genome-wide analysis can
allow a thorough understanding of disease processes that
are relevant in intensive care, such as acute trauma, sepsis,
acute respiratory distress syndrome, and multiple organ
dysfunction syndrome.[52] Understanding critical illness at
the genomic level may allow more effective stratification of
patient subclasses and targeted, patient-specific therapy.[53]
The related fields of pharmacogenomics and pharmaco-
genetics hold the promise of improved drug development
and the tailoring of drug therapy based on an individual’s
drug metabolism profile. The “genotyping” of critically ill
patients will allow us to ascertain individual cytoprotective
mechanisms that are crucial to organ and tissue protec-
tion in these patients.[54] It is important that in future all
physicians caring for critically ill patients be familiar with
advances in genomic technologies and applications in clini-
cal medicine. Developments in genomics and related fields
relevant to the care of critically ill patients are discussed
separately in this volume (see also Chapter 48).
C A R D I OVA S C U L A R D I S E A S E S ,
I N C LU D I N G S U D D E N D E AT H
Sudden death is a major public health concern. It has inevi-
table social, personal, and economic consequences. The
extent of personal grief and long-term psychological effects
associated with sudden death are impossible to assess. This
is particularly true in relation to the unexplained death of an
infant (“cot death”) or a young person, which could be due
to anas-yet-unknown genetic, metabolic, or cardiac disease.
Sudden death in an adult is often due to an underlying car-
diac disease, referred to as “sudden cardiac death,” or SCD.
Apart from established Mendelian disorders (Marfan
syndrome, hypertrophic cardiomyopathy, long QT syn-
drome, etc.), SCD is commonly used to refer to coronary
heart disease and heart failure. Advances in genomic sci-
ence applicable to sudden death with particular reference
to SCD are reviewed in Chapter 21. The chapter discusses
novel bioinformatics approaches in identifying candi-
date genes/pathways and their functional significance.
This chapter also discusses the possibility of applying
high-density genome-wide SNP analysis in organizing
community-based screening for genetic susceptibility to
common heart diseases.
D I A B ET E S M E L L IT US A N D R E L AT E D
M ETA B O L I C D I S E A S E S
Obesity is endemic in the developed world and is rapidly
reaching epidemic proportions in the developing world.
Obesity is associated with hypertension, coronary heart
disease, and type 2 diabetes mellitus(T2DM). Only a small
proportion of obesity is related to genetically determined
causes; in such cases, it is often accompanied by involvement
of other body-systems. In the majority of cases, obesity is
related to environmental factors. However, the severity and
clinical outcome are modulated by multiple genetic factors
consistent with polygenic/multifactorial inheritance.[55] An
association of obesity with type 2 diabetes mellitus is well
recognized. In addition, obesity increases the risk of hyper-
tension and coronary heart disease.
In contrast to the immunologically determined
insulin-dependent type 1 diabetes mellitus(IDDM;
T1DM), type 2 non-insulin DM(NIDDM; T2DM) is
genetically heterogeneous. Several candidate genes are impli-
cated in the pathogenesis of T2DM. A genomic approach is
essential in analyzing an individual’s risk for T2DM.[56]
Obesity, hypertension, and T2DM are good examples
where specifically designed microarrays could eventually
be very effective in the screening of “at-risk” individuals
and in the identification of those who might benefit from
appropriate lifestyle changes and prophylactic pharmaco-
logical interventions.[57] Chapter 22 reviews the genetic and
genomic aspects of diabetes mellitus and related metabolic
diseases.
B RO N C H I A L A S T H M A A N D C H RO N I C
O B S T RU C T I VE LU N G D I S E A S E S
Asthma is a complex genetic disorder with a heterogeneous
phenotype resulting from interactions among many genes
and the environment. Numerous loci and candidate genes
have been reported to show linkage and association with
asthma and the asthma-associated phenotypes.[58] These
include microsatellite markers and single-nucleotide poly-
morphisms associated with specific cytokine/chemokine
and immunoglobulin E (IgE) regulating genes. Although
significant progress has been made in the field of asthma
genetics in the past decade, the clinical implications of the
genetic variations within the numerous candidate asthma
G enetic and G enomic A pproaches to C linical M edicine   •   2 8 1
genes remain largely undetermined. However, new infor-
mation has recently emerged from post-genomic research
with the cloning of new asthma genes, such as ADAM33
and PHF11. Chapter  30 highlights recent developments
in genetic, genomic and proteomic research in asthma and
related respiratory diseases.
C H RO N I C I N FL A M M ATO RY D I S O R D E R S
The complexity of the immune system is related to gene
expression in the tissues, cells, and biological systems. An
analysis of selected gene systems, using the high-throughput
whole-genome-screening approach, has given us insight
into these complex systems. Development of sophisticated
methodologies, such as microarray technology, allows an
open-ended survey to identify comprehensively the fraction
of genes that are differentially expressed between different
biological samples. New developments in genomics have
helped us improve our understanding of basic and applied
aspects of immunologically determined disorders. For
instance, improved understanding of the molecular basis of
inflammatory bowel disease (IBD) has enabled the develop-
ment of new therapeutic agents (see Chapter 28).
Rheumatic disorders comprise several heterogeneous
diseases that impose a heavy burden on health care services
because of the associated significant long-term morbidity
and disability. A  small number of these conditions result
from single-gene connective tissue diseases; for example,
Marfan syndrome, Ehlers-Danlos syndrome, and uncom-
mon inherited metabolic diseases. However, the etiology
and pathogenesis in a large number of these diseases are
poorly understood, except for some evidence of autoim-
mune pathogenesis supporting multifactorial/polygenic
inheritance. The post-genomic advances are enhancing our
understanding of the pathogenesis in several rheumatic
diseases, such as rheumatoid arthritis and osteoarthritis.
Chapter  29covers all these developments and highlights
future methods for diagnosing rheumatic diseases and pos-
sible therapeutic interventions.
N EU RO -P S YC H I AT R I C D I S E A S E S
Although rapid progress has been made in mapping and
characterizing genes for several monogenic neurological dis-
eases, the pathogenesis in a large number of neuro-psychiatric
disorders remains unexplained. However, there is evidence,
albeit limited, that genetic factors play a significant role in the
causation of these disorders, interacting with environmental
factors. The list of these conditions is long and includes sei-
zure disorders (Chapter 31), multiple sclerosis (Chapter 32),
Parkinson’s disease, Alzheimer’s dementia (Chapter  33),
schizophrenia and bipolar disorders (Chapter 34) and autism
and learning disorders (Chapter 35).
C O M MO N C A N C E R S
Cancer genetics is a relatively new but rapidly developing
field that has acquired a prominent place in clinical genetics.
However, it has largely focused on uncommon developmen-
tal malformation syndromes with malignancy, familial breast,
ovarian, and colorectal cancers, and some other uncommon
Mendelian familial cancer syndromes. Major clinical genetic
centers are now equipped with laboratory facilities offering
diagnostic and pre symptomatic genetic testing in selected
conditions and situations. There are well-established clini-
cal protocols dealing with clinical referrals, risk assessment,
genetic counseling, genetic testing, early detection of cancer
(screening) in family members at increased lifetime risk, and
including provision of follow-up and long-term support.
However, there are no clinically validated protocols deal-
ing with isolated common cancers, such as cancers of the
lung, skin, breast, bowel, and prostate. The etiology in these
malignancies is not clearly known, probably following multi-
factorial pattern, with multiple genes or polymorphisms con-
ferring a genetic predisposition.[59]
Completion of the human genome sequence and dissem-
ination of high-throughput technology will provide oppor-
tunities for systematic analysis of cancer cells. Genome-wide
mutation screens, high-resolution analysis of chromosomal
aberrations, and expression profiling all give comprehen-
sive views of genetic alterations in cancer cells. These analy-
ses will facilitate the compilation of a complete list of the
genetic changes causing malignant transformation and of the
therapeutic targets that may be exploited for clinical benefit.
It has been suggested that utilization of single-nucleotide
polymorphisms(SNPs) will aid in identifying individuals
at high risk of developing certain cancers, and will also help
researchers develop tailored medication or identify genetic
profiles of specific drug action and toxicity. This is facilitated
by introduction of new concepts, such as epigenomics, in
developing targeted therapeutic tools (Chapter 4). The sig-
nificance and challenges of genomics-based technologies in
the diagnosis and treatment of cancer(see Chapter 36).
C L I N I C A L H E M ATO L O GY
Genetic blood diseases make up a significant part of the
workload of a busy clinical hematology service. These
mainly include single-gene diseases such as hemophilia
A, Christmas disease, von Willibrand disease, sickle-cell
2 8 2   •   G enomics in C linical P ractice
disease, and other hemoglobinopathies. Genetic factors
play a significant role in the causation of several other
hematological disorders, including auto-immune hemo-
lytic anemias, platelet disorders, and complex thrombosis
and bleeding disorders. Genetic factors play a crucial role
in both the etiology and the therapeutic outcome of various
kinds of hematological malignancies. Advances in genomics
and applications of genomics-based technology have made
promising contributions to the development of powerful
diagnostic and therapeutic tools for dealing with complex
hematological diseases; for example, deep-vein thrombo-
sis (DVT), disseminated intravascular coagulation (DIC),
auto-immune thrombocytopenias, auto-immune hemolytic
anemias, and hematological malignancies.
Some advances in genetics are currently in use in clini-
cal hematology. For example, screening for Factor V Leiden
heterozygous status can help identify persons at risk for
thrombophilia, which can clinically manifest with poten-
tially life-threatening complications of DVT and pulmo-
nary embolism.[60] Approximately 4% of the population
could be heterozygous (carriers) for this mutation. Similarly,
confirmation of the homozygous status for the gene encod-
ing thiopurine-S methyltransferase, an enzyme that inacti-
vates the chemotherapeutic drug mercaptopurine, can help
clinicians in selecting an alternative therapy or reducing the
maintenance dose for children suffering from acute lym-
phoblastic leukemia. It is well known that about 1 in 300
children develops a serious, sometimes lethal, adverse reac-
tion to mercaptopurine therapy.
Chapters on hemostasis and thrombosis (Chapter 25),
inherited hemoglobin disorders (Chapter 26) and hemato-
logical malignancies (Chapter 27) provide comprehensive
reviews of genomics-associated developments and clinical
applications.
C L I N I C A L P E D I AT R I C S
Advances in microarray technology have made a significant
contribution to the diagnosis and management of children
with developmental disorders (see Chapter39). Examples
of potential uses of this technology in clinical pediatrics
include disease classification, risk stratification, pathogen
detection, pathogen subtyping, antibiotic-resistance analy-
sis, newborn screening, and prediction of drug responses
and adverse reactions.[61]
The technique of comparative genomic hybridization
(CGH) has now made it possible to carry out an in-depth
genomic analysis in a child with unexplained develop-
mental and physical disability, often called “unknown
dysmorphic syndrome.”[62] Applications of microarrays
in pediatric oncology are being used for specific tumor
sub-typing, which yields prognostic information and helps
doctors plan therapy.[63] For instance, gene-expression
analysis in medulloblastoma can help clinicians distinguish
between various histologically distinct tumor subtypes. This
can be aid in advising prognosis and the outcome of therapy.
The identification of genes that are differentially expressed
in patients with poor prognoses might help in developing
more effective therapies for these children.
Microarrays are being used for sensitive detection of
pathogens without the need for cell culture. Such a method
is of particular use in the rapid and reliable diagnosis of
Mycobacterium tuberculosis. This can allow institution
of  prompt anti-tuberculous therapy.[64] Various subtypes
of mycobacterium species can be reliably detected that can
be useful in making an appropriate choice of antimicrobial
drug and chemotherapy. This approach has been used in
detecting rifampicin resistance, a major problem in myco-
bacterial therapy. In addition, the outcome of therapy can be
accurately assessed in critically ill children with tuberculous
meningitis. A 32-fold amplification of pathogen sequences
of E. coli O157:H7 strain, using polymerase chain reaction
(PCR), could allow the rapid and sensitive identification of
pathogens and antibiotic-resistance genes so that appropri-
ate therapy can be more quickly instituted.[65]
Genetic analyses are an obvious application of microar-
ray technology. The simultaneous identification of specific
mutations in multiple genes is now possible. This technique
has tremendous potential in the care of a sick newborn. Such
microarrays may be useful for newborn screening, covering
several genes. For instance, a single blood sample from a baby
with neonatal cholestasis might be used to detect mutations
present in tyrosinemia, galactosemia, the various forms of
familial intrahepatic cholestasis, α-1 antitrypsin deficiency,
cystic fibrosis, and others.[66] A similar technique can be used
in determining the risk of polygenic diseases. This employs
preparing the whole-genome profile of single-nucleotide
polymorphisms (SNPs). This technique could also be used
in assessing an individual’s response to medication and in
the selection of the most efficacious drugs with the least
risk of adverse reaction for a given patient and disorder. The
application of genomic techniques in clinical pediatrics has
begun and is now poised for expansion beyond the research
setting into clinical care.
OBSTETRICS AND GYNECOLOGY,
INCLUDING REPRODUCTIVE MEDICINE
Developments in genomic medicine have far-reaching
implications for all aspects of clinical medicine. Numerous
G enetic and G enomic A pproaches to C linical M edicine   •   2 8 3
genomic studies concerned with clinical obstetrics and
gynecology have expanded the profile of genomic medi-
cine.[67,68] The impact of genomics-related research is evi-
dent in the way the practice of reproductive medicine has
rapidly changed during the last few years. Genomic micro-
array technology is being used to improve our understand-
ing of different aspects of reproductive medicine, including
physiological processes, disease diagnosis, and drug devel-
opment (see Chapter 46).
Gene-expression microarray studies on endometrial
receptivity and implantation in both mouse and human
have increased our understanding of the physiology of
implantation.[69] Such studies have been helpful in deter-
mining the causes of, and treating, implantation failure.
Similar techniques have been applied in studying endome-
trial decidualization, ovarian follicle development, labor,
and normal placentation.[67]
Genomic microarrays can be useful in studying specific
gynecological diseases. For example, the role of microar-
rays in ovarian cancer has been the subject of a number of
reports describing therapeutic targets and diagnostic mark-
ers.[70] Similarly, both endometrial and cervical cancers
have been studied using microarrays. A  specific subtype
of cervical cancer with resistance to radiotherapy is attrib-
uted to genes conferring such resistance. Genomic studies
in other obstetrical and gynecological conditions include
pre-eclampsia, trisomy-21 pregnancies, and endometriosis.
Gene-expression studies in endometriosis sufferers have
been helpful in improving our understanding of the causes
of implantation failure using eutopic and ectopic endome-
trium.[71,72] Similar approaches have helped us understand
the pathophysiology of fibroid growth by studying differ-
ential gene expression in fibroid and adjacent normal myo-
metrium.[73] Finally, the availability of microarray chips for
infectious diseases will enable prompt detection of patho-
genic organisms in the reproductive tract that account for
a significant proportion of secondary reproductive failure.
In the future, the practice of obstetrics and gynecol-
ogy could dramatically change with the availability of
genomic profiling using SNPs. This would enable specific
drug development targeted at an individual’s genomic
profile, thus avoiding the risk of serious iatrogenic effects;
for example, ovarian hyper stimulation syndrome. Each
patient attending an antenatal clinic in the future may be
offered DNA testing using the buccal swab. This might
enable the attending clinician to predict the likelihood
of a variety of pregnancy-related disorders, including
pre-eclampsia and preterm premature rupture of mem-
branes (PPROM). The DNA sample could be analyzed
on a microarray chip, designed to identify a number of
genetic polymorphisms known to confer an increased risk cloning and making “designer babies.”Various theoretical
of these disease.[67] For example, recently a SNP in the pro- and technical aspects of stem-cell genomics are discussed
moter region of the gene for matrix-metalloproteinase-9 in Chapter47, including ethical and legal issues related to
was found to be associated with PPROM.[74] It is likely stem-cell genomics and cell-based therapy.
that similar developments will transform the practice of
obstetrics and gynecology. This would make it necessary
P R E D I C T I VE G E N O M I C M E D I C I N E
for obstetricians and gynecologists to be educated and
trained in the application and delivery of genomic-related The rapid progress of genetic profiling technologies, in a
methods in the diagnosis and management of various whole-genome context, is creating some of the fundamen-
obstetrical and gynecological conditions. tal prerequisites for a new, much-heralded era of predictive
genomic medicine.[78] Nevertheless, formidable conceptual
and practical obstacles must still be addressed by phar-
GENE AND CELL THERAPY
maceutical, biotech, academic, and government research
Many people falsely assume that germline gene therapy is organizations before clinical applications of these tech-
already taking place with regularity. News reports of par- nologies become commonplace. Many of these effectively
ents selecting a genetically tested egg for implantation or translate into information technology challenges, including
for choosing the sex of their unborn child can mislead the privacy issues related to the use of genetic data, and new
public into believing that this is “gene therapy.” Actually, in data-analysis approaches in dealing with complex, hetero-
these cases, genetic information is being used for selection. geneous phenotypes.[79]
A recent review provides a detailed account of the develop- An integral part of the practice of clinical genetics
ments in somatic gene transfer and the associated risks and is genetic testing that involves chromosomal, molecular,
ethical issues.[75]. Regular updates are also available on the or biochemical testing to establish the disease-status sus-
public domain (http://www.ornl.gov/hgmis/medicine/​ pected on the basis of the patient’s clinical phenotype. An
genetherapy.html.) important aspect of clinical genetics involves assessing and
The term “gene therapy” encompasses at least four discussing with an individual, usually a family member, pre-
types of application of genetic engineering for the inser- sumed to be “atrisk” on the basis of family history, or fol-
tion of genes into humans. The scientific requirements lowing investigations. Predictive or presymptomatic genetic
and the ethical issues associated with each type are dis- testing is carried out in the absence of any symptoms, solely
cussed next. with the aim of verifying the genetic risk.
Somatic cell gene therapy is technically the simplest and Genetic or genomic screening (Chapter  14) is aimed
ethically the least controversial. The first clinical trials were at selecting individuals at highrisk for developing a specific
undertaken in 1986–1987. genetic disease from a selected population subgroup. These
Germline gene therapy will require major advances in individuals are then referred to a clinical genetics service or
our present knowledge, and it raises ethical issues that are to an appropriate clinical service for confirmatory genetic
now being debated. In order to provide guidelines for deter- testing and genetic counseling. Examples include antena-
mining when germline gene therapy would be ethical, the tal screening for neural tube defects and Down syndrome.
three criteria that should be satisfied prior to a human clini- Neonatal genetic screening programs include phenylketon-
cal trial.[76] uria, galactosemia, cystic fibrosis, and Duchenne-type mus-
Enhancement genetic engineering presents significant, cular dystrophy. Developments in genomics now open the
and troubling, ethical concerns. Except where this type of door for large-scale population screening for complex traits
therapy can be justified on the grounds of preventive medi- in selected high-risk subgroups, such as coronary artery dis-
cine, enhancement engineering should not be performed. ease, hyperlipidemia, bronchial asthma, bipolar depression,
The fourth type, eugenic genetic engineering, is impossible schizophrenia, and susceptibility for infectious diseases.[80]
at present and will probably remain so for the foresee- The prediction of an individual’s genetic risk for one
able future, despite the widespread media attention it has of the 1,500 so-called Mendelian disorders (see Online
received.[77] Mendelian Inheritance in Man, OMIM) is possible on
The whole field of stem-cell genomics carries great finding a mutation in a single gene. But genetic testing for
potential in developing powerful therapeutic tools and these single-gene disorders, of which autosomal-dominant
increasing the availability of a wide range of transplant- Huntington’s disease is a prime example, has had only lim-
able tissues. There is also the perceived danger of human ited benefit in overall health care. This is due to several

2 8 4   •   G enomics in C linical P ractice

factors, including the non-availability of effective therapeu- Table 19.4  HERITABILITY ESTIMATES IN COMMON
tic or preventive interventions, except for contraception or POLYGENIC/MULTIFACTORIAL DISEASES (COM-
termination of a “high-risk” pregnancy. These conditions PILED FROM DIFFERENT SOURCES AND PUBLISHED
DATA)
only account for about 5% of the total disease burden. The
vast majority of illnesses are common multifactorial disor- DISORDER FREQUENCY HERITABILITY
ders, such as coronary heart disease, diabetes mellitus, bron- (%)

chial asthma, arthritis, and major depression.[81] Several Schizophrenia 1 85

studies support the “gene–environment” interaction as the Bronchial asthma 4 80
conceptual basis for the etiology of these disorders. Such Cleft lip± palate 0.1 76
studies, known as genetic association studies, have been rep- Pyloric stenosis 0.3 75
licated on several occasions by substantial genetic contribu- Ankylosing spondylitis 0.2 70
tion and interaction with environmental factors. Talipes (club foot) 0.1 68
Genetic association studies help in estimating the Coronary artery disease 3 65
genetic contribution to the etiology of polygenic/multi- Essential hypertension 5 62
factorial disorders. This is commonly referred to as their
Congenital dislocation—hip 0.1 60
“heritability.” Heritability is either expressed as a fraction
Neural tube defect (spina bifida) 0.3 60
of 1 or as a percentage figure. Since, currently, heritability
Type 2 diabetes mellitus (T2DM) 5 50
is measured on the basis of the phenotype, this represents
the phenotypical variance attributed to the genetic factors. Peptic ulcer 4 37
Heritability estimates for common multifactorial disorders Congenital heart disease 0.5 35
range from 39–80% (Table 19.4). Identification of genetic Type 1 diabetes mellitus (T1DM) 0.4 15
factors in these disorders is a prerequisite for developing
predictive genomic tests.[82]
Identification of genes for a few single-gene dominantly genetic susceptibility for colorectal cancer includes several
inherited disorders, such as familial polyposis coli (FAP) for genes triggered by mutations in the familial adenomatous
colorectal cancer and BRCA1 and BRCA2 for breast/ovar- polyposis gene (APC) (Figure 19.3). Conservative esti-
ian cancers, that are strongly associated with disease risk, is mates of the number of susceptibility alleles for major can-
a good model for developing new genetic tests for reliable cers range between tens and hundreds, all of which increase
prediction of the risk in common diseases. The main strat- disease risk only modestly, depending on their interactions
egy used so far has been to look for associations between with other genes and environmental factors.[87]
a disorder and either genetic markers or specific candidate The prediction of disease risk will depend on various
alleles. Typically, in both approaches, the aim is to compare factors:  the number of genes influencing each condition,
a case-control design with a set of specific alleles, genotypes, the frequency of susceptibility alleles in the population, the
or genetic markers (for example, SNPs) in individuals who penetrance of these alleles, the predictive power of these
have the disease (a defined phenotype) with matched con-
trols (population, age and gender)[83] Unfortunately, most
of the results have been disappointing. Meta-analyses of Normal
association studies to identify susceptibility alleles for heart 1 LOH 5q (APC)
disease, cancers, depression, asthma, and diabetes have
K-ras
shown that many initially positive findings have not been 2
replicated in later studies.[84] However, modest replications
LOH 18q(SMAD4)
of such studies indicate that people who share these suscep- 3
tibility alleles are 1.2 to 1.5 times more likely to develop LOH 17p(p53)
these disorders.[85] A  unsuccessful outcome of association
4
studies is due to several factors, including a small sample,
poor study design, inadequate or inappropriate selection of
genetic markers, and a publication bias against the report- 5
ing of negative results.[86] Therefore, it is likely that a few Figure 19.3 
Multiple susceptibility genes in pathogenesis of colorectal
rare alleles confer susceptibility in conjunction with several cancer. Courtesy of Dr. Ian Frayling, Institute of Medical Genetics, University Hospital of Wales,
other genes with only modest susceptibility. For example, Cardiff, Wales, U.K.

G enetic and G enomic A pproaches to C linical M edicine   •   2 8 5

alleles, how these alleles interact with each other and under
different genetic backgrounds, and, finally, interactions
between these alleles and other risk factors.[88] Pessimists
argue that predictive genomic medicine does not have any
future, as the prediction of risk for most polygenic disorders
is not feasible. However, optimists point to the “common
disease, common variant”(CDCV) hypothesis, which states
that susceptibility alleles for common diseases reflect muta-
tions that occurred in the human population 100,000 years
ago (Balmain et al., 2003) and can therefore be identified in
large association studies with 1,000 to 5,000 cases and con-
trols.[59,89] Prospects for predictive genomic medicine thus
depend heavily on validation of the CDCV hypothesis. It
is true that association studies would need to employ care-
fully matched patients and controls and candidate alleles or
genetic markers with reasonable population frequency.[88]
There is continued anxiety about the feasibility of
population-based predictive genomic screening for com-
mon diseases. Fewer than 5% of the available predictive
genetic tests are applicable to common diseases.[90] Most of
these tests employ alleles that carry a high predictive power.
Thus the use of single or a few alleles will not offer good
prediction unless the lifetime risk of the disease is 5% or
more and the genotype is either rare or increases disease risk
20 times or more.[91,92] Others argue that it will simply be
economically unviable for a country’s healthcare system to
screen the whole population for susceptibility alleles to pre-
vent only a small number of these disorders. Nevertheless, a
better prediction of future disease risk will be made possible
by employing multiple genetic variants. Results from several
studies could be combined to devise a risk-scoring system
for use in population-based genomic screening.[93]
The efficiency of genomic screening could be further
enhanced if the decision to test for multiple susceptibility
alleles is based on a person’s family history of the disease.
Family histories routinely carried out in any clinical setting
help in categorizing the risk group: average risk, the same as
the general population; moderate risk, has 1 or 2 affected
close relatives; and high risk, has 2or 3 affected family mem-
bers, either first-degree relatives, or an earlier onset of the
disorder. For example, most of the family history clinics
for cancer employ this strategy in triaging the family his-
tory for further action. It is estimated 30–50% of the family
history would fall into the moderate-risk group, 10% into
the high-risk group, and about 40–60% in the average-risk
group.[88] This approach would eliminate about half the
people originally selected for genetic screening, thus
improving the efficiency of available testing. Apart from
the family history, other epidemiological factors could be
relevant in selecting people for genetic testing. Such studies
2 8 6   •   G enomics in C linical P ractice
and surveys are currently in progress and likely to add valu-
able information.
The success of predictive genomic medicine will depend
on effective communication, provision of evidence-based
effective intervention for prevention of disease, and safe-
guarding that screening does not cause any kind of social,
economic, or psychological disadvantage to the persons
involved. Some researchers have expressed concerns that
inappropriate communication of risks may instead result
in demoralization and reduce a person’s self-confidence,
compromising their ability to change their lifestyle and
make effective use of the available prevention methods
or treatment. The outcome of disease-risk reduction will
depend not only on the availability of treatment methods
and medications, but also on the person’s own perception of
the risk, acceptance of the risk, and motivation to make life-
style changes and modify their health behavior. This would
largely depend on how the information was delivered to
the person. The aim and method of genetic counseling in
this situation will be different from that employed with a
specific Mendelian disease or a chromosomal abnormality.
It remains to be established whether the traditional genetic
counseling approach should be adapted when used in pre-
dictive genomic medicine.
Advocates of predictive genomic medicine have
expressed concerns that widespread genomics-based screen-
ing might adversely influence public health policies aimed
at reducing the overall health burden on the population.
The related public health strategies would include recom-
mendations to promote a reduction in smoking and a per
capita reduction in alcohol consumption, and to promote
healthy eating and exercising regularly to reduce the risks for
high blood pressure, diabetes, and heart and lung diseases.
Genomic screening would enable selection of the high-risk
group and targeting them with appropriate health advice
and interventions.[94] Public health policies would need to
address these issues and ensure effective and efficient imple-
mentation. This will be a prerequisite before any form of
predictive genomic screening is offered to a population.
Concerns about the social, ethical, and moral implica-
tions of predictive genomic medicine are largely based on
experiences and issues surrounding predictive genetic test-
ing in late-onset Mendelian disorders such as Huntington’s
disease(HD). Mutations that cause HD and many other
serious late-onset disorders carry a strong predictive power.
Thus, predictive genetic testing in closely related fam-
ily members poses a potential threat that the information
could be misused, leading to discrimination in their careers,
and affecting their employment prospects and financial
planning. In addition, they might also experience personal
and social difficulties. By contrast, genomics-based screen-
ing will not have the benefit of a strong predictive power for
common polygenic diseases. It is not clear what impact the
outcome of genomic screening will have on an individual’s
prospects for career choice, employment, life insurance, and
other personal interests.
Another important factor in successful implementa-
tion of the predictive genomic medicine program would be
public awareness and education. This is fast changing with
the help of wide media coverage and online availability of
a massive amount of data and information. It is feared that
the current level of understanding of genetics and genom-
ics is not high among the general public.[95] However, it
would be wrong to assume that the public will only be pas-
sive consumers of information and genetics services. Some
remain optimistic that the public will have clear perceptions
and be able to distinguish between, for example, gene-X
that is likely to be associated with the disease and gene-Y,
which is less likely to be associated with developing disease
symptoms.[3]‌ Nevertheless it is essential that educational
campaigns to improve public awareness on genetics and
genomics be launched to provide an opportunity for under-
standing and appreciating realistic applications of predic-
tive genomic medicine.
FORENSIC MEDICINE AND GENOMICS
DNA “fingerprinting” is regarded as one of the major dis-
coveries of genetics. It is now widely used in forensic sci-
ence and is admissible as evidence in criminal and civil legal
cases all over the world. This was fully developed prior to
the sequencing of the human genome. DNA fingerprinting
these days is a polymerase chain reaction (PCR)-based tech-
nique that uses highly variable regions known as short tan-
dem repeats(STRs) to construct a profile of an individual’s
DNA. This is occasionally supported by carrying out mito-
chondrial DNA (mtDNA) typing, particularly when the
nuclear DNA is degraded. mtDNA is inherited only from
the mother, and there may be several thousand copies in a
single cell. There is now a range of commercially available
kits for forensic use.
New methods are on the horizon, products of the flood
of new information and new technology arising from
genomics research, such as the use of SNPs, microarrays,
and robotics. The new genomic technology can be used to
build up the physical profile of an individual, called “phe-
notyping.” Phenotyping is a hot topic in forensic science;
for example, an individual’s DNA left at the scene of a crime
can be used not only for matching with the DNA database
but also in describing various physical characteristics such as
G enetic and G enomic A pproaches to C linical M edicine   •   2 8 7
skin, hair and eye color, build, and ethnic origin. However,
it will require considerable time before phenotyping is
accepted as reliable legal evidence.
Genomics of other species, such as flies, slugs, algae,
and small plants, can also be used for forensic purposes
and could help in locating the scene of a crime or in collat-
ing the non-human evidence. Thus the impact of genom-
ics on forensic medicine is likely to be considerable. This is
not addressed in any more detail in this book, as forensic
genomics and its scope are outside the remit of this book.[96]
ET H I C A L , L E G A L , A N D S O C I ETA L
ISSUES ( ELSI)
Dr.  Francis Collins, director of the National Institute
for Human Genome Research, spoke at an American
Association for the Advancement of Science event on the
day President Bill Clinton signed an executive order prohib-
iting federal government agencies from obtaining genetic
information from employees or job applicants or from using
genetic information in hiring and job promotion decisions.
Collins noted:
But genetic information and genetic technology
can be used in ways that are fundamentally unjust.
Genetic information can be used as the basis for
insidious discrimination. Already, with but a hand-
ful of genetic tests in common use, people have lost
their jobs, lost their health insurance, and lost their
economic well-being because of the misuse of genetic
information. It is estimated that all of us carry dozens
of glitches in our DNA—so establishing principles of
fair use of this information is important for all of us.
The practice of modern clinical genetics is inseparable
from ethical, legal, and social issues(ELSI) (see Chapter 17).
Ethics in the new genomics era will be even more complex
than at present, where already it often arouses passion and
confusion.[97] Several challenges have surfaced from the sci-
entific developments in genomics, including professional
responsibility, liability, and issues regarding processing and
management of genetic information, as they relate to core
principles of modern ethics, such as autonomy, beneficence,
non-maleficence, and justice. However, it is anticipated that
genomic medicine will diminish rather than enhance exist-
ing sex, race, and socioeconomic inequalities in healthcare
access and delivery. Pertinent aspects of the social, ethical,
moral, and legal implications of genomics in clinical prac-
tice are separately discussed in ­chapter 17.
 GENOMICS AND MEDICAL
E D U C AT I O N
The curriculum for both undergraduate and postgraduate
medical courses should reflect the current state of the sci-
ence and art of clinical practice, and include prospective
developments that are likely to shape future medical prac-
tice (see Chapter 16). Medical genetics has had a significant
impact on the way we now consider disease causation and
strive for new therapeutic avenues. It has turned a corner
and has moved from the study of rare conditions to the
study of common diseases that affect the entire spectrum of
medical practice and is likely to change as we rapidly move
into the genomics era.
The future generation of medical practitioners will
be expected to be equipped with a new set of genetics-
or genomics-related basic principles and clinical skills.
Currently, genetics is commonly taught as a basic science,
sometimes as a freestanding course, but often as part of
a larger course, such as biochemistry or cell biology.[98]
Unfortunately, genetics is typically but a small component
of clinical teaching, leading the medical students to believe
that genetics is irrelevant in medical practice. However,
there has been a shift from this view, and some medical
school curricula incorporate genetics in the final stages of
the undergraduate medical course, such as during the pedi-
atric and obstetrics rotations. There is a huge disparity in
the curricula across medical schools. The gap is particularly
obvious and alarming in the developing or underdeveloped
countries.[99]
Even before the completion of the Human Genome
Project, ambitious predictions were made of wide-ranging
changes in medical practice. It has been proposed that
physicians will use genetic testing routinely to determine
disorders that their patients will someday develop, in order
to prescribe medications or recommend changes of life-
style that prevent these conditions. Understandably, some
clinicians have expressed concerns that this approach will
create a new class of underprivileged individuals, who
will be deprived life or health insurance coverage, or be
restricted in making appropriate career choices. The power
of assisted-reproduction techniques and prenatal testing
has been extended to the selection of gender, physical, and
mental traits in the unborn child. At the other extreme lies
the perceived danger of “designing” children with desired
physical or mental traits using “human cloning” tech-
niques. Society will need appropriately trained and skilled
physicians to guide and supervise these new developments
for the benefit of patients, families, and the community at
large.
2 8 8   •   G enomics in C linical P ractice
It is not essential that all medical practitioners should
master all of the basic principles of genetics to be able to
apply them in their clinical practice. For example, any phy-
sician can request blood gases or electrolyte levels without
understanding the chemistry and methodology that is used
in the assays. However, the clinician is expected to interpret
the results and apply these to clinical practice. Similarly,
an understanding of the common genetic tests should be
acquired by all clinicians. These are a set of concepts and
skills that will be necessary if genetics tools are to be used
wisely and efficiently.
Among many views and recommendations,[98] the fol-
lowing are cited for genetics education targeting the future
class of medical practitioners:
1. The ability to obtain a family history and to recognize
the major patterns of genetic transmission.
2. Awareness of the indications for chromosome analysis
and the ability to interpret the reports of chromosomal
abnormalities.
3. An understanding of molecular genetic testing and an
ability to interpret test results.
4. Appreciation of the major approaches to prenatal
diagnosis, including indications and limitations.
5. Knowledge of the role of genetics in the pathogenesis
of cancer, and opportunities for genetic testing to refine
the estimation of risk based on family history.
6. Awareness of population-screening programs, including
both newborn screening for metabolic disorders and
carrier-screening programs.
7. Understanding the role of clinical geneticists and
genetic counselors.
8. Sensitivity to the issues of privacy, discrimination, and
the family that arise in dealing with genetic conditions.
9. Awareness of where to go or whom to approach for
up-to-date information on a genetic disease and genetic
testing that can be applied at the point of clinical care.
T H E F U T U R E —P E R S O N A L I Z E D
MEDICINE
Although 99.9% of the human genome sequence is similar
in any two individuals, sequence variation in the remainder
of the genome could be linked to functional information
relevant to an individual’s unique genetic constitution. This
“personalized sequence variation” has major applications in
genomic medicine. Functional annotation for individual
sequence variation, when complete, will be of fundamental
importance in diagnosing and selecting appropriate thera-
peutic agents. This is likely to be vastly improved with the
availability of targeted sequencing of selected genes, exons,
or promoters. There are fewer variants in protein coding
than non-coding sequences.[100]
Variants that cause amino acid changes, and thus altered
protein product, are in general dissimilar (non-synonymous)
compared to those that lack such an association (syn-
onymous). If an excess of non-synonymous substitution is
observed for one particular coding region, then this can
be taken as an indicator of diversifying (positive) selec-
tion. With the help of newer technologies, more and more
variants are being characterized and sequence annotations
made available. Ultimately, a fuller picture will emerge of
the variants that alter genome function, and this will enable
the selection of those that contribute to health and disease.
The success of genomic medicine will depend upon our
ability to sequence an individual’s full genome. With the
benefit of new technologies, it is possible to generate giga-
bases of data as short-sequence reads, and to assemble the
data accurately using the finished sequence as a template.
This will provide the essential database of human genome
variation for a given population. Comparison of these data
sets will provide a full profile of common genome variation
along each chromosome. Detection of each variant will
help in estimating the recombination rates and correlation
along each chromosome. This approach could give impor-
tant baseline information on healthy tissue compared to
pathological tissue. For example, a comparison of the can-
cer genome sequences could allow monitoring the DNA
changes on a genome-wide basis for cancer development.
A similar approach could also be applied in other diseases.
This genomic information on both healthy and diseased tis-
sue could be used in screening an individual’s disease risk
and devising appropriate therapy and medical advice, pav-
ing the way for personalized medicine.
As human genome functional annotation becomes avail-
able, the prospects of personalized medicine will improve.
A  hypothetical scenario is described[100], where variation
in the PPAR-γ gene, one of the susceptibility genes in type
2 diabetes mellitus, is employed in selection of the most
appropriate oral hypoglycemic agent.
The Figure 19.4 illustrates the above mentioned hypo-
thetical scenario. The chromosome 3 region (i) (12,300–
12,450 kb, numbering as in build 34; seehttp://www.
ensembl.org) contains the PPAR-γgene structure
(dark blue) with an alternative promoter (light blue),
G enetic and G enomic A pproaches to C linical M edicine   •   2 8 9
hypothetical noncoding functional variants (green shaded
boxes), and functional variants (red). Magnification of the
variant segments (ii) shows the translated sequence with
nucleotide changes (functional variants highlighted in
blue) and amino-acid changes (pink, iii). The amino-acid
variant results in variation in the protein molecule, as
confirmed from a linked database. This variant protein
molecule contains a specific drug-binding site (blue) for
antidiabetic thiazolidinedione, an oral hypoglycemic
agent. A  number of biological consequences—biochemi-
cal, medical, and pharmacological—can be predicted
using linked database information (iv). This information
can be regularly updated and curated, allowing a detailed
listing of the likely consequences. A  small subset of this
information would define the disease or drug outcome or
side effect associated with each variant, and would enable
the clinician to provide specific risk information in clini-
cal consultation. This information could be made avail-
able in the public domain subject to stringent review and
including only those data for which medical relevance was
established.
The use of personal genetic information in a clinical
setting could be requested from, and consented to, by the
individual concerned. The individual sequence acquired
could be restricted to one or two genotypes, or as much
as a complete genome sequence. The information thus
acquired would be exclusive and private and wholly owned
by the individual. It could be stored electronically, pro-
tected by a high-security code requiring unique personal
identifiers, such as multiple finger-print or iris-pattern,
for access only with the consent of the individual (v). The
information might be taken either before consultation or
afterwards, and in either case would be accompanied by
counseling from the medical practitioner and consent by
the individual.
The clinical consultation could initiate a specific investi-
gation (vi). The personal genetic information would then be
supplied by the individual, for interpretation with respect
to an agreed set of variants and/or a specific phenotype.
The clinician would use the available risk information con-
cerning each variant to provide a genetic assessment for the
individual (vii). In the case illustrated, the individual has
the heterozygous genotype TC at position 3:12,450,610.
This corresponds to having both Pro 495 and Ala 495 forms
of the protein PPAR-γ. This genotype confers an increased
risk of insulin-resistant diabetes mellitus on the individual,
and also resistance to the thiazolidinedione class of anti-
diabetic drugs. Combining this with risk information for
other genotypes would help the patient and doctor make
informed subsequent clinical decisions (viii).
 Human genome project (E) Individual human sequence
Functional annotation Human resequencing
PGI i.d.: 591032-61215923014
Other genomes HapMap
(A) PPar-y

3: 12,300 3: 12,400 (kb)

1 2 3 4 5 6 7 8 9 10 11 12
Genomic information

(B) 3: 12,433,320 3: 12,433,340 3: 12,450,610

13 14 15 16 17 18 19 20 21 22 23 24

(F) Personal genetic information (owned by individual)

Gln Gly Cys Gln Phe Arg Ser Val Glu Ala . . . . . . Leu His Pro Leu Leu
Pro His Met Leu

314 316 318 495 Genetic counselling

Functional variants Patient consent

(C) Risk information selected and

Drug binding used to interpret PGI of patient
(thiazolidinedione)
Gln 314 Pro TC at 3: 12,450,610
Risk information
Arg 316 His

Pro 495 Leu

Val 318 Met Medical consultation
(G)
Structural context PGlkd.: 5910322–61215923014
Disease: Diabetes (type2)
(D) Genome base 3: 12,450,610 Nucleotide Risk Individual Medical Other
position genotype risk risk risks/effects
Genotype t/c
INS-resistance Thiazolidinedione
3: 12,450,610 TC
Individual information Pro/Leu Hypertension resistance

Biochemical consequence 6: 149,031,974 GG

Modelled: slight altered conformation of peptide

Personal genetic assessment
Stuctural consequence
backbone; increased local hydrophobicity.
(H)
Order further tests
Known: associated with severe insulin resistance,
Medical consequence Provide genetic counselling
diabetes mellitus and hypertension,
Recommend exposure avoidance
Pharmacological Prescribe or change medication
Known: resistant to thiazolidinediones.
consequence
Clinical decision
Biological consequences

Figure 19.4  A hypothetical model of “personalized medicine.” Adapted from Nature,with permission.[100]

Thus, with easy access to a well-annotated human To reach these key long-term goals, NIH-USA, the
genome and the availability of cheap, accurate, National Institute of Health Research (NIHR-UK),and
whole-genome-sequencing technology, an individual could many other organizations are actively pursuing and pro-
acquire either a specific or a complete genetic health profile, moting research in the above areas. These organizations
including risk and resistance factors. The information could are strategically investing in research to further our under-
then be used to improve and guide important medical deci- standing of the fundamental causes of diseases at their
sions, to assess the risk of possible future exposures, and to earliest genetic, genomic, and molecular stages. The cen-
select preventive treatments for improved health.[100] tral theme of the personalized medicine is based on the
In brief, the practice of personalized or specifically simple basic concept that individuals respond differently
individualized medicine will become the central focus of to environmental factors, including therapeutic inter-
the future practice of clinical medicine. However, this will ventions, according to their genetic/genomic endow-
demand lot of commitment, perseveration, and investment ment and their own behavior and lifestyle. In the future,
at the personal, family, community, and public or state lev- applied and translational genomic and molecular research
els. Inevitably and understandably, this approach will raise will allow us to predict how, when, and in whom a dis-
several ethical and social concerns: the fears of inequity, dis- ease will develop.[102] We can envision a time when we
crimination (primarily due to enormous costs and afford- will be able to precisely target or stratify treatment on a
ability), and potential misuse or abuse (malpractice). The personalized(individualized) basis to those who need it,
practice of personalized medicine shall not be allowed to avoiding treatment of those who do not. Ultimately, this
develop without relevant professional and statutory safe- individualized approach will allow us to preempt disease
guards put in place. This approach should be one of the other before it occurs, utilizing the participation of individuals,
major ingredients of clinical practice pathway, what is often communities, and healthcare providers in a proactive fash-
referred to The 4 P’s of Medicine: medicine that will be more ion, as early as possible, and throughout the natural cycle
Predictive, Personalized, Pre-emptive, and Participatory.[101] of a disease process.[103]

2 9 0   •   G enomics in C linical P ractice

 S U M M A RY
An important milestone in the history of medical science was
the recent completion of the human genome sequence. The
progress on identifying approximately 23,000 genes and their
regulatory regions provides the framework for understanding
the molecular basis of disease. This advance has also laid the
foundation for a broad range of genomic tools that can be
applied to medical science. These developments in gene and
gene-product analysis across the whole genome have opened
the way for targeted molecular genetic testing in a number
of medical disorders. This is destined to change the practice
of medicine:  future medical practice will be more focused
and individualized, what is often referred to as “personalized
medicine.” However, despite these exciting advances, many
practicing clinicians perceive the role of molecular genetics,
in particular that of medical genomics, as confined to the
research arena with limited clinical applications. Genomic
medicine applies the knowledge and understanding of all
genes and genetic variation in human disease. This chapter
introduces genomics-based advances in disease susceptibility
screening, diagnosis, prognostication, therapeutics, and pre-
diction of treatment outcome in various areas of medicine.
Finally, the art and science of the practice of medicine
at all times are true reflections of the dynamic adjustment
of the physical state of the human body and environmen-
tal pressures. In this context, the innate characteristics con-
ferred by the genetic and genomic constitution provide
the framework on which a range of lifetime environmental
experiences and pressures would act and manifest in either
positive or morbid (disease) states. This was echoed clearly
over 100  years ago across the medical community in one
of the classic Harveian Orations of the Royal College of
Physicians in London, England:
It was in Padua that medicine, long degraded and
disguised, was now to prove her lineage as the
mother of natural science, and the truth of the say-
ing of Hippocrates, that to know the nature of man
one must know the nature of all things.
—SirClifford Allbutt, Regius Professor of Physic, Harveian
Oration (1900).[104]
REFERENCES
1. Evans, G.E., and H.P. Rasmussen, Chromosome counts in three culti-
vars of Juniperus l. Botanical Gazette, 1971.132(4): p. 259–262.
2. Bodmer, W.F., The Book of Man: The Human Genome Project and
the Quest to Discover Our Genetic Heritage. 1997: Oxford University
Press, Oxford.
G enetic and G enomic A pproaches to C linical M edicine   •   2 9 1
3. Kumar, D., Clinical medicine in the genome era:  an introduction.
Genomics and Clinical Medicine, 2008; p. 145. Oxford University
Press, New York.
4. Anderson, S., et  al., Sequence and Organization of the Human
Mitochondrial Genome.Nature,1981.290(5806): p. 457–465.
5. Selander, R.K., et al., DNA Sequence Analysis of the Genetic Structure
of Populations of Salmonella Enterica and Escherichia Coli, in
Bacterial Diversity and Systematics. 1994, Springer. p. 17–49.
6. Guttmacher, A.E., and F.S. Collins, Welcome to the genomic era. New
England Journal of Medicine, 2003. 349(10): p. 996–998.
7. Roos, D.S., Bioinformatics—trying to swim in a sea of data. Science,
2001. 291(5507): p. 1260–1261.
8. Venter, J.C., et al., The sequence of the human genome. Science, 2001.
291(5507): p. 1304–1351.
9. Bell, J.I., The double helix in clinical practice. Nature, 2003.
421(6921): p. 414–416.
10.
Wessels, M.W., and P.J. Willems, Genetic factors in
non-syndromic congenital heart malformations. Clinical Genetics,
2010.78(2): p. 103–123.
11. Lupski, J.R., and P. Stankiewicz, Genomic disorders: molecular mech-
anisms for rearrangements and conveyed phenotypes. PLoS Genetics,
2005. 1(6): p. e49.
12. Kumar, D., Disorders of the genome architecture: a review. Genomic
Medicine, 2008. 2(3–4): p. 69–76.
13. Gomase, V.S., and S. Tagore, Toxicogenomics. Current Drug
Metabolism, 2008. 9(3): p. 250–254.
14. Clarke, P.A., et al., Gene expression microarray analysis in cancer biol-
ogy, pharmacology, and drug development:  progress and potential.
Biochemical Pharmacology, 2001. 62(10): p. 1311–1336.
15. Sandovici, I., et al., The dynamic epigenome: the impact of the envi-
ronment on epigenetic regulation of gene expression and developmental
programming. Epigenetics, 2008: p. 344–370.
16. Vlaanderen, J., et al., Application of OMICS technologies in occupa-
tional and environmental health research; current status and projec-
tions. Occupational and Environmental Medicine. 2010. 67(2):
p. 136–143.
17. Heijne, W.H.M., et  al., Systems Toxicology:  Applications of
Toxicogenomics, Transcriptomics, Proteomics and Metabolomics in
Toxicology.Expert review of Proteomics, 2005. 2(5):  p.  767–780;
doi:10.1586/14789450.2.5.767
18. Patti, G.J., O. Yanes, and G. Siuzdak, Innovation: Metabolomics: the
apogee of the omics trilogy. Nature Reviews Molecular Cell Biology,
2012. 13(4): p. 263–269.
19. Harrigan, G.G., and R. Goodacre, Metabolic Profiling:  Its Role in
Biomarker Discovery and Gene Function Analysis. 2003: Springer, Berlin.
20. Bernstein, L.H., Metabolomics, metabonomics and functional nutri-
tion:  the next step in nutritional metabolism and biotherapeutics.
Journal of Pharmacy and Nutrition Sciences, 2012. 2: p. 1–14.
21. Afman, L., and M. Müller, Nutrigenomics: from molecular nutrition
to prevention of disease. Journal of the American Dietetic Association,
2006. 106(4): p. 569–576.
22. Flordellis, C.S., The Emergence of a New Paradigm of Pharmacogeno
mics.Pharmacogenomics, 2005.6(5): p. 515–526.
23. de Leon, J., Pharmacogenomics:  the promise of personalized
medicine for CNS disorders. Neuropsychopharmacology, 2008.
34(1): p. 159–172.
24. Chan, I.S., and G.S. Ginsburg, Personalized medicine: progress and
promise. Annual Review of Genomics and Human Genetics, 2011.
12: p. 217–244.
25. Attwood, T.K., and C.J. Miller, Progress in bioinformatics and the
importance of being earnest. Biotechnology Annual Review, 2002.
8: p. 1–54.
26. Frazer, K.A., et al., Human genetic variation and its contribution to
complex traits. Nature Reviews Genetics, 2009. 10(4): p. 241–251.
27. Browning, S.L., Human Genetic Variation with Implications
for Healthcare in Ethiopian Populations.Doctoral Thesis, 2010,
University College London.
 28. Yaspo, M.-L., Taking a functional genomics approach in molecular
medicine. Trends in Molecular Medicine, 2001. 7(11): p. 494–501.
29. Pennington, S., and M.J. Dunn, Proteomics: From Protein Sequence
to Function. 2001: Taylor & Francis, Oxford.
30. Kim, J.H., Bioinformatics and genomic medicine. Genetics in
Medicine, 2002. 4: p. 62S–65S.
31. Elkin, P.L. Primer on medical genomics.Part V: Bioinformatics. Mayo
Clinic Proceedings, 2003. Elsevier.78(1):p. 57–64.
32. Maojo, V., and C.A. Kulikowski, Bioinformatics and medi-
cal informatics:  collaborations on the road to genomic medicine?
Journal of the American Medical Informatics Association, 2003.
10(6): p. 515–522.
33. Bieber, T., Stratified medicine: a new challenge for academia, industry,
regulators and patients. Stratified Medicine:  p.  3.Future Medicine.
London, 2013.
34. Trusheim, M.R., E.R. Berndt, and F.L. Douglas, Stratified medicine:
strategic and economic implications of combining drugs and clinical bio-
markers. Nature Reviews Drug Discovery, 2007. 6(4): p. 287–293.
35. Silver, M.A., et al., BNP Consensus Panel 2004: A clinical approach
for the diagnostic, prognostic, screening, treatment monitoring, and
therapeutic roles of natriuretic peptides in cardiovascular diseases.
Congestive Heart Failure, 2004. 10(s5): p. 1–30.
36. (UK), The Academy of Medical Sciences, Realizing the Potential of
Stratified Medicine. 2013. (www.academsci.ac.uk)
37. Khoury, M.J., et al., Multilevel research and the challenges of imple-
menting genomic medicine. Journal of the National Cancer Institute
Monographs,2012(44): p. 112–120.
38. (UK), The Academy of Medical Sciences, Realizing the Potential of
Stratified Medicine: Case Studies. 2013.(www.academsci.ac.uk)
39. Pao, W., and V.A. Miller, Epidermal growth factor receptor muta-
tions, small-molecule kinase inhibitors, and non–small-cell lung
cancer: Current knowledge and future directions. Journal of Clinical
Oncology, 2005. 23(11): p. 2556–2568.
40. Harris, T.J.R., and F. McCormick, The molecular pathology of cancer.
Nature reviews Clinicaloncology, 2010. 7(5): p. 251–265.
41. Tsuboi, M., et al., The present status of postoperative adjuvant chemo-
therapy for completely resected non-small cell lung cancer. Annals of
thoracic and cardiovascular surgery, 2007. 13(2): p. 73.
42. Azzoli, C.G., Dx/Rx: Lung Cancer: Lung Cancer: Jones & Bartlett
Publishers, Boston, Mass.
43. Baer-Dubowska, W., A. Majchrzak-Celiñska, and M. Cichocki,
Pharmocoepigenetics: a new approach to predicting individual drug
responses and targeting new drugs. PharmacologyReport, 2011.
63(2): p. 293–304.
44. Jaenisch, R., and A. Bird, Epigenetic regulation of gene expres-
sion: how the genome integrates intrinsic and environmental signals.
Nature genetics, 2003. 33: p. 245–254.
45. Tambuyzer, E., Towards a framework for personalized healthcare: les-
sons learned from the field of rare diseases. Personalized Medicine,
2010. 7(5): p. 569–586.
46. López, C., et al., Mechanisms of geneticallybased resistance to malaria.
Gene, 2010. 467(1): p. 1–12.
47. Horuk, R., et  al., A receptor for the malarial parasite Plasmodium
vivax: the erythrocyte chemokine receptor. Science, 1993. 261(5125)
p. 1182–1184.
48. O’Brien, S.J., and J.P. Moore, The effect of genetic variation in che-
mokines and their receptors on HIV transmission and progression to
AIDS. Immunological reviews, 2000. 177(1): p. 99–111.
49. Hummel, S., et  al., Detection of the CCR5-Delta32HIV resistance
gene in Bronze Age skeletons. Genes and immunity, 2005. 6(4):
p. 371–374.
50. Miesel, L., J. Greene, and T.A. Black, Genetic strategies for anti-
bacterial drug discovery. Nature Reviews Genetics, 2003. 4(6):
p. 442–456.
51. Fleischmann, R.D., et  al., Whole-genome random sequencing and
assembly of Haemophilus influenzae Rd. Science, 1995. 269(5223):
p. 496–512.
2 9 2   •   G enomics in C linical P ractice
52. Chung, L.P., and G.W. Waterer, Genetic predisposition to respiratory
infection and sepsis. Critical reviews in clinical laboratory sciences,
2011. 48(5–6): p. 250–268.
53. Wheeler, D.S., and H.R. Wong, The impact of molecular biology on
the practice of pediatric critical care medicine. Pediatric Critical Care
Medicine, 2001. 2(4): p. 299–310.
54. Hudson, V.M., Rethinking cystic fibrosis pathology:  the criti-
cal role of abnormal reduced glutathione (GSH) transport caused
by CFTR mutation. Free radical biology and medicine, 2001.
30(12): p. 1440–1461.
55. Eckel, R.H., Grundy, S.M., Zimmet, P.Z.The metabolic syndrome.
Lancet, 2005 Apr 16–22. 365(9468): p. 1415–1428.
56. Sladek, R., et al., A genome-wide association study identifies novel risk
loci for type 2 diabetes. Nature, 2007. 445(7130): p. 881–885.
57. Blankenberg, S., McQueen, M.J., Smieja, M.Comparative impact of
multiple biomarkers and N-terminal pro-brain natriuretic peptide in
the context of conventional risk factors for the prediction of recurrent
cardiovascular events in the Heart Outcomes Prevention Evaluation
(HOPE) study.Circulation,2006.114: p. 201–208.
58. Hoffjan, S., and C. Ober, Present status on the genetic studies of
asthma. Current opinion in immunology, 2002. 14(6): p. 709–717.
59. Balmain, A., J. Gray, and B. Ponder, The genetics and genomics of
­cancer. Nature genetics, 2003. 33: p. 238–244.
60. Torbicki, A., etal., Guidelines on the diagnosis and management
of acute pulmonary embolism:  The Task Force for the Diagnosis
and Management of Acute Pulmonary Embolism of the European
Society of Cardiology (ESC). European heart journal, 2008. 29(18):
p. 2276–2315.
61. Bates, M.D., The potential of DNA microarrays for the care of chil-
dren. The Journal of pediatrics, 2003. 142(3): p. 235–239.
62. Miles, J.H., Autism spectrum disorders—a genetics review. Genetics in
Medicine. 2011; 13(4): p. 278–294.
63. Wadlow, R., and S. Ramaswamy, DNA microarrays in clinical cancer
research. Current molecular medicine, 2005. 5(1): p. 111–120.
64. Palomino, J.C., Nonconventional and new methods in the diagno-
sis of tuberculosis: feasibility and applicability in the field. European
Respiratory Journal, 2005. 26(2): p. 339–350.
65. Perna, N.T., et al., Genome sequence of enterohaemorrhagic Escherichia
coli O157: H7. Nature, 2001. 409(6819): p. 529–533.
66. Venigalla, S., and G.R. Gourley. Neonatal cholestasis. In Seminars in
Perinatology, 2004.28(5): p. 348–355.
67. Weston, G.C., et  al., Genomics in obstetrics and gynaecology.
Australian and New Zealand Journal of Obstetrics and Gynaecology,
2003. 43(4): p. 264–272.
68. Chen, H., and C. Tzeng, Applications of microarray in reproductive
medicine. Chang Gung medical journal, 2006. 29(1): p. 15.
69. Wang, H., and S.K. Dey, Roadmap to embryo implantation: clues from
mouse models. Nature Reviews Genetics, 2006. 7(3): p. 185–199.
70. Reiher, F.K., et al., Inhibition of tumor growth by systemic treatment
with thrombospondin-1 peptide mimetics. International journal of
cancer, 2002. 98(5): p. 682–689.
71. Kim, J.J., et  al., Altered expression of HOXA10 in endometrio-
sis: potential role in decidualization. Molecular human reproduction,
2007. 13(5): p. 323–332.
72. Horcajadas, J.A., A. Pellicer, and C. Simon, Wide genomic analysis of
human endometrial receptivity: new times, new opportunities. Human
Reproduction Update, 2007. 13(1): p. 77–86.
73. Leppert, P.C., W.H. Catherino, and J.H. Segars, A new hypothesis
about the origin of uterine fibroids based on gene expression profiling
with microarrays. American journal of obstetrics and gynecology,
2006. 195(2): p. 415–420.
74. Ferrand, P.E., et al., A polymorphism in the matrix metalloproteinase-9
promoter is associated with increased risk of preterm premature rupture
of membranes in African Americans. Molecular human reproduction,
2002. 8(5): p. 494–501.
75. Kimmelman, J., Recent developments in gene transfer: risk and ethics.
BMJ: British Medical Journal, 2005. 330(7482): p. 79.
 76. Shannon, T.A., and J.J. Walter, The New Genetic Medicine: Theological
and Ethical Reflections. 2003: Rowman & Littlefield, London.
77. Anderson, W.F., Human gene therapy:  scientific and ethi-
cal considerations. Journal of Medicine and Philosophy, 1985.
10(3): p. 275–292.
78. Khoury, M.J., et al., The continuum of translation research in genomic
medicine: how can we accelerate the appropriate integration of human
genome discoveries into health care and disease prevention? Genetics
in Medicine, 2007. 9(10): p. 665–674.
79. Hood, L., et  al., Systems biology and new technologies enable
predictive and preventative medicine. Science Signaling, 2004.
306(5696): p. 640.
80. Green, E.D., and M.S. Guyer, Charting a course for genomic medicine
from base pairs to bedside. Nature, 2011. 470(7333): p. 204–213.
81. Murray, C.J.L., and A.D. Lopez, Alternative projections of mortality
and disability by cause 1990–2020: Global Burden of Disease Study.
The Lancet, 1997. 349(9064): p. 1498–1504.
82. Ginsburg, G.S., and H.F. Willard, Genomic and personalized medi-
cine:  foundations and applications. Translational Research, 2009.
154(6): p. 277–287.
83. Schork, N.J., D. Fallin, and J.S. Lanchbury, Single nucleotide poly-
morphisms and the future of genetic epidemiology. Clinical genetics,
2000. 58(4): p. 250–264.
84. Lohmueller, K.E., et  al., Meta-analysis of genetic association studies
supports a contribution of common variants to susceptibility to common
disease. Nature genetics, 2003. 33(2): p. 177–182.
85. Ioannidis, J., Genetic associations: false or true? Trends in molecular
medicine, 2003. 9(4): p. 135–138.
86. Colhoun, H.M., P.M. McKeigue, and G.D. Smith, Problems of
reporting genetic associations with complex outcomes. The Lancet,
2003. 361(9360): p. 865–872.
87. Ponder, B.A.J., Cancer genetics. Nature, 2001. 411(6835): p. 336–341.
88. Hall, W.D., K.I. Morley, and J.C. Lucke, The prediction of disease risk
in genomic medicine. EMBO reports, 2004. 5: p. S22–S26.
89. Zondervan, K.T., and L.R. Cardon, The complex interplay among fac-
tors that influence allelic association. Nature Reviews Genetics, 2004.
5(2): p. 89–100.
90. Wright, C.F., and M. Kroese, Evaluation of genetic tests for susceptibil-
ity to common complex diseases: why, when and how? Human genet-
ics, 2010. 127(2): p. 125–134.
G enetic and G enomic A pproaches to C linical M edicine   •   2 9 3
91. Khoury, M.J., et al., Do we need genomic research for the prevention
of common diseases with environmental causes? American Journal of
Epidemiology, 2005. 161(9): p. 799–805.
92. Holtzman, N.A., and T.M. Marteau, Will genetics revolution-
ize medicine? The New England Journal of Medicine, 2000.
343(2): p. 141.
93. Horne, B.D., et al., Generating genetic risk scores from intermediate
phenotypes for use in association studies of clinically significant end-
points. Annals of human genetics, 2005. 69(2): p. 176–186.
94. Qureshi, N., B. Modell, and M. Modell, Raising the profile of
genetics in primary care. Nature Reviews Genetics, 2004. 5(10):
p. 783–790.
95. Metcalfe, S.A., M. Aitken, and C.L. Gaff, The importance of pro-
gram evaluation: how can it be applied to diverse genetics education
settings? Journal of genetic counseling, 2008. 17(2): p. 170–179.
  96. Giardina, E., et al., Whole genome amplification and real-time PCR
in forensic casework.BioMedCentral genomics, 2009. 10(1): p. 159.
97. Guttmacher, A.E., F.S. Collins, and E.W. Clayton, Ethical, legal,
and social implications of genomic medicine. New England Journal
of Medicine, 2003. 349(6): p. 562–569.
98. Korf, B.R., Integration of genetics into clinical teaching in medical
school education. Genetics in Medicine, 2002. 4: p. 33S–38S.
99. Ghosh, K., and D. Mohanty, Teaching of medical genetics in the
medical colleges of India—way ahead. Indian Journal of Human
Genetics, 2002. 8(2): p. 43.
100. Bentley, D.R., Genomes for medicine. Nature, 2004. 429(6990):
p. 440–445.
101. Galas, D.J., Hood, L.Systems biology and emerging technologies will
catalyze the transition from reactive medicine to predictive, personal-
ized, preventive and participatory (P4) medicine. Interdisciplinary
Bio Central, 2009.1: p. 0006.
102. 
Xu, L.H., et al., The Re-emerging Concept of Personalized
Healthcare. Personalized Medicine, 2008. 5(5): p. 457–469,
doi:10.2217/17410541.5.5.457
103. Yang, Y.T., E. Wiley, and J. Leppard, Individualized medicine and
pharmacogenomics:  ethical, legal and policy challenges. Journal of
Medicine and the Person. 2011; 9(2): p. 48–57.
104. Allbutt, T.C., The Harveian Oration on Physiological Darkness
before Harvey:  Delivered before the Royal College of Physicians on
October 18th. British medical journal, 1900. 2(2078): p. 1271.
 20.
GENETIC AND GENOMIC TAXONOMY
OF HUMAN DISEASE
Dhavendra Kumar
T
he philosophy of disease is complex and reflects the
way an abnormal body state is perceived and under-
stood against the background of sociocultural, psy-
chological, and biomedical interpretation. During the late
phase of the last millennium, new scientific discoveries in
genetics offered entirely new approaches[1]‌. Nevertheless,
environmental and sociocultural factors remain important
in influencing the outcome of disease. The psychological
and biomedical factors provide the innate framework for
generating and developing symptoms and signs of the dis-
ease, depending on the pathology. On this basis, a “disease”
could be best defined as an “overall perception of an abnor-
mal body state and appreciation of the ensuing psychologi-
cal and physical impact.”
Historically, clinical practice has always faced its inabil-
ity to differentiate events that mediate the disease process
from resulting clinical, biochemical, and pathological
changes. Rapid advances in biomedical sciences, particu-
larly genomics, have opened entirely new horizons[2]‌[3].
However, despite having made tremendous advances, cli-
nicians continue to rely on phenotypical manifestations of
the disease process to define most diseases. Inevitably, this
approach often obscures the underlying mechanisms; thus
a clinician may fail to identify significant heterogeneity.
Concerns have been raised that most human disease pro-
vides only “insecure and temporary conceptions”[4]. Apart
from infectious diseases, there are alarmingly few diseases
that have a truly mechanism-based nomenclature.
The classification of human disease depends on several
factors, ranging from perception and analysis of symptoms,
sociocultural interpretation of varied manifestations of the
disease, and biological considerations,to therapeutic inter-
ventions. Conventionally, “disease” refers to a particular
organ or system dysfunction resulting from one or more caus-
ative factors such as physical trauma, infection, exposures to a
toxic substance, or malnutrition. A large number of diseases
remain unaccounted for due to the lack of a clear explanation
294
for their underlying mechanisms. Terms like degenerative or
autoimmune disorders are not uncommonly used to describe
an organ-system dysfunction. However,whether there is a
“cause and effect” relationship remains unclear. Nevertheless,
associated pathological changes often provide a firm and pre-
cise basis to categorize a disease or disorder. This has helped
in delineating distinct categories of diseases such as immuno-
logical and metabolic diseases.
Developments in genetics and molecular biology have
provided a vast amount of data and information to support
the view that most human diseases have a significant genetic
component. Characterization of the genetic determinants
of disease provides remarkable opportunities for clinical
medicine through an improved understanding of patho-
genesis, diagnosis, and therapeutic options. An understand-
ing of the genetic basis of human disease has opened the
way for a new taxonomy of human disease that will be free
from limitations and bias in developing diagnostic criteria
related to events that are often secondary and peripheral to
its cause[3]‌. For instance, genetic information has allowed us
to identify distinct forms of diabetes mellitus: defining an
auto-immune form associated with highly diverse and com-
plex human leukocyte antigens (HLA) and other inherited
factors that affect both expression and modification of
gene products in mediating the adult form of the disease[5].
Similarly, many genetically determined molecules and path-
ways have been characterized that are crucial in the patho-
genesis of bronchial asthma[6]. It is now widely believed that
a clearer understanding of the mechanisms and pathways of
a disease will assist us in delineating distinct disease subtypes
and may resolve many questions relating to variable disease
symptoms, progression, and response to therapy. This might
help in revising the current diagnostic criteria. Eventually,
genetics may contribute a new taxonomy of human disease
in clinical practice.
Although genetics is acknowledged to be an important
aspect of understanding the pathogenesis of disease, the
 Malformations Childhood diseases Adult diseases Infections Trauma

100% 0%
Proportion of genetic factors

Figure 20.1  Genetic factors in human disease.

genetic classification of human disease has not yet received
full recognition. There is ample evidence in support of the
argument that genetic factors are probably associated with
all human diseases except for trauma (Figure 20.1). However,
underlying genetic and genomic factors such as genetically
determined connective-tissue disorders, host-response to
infection, and tissue damage or inflammation could influ-
ence the outcome of trauma. Various categories of genetic
disorders are considered to be rare, with a tendency to be
included under the broad title of “organ-system diseases.”
Often these are listed as simply etiological factors rather than
as a distinct disease category. This concept and approach is
now rapidly being outdated, however, and replaced with
new classes of diseases. This progress is seriously hampered
by the lack of formal education at all levels and integration
of appropriate technologies into the modern medical diag-
nostic and therapeutic infrastructure.
Traditionally, genetic diseases are classified as chromo-
somal (numerical or structural), Mendelian or single-gene
disorders, multifactorial/polygenic complex diseases or
congenital anomalies, and diseases associated with specific
mitochondrial gene mutations (Table 20.1). Apart from
chromosomal disorders, essentially all genetic disorders
result from some form of alteration or mutation occurring
in a specific gene (single-gene diseases) or involving multiple
loci spread across the human genome (polygenic disorders).
The major impact of chromosomal disorders occurs before
birth and inflicts a serious health burden throughout child-
hood and during the early years of life (Figure 20.2). On
the other hand, single-gene diseases can pose a real medical
and health burden from the perinatal period to adult age,
with a peak around mid-childhood. In contrast, the poly-
genic/multifactorial diseases tend to present late, except for
developmental anomalies that will require active multidis-
ciplinary care during a child’s early life. A brief description
of the major types of genetic diseases is included here. Any
leading medical genetics textbook will contain a detailed
description of all these group of genetic disorders.
Table 20.1 THE CLASSIFICATION OF GENETIC
DISORDERS
Chromosomal: Numerical—aneuploidy
Structural—deletion, duplication, inversion,
isochromosome
Ring chromosome, reciprocal or Robertsonian
translocation
Mendelian: Autosomal recessive
Autosomal dominant
X-linked recessive
X-linked dominant
Epigenetic: Imprinting/parent of origin effect; indirect
influence on gene function
Oligogenic: Distinct phenotype due to 2 or more genes
Polygenic: Environmental interaction with several hun-
dreds of low-risk alleles, genetic polymor-
phisms, and genomic copy number variations
Mitochondrial: Deletion; point mutations; polymorphic vari-
ants in mtDNA
Genomic variation: Copy number variation; single-nucleotide
polymorphisms
C H R OMOSOMA L DISO R DE R S
The entire human genome is spread around 23 pairs of
chromosomes, including one pair specifically assigned
to male (XY) or female (XX) gender, designated the
“sex-chromosome pair.” The chromosomal constitution of
man is complex and comprises variable amounts of euchro-
matin and heterochromatin that exhibit with a character-
istic “banding-pattern,” and is essential for the physical
and distinctive appearance of a particular chromosome.
Typically, a chromosome pair includes two homologues,
each comprising a short arm (p) and a long arm (q) sepa-
rated by the central heterochromatin-G-C–rich region
designated the “centromere.” A  detailed account of the
chromosome structure and fundamental changes that occur

G e n et i c a n d G e n o m i c Taxo n o m y o f H u m a n D i s e a s e   •   2 9 5

Chromosomal
Multifactorial
Single gene
(Mendelian)
Birth Puberty
Figure 20.2  Distribution of different genetic disorders in various age
groups. Adapted with permission from Principles of Medical Genetics by Thomas D. Gelehrter, Francis
S. Collins, and David Ginsburg ‌.
[1]
during meiosis and mitosis can be found in any leading text-
book on basic genetics.
Chromosomal disorders are essentially disorders of the
genome resulting from either loss or addition of a whole
chromosome (aneuploidy) or parts of chromosomes
(structural). A chromosome abnormality results in major
disturbance in the genomic arrangement, since each chro-
mosome or part thereof consists of thousands of genes
and several non-coding polymorphic DNA sequences.
The physical manifestations of chromosome disorders are
often quite striking, characterized by growth retardation,
developmental delay, and a variety of somatic abnormali-
ties. Many chromosomal syndromes are now recognizable.
The diagnosis and genetic management of these disor-
ders fall within the scope of the sub-specialty “clinical
cytogenetics.”
The management of chromosomal disorders requires a
coordinated and dedicated team approach involving a wide
range of clinicians and health professionals. A typical exam-
ple is Down syndrome, resulting from either three copies
of chromosome 21 (trisomy) (Figure 20.3) or an addition
to the long arm of chromosome 21, usually resulting from
Adult
an unbalanced meiotic rearrangement of a parental chro-
mosomal translocation between chromosome 21 and one
of the other acrocentric (centromere located at the end)
chromosomes (Robertsonian translocation). Down syn-
drome occurs in about one in 800 live births and increases
in frequency with advancing maternal age. It is character-
ized by growth and developmental delay, moderate to severe
mental retardation, and the characteristic facial appearance
recognized with upward-slanting eyes. A  major cause of
death in these individuals is associated congenital heart
defects that can complicate the clinical management in a
significant proportion of Down syndrome cases. Prenatal
diagnosis and prenatal assessment of the maternal risk for
Down syndrome employing a variety of imaging and bio-
chemical parameters is now established clinical and public
health practice in most countries.

1 2 3 4 5

6 7 8 9 10 11 12

13 14 15 16 17 18

19 20 21 22 XX

Figure 20.3  Karyotype of a female (XX) with Down syndrome—note trisomy 21.

2 9 6   •   G e n o m i c s i n Cl i n i c a l P r ac t i c e
 Clinically significant chromosome abnormalities occur
in nearly 1% of live-born births and account for about 1%
of pediatric hospital admissions and 2.5% of childhood
mortality[7]‌. The loss or gain of whole chromosomes is
often incompatible with survival, and such abnormalities
are a major cause of spontaneous abortions or miscarriages.
Almost half of the spontaneous miscarriages are associated
with a major chromosomal abnormality. It is estimated that
about a quarter of all conceptions may suffer from major
chromosome problems, because approximately 50% of all
conceptions may not be recognized as established pregnan-
cies, and 15% of these end in a miscarriage. Essentially, the
major impact of chromosomal disorders occurs before birth
or during early life (Figure 20.2).
The delineation of rare and uncommon chromosomal
disorders has been crucial in the gene-mapping of several
Mendelian (single-gene) disorders such as the X-linked
Duchenne muscular dystrophy and type 1 neurofibromato-
sis. The chromosomal regions involved in deletion, dupli-
cation, inversion, and break points involved in a complex
chromosomal rearrangement provide an important clue
and assist the keen researcher in focusing on genes located
within the chromosomal segment.
MENDE L IAN ( SIN G L E - G ENE )
DISO R DE R S
About 4,000 human diseases are caused by mutations in
single genes, and these constitute a major health burden.
Single-gene disorders account for approximately 5–10%
of pediatric hospital admissions and childhood mortality.
The major impact of these disorders occurs in the newborn
period and early childhood. However, these also constitute
a significant proportion of adulthood diseases, notably
late-onset neurodegenerative diseases and various forms of
familial cancer. Although the majority of single-gene dis-
eases are rare, some are relatively common and pose a major
health problem. For example, familial hypercholesterolemia,
a major predisposing factor in premature coronary artery
disease, occurs in one in 500 people. Other good examples
would be familial breast and colorectal cancers, which affect
approximately one in 300. Some single-gene disorders are
specific for certain populations, like Tay-Sachs disease
among Ashkenazi Jews, or cystic fibrosis in Caucasians,
thalassemias among people from Southeast Asia and the
Mediterranean countries, and sickle-cell disease in people
of Western African origin. Techniques in molecular biology
have enabled the characterization of a number of mutated
genes. Sickle-cell disease was the first single-gene disorder
G e n et i c a n d G e n o m i c Taxo n o m y o f H u m a n D i s e a s e   •   2 9 7
to be defined at the molecular level. This has revolution-
ized the diagnosis and management of these disorders. The
single-gene disorders are inherited in a simple Mendelian
manner, and hence justifiably called “Mendelian disorders.”
The genetic transmission of altered genes or traits follows
principles set out by the Austrian monk Gregor Mendel in
1865, based on his seminal work on garden pea plants[8]‌.
Mendel inferred that “those characteristics that are trans-
mitted entire, or almost unchanged by hybridization, and
therefore constitute the characters of the hybrid, are termed
dominant, and those that become latent in the process,
recessive.”
The nomenclature of these disorders reflects their
gender-specific transmission and is supported by localiza-
tion of an altered gene on either an autosome (1–22) or the
X chromosome. Mendelian disorders are described as auto-
somal dominant, autosomal recessive, and X-linked reces-
sive (Figure 20.4) or X-linked dominant (Figure 20.5). The
latter pattern differs from the X-linked recessive by having
an excess of affected females in a family because the hetero-
zygous mutation on the X chromosome can be transmit-
ted to the daughter from an affected mother as well as the
affected father. Sporadic X-linked dominant diseases are
predominantly encountered in a female rather than a male
due to being lethal in the latter. A detailed family history
and careful interpretation of the pedigree are essential pre-
requisites in the diagnosis of a Mendelian disease. Accurate
risk estimates, for use in genetic counseling, are impossible
without a reliable and comprehensive pedigree. The major
features of the individual inheritance pattern are described
in leading genetic textbooks[1]‌. All human disorders and
traits that follow the Mendelian principles are listed in a
major resource—“McKusick’s Catalogue of Mendelian
Inheritance of Man.” An online version (Online Mendelian
Inheritance in Man; www.OMIM.​org) is available and reg-
ularly updated.
P O LYG ENIC O R
MU LTIFACTO R IA L DISO R DE R S
This group of disorders includes the most common and least
understood human genetic diseases. These diseases result
from the interaction of certain environmental factors with
multiple genes, some of which may have a major effect, but
the majority of which carry only a relatively minor effect.
The minor additive effect of these multiple loci lowers the
“threshold” of an organ or body system’s ability to with-
stand environmental pressures, resulting in either a devel-
opmental anomaly or an abnormal disease state. Examples
 (A) Autosomal dominant

I
1 2

II
1 2 3

III
1 2 3 4 5 6 7 8

(B) Autosomal recessive

I
1 2 3

II
1 2 3 4 5 6 7 8

(C) X-linked recessive

I
1 2

II
1 2 3 4 5 6

III
1 2 3 4 5 6 7 8

Figure 20.4  Typical pedigree appearances in Mendelian inheritance. Key to symbols: blank square = unaffected male; open circle = unaffected female; black-filled = affected (homozygous);
half black-filled = carrier (heterozygous)

II 3 3

III 6 2 3 4 6

Figure 20.5 
A pedigree with an X-linked dominant disorder—note absence of “male–male” transmission; all daughters of an affected male would be
heterozygous and thus could be symptomatic. Adapted with permission from Principles of Medical Genetics by Thomas D. Gelehrter, Francis S. Collins, and David Ginsburg ‌. [1]

2 9 8   •   G e n o m i c s i n Cl i n i c a l P r ac t i c e
include common congenital anomalies such as cleft lip, cleft
palate, neural tube defects, and most congenital heart dis-
eases. The common chronic medical diseases fall within this
category of genetic disorders, including diabetes mellitus,
coronary heart disease, hypertension, arthritis, and schizo-
phrenia. Understanding the genetic basis of common dis-
eases remains the major challenge facing modern genetics
and genomics.
The clinical impact of multifactorial diseases is signifi-
cant both in the neonatal period as well as in adult life. It
is estimated that about 25–50% pediatric hospital admis-
sions are related to these groups of disorders and are associ-
ated with 25–35% of childhood mortality. There is an even
greater medical and health burden from these disorders
during adult life due to the sufferers’ chronic natural history
of resulting medical diseases. For instance, diabetes melli-
tus and obesity account for about 40% of the adult medical
problems in the developed and developing world.
Identification of any such disorder or condition is
important in assessing risks to close relatives. A  compari-
son of general population and multiple cases in a family
would indicate a shift of the bell-shaped Gaussian curve to
the right, reflecting a lowered threshold with an increased
incidence (Figure 20.6). The precise additional risk would
dependon the degree of relationship with the index case in
the family. In addition, the gender of the index case is also
important in assessing the liability. The genetic liability is
estimated to be greater if the index case is of the gender with
lowest incidence. For example, in the case of pyloric steno-
sis, greater risk would be applicable if the index case were
a female, which carries the lowest birth prevalence.Finally,
Liability of
general
population
Number of individuals
Liability of
first degree
relatives
Liability
Figure 20.6 
The “Gaussian” bell-shaped curve to illustrate “genetic threshold,”indicated by liability in the general population (shown in black). A shift
‌.
to the right (in gray) indicates increased liability in first-degree relatives with an increased risk of recurrence. With permission from Oxford University Press, U.K.[9]
G e n et i c a n d G e n o m i c Taxo n o m y o f H u m a n D i s e a s e   •   2 9 9
recurrence risks for a given population group are estimated
to equal the square root of the birth incidence. For instance,
birth incidence of ventricular septal defect is approximately
three per 1000, the recurrence risk to a first-degree relative,
such as the next child, would be the square root of 0.003,
or 3%. These figures are useful in giving a family genetic
counseling following the birth of a child with a congenital
anomaly.
This group of diseases poses the challenge of working
out the mechanisms that determine the additive or inter-
active effects of many genes creating predisposition to
diseases, which in turn manifest only in the presence of
certain environmental factors. It is hoped that a combina-
tion of molecular genetic approaches, gene mapping, and
functional genomics will enable a clearer definition of these
genetic diseases. Several sections in this book will address
this issue at length and focus on specific disease groups and
systems.
MITOC H OND R IA L G ENETIC
DISO R DE R S
Apart from nuclear DNA (nDNA), a small proportion of
DNA is also found in mitochondria in the cytoplasm of cells
(mtDNA). Each cell contains 2–100 mitochondria, each
of which contains 5–10 circular chromosomes. The 16.5kb
mtDNA molecule is free from any non-coding intronic
regions and encodes two ribosomal RNA (rRNA) genes, 22
transfer RNAs (tRNA), and 13 polypeptides that are parts of
multi-subunit enzymes involved in oxidative phosphorylation
Threshold
Affected
individuals
(see also Chapter  9; Figure 20.7). In comparison to the
nuclear DNA, the mtDNA is 20 times more prone to recur-
rent mutations, resulting in generation of mutagenic oxygen
radicals in the mitochondria. The inheritance of mtDNA
is exclusively maternal, due to its cytoplasmic location. The
mature sperm head contains very little mtDNA, since it is
almost completely lost during the fertilization process, appar-
ently with the loss of the tail that carried the bulk mtDNA
in the cytoplasm. Due to the wholly maternal cytoplasmic
location, only females can transmit mitochondrial diseases to
their offspring of either gender (see Figure 20.7).
Since mtDNA replicates separately from the nuclear
DNA, and mitochondria segregate in daughter cells inde-
pendently of the nuclear chromosomes (replicative seg-
regation), the proportion of mitochondria carrying the
mtDNA mutation can differ among somatic cells. This
mitochondrial heterogeneity is also called heteroplasmy and
plays an important part in the variable and tissue-specific
phenotype of mitochondrial disease. Since different tis-
sues have varying degrees of dependence on oxidative phos-
phorylation, with heart, muscle, and central nervous system
being the most dependent, the common manifestations of
mitochondrial disease include cardiomyopathy, myopathy,
and encephalopathy (see Figure 20.1). Furthermore, oxida-
tive phosphorylation declines with age, probably related to
the accumulation of successive mtDNA mutations. Thus
the clinical phenotype in a mitochondrial disease is not
OH
12S
S
16
Deafness LHON
(1555) (15,257)
MELAS (3243)
ND-1
LHON (3460)
LHON
(11,778)
N D-
OL
2
NARP
MERRF or Leigh
(8344) (8993)
CO
I ND
CO I I A8 A6
tion
5-kb dele
Figure 20.7  The human mitochondrial DNA molecule with examples of point mutations with their associated clinical phenotypes. Adapted from Neurogenetics
by Stefan-M. Pulst, Oxford University Press, New York, 2000, with permission[10].
3 0 0   •   G e n o m i c s i n Cl i n i c a l P r ac t i c e
simply or directly related to mtDNA genotype, but reflects
several factors, including the overall capacity for oxidative
phosphorylation determined by mtDNA and nuclear DNA
genes, the accumulation of somatic mtDNA mutations and
degree of heteroplasmy, tissue-specific requirements of oxi-
dative phosphorylation, and age.
Several mitochondrial diseases have now been char-
acterized (Table 20.2). One of the best-characterized is
Leber’s hereditary optic neuropathy (LHON), which
exclusively affects males. There is loss of central vision sec-
ondary to optic nerve degeneration. The vision loss usu-
ally occurs in the patient’s 20s and can progress rapidly in
some men. Eleven different missense mtDNA mutations
in three different mitochondrial genes encoding respira-
tory chain enzyme subunits have been described. The phe-
notype in other mitochondrial diseases tends to include a
combination of heart, muscle, and central nervous system
manifestations, with considerable intra-/inter-familial vari-
ability for the same mtDNA mutation. In addition, mito-
chondrial dysfunction can be part of the phenotype in
some Mendelian diseases where the mutant gene-product
presumably has a pathogenic influence on the mitochondri-
ally mediated metabolic pathway. Examples of this are the
autosomal recessive respiratory enzyme disorders. Genetic
counseling and decision for prenatal diagnosis can be diffi-
cult in mitochondrial disorders due to difficulty in predict-
ing the phenotype in the affected pregnancy.
CY
T-b
N
D
-6
Ribosomal RNA genes
ND-
LHON Transfer RNA genes
5
(14,484)
Complex-1 genes
(NADH dehydrogenase)
Complex-III genes
(ubiquinol:cytochrome-c oxidoreductase)
-4
Complex-V genes
ND
(ATP synthase)
Complex-IV genes
-4L
ND
-3
(cytochrome-c oxidase)
III
CO
Table 20.2  GENETIC CLASSIFICATION OF MITOCHONDRIAL DISORDERS

DISORDER MAJOR CLINICAL FEATURES TYPE OF GENE MITOCHONDRIAL DNA MUTATION

Chronic progressive External ophthalmoplegia, bilateral ptosis, tRNA • A3243G, T8356C
external ophthalmoplegia mild proximal myopathy • Rearrangement
(CPEO) (deletion/ duplication)
Kearns-Sayre syndrome Progressive external ophthalmoplegia (PEO) Rearrangement
(KSS) onset <20 years, pigmentary retinopathy, (deletion/ duplication)
cerebellar ataxia, heart block, CSF protein
>1g/L
Pearson syndrome Sideroblastic anemia of childhood, Rearrangement
­pancytopenia, renal tubular defects, exocrine (deletion/ duplication)
­pancreatic deficiency
Diabetes and Deafness Diabetes mellitus, sensorineural hearing loss tRNA • A3243G, C12258A
• Rearrangement
(deletion/ duplication)
Leber’s hereditary optic Subacute painless bilateral visual loss, age Protein encoding G11778A, T14484C, G3460A
neuropathy (LHON) of onset 24 years, males>females (~4:1),
­dystonia, cardiac pre-excitation syndromes
Neurogenic ataxia with ret- Late-childhood or adult-onset peripheral Protein encoding T8993G/C
initis pigmentosa (NARP) neuropathy, ataxia, pigmentary retinopathy
Leigh syndrome (LS) Subacute relapsing encephalopathy, cerebellar Protein encoding T8993G/C
and brainstem signs, infantile onset
Exercise intolerance and Exercise-induced myoglobulinuria Protein encoding Cyt B mutations
myoglobulinuria
Mitochondrial encepha- Stroke-like episodes before 40 years, seizures tRNA A32343G, T3271C, A3251G
lomyopathy with lactic and/or dementia, ragged-red fibers and/or
acidosis and stroke-like lactic acidosis, diabetes mellitus, cardiomy-
episodes (MELAS) opathy (HCM/DCM), deafness, cerebellar
ataxia
Myoclonic epilepsy Myoclonus, seizures, cerebellar ataxia, tRNA A8344G, T8356C
with ragged-red fibers myopathy, dementia, optic atrophy, bilateral
(MERRF) deafness, peripheral neuropathy, spasticity,
multiple lipomata
Cardiomyopathy Hypertrophic cardiomyopathy (HCM) pro- tRNA A3243G, A4269G
gressing to dilated cardiomyopathy (DCM)
Infantile myopathy/ Early-onset progressive muscle weakness with tRNA T14709C, A12320G, G1606A, T10010C
encephalopathy developmental delay
Nonsyndromic sensori- Early-onset, progressive, bilateral, moderate rRNA A7445G
neural deafness to severe sensorineural hearing loss
Aminoglycoside-induced Early-onset, non-progressive sensorineu- rRNA A1555G
nonsyndromic deafness ral deafness secondary to aminoglycoside
administration

Finally, a high degree of sequence variation (polymor- G ENOMIC DISO R DE R S

phism) is known to occur in the non-coding region of the
mitochondrial chromosome (the D-loop). This polymor-
phism has been used in anthropological and evolutionary
studies to trace the origins and links of human popula-
tions. In addition, this information has been applied in
forensic analysis as well, to match maternal grandparents’
mtDNA with an orphaned child whose parents have “dis-
appeared” during war, a natural disaster, or in mysterious
circumstances.
Recent advances in molecular genetics have enabled us to
identify specific groups of disorders that result from char-
acteristic mechanisms involving specific areas of the human
genome. Often, these do not conform to the standard basic
principles of genetics.A broad term,genomic disorders, has
been coined to describe these conditions (Table 20.3)[11].
A number of hereditary disorders present with com-
plex genetic pathology that do not follow the conventional

G e n et i c a n d G e n o m i c Taxo n o m y o f H u m a n D i s e a s e   •   3 0 1
Table 20.3 CLASSIFICATION OF GENOMIC
DISORDERS
Disorders of genomic imprinting (epigenetic diseases)
Disorders of genome architecture (loss or gain of variable genomic
segments)
Tri-nucleotide repeat disorders (variable number of nucleotide
repeats with effect on gene function/ expression)
Genomic variation (copy number variation; single-nucleotide
polymorphisms)
principles of inheritance as outlined in the previous sec-
tions. There is now overwhelming evidence within these
disorders that indicates unusual mechanisms suggesting
“nontraditional inheritance.” The mechanisms involve cer-
tain genomic regions that directly or indirectly influence
regulation and expression of one or more genes manifesting
in complex phenotypes. Currently, some of these disorders
are listed either as chromosomal or as single-gene disorders.
DISORDER S OF GENOMIC
I M P R I N T I N G : E P I G E N ET I C D I S E A S E S
The term epigenetics refers to heritable factors that affect gene
expression without any change in the gene coding-sequence.
These factors could be operational either during meiosis or
mitosis and are often selective and preferential on the basis
of their “parent of origin.” The term imprinting is commonly
used to describe this important biological mechanism that
is recognized to influence wide-ranging physical and molec-
ular phenotypes. Numerous human diseases have now been
confirmed to result from epigenetic changes in various
parts of the genome. The term epigenetic diseases(or genomic
imprinting disorders) refers to this group of diseases.Basic
mechanisms related to the phenomenon of epigenetics or
epigenomics are reviewed separately (see also Chapter 4).
Epigenetic initiation and silencing is regulated by the
complex interaction of three systems, including DNA meth-
ylation, RNA-associated silencing, and histone modifica-
tion[12]. The relationship between these three components
is vital for the expression or silencing of genes (Figure 20.8).
Disruption of one or another of these interacting systems
can lead to inappropriate expression or silencing of genes,
leading to “epigenetic diseases.” Methylation of the C5
position of cytosine residues in DNA has long been recog-
nized as an epigenetic silencing mechanism of fundamental
importance[13]. The methylation of CpG sites within the
human genome is maintained by a number of DNA meth-
yltransferases (DNMTs) and plays multifaceted roles in
the silencing of transportable elements, for defense against
3 0 2   •   G e n o m i c s i n Cl i n i c a l P r ac t i c e
1
I
1 2 3 4
II
2
1 3 4 5 6 7
III
MITOCHONDRIAL
MYOPATHY DEAFNESS
MYOCLONIC EPILEPSY, DEMENTIA
ABNORMAL EEG
BLACK = SEVERE RED = MILD
Figure 20.8 
Pedigree of a family with mitochondrial encephalopathy with
ragged-red muscle fibers (MERRF)—note segregation of different
features with variable severity in the affected family members.
viral sequences, and for transcriptional repression of certain
genes. A  strong suppression of the CpG methyl-acceptor
site in human DNA results from mutagenic changes
in 5-methylcytosine, causing C:G to T:A transitions.
Normally, CpG islands, which are GC-rich evolutionarily
conserved regions of more than 500 base pairs, are kept
free of methylation. These stretches of DNA are located
within the promoter region of about 40% of mammalian
genes and, when methylated, cause stable, heritable tran-
scriptional silencing. Aberrant de novo methylation of CpG
islands is a hallmark of human cancers and is found early
during carcinogenesis[14].
In addition to DNA methylation, histone modifi-
cations have also been found to have epigenetic effects.
Acetylation and methylation of conserved lysine residues
of the amino-terminal tail domains are the key elements in
histone modification. Generally, the acetylation of histones
marks active, transcriptionally competent regions, whereas
hypoacetylation histones are found in transcriptionally
inactive euchromatic and heterochromatic regions. On the
other hand, histone methylation can be a marker for both
active and inactive regions of chromatin. Methylation of
lysine residue 9 on the N terminus of histone 3 (H3-K9) is
a hallmark of silent DNA and is evenly distributed through-
out the heterochromatic regions such as centromeres and
telomeres, including the inactive Xchromosome. In con-
trast, methylation of lysine 4 of histone 3 (H3-K4) denotes
activity and is predominantly found at promoter regions of
active genes[15]. This constitutes a “histone code,” which can
be read and interpreted by different cellular factors. There
is evidence that DNA methylation depends on methyla-
tion of H3-K9 and can also be a trigger for its methylation.
Recently, evidence has accumulated on the role of RNA
in post-transcriptional silencing. In addition, RNA in the
form of antisense transcripts (Xist or RNAi) can also lead to
mitotically heritable transcriptional silencing by the forma-
tion of heterochromatin. For example, transcription of anti-
sense RNA led to gene silencing and to the methylation of
the structurally normal α-globin gene in patients with alpha
thalassemia. This could be one of the many human diseases
resulting from epigenetic silencing due to antisense RNA
transcripts[16].
Mutations in genes that affect genomic epigenetic pro-
files can give rise to human diseases that can be inherited or
somatically acquired (Table 20.4). These epigenetic muta-
tions can be due either to hypermethylation (silencing) of
a regulating gene or to loss of methylation (LOM) (activa-
tion) of another gene that has a positively modifying effect
on the phenotype. The parental imprinting effect can be
inferred by demonstrating the parental origin of the mutant
allele. Similarly, either a loss or a gain of a chromosomal
segment can result in the same situation. Confirmation
of a specific chromosomal deletion or duplication is usu-
ally possible by using the fluorescent insitu hybiridization
(FISH) method. The paternal imprinting in this situation
is commonly demonstrated by genotyping a set of polymor-
phic markers located within the chromosomal segment.
Inheritance of the whole chromosomal homologue from
one parent effectively confirms imprinting phenomenon,
Table 20.4  RECOGNIZABLE EPIGENETIC DYSMORPHIC SYNDROMES [12]
DISEASE MAIN FEATURES
ATR-X syndrome α-thalassemia, facial dysmorphic features,
neurodevelopmental disabilities
Fragile-X syndrome Chromosome instability, physical and learn-
ing/ behavioral difficulties
ICF syndrome Chromosome instability, immunodeficiency
Angelman syndrome Seizures and intellectual disabilities
Prader-Willi syndrome Obesity, intellectual disabilities
Beckwith-Wiedemann (BWS) Organ overgrowth, childhood tumors
Russel-Silver syndrome Growth delay, body asymmetry
Rett syndrome Seizures, intellectual disabilities
Rubinstein-Taybi syndrome Facial dysmorphism, intellectual disabilities
Coffin-Lowry syndrome Facial dysmorphism, developmental delay
Abbreviations: ATR-X—α-thalassemia, X-linked mental retardation; UT—untranslated region; ICF—immunodeficiency, chromosome instability, facial anomalies;
CREB—cAMP-response-element-binding protein
G e n et i c a n d G e n o m i c Taxo n o m y o f H u m a n D i s e a s e   •   3 0 3
since the regulatory gene sequences for the pathogenic gene
would be missing from the other parent. This characteris-
tic abnormality is commonly referred to as “uni-parental
disomy” or UPD. This could either be isodisomy (simi-
lar parental homologues) or heterodisomy (parental and
grandparental homologues) (Figure 20.9). The origin of
UPD is believed to result from the loss of the additional
chromosomal homologue, failing which the conceptus
would be trisomic. This mechanism is also called “trisomic
rescue.”
For a maternally imprinted disorder, paternal UPD
would be confirmatory and maternal UPD diagnostic for
the paternally imprinted condition. For example, mater-
nal UPD is diagnostic for Prader-Willi syndrome, and
paternal UPD for Angelman syndrome, both conditions
being associated with a microdeletion of the 15q11 region.
The parental origin of the 15q microdeletion follows the
expected epigenetic pattern and is in keeping with the clini-
cal diagnosis. Recurrence risk estimates vary, depending on
the specific epigenetic pattern. This information is crucial
to obtain in order to offer accurate genetic counseling in any
genomic imprinting disorder.
Many epigenetic diseases are associated with chromo-
somal alterations and manifest with physical and learning
difficulties. For example, mutations in X-linked mental retar-
dation with the alpha thalassemia phenotype (ATRX) result
in consistent changes in the methylation pattern of ribosomal
DNA, Y-specific repeats, and subtelomeric repeats. Another
X-linked recessive mental retardation syndrome, associated
EPIGENETIC MECHANISM
Mutations in ATRX gene; hypomethylation of repeat and
satellite sequences
Expansion and methylation of CGG repeat in FMR1 5′
UTR, promoter methylation
DNMT3 mutations; DNA hypomethylation
Deregulation of one or more imprinted genes at 15q11-13
(maternal)
Deregulation of one or more imprinted genes at 15q11-13
(paternal)
Deregulation of one or more syndrome imprinted genes at
11p15.5 (IGF2, CDKN1C, KvDMR1,etc.)
Deregulation of one or more imprinted genes at 7p
(maternal)
MeCP2 mutations
Mutation in CREB-binding protein (histone acetylation)
Mutation in RSk-2 (histone phosphorylation)
 Eye
Heart Optic neuropathy
Conduction disorder Ophthalmoplegia
Wolft-Parkinsons-White Retinopathy
Skeletal muscle
Weakness Cardiomyopathy
Fatigue
Myopathy
Neuropathy Liver
Hepatopathy

Brain
Seizures Kidney
Myoclonus Fanconi's syndrome
Ataxia Glomerulopathy
Stroke
Dementia
Migraine
ATP
Nuclear DNA
Pancreas
subunits OX PHOS
mt DNA Diabetes mellitus
O2 H O
2

Defects in intergenomic
communication
Multiple mtDNA deletions and
mtDNA depletion

Blood
Pearsson's syndrome

Inner ear
Colon
Sensorineural
Pseudo-obstruction
hearing loss
OX PHOS = OXIDATIVE PHOSPHORYLATION

Figure 20.9 
The origin of uniparental disomy 15 in Prader-Willi syndrome through trisomic rescue during early embryogenesis—note different
homologues (maternal heterodisomy).

with a visible “fragile site” on the terminal part of the long arm
of the X chromosome (fragile-X syndrome), results from de
novo silencing of the pathogenic gene FMR1. The syndrome
is characteristically associated with an abnormal expansion of
CGG triplet repeats in the FMR1 5′ untranslated terminal
region. Methylation of the expansion leads to silencing of the
FMR1 gene and under certain cultural conditions creates the
visible “fragile site” on the X chromosome.
Epigenetic silencing is probably also significant in other
neurodevelopmental disorders. For example, in Rett syn-
drome, a common cause of intellectual disability in young
girls, mutations of the MeCP2 gene are seen in about 80%
of cases. The MeCP protein binds to methylcytosine resi-
dues and causes de-repression of genes normally suppressed
by DNA methylation. Despite the lack of firm evidence, it
is thought likely that MeCP2 might have a key role in the
control of neuronal gene activity resulting in the pathology
of Rett syndrome[17]. Interaction with another pathogenic
gene (CTKL5 or STK9) in Rett syndrome is likely to be
important in the pathogenesis of this neurodevelopmental
disorder[18]. On a wider genomic level, mutations in the
DNMT3b gene, causing the ICF (immunodeficiency,
centromeric region instability, and facial anomalies) syn-
drome, result in deregulation of DNA methylation pat-
terns. A notable example is that of Beckwith-Wiedemann
syndrome (BWS), an overgrowth syndrome predisposing
to Wilms’ tumor and other childhood tumors, which is
associated with duplications and rearrangements of a small
chromosomal region on the short arm of the chromosome
(11p15.5). This region contains a cluster of genes, which
is susceptible to a number of epigenetic alterations, mani-
festing with the BWS phenotype and tumorigenesis, par-
ticularly Wilms’ tumor and other childhood embryonal
tumors (Figure 20.10). Loss of methylation in imprinting
control regions (such as KvDMR1) can cause deregulation
of imprinting, and either biallelic expression (IGF2 and
H19) or silencing (such as CDKN1C) of imprinted genes,
which is seen in most sporadic BWS cases[19].
The epigenetic phenomenon is probably significant for
the phenotypical manifestations in some other hereditary

3 0 4   •   G e n o m i c s i n Cl i n i c a l P r ac t i c e
 species have enabled investigators to view genetic informa-
RNA tion in the context of the entire genome. As a result, it is now
possible to recognize mechanisms of some genetic diseases
at the genomic level. Amongst the several biological pro-
cesses, duplication of genes, gene segments, and repetitive
Gene gene clusters have helped in the evolution of mammalian
genomes[22]. This aspect of genome architecture provides
Histone modification DNA methylation recombination hot spots between non-syntenic regions of
chromosomes that are distributed across the whole genome.
These genomic regions become susceptible to further DNA
Figure 20.10 
The cluster of genes on 11p15.5 associated with the rearrangements that may be associated with an abnormal
phenotype of Beckwith-Wiedemann syndrome. The methylated region phenotype. Such disorders are collectively grouped under
KvDMR1 is indicated by the gray box within the gene KCNQ1OT1
and marked CH3 on the maternal homologue. The methylated region
the broad category of “genome architecture disorders”[11].
between the IGF2 and H19 genes is indicated by the hatched box and The term genome architecture disorder refers to a disease
marked CH3 on the paternal homologue. With permission from Oxford University that is caused by an alteration of the genome that results
Press[23].
in complete loss, gain, or disruption of the structural integ-
rity of a dosage sensitive gene(s) (Figure 20.12). Notable
examples include a number of chromosome deletion/
tumors. For example, transmission of autosomal dominant
duplication syndromes (Table 20.5). In these conditions,
familial chemodectomas (non-chromaffin paragangliomas
there is a critical rearranged genomic segment flanked by
or glomus tumors) is exclusively via the paternal line (Figure
large (usually >10 kb), highly homologous low copy repeat
20.11)[20]. The maternally derived gene is inactivated during
(LCR) structures that can act as recombination substrates.
oogenesis and can be reactivated only during spermatogen-
Meiotic recombination between non-allelic LCR copies,
esis. This genetically heterogeneous cancer family syndrome
also known as non-allelic homologous recombination, can
is associated with germline mutations in succinate dehydro-
result in deletion or duplication of the intervening segment.
genase subunits B (SDHB) and D (SDHD)[21].
Similarly, other chromosomal rearrangements, includ-
Thus epigenetic changes are probably significant in a
ing reciprocal, Robertsonian, and jumping translocations;
number of other complex phenotypes, particularly those
inversions; isochromosomes; and small marker chromo-
associated with cancer and a number of degenerative dis-
somes, may also involve susceptibility to rearrangement
eases (see “Complex Genomic Diseases”).
related to genome structure or architecture. In several cases,
LCRs, A-T–rich palindromes, and pericentromeric repeats
D I S O R D E R S O F G E N O M E A RC H IT E C T U R E are located at such rearrangement breakpoints. This suscep-
tibility to genomic rearrangements is implicated not only in
Recent completion of the Human Genome Project and
disease etiology, but also in primate genome evolution[25].
sequencing of the total genomes of yeast and other bacterial
An increasing number of Mendelian diseases
(Table  20.6) are recognized to result from recurrent
inter- and intra-chromosomal rearrangements involving
Disomy 15 Haploid unstable genomic regions facilitated by low-copy repeats
oocyte sperm
(LCRs)[26]. These genomic regions are predisposed to
non-allelic homologous recombination (NAHR) between
Trisomy 15 conceptus
paralogous genomic segments. LCRs usually span approxi-
mately 10–400 kb of genomic DNA, share 97% or greater
Loss of one homologue
sequence identity, and provide the substrates for NAHR,
thus predisposing to rearrangements. LCRs have been
Maternal shown to facilitate meiotic DNA rearrangements asso-
or heterodisomy 15
(Prader-Willi syndrome)
ciated with several multiple malformation syndromes
Biparental Biparental Uniparental and some disease traits (Table 20.6). Seminal examples
include microdeletion syndromes (Williams-Beuren
Figure 20.11 
Pedigree showing paternal transmission of paraganglioma
in a family: note no maternal transmission among “at-risk” family syndrome[7q11del], DiGoerge syndrome[22q11del]);
members[20]. autosomal dominant Charcot-Marie-Tooth disease type

G e n et i c a n d G e n o m i c Taxo n o m y o f H u m a n D i s e a s e   •   3 0 5
 Centromeric domain Telomeric domain

Maternal

CDKNIC KCNQ1 TSSC4 PHEMX IGF2 HI9 Telomere

CH3
Paternal
KCNQ1OT1

Expressed
Silenced
kvDMR
H19DMR

Figure 20.12 
Molecular mechanisms for genomic disorders—dashed lines indicate either deleted or duplicated region; the rearranged genomic interval
is shown in brackets; gene is depicted by filled horizontal rectangle; regulatory gene is shown as horizontal hash-marked rectangle; asterisks denote
point mutations[24].

1A (PMP22 gene duplication); hereditary neuropathy of
pressure palsy (HNPP: PMP22 gene deletion) mapped to
17p11.2; and Smith-Magenis, a contiguous gene syndrome
(CGS) with del (17)(p11.2p11.2). Dominantly inherited
male infertility related to AZF gene deletion follows a
similar mechanism. In addition, this LCR-based complex
genome architecture appears to play a major role in primate
karyotype evolution, the pathogenesis of complex traits,
and human carcinogenesis.
A notable example includes genetically heterogeneous
Charcot-Marie-Tooth disease (CMTD). The disorder is
also known as “hereditary motor and sensory neuropathy”
(HMSN) by virtue of being a peripheral neuropathy due to
involvement of either the axonal or myelinated segments
of the peripheral nerve. Genetically autosomal dominant,
autosomal recessive, and X-linked dominant types are rec-
ognized. The disorder is not uncommon, affecting approxi-
mately one in 2,500 of the adult population. This could be
an underestimate, since medically the condition is benign,
often not requiring any medical or surgical intervention.
However, some affected individuals experience increasingly
progressive neuromuscular weakness of distal muscles of
lower legs, feet, distal forearms, and hands, with onset in
the early teens, and causing severe locomotor restrictions.
An affected person usually presents late with relative
hypertrophy of the upper calf muscles, described as an

Table 20.5 CONTIGUOUS GENE SYNDROMES AS GENOMIC DISORDERS [11]

DISORDER (OMIM) INHERITANCE LOCUS GENE REARRANGEMENT RECOMBINATION

PATTERN SUBSTRATES
ORIENTATION
TYPE SIZE (KB) REPEAT % IDENTITY
William-Beuren syndrome (194050) AD 7q11.23 ELN del;inv 1600 >320 98 C
Prader-Willi syndrome (176270) AD 5q11.2q13 ? del 3500 >500 C
Angelman syndrome (105830) AD 15q11.2q13 UBE3A del 3500 >500 C
Dup(15)(q11.2q13) 15q11.2q13 ? dup 3500 >500 C
Triplication 15q11.2q13 15q11.2q13 ? trip >500 C
Smith-Magenis syndrome (18290) 17p11.2 RA13 del 4000 ~250 98 C
Dup(17)(p11.2p11.2) AD 17p11.2 PMP22 dup 4000 ~250 98 C
DiGoerge/VCFS (188400) AD 192430 22q11.2 TBX1 del 3000/1500 ~225–400 97–98 C
Male infertility (415000) YL Yq11.2 DBY, del 800 ~10 D
AZFa microdeletion USP9Y
AZFc microdeletion 400024 YL Yq11.2 RBMY del 3500 ~220 99.9 C
DAZ?
Abbreviations: del—deletion; dup—duplication; inv—inversion; D—direct; C—complex

3 0 6   •   G e n o m i c s i n Cl i n i c a l P r ac t i c e
Table 20.6 MENDELIAN GENOMIC DISORDERS [11]

DISORDERS INHERITANCE CHROMOSOME GENE(S) REARRANGEMENT RECOMBINATION SUBSTRATES

LOCATION
TYPE SIZE(KB) REPEAT % IDENTITY
ORIENTATION
Barter syndrome type III AD 1p36 CLCNKA/8 del 11 91 D
Gaucher disease AR 1q21 GBA del 16 14 D
Familial juvenile AR 2q13 NPHP1 del 290 45 >97 D
nephronophthisis
Facioscapulohumeral AD 4q35 FRG1? del 25–222 3.3 D
muscular dystrophy
Spinal muscular dystrophy AR 5q13.2 SMN inv/dup 500 I
Congenital ­adrenal AR 6p21.3 CYP21 del 30 96–98 D
­hyperplasia–21
­hydroxylase deficiency
Glucocorticoid remedi- AD 8q21 CYP11B1/2 dup 45 10 95 D
able aldosteronism
β-thalassemia AR 11p15.5 β-globin del 4,(7)? D
α-thalassemia AR 16p13.3 α-globin del 3,7,4.2? 4 D
Polycystic kidney disease AD 16p13.3 PKD1 50 95
type 1
Charcot-Marie-Tooth AD 17p12 PMP22 dup 1400 24 98.7 D
(CMT1A)
Hereditary neuropathy AD 17p12 PMP22 del 1400 24 98.7 D
with liability to pressure
palsy(HNPP)
Neurofibromatosis type AD 17q11.2 NF1 del 1500 85 D
1(NF1)
Pituitary dwarfism AR 17q23.3 GH1 del 6.7 2.24 99 D
CYP2D6-pharmacogenetic AR 22q13.1 CYP2D6 del/dup 9.3 2.8
trait
Ichthyosis XL Xq28 STS del 1900 20 D
Red-green color blindness XL Xq28 RCP/GCP del 0 39 98 D
Incontinentia pigmenti XL Xq28 NEMO del 10 0.870 D
Hemophilia A XL Xq28 F8 inv 300–500 9.5 99.9 I
Emery-Dreifuss muscular XL Xq28 Emerin/FLN1 del/dup/ 48 11.3 99.2 I
dystrophy (EMD) inv
Hunter syndrome XL Xq28 IDS inv/del 20 3 >88
Abbreviations: del—deletion; dup—duplication; inv—inversion; D—direct; C—complex; I—inverted

“inverted Champagne bottle”appearance (Figure 20.13),
associated with pes cavus due to wasting of the small muscles
of the feet. Similarly, wasting of the small muscles of hand
leads to “clawhands.” Neurophysiological studies remain an
essential method of differentiating the two major types of
CMTD. A  reduced motor-nerve-conduction velocity of
less than 35 m/sec helps in differentiating type 1 CMTD
from type 2 CMTD, in which the motor-nerve-conduction
velocity is usually normal but the sensory-nerveconduction
is often slow. Whilst this distinction is undoubtedly helpful
in determining clinical management, application for genetic
counseling is limited because both types are genetically het-
erogeneous. For instance, molecular characterization and
gene mapping have confirmed the existence of at least four
types of type 1 CMTD: autosomal dominant types 1a, 1b,
and 1c, and the X-linked type (XCMT). Similarly, there are
distinct genetic types within the type 2 CMTD group.
Approximately two-thirds of cases of CMT1 have a
detectable  1.5 Mb duplication within a proximal chro-
mosomal segment of the short arm of chromosome 17
(17p12)[27]. This duplicated chromosomal segment contains
a gene for peripheral myelin protein called PMP22. This

G e n et i c a n d G e n o m i c Taxo n o m y o f H u m a n D i s e a s e   •   3 0 7
 4

2 4 2 3

1 4 2 2 1 3 1

Figure 20.13  Lower legs and feet in Charcot-Marie-Tooth disease—note characteristic lower-leg appearance and pes cavus.

duplication results in the disruption of the gene, leading to
abnormal myelination of the peripheral nerves, an essential
molecular pathological step resulting in the CMT1 phe-
notype designated as CMT1A. The CMT1A duplication
was visualized by multiple molecular methods, including
fluorescence in-situ hybridization (FISH), pulsed-field gel
electrophoresis (PFGE), and dosage differences of het-
erozygous alleles by restriction-fragment-length polymor-
phisms (RFLPs) (Figure 20.14). This finding led to further
molecular studies on the origin of the 1.5 Mb duplicated
17p12 segment[28].
Studies by several investigators have revealed a sig-
nificant variation in the size of marker alleles flanking the
duplicated 17p12 region. It soon became apparent that a 500
kb allele co-segregated with 17p duplication in all affected
individuals. This suggested a stable mutation and followed
a precise recombination mechanism. However, in de novo
duplication, the presence of repeated flanking marker
alleles indicated the mechanism of unequalcrossing-over
leading to duplication. Indeed, this was confirmed when
a highly homologous >20 kb–size repeat sequence was
confirmed flanking the 17 p duplication. It was appropri-
ately named “CMT1A-REP.” As predicted by the unequal
crossing-over model, CMT1A-REP was found to be pres-
ent in three copies on the CMT1A duplication-bearing
chromosome. Interestingly, the presence of only one copy

(A) Gene dosage

del/dup
(B) Gene interruption

(C) Gene fusion

(D) Position effect

(E) Unmasking recessive allele

or * or *
Functional polymorphism

(F) Transvection effect

The 1.5 Mb duplicated chromosomal region of 17p12 including the PMP22 gene—note 500 Kb junction fragment allele flanking the
Figure 20.14 
CMT1A gene detected by PFGE and Southern analysis. Note additional 17p segment (red color) by metaphase (top two pictures) and interphase
(lower two pictures) FISH[11].

3 0 8   •   G e n o m i c s i n Cl i n i c a l P r ac t i c e
Figure 20.15  The unequal meiotic recombination (crossing-over) resulting in duplication (CMT1A) and deletion (HNPP)[11].

was soon demonstrated in another peripheral nervous D I S O R D E R S WI T H T R I-N U C L EOT I D E

system disorder, known as “hereditary neuropathy with ( T R I P L ET ) R E P E ATS
liability to pressure” (HNPP)[29]. Most clinically affected
Several disorders are recognized to have a phenomenon
individuals with HNPP present with mild to moderate
of earlier age-at-onset of disease in successive generations.
episodic weakness of the lower limbs and occasionally of
This is known as “anticipation.” This observation failed to
upper limbs when subjected to prolonged pressure, such as
secure a valid biological explanation and had been put aside
sitting or sleeping. The disorder is dominantly inherited in
simply on the basis of biased ascertainment of probands
an autosomal dominant manner. This is generally a clini-
or random variations in the age of onset. With the iden-
cally mild and benign hereditary neuropathy. The presence
tification of unstable DNA repeats distributed across the
of only one copy results from a reciprocal deletion follow-
genome, a molecular basis has been found for the phenom-
ing unequal crossing-over involving the CMT1A-REP
enon of anticipation. These unstable DNA repeats tend to
repeat (Figure 20.15).
increase in size during meiosis over successive generations.
Similar observations were also made in relation to
The abnormal expansion is correlated with reducing age
Smith-Magenis syndrome (SMS), a contiguous gene syn-
of onset and increased severity with further expansion of
drome associated with a microdeletion of the 17p11.2 seg-
DNA repeats. The characteristic pattern of the DNA repeat
ment (Greenberg et  al. 1991). Affected children present
involving a set of three nucleotides is commonly referred to
with facial dysmorphic features, severe speech delay, and
“tri-nucleotide” or “triplet” repeats[31]. This soon became
behavioral problems, with signs of self-harm. A  specific
established as a novel class of mutation, and it offered a
junction fragment was detected by PFGE (SMS-REP)
plausible explanation for the phenomenon of anticipation
that was involved in recurrent rearrangement resulting in
and variable clinical severity in a number of neurodegenera-
either SMS or reciprocal 17p11.2 duplication. Pathogenic
tive diseases (Table 20.7).
mutations in RAI1 gene, mapped to the 17p11.2 chromo-
The X-linked recessive spinal bulbar atrophy (SBA)
somal region, are now shown to be etiologically linked
was one of the first hereditary neurological disorders rec-
with SMS[18]. It is also possible to have both duplication
ognized to be associated with CAG triplet repeats. The
and deletion at the same time, resulting from DNA rear-
expanded region can occur anywhere in the gene and
rangements on both homologues of chromosome 17. This
thus can disrupt the expression of the gene. In the case of
was demonstrated in a patient with mild delay and a family
X-linked fragile-X syndrome (FRAXA), the CGG repeats
history of autosomal dominant carpel-tunnel syndrome[30].
are found in the 5′-untranslated region of the first exon of
The occurrence of both the 17p11.2 duplication and the
FMR1, the pathogenic gene for FRAXA (Figure 20.16).
HNPP deletion in this patient reflects the relatively high
However, in the case of Friedreich’s ataxia (FA), an auto-
frequency at which these abnormalities arise and the
somal recessive form of spinocerebellar ataxia (SCA), the
underlying molecular characteristics of the genome in this
expanded triplet repeat allele (GAA) occurs in the first
region.
intron of X25, the gene encoding frataxin. In Huntington
It is perfectly reasonable to accept the argument that
disease (HD) and other inherited neurodegenerative dis-
similar molecular mechanisms apply in causing other disor-
orders, the CAG triplet repeats occur within exons and
ders (Table 20.6). The human genome has evolved an archi-
encode an elongated polyglutamine tract (Figure 20.17).
tecture that may make us as a species more susceptible to
However, the expanded CTG triplet repeats of myotonic
rearrangements causing genomic disorders[28].
dystrophy (DM) are found in the 3′-untranslated region

G e n et i c a n d G e n o m i c Taxo n o m y o f H u m a n D i s e a s e   •   3 0 9
 Table 20.7 DISORDERS WITH TRINUCLEOTIDE (TRIPLET) REPEAT EXPANSION [11]

DISORDER TRIPLET LOCATION NORMAL# MUTATION #

Fragile-X syndrome CGG 5′UTR 10–50 200–2000
Friedreich’s ataxia GAA Intronic 17–22 200–900
Kennedy disease (SBMA) CAG Coding 17–24 40–55
Spinocerebellar ataxia 1 (SCA1) CAG Coding 19–36 43–81
Huntington disease CAG Coding 9–35 37–100
Dentatorubral-Pallidoluysian CAG Coding 7–23 49->75
Atrophy (DRPLA)
Machado-Joseph disease (SCA3)CAG Coding 12–36 67->79
Spinocerebellar ataxia 2 (SCA2) CAG Coding 15–24 35–39
Spinocerebellar ataxia 6 (SCA6) CAG Coding 4–16 21–27
Spinocerebellar ataxia 7 (SCA7) CAG Coding 7–35 37–200
Spinocerebellar ataxia 8 (SCA8) TG UTR 16–37 100->500
Myotonic dystrophy CTG 3′UTR 5–35 50–4000
Fragile site E (FRAXE) CCG Promoter 6–25 >200
Fragile site F(FRAXF) GCC ? 6–29 >500
Fragile site 16 A (FRA16A) CCG ? 16–49 1000–2000
Abbreviation: UTR—untranslated region

of the last exon of the DM protein kinase (myotonin)
gene (DM).
Each class of trinucleotide repeats exists in normal indi-
viduals. A pathogenic expansion is the one that is seen in
clinically symptomatic individuals. Carriers for an X-linked
disease also have an expanded allele (pre-mutation), which
does not usually result in an abnormal phenotype. However,
it is likely that some carrier females might exhibit some
manifestations as in fragile-X syndrome. An expanded allele
in the pre-mutation range in a male would not be associ-
ated with any clinical manifestations (normal transmitting
male NTM), but this could further expand, resulting in all
his daughters’ being carriers. However, recent studies have
provided data on the existence of late-onset gait ataxia in
NTMs[32]. On the other hand, a normal-size CGG repeat
in a normal male could undergo further expansion during

CMT1A
junction fragment CMT1A/HNPP
Analysis
PMP22

1.5 Mb
17p12

17

Figure 20.16  Location of four classes of triplet repeats in human diseases. Exons are shown in light pink with intervening introns as a pink solid line.
The translation site AUG and termination signal TAA are indicated by red vertical bars. Adapted with permission from Principles of Medical Genetics by Thomas D. Gelehrter,
Francis S. Collins, and David Ginsburg[1]‌.

3 1 0   •   G e n o m i c s i n Cl i n i c a l P r ac t i c e

Deletion
PMP22 HNPP
PMP22
Duplication
CMT1A
Figure 20.17 
Schematic diagram of the polyglutamine tract resulting from
abnormal expansion of CAG trinucleotide repeats[33].
meiosis, leading to a carrier daughter. This usually comes to
light when a symptomatic grandson is confirmed to have
pathogenic FRAXA expansion. Prior to availability of the
molecular testing in FRAXA, this kind of unusual pedi-
gree pattern in fragile-X syndrome was called the “Sherman
paradox” (Figure 20.18). Detailed molecular studies in the
family are often necessary to offer accurate genetic coun-
selling to “at-risk” carrier females. Carrier females are at an
additional risk for developing premature ovarian failure,
usually diagnosed when investigated for secondary infertil-
ity (see Chapter 46).
Genetic counselling in other neurodegenerative disorders
with triplet repeats is often complicated. In particular, the
clinical prediction in “borderline” expanded triplet repeats
(intermediate allele) in HD is extremely difficult due to lack
of reliable data. However, recent studies have produced some
data that are likely to be helpful in genetic counselling.
COMPLEX GENOMIC DISEASES
All inherited disorders have a genetic abnormality present
in the DNA of all cells in the body, including germ cells
(sperm and egg), and can be transmitted to subsequent gen-
erations. In contrast, a genetic abnormality present only in
specific somatic cells could not be transmitted. The genetic
abnormality in a somatic cell can occur any time from the
post-conception stage to late adult life. The paradigm of
somatic cell genetic disorder is cancer, where the develop-
ment of malignancy is often the consequence of mutations
in genes that control cellular growth. There are several such
genes, and these are designated oncogenes. It is now accepted
that all human cancer results from mutations in the nuclear
DNA of a specific somatic cell, making it the most com-
mon genetic disease. The various genetic mechanisms that
can result in cancer are discussed in the chapter on cancer
genomics (see Chapter 36).
The clinical course and outcome of treatment in a num-
ber of acute and chronic medical conditions depend upon
various factors. For instance, there is overwhelming evi-
dence that highly polymorphic cytokine, interferon, and
interleukin families of complex proteins influence the host’s
response to acute infection and physical injury or inflam-
mation. Several genes encode these inflammatory pathway
proteins. Similarly, association of human leucocyte antigens
in the pathogenesis of a number of acute and chronic medi-
cal disorders is well known (see Chapter 38). In addition,
interaction of mutations within these genes and with several
other genomic polymorphisms, such as single-nucleotide
polymorphisms (SNPs) is probably important in several
acute medical conditions, including trauma. This will
have a major impact in critical care and acute medicine
(see Chapter  48). The role of SNPs in modulating com-
plex medical disorders, such as diabetes mellitus, coronary
heart disease, hypertension, and various forms of cancer, is
unclear. However, the complexity of interaction of SNPs
with other genetic traits and loci is probably important in

AUG TAA
CGG GAA CAG CTG
gln

FRAGILE X SYNDROME FRIEDREICH ATAXIA SPINAL AND BULBAR MYOTONIC DYSTROPHY

MUSCULAR ATROPHY

SPINOCEREBELLAR
ATAXIA TYPE I

HUNTINGTON DISEASE

DENTATORUBRAL-
PALLIDOLUYSIAN ATROPHY
(HAW RIVER SYNDROME)

MACHADO-JOSEPH DISEASE

Figure 20.18  The Sherman paradox: a hypothetical pedigree showing affected members (red) and carrier females (pink); individual III.1 is a normal
transmitting male; the % risk for mental retardation is given for respective size of the triplet (CGG) repeats. Adapted with permission from Principles of Medical
Genetics by Thomas D. Gelehrter, Francis S. Collins, and David Ginsburg[1,2].

G e n et i c a n d G e n o m i c Taxo n o m y o f H u m a n D i s e a s e   •   3 1 1
 1
I

50–60 CGG REPEATS

1 2 3
II
0% 0%

60–70
III 1 2 3 4
T

60–70 5% 9%
70–90
1 2 3 4
IV
0% 40% 16%
70–90 >90

1 2 3 4
V
40% 16% 50% 28%

Figure 20.20 
Schematic diagram of the polyglutamine tract resulting from
abnormal expansion of CAG trinucleotide repeats. Adapted from Perutz et al.,
1994, with permission.

of human cancer[12]. A  high percentage of patients with

4.8 Å
–Cα, –C, –N, – not only in the tumor, but also in normal somatic tissues,
Figure 20.19 
The Sherman paradox: a hypothetical pedigree showing
affected members (red) and carrier females (pink); individual III.1 is a
normal transmitting male; the % risk for mental retardation is given for
respective size of the triplet (CGG) repeats. (Fu et al., 1991)
the prognosis of these disorders, in particular the outcome
of therapeutic interventions. This argument probably jus-
tifies separating some of these disorders under the title of
“complex genomic diseases.”
Various cancers and degenerative diseases occur with
increasing frequency in old age. However, these may also
present at a younger age, such as childhood leukemias. The
molecular mechanisms in these diseases are not entirely
clear, but probably include defects in DNA repair mecha-
nisms, accelerated apoptosis, deregulation of imprinted
genomic regions, and de novo chromosome rearrangements
involving specific genomic regions. Although these disor-
ders can be arguably included under the broad category of
“multi-factorial/polygenic diseases,” the pattern of distribu-
tion and recurrence does not follow the agreed principles
of multi-factorial/polygenic inheritance as discussed else-
where in this chapter.
As described in the previous section on epigenetics,
epigenetic changes play a major role in the development
sporadic colorectal cancer (CRC) possess microsatel-
lite instability and show methylation and silencing of
the gene encoding MLH1. It is thus likely that epigen-
etic changes also predispose to genetic instability. In some
cases, promoter-associated methylation of MLH1 is found
including spermatozoa. These germline “epimutations” pre-
dispose individuals carrying abnormal methylation patterns
to multiple cancers. Indeed, disruption of pathways that
lead to cancer is often caused by the de novo methylation
of the relevant gene’s promoters[14]. Epigenetic silencing has
been recognized as a third pathway satisfying Knudson’s
“two-hit” hypothesis for the silencing of tumor-suppressor
genes[34].
Chromosomal rearrangements have long been associ-
ated with human leukemias. These result in the formation
of fusion proteins, including histone acetyltransferases and
histone methyltransferases, that influence upregulation of
target genes. In acute promyelocytic leukemia, the onco-
genic fusion protein PML-RARα (promyelocytic leuke-
mia–retinoic acid receptor-α) causes the repression of genes
that are essential for the differentiation of hematopoietic
cells. Similarly, in acute myeloid leukemia, AML-ETO
fusions recruit the repressive N-COR-Sin3-HDAC1
complex and inhibit myeloid development[35]. There are
further examples of complex genomic arrangements that
result in other cancers, and that can modify the therapeutic
response. For example, mutations in genes for ATPase com-
plex are associated with poorer prognosis in patients with
non–smallcell lung cancer[36].

3 1 2   •   G e n o m i c s i n Cl i n i c a l P r ac t i c e
Table 20.8 TAXONOMY OF HUMAN DISEASE: CORRELATION OF CLINICAL PHENOTYPES, MOLECULAR
PATHWAYS, AND GENOTYPES

I: Phenotypes:
Clinical—symptoms and physical signs
I: Correlation and interpretation of the above with one or more of the following parameters with the aim of arriving at a diagnosis or most
likely underlying mechanism of the disease:
Biochemical; e.g., urea/electrolytes, blood gases
Metabolic; e.g., sugar/lipid/endocrine profiles
Radiological—X-rays, ultrasound, CT/MRI scans, magnetic resonance spectroscopy, radioisotope, etc.
Pathological—histopathology, histochemistry, immune-histology, fluorescence microscopy, and electron microscopy
Hematological—hemoglobin, hematocrit, coagulation profile, etc.
Immunological—immunoglobulins (IgG, IgM, IgA, etc.); antibody profiles, e.g., lupus; specific immunological investigation
Microbial/ Pathogens—battery of tests for bacterial, viral, protozoal, parasitic, and fungal infections, including specific pathogen profiles
Toxicology—poisons, alcohol, therapeutic and recreational drugs
Environmental (Ecological)—Temperature extremes, high altitude, supersonic flying, and space travel
II: Correlation of the above phenotypes with one or more of the biological/ molecular pathways:
Growth factors/ growth factor receptors (e.g., EGF/EGFRs, FGF/FGFRs, TGF/TGFRs,and VGF/VGFRs)
Dynamic cell/tissue factors (e.g., protein kinase families such as P13, RAS/MAPK; tumor suppressor systems, etc.)
Respiratory chain/ oxidative pathways (e.g., mitochondrial and cyclooxygenase systems)
Cell/ tissue response systems (e.g., cytokines, interleukins, tissue necrosis factors, complement factors, etc.)
Apoptosis/ senescence systems (e.g., apoptotic pathways)
Scavenger/ housekeeping systems (e.g., lysosomal enzymes, alpha 1 antitrypsin, DNA repair genes, etc.)
Metabolic regulatory systems (e.g., insulin/glycemic regulation, lipids/ hepato-biliary systems, Krebs cycle, etc.)
Energy regulation (e.g., temperature regulation, energy conservation, nutritional state, etc.)
Hormonal regulation (e.g., endocrine pathways, autocrine and paracrine pathways)
Vascular pathways (e.g., angiogenetic systems, clotting/ coagulation pathways, and platelet factors)
III: Correlation and/or interpretation with ALL of the above with one (or more) individual’s genetic/ genomic pathology:
Chromosomal aberration—aneuploidy (e.g., trisomy 21, 18, 13; triploidy); structural changes (micro-deletion/duplication, inversion, ring
chromosome, etc.)
OR
Specific gene mutations in a Mendelian disorder (e.g., betal thalassemia, cystic fibrosis, Duchenne muscular dystrophy)
OR
Mutations in 2 or more genes (oligo- or multigenic) belonging to a gene/ molecular family (e.g., sarcomere genes in hypertrophic
cardiomyopathy)
OR
Interaction of several hundreds and thousands of low risk alleles/genes with one or more environmental factors, including the lifestyle—
polygenic/ multifactorial
OR
Mitochondrial gene mutations and/or polymorphisms—several multisystem disorders that follow matrilineal inheritance pattern.

DISEASE S P ECT RUM , B IO L O G ICA L
PAT H WAYS , AND G ENOTY P ES
In modern medicine, diagnosis of any disease or morbid
state relies on establishing the phenotype along the lines
of agreed parameters (Table 20.8). The next logical step is
to find evidence for likely pathophysiological changes that
could be logically linked with one or more phenotypes. This
could be demanding and challenging, as it might involve
in-depth analysis and understanding of the complex bio-
logical (e.g., metabolic or molecular) pathways implicated
in the disease process (Table 20.9). Once this was achieved,
then a correlation could be looked for with specific protein
or enzyme systems recognized to be essential component(s)

G e n et i c a n d G e n o m i c Taxo n o m y o f H u m a n D i s e a s e   •   3 1 3
Table 20.9  GENETIC AND GENOMIC PATHOLOGY IN HUMAN DISEASE

The following genetic/ genomic pathology might be associated with one or more clinical phenotypes. Interpretation and precise diagnosis
would depend on the natural history, family history, and sensitivity/ specificity of genetic/genomic analyses:
• Epigenetic/ epigenomics changes—mutations/ deletions/ duplication/ inversion of genes or genomic segments adjacent to the promoter
region of certain genes; genetic imprinting abnormality involving specific genes demonstrating “parent of origin effect,” including com-
plete, partial, or mosaic uniparental disomy.
• Genome-wide abnormalities—pathogenic or disease-modification effect of structural variation across the genome; for example,
single-nucleotide polymorphisms, copy number variations, deletions/ duplications, nucleotide repeats (e.g., trinucleotide repeats).
• Gene function/ expression—specific “gain of function” or “loss of function” gene mutations (e.g., increased risk of cancer/tumor due to
mutation in a tumor-suppressor gene); mutations in transcription factors associated with a range of developmental anomalies; abnormali-
ties in RNA interference system associated with exaggerated or blunted therapeutic response; post-translational modification/ changes in
the gene product associated with one or more phenotypes consistent with a disease diagnosis.

of the core and successive biological pathways. Finding
structural or functional abnormalities of any given protein
or enzyme system would require undertaking investigations
targeted at specific gene(s) or genomic regions harboring
particular gene(s). Establishing the precise genotype would
then be the final piece in the complex jigsaw puzzle that is
collectively labeled as a disease or syndrome. The individual
genotype could be in any form (Table 20.8,III) including
gross chromosomal changes, specific genes or gene clus-
ters, and extremely small segments of the genome. Thus the
whole landscape of the disease or diagnosis involves a closely
linked network of three domains in the order of genotype–
pathway–phenotype. In other words, a diagnosis of any dis-
ease or morbid state (including the mortal state) would be
a cumulative process that should take into account all three
of these domains: disease= genotype + molecular pathway
+ genotype.
explain complex pathogenesis in some disorders. The spec-
trum of these disorders is wide and includes both acute and
chronic medical and surgical diseases. Perhaps it is reasonable
to identify these disorders on the basis of their underlying
molecular pathology, including genomic imprinting, genomic
rearrangements, and gene–environment interactions involv-
ing multiple genes and genomic polymorphisms. This chapter
has reviewed the genetic and genomic approaches in the clas-
sification of human disease. A stepwise approach is presented
based on correlations of the clinical phenotype, supporting
investigative phenotypes and specific evidence from targeted
genetic and genomic analyses. This approach would enable a
modern clinician to finally arrive at the final determining fac-
tor in the causation of human disease. The new taxonomy of
human disease is likely to have a major impact on the practice
of clinical medicine in the future.

SUMMA RY
Developments in genetics and subsequently the sequencing
of the human genome have provided us with an opportunity
to review the taxonomy of human disease. Conventionally,
the causation of human disease includes malformations,
trauma, infection, immune dysfunction, metabolic abnor-
mality, malignancy, and degenerative conditions associated
with aging. Genetic factors have long been recognized in
all of these disease groups. The traditional genetic catego-
ries of diseases include chromosomal disorders, single-gene
or Mendelian diseases, and several forms of multifactorial/
polygenic conditions. In addition, somatic genetic changes
and mutations of the mitochondrial genome probably
account for a small, albeit important, number of diseases.
These groups of disorders are well recognized and have an
established place in the classification of human disease.
Recent developments in genome research have provided
a wealth of data indicating different genomic mechanisms to
R EFE R ENCES
1. Gelehrter, T.D., F.S. Collins, and D. Ginsburg, Principles of medical
genetics. 1998: Lippincott Williams & Wilkins, Philadelphia.
2. Kumar, D., Clinical medicine in the genome era:  an introduc-
tion. Genomics and Clinical Medicine, Oxford University Press,
New York. 2008: p. 145.
3. Bell, J.I., The double helix in clinical practice. Nature, 2003.
421(6921): p. 414–416.
4. Pearce, J.M.S., Sir Thomas Lewis MD, FRS(1881-1945). Journal of
neurology, 2006. 253(9): p. 1246–1247.
5. Ioannidis, J.P.A., and F.K. Kavvoura, Concordance of functional in
vitro data and epidemiological associations in complex disease genetics.
Genetics in Medicine, 2006. 8(9): p. 583–593.
6. Daniels, S.E., et al., A genome-wide search for quantitative trait loci
underlying asthma. Nature, 1996. 383(6597): p. 247–250.
7. Hall, J.G., et al., The frequency and financial burden of genetic disease
in a pediatric hospital. American journal of medical genetics, 1978.
1(4): p. 417–436.
8. Henig, R.M., The monk in the garden:  the lost and found genius
of Gregor Mendel, the father of genetics. 2000:  Houghton Mifflin
Harcourt, London, UK.
9. Weatherall, D.J., The new genetics and clinical practice. Vol. 12.
1991: Oxford University Press, Oxford, UK.
10. Pulst, S.M., Neurogenetics: XA-GB. 2000: Oxford University Press,
New York.

3 1 4   •   G e n o m i c s i n Cl i n i c a l P r ac t i c e
 11. Kumar, D., Disorders of the genome architecture: a review. Genomic
medicine, 2008. 2(3–4): p. 69–76.
12. Egger, G., et al., Epigenetics in human disease and prospects for epigen-
etic therapy. Nature, 2004. 429(6990): p. 457–463.
13. Jaenisch, R., and A. Bird, Epigenetic regulation of gene expres-
sion: how the genome integrates intrinsic and environmental signals.
Nature genetics, 2003. 33: p. 245–254.
14. Jones, P.A., and S.B. Baylin, The fundamental role of epigenetic events
in cancer. Nature Reviews Genetics, 2002. 3(6): p. 415–428.
15. Lachner, M., R.J. O’Sullivan, and T. Jenuwein, An epigenetic road
map for histone lysine methylation. Journal of cell science, 2003.
116(11): p. 2117–2124.
16. Lapidot, M., and Y. Pilpel, Genome-wide natural antisense transcrip-
tion:  coupling its regulation to its different regulatory mechanisms.
EMBO reports, 2006. 7(12): p. 1216–1222.
17. Martinowich, K., H. Manji, and B. Lu, New insights into BDNF
function in depression and anxiety. Nature neuroscience, 2007.
10(9): p. 1089–1093.
18. Carmona-Mora, P., et al., Mouse models of genomic syndromes as tools
for understanding the basis of complex traits:  an example with the
Smith-Magenis and the Potocki-Lupski syndromes. Current genom-
ics, 2009. 10(4): p. 259.
19. Murrell, A., et  al., An association between variants in the IGF2
gene and Beckwith-Wiedemann syndrome:  interaction between
genotype and epigenotype. Human Molecular Genetics, 2004.
13(2): p. 247–255.
20. Petropoulos, A.E., et al., Genetic analysis in the diagnosis of familial
paragangliomas. The Laryngoscope, 2000. 110(7): p. 1225–1229.
21. Dannenberg, H., et  al., Frequent germ-line succinate dehydroge-
nase subunit D gene mutations in patients with apparently sporadic
parasympathetic paraganglioma. Clinical cancer research, 2002.
8(7): p. 2061–2066.
22. Ovcharenko, I., et al., Evolution and functional classification of verte-
brate gene deserts. Genome research, 2005. 15(1): p. 137–145.
23. Weksberg, R., et al., Tumor development in the Beckwith-Wiedemann
syndrome is associated with a variety of constitutional molecular 11p15
G e n et i c a n d G e n o m i c Taxo n o m y o f H u m a n D i s e a s e   •   3 1 5
alterations including imprinting defects of KCNQ1OT1. Human
Molecular Genetics, 2001. 10(26): p. 2989–3000.
24. Lupski, J.R., and P. Stankiewicz, Genomic disorders: molecular mech-
anisms for rearrangements and conveyed phenotypes. PLoS genetics,
2005. 1(6): p. e49.
25. Gu, W., F. Zhang, and J.R. Lupski, Mechanisms for human genomic
rearrangements. Pathogenetics, 2008. 1(1): p. 4.
26. Zhang, F., C. Carvalho, and J.R. Lupski, Complex human chro-
mosomal and genomic rearrangements.Trends in Genetics, 2009.
25(7): p. 298–307.
27. Garcia, C.A., et al., Clinical variability in two pairs of identical twins
with the Charcot-Marie-Tooth disease type 1A duplication. Neurology,
1995. 45(11): p. 2090–2093.
28. Lupski, J.R., Genomic disorders:  recombination-based disease result-
ing from genome architecture. American journal of human genetics,
2003. 72(2): p. 246.
29. Murakami, T., et al., Charcot-Marie-Tooth disease and related inher-
ited neuropathies. Medicine, 1996. 75(5): p. 233–250.
30. Potocki, L., et al., DNA rearrangements on both homologues of chro-
mosome 17 in a mildly delayed individual with a family history of
autosomal dominant carpal tunnel syndrome. American Journal of
Human Genetics, 1999. 64(2): p. 471–478.
31. Margolis, R.L., et al., Trinucleotide repeat expansion and neuropsychi-
atric disease. Archives of general psychiatry, 1999. 56(11): p. 1019.
32. Greco, C.M., et  al., Neuropathology of fragile X–associated tremor/
ataxia syndrome (FXTAS). Brain, 2006. 129(1): p. 243–255.
33. Chen, Y.W., Local protein unfolding and pathogenesis of
polyglutamine-expansion diseases. Proteins: Structure, Function, and
Bioinformatics, 2003. 51(1): p. 68–73.
34. Jones, P.A., and P.W. Laird, Cancer-epigenetics comes of age. Nature
genetics, 1999. 21(2): p. 163–167.
35. Jones, L.K., and V. Saha, Chromatin modification, leukaemia and
implications for therapy. British journal of haematology, 2002.
118(3): p. 714–727.
36. Reisman, D., S. Glaros, and E.A. Thompson, The SWI/SNF complex
and cancer. Oncogene, 2009. 28(14): p. 1653–1668.
 21.
GENOMICS OF COMPLEX CARDIOVASCULAR DISEASE
Foram N. Ashar and Dan E. Arking
INTRODUCTION
The genomics revolution, including the complete sequence
of the human genome, has greatly advanced our understand-
ing of the molecular basis of Mendelian disorders, which
are largely the result of necessary and sufficient mutations
that are individually rare. As fruitful as genome-targeted
efforts have been in the field of single-gene disorders, prog-
ress in understanding the genetic architecture of complex
traits has been slower and more tenuous. These traits are
governed by the interplay between genes, epigenetic fac-
tors, and the environment. In this scenario, complex traits
are likely to be modulated by multiple genes (and possi-
bly by multiple genetic variants within a gene), which are
individually neither necessary nor sufficient to determine a
trait. In the field of cardiovascular genetics, the multitude
of different outcomes assessed, often using slightly differ-
ent definitions of the phenotype in question, add to the
complexity. Thus, reproducibly identifying genetic variants
that impact these traits was a daunting task until recently,
when the popularization of microarray technology made
it possible to survey large cohorts at a large number of loci
in order to scan the genome for variants associated with
disease. Genetic studies of cardiovascular disease (CVD)
in particular are largely dichotomized into studies of genes
that influence CVD risk factors, such as lipids and blood
pressure, and studies that focus on clinical outcomes.
In this chapter, we summarize the progress made in the
search for genes affecting cardiovascular disease risk both
by candidate-gene studies and genome-wide association
studies (GWAS), while focusing on three specific CVD
events (myocardial infarction, stroke, and sudden cardiac
death), with the idea that genetic variants that directly
affect clinical outcomes would not necessarily be identi-
fied only through studies of CVD risk factors, and often
require a direct interrogation of the clinical outcome of
interest, allowing for more accurate risk stratification. This
is particularly important in sudden cardiac death, where
316
two-thirds of victims do not have CVD risk profiles that
would warrant intervention under current guidelines. The
results of this search for genetic variants of interest are sum-
marized in Table 21.1.
M YO C A R D I A L I N FA R C T I O N
( M I ) A N D C O R O N A RY H E A RT
DISEASE (CHD)
Myocardial infarction (MI) is an often fatal manifesta-
tion of coronary artery disease (CAD), with an annual
incidence of approximately 935,000 cases in the United
States, 172,000 of whom will die from the disease (Roger
et al., 2012). The proximal cause of MI is believed to be
a thrombosis event triggered by an atherosclerotic plaque
rupture, which occludes a coronary artery and leads to
necrosis of the myocardium. In a long-term follow-up
study of ~21,000 monozygous (MZ) and dizygous (DZ)
twins, heritability of fatal coronary events was 57% and
38%, respectively, for men and women (Zdravkovic et al.,
2002), underlining the importance of genetic factors in
susceptibility to MI. Several clinically relevant risk factors
have been identified, many of which display significant
heritability and thus are, at least in part, under genetic
influence (Table 21.2). Indeed, rare Mendelian mutations
that affect these risk factors can lead to premature coro-
nary artery disease (Table 21.3) and have been signposts
of important pathways that have directly led to success-
ful drug therapies, including the cholesterol metabolism
pathway that emerged from studies of familial hyperchol-
sterolemia (FH).
The majority of research efforts prior to 2007 were
focused on identifying genes that modulate common varia-
tion in traditional cardiovascular risk factors. A major break-
through was made with the popularization of GWAS and
the ability to genotype a large patient population for a mil-
lion single-nucleotide polymorphisms (SNPs) or more. The
Table 21.1  COMMON GENETIC VARIANTS THAT INFLUENCE SUSCEPTIBILITY TO CARDIOVASCULAR EVENTS

EVENT CHROMOSOMAL LOCUS/GENE NAME (SYMBOL) ALLELE(S) COMMENT

MI/IHD: Apolipoprotein E (APOE) E2, E3, E4 Associated with LDL-C levels, gene–environ-
ment interactions
Factor V Leiden Weak association with CAD
Prothrombin G20210A Marginal association with CAD
9p21 (CDKN2A/CDKN2B) Strongest signal from association studies, also
associated with T2D, various cancers
Sortilin 1 (SORT1) rs599839 Medicates LDL-C levels by disruption of
enhancer binding site
Stroke: Factor V Leiden Increased thrombin generation leads to
increased susceptibility
Prothrombin G20210A Mediates stroke risk by mechanism similar to
factor V Leiden
Methylenetetrahydrofolate reductase (MTHFR) C677T Associated with elevated homocysteine levels
Angiotensin-converting enzyme (ace) I/D Polymorphism affects circulating levels of ACE,
mechanism of action unknown
Phosphodiesterase 4D (PDE4D) AX, G0, GX Identified by linkage and association
Arachidonate 5-lipoxygenase-activating protein HapA, HapB Also associated with MI
(ALOX5AP)
Paired-like homeodomain transcription factor 2 Identified by association studies for ischemic
(PITX2) stroke, also associated with AF
SCD/QT Nitric oxide synthase 1 adaptor protein (NOS1AP) rs12143842 Replicated in multiple studies, nominal direct
interval: association with SCD
α Nav1.5 subunit (SCN5A) S1102Y, H558R Only S1102Y directly associated with SCD
SCD: 21q21 (CXADR) rs2824292 Association seen in individuals with MI and VF
2q24 (BAZ2B) rs4665058 Genome-wide significant in the largest GWAS
to date
QRS interval:* TKT/CACNA1D/PRKCD rs4687718 QRS-prolonging allele protective for risk of
SCD
*QRS interval is included as a subclinical phenotype for SCD, due to the limited direct studies of SCD.

efforts to identify common genetic variants associated with
CAD and MI have produced a list of over 30 loci showing
robust association that has been replicated across multiple
studies and cohorts (Kathiresan et al., 2009; Samani et al.,
2007; Schunkert et al., 2011).
We highlight the progress made in the search for genes
associated with MI by both the candidate gene approach
and the genome-wide association approach, by discussing
four specific examples: apolipoprotein E (APOE), coagula-
tion proteins, 9p21, and SORT1, in detail.
A P O L I P O P ROT E I N E (A P O E)
Apolipoprotein E, a low-density lipoprotein (LDL) receptor
ligand, is an important player in the metabolism of cholesterol
and triglycerides, where it mediates the clearance of chylomi-
cron and very low-density lipoprotein (VLDL) from plasma.
Utermann and colleagues (Utermann et  al., 1977)  first
described the effects of three common allelic variants of
APOE (termed E2, E3, and E4, and with frequencies of 8%,
77%, and 15%, respectively, in Caucasian populations) on
type III familial hyperlipoproteinemia, in which more than
95% of affected individuals were homozygous for the rare
allele. The common genotype, E3/E3, is used as the reference
group in most studies, and individuals who carry the E2 allele
have ~14 mg/dl lower LDL levels, and E4 carriers have ~7
mg/dl higher LDL levels (Motulsky and Brunzell, 2002).
Numerous studies have examined the association
between the E2/E3/E4 variants and coronary heart disease
(CHD), including a large-scale meta-analysis incorporating
121 studies with 37,850 cases and 82,727 controls (Bennet
et  al., 2007). The meta-analysis, which was stratified by
the number of participants in an individual study at a cut-
off of at least 1,000 healthy controls and 500 cases, dem-
onstrated a moderate increased risk for E4 carriers (odds
ratio [OR] 1.06, 95% confidence interval [CI], 0.99–1.13),
and a significant decreased risk for E2 carriers (OR 0.80,
95% CI 0.70–0.90) in the group with larger study size. In

G enomics of C om p le x C ardio vascular D isease   •   3 1 7

Table 21.2  RISK FACTORS FOR MYOCARDIAL report conflicting results (Humphries et  al., 2001; Lahoz
INFARCTION* et al., 2001; Volcik et al., 2006; Ward et al., 2009).
Risk Factors with a Significant Genetic Component (Heritability)
Total cholesterol (40–60%) C OAGU L AT I O N A N D FI B R I N O LY T I C
HDL-cholesterol (45–75%) GENES
Total triglycerides (40–80%)
The most common pathogenetic pathway of acute myocar-
Body mass index (25–60%)
dial infarction is through thrombosis, generally triggered
Blood pressure (50–70%)
by atherosclerotic plaque rupture. Thus, a great number of
Lp(a) levels (90%) candidate gene studies have involved the examination of
Homocysteine levels (45%) genetic variants in genes involved in coagulation and fibri-
Type 2 diabetes (40-80%) nolytic pathways. Ye and colleagues (Ye et  al., 2006)  per-
Fibrinogen (20–50%) formed a meta-analysis of 191 studies to determine the
C-reactive protein (20–50%) relationship between CAD and variants in seven genes
Gender involved in the thrombotic process:  factor V Leiden, fac-
Age tor VII G10976A, prothrombin G20210A, plasminogen
Environmental risk factors: activator inhibitor-1 (PAI-1) [-675] 4G/5G, and three
Smoking
platelet glycoprotein (GP) variants (GPIa, C807T, GPIba
T[-5]C, GPIIIa C1565T). In contrast to an earlier study
Diet
(Boekholdt et  al., 2001), which examined the association
Exercise
of several of the same variants with MI and found that asso-
Infection
ciations for these genetic variants were either weak (PAI-1,
*Adapted from Lusis et al., 2004.
fibrinogen) or absent (factor V, prothrombin), Ye and col-
leagues reported a mild association with factor V Leiden
(OR 1.12, 95% CI 0.91–1.36) and prothrombin (OR 1.91,
addition, the study also established a nearly linear relation- 95% CI 0.91–1.55). Despite this weak and conflicting asso-
ship between APOE status and both LDL cholesterol levels ciation with MI, stronger, consistent associations are seen
and risk for CVD, putting forward the possibility that the with stroke, and these variants are discussed in greater detail
differential risk associated with the different APOE isoforms later in the chapter in that context.
may be mediated by LDL cholesterol, a well-established risk Given the difficulty in identifying genetic variants that
factor for CAD. Indeed, while most of the studies in the play a significant role in susceptibility to MI/IHD through
meta-analysis did not report odds ratios adjusted for lipids, a candidate gene approach, a problem observed for most
and thus the meta-analysis did not determine whether the complex diseases, family-based strategies have also been
APOE variants influence risk for CHD independently of implemented. While many regions have been implicated
lipids, several studies that have examined the data for an through linkage analysis, these studies often result in large
association of the E4 allele after adjusting for lipid levels candidate regions, often comprising hundreds of genes.

Table 21.3  MENDELIAN DISEASES THAT EXHIBIT PREMATURE CORONARY ARTERY DISEASE*

DISEASE GENES EFFECT OF MUTATIONS

Familial hypercholesterolemia LDLR Defective binding of LDL by receptor
Familial defective APOB APOB Reduced binding affinity of APOB to LDLR
Sitosterolemia ABCG5, ABCG8 Increased absorption of plant sterols
Autosomal recessive hypercholesterolemia ARH Defective endocytosis of LDLR
APOA1 deficiency APOA1 Deletion or loss-of-function mutations that lead to very low
HDL
Tangier disease ABCA1 Impaired cholesterol efflux in macrophages
Homocystinuria CBS Leads to increased thrombotic tendency
*Adapted from Watkins et al., 2006.

3 1 8   •   G enomics in C linical Practice

Therefore, their utility in identifying specific gene vari-
ants associated with disease is limited, though some suc-
cesses have been reported, for instance in the association of
genes ALOX5AP and PDE4D. with stroke (details in next
section). With both a rapid reduction in genotyping costs
and a vast increase in throughput, the focus has turned to
genome-wide association studies.
Cardiovascular disease has proven to be an excellent
example to illustrate the power of genome-wide association
studies to find loci that contribute to common complex
disorders. The high incidence of CVD, combined with the
ready availability of large cohorts with detailed data on tra-
ditional cardiovascular risk factors, has led to several large
association studies that have brought the list of loci associ-
ated with either CVD or MI to more than 33. A majority
of these loci come from three large multi-cohort studies.
Samani and colleagues used data from a combined 2801
cases and 4582 controls, replicated the association of the
9p21 region, and identified six additional loci associated
with coronary artery disease (Samani et al., 2007). Four of
these loci were also shown to be associated with myocardial
infarction in a large-scale four-stage study by Kathiresan
and colleagues (Kathiresan et al., 2009), who also reported
three additional novel loci. In the third and largest associa-
tion study for CAD, Schunkert and colleagues (Schunkert
et  al., 2011)  carried out a meta-analysis of 14 studies to
obtain a sample size of 22,233 cases and 64,762 controls of
European descent. In addition to confirming the 10 previ-
ously associated loci, the study also identified 13 novel loci
associated with CAD, only three of which showed any
significant association with traditional CAD risk factors.
From this long list of loci (see Table 21.4), we will focus on
9p21 and SORT1 as specific examples of GWAS discoveries
resulting in interesting insights about biology and disease.
C H RO MO S O M E 9p21
One of the most strongly associated signals from GWAS
has been in the noncoding region on chromosome 9p21.3.
After its initial discovery in a cohort of early-onset MI
patients from the Ottawa Heart Study and deCode proj-
ect (Helgadottir et al., 2007; McPherson et al., 2007), the
region was fine-mapped to a 58kb locus with multiple
tagged SNPs in tight linkage disequilibrium (LD), and it
has shown consistent association with disease independent
of traditional CAD risk factors, in multiple cohorts of
various ethnic backgrounds (Cheng et al., 2011; Gori et al.,
2010; IBC 50K CAD Consortium, 2011; Shen et al., 2008;
Takeuchi et al., 2012; Xie et al., 2011). In a meta-analysis of
47 studies comprising 35,872 cases and 95,837 controls of
G enomics of C om p le x C ardio vascular D isease   •   3 1 9
either European or Asian descent, Palomaki and colleagues
surveyed data from one of three common lead SNPs in
the region (rs1333049, rs10757274, rs2382207) that are
in tight LD, to confirm that the region had a small but
significant effect on risk for MI (OR 1.25, 95% CI 1.21–
1.29) (Palomaki et  al., 2010). It is important to mention
that while these results have been replicated in a number
of European and Asian cohorts, the results from studies
including African-Americans have been more conflicting
(Beckie et  al., 2011; Kral et  al., 2011; Patel et  al., 2011;
Yamagishi et al., 2009). Although the risk locus is devoid of
a protein-coding gene, it lies within a well-described non-
coding RNA in the INK locus (ANRIL) (Pasmant et  al.,
2007) and overlaps with upstream cyclin-dependent kinase
inhibitor genes CDKN2B and CDKN2A, which have been
studied for their role as cell cycle regulators and tumor sup-
pressors. While the genes lie in the same LD block, they
are also approximately 100 kb upstream of the implicated
risk region, and the direct evidence for their putative role
in the modulation of cardiovascular disease did not emerge
until Axel and colleagues developed a mouse model with
a homozygous deletion of the orthologous region on the
mouse chromosome 4, which showed reduced levels of
cardiac expression of both genes (Visel et al., 2010). There
have been numerous studies that have looked at the direct
effect of expression of the CDKN2A, CDKN2B, and
ANRIL (Cunnington et  al., 2010; Folkersen et  al., 2009;
Harismendy et  al., 2011; Jarinova et  al., 2009; Liu et  al.,
2009) with conflicting results on the direction and magni-
tude of the effect of region in the development of CAD.
To add to the confusion, 9p21 has turned out to be
somewhat of a hot spot for GWAS hits for a number of
conditions, including type 2 diabetes (T2D) (Diabetes
Genetics Initiative of Broad Institute of Harvard and MIT,
Lund University, and Novartis Institutes of BioMedical
Research et  al., 2007; Scott et  al., 2007; Zeggini et  al.,
2007; Zeggini et  al., 2008), intracranial aneurysm and
abdominal aortic aneurysms (AAA) (Helgadottir et  al.,
2008), and a number of different cancers (Amos et  al.,
2011; Antoniou et  al., 2012; Enciso-Mora et  al., 2012;
Rajaraman et  al., 2012; Sherborne et  al., 2010; Shete
et al., 2009; Yang et al., 2010). While the link to cancer
can be explained by the presence of CDKN2A/B, two
well-established tumor suppressor genes, the presence of a
strong association to T2D, a traditional CAD risk factor,
paved the ground for the tempting possibility of a shared
biological function of the region that would affect both
conditions. In an elegant study, Helgadottir and colleagues
showed that the region was in fact that tagged by two sep-
arate lead SNPs (rs10757278 for CAD, and rs10811661
Table 21.4  CANDIDATE LOCI FROM GWAS OF EITHER MYOCARDIAL INFARCTION OR CORONARY ARTERY
DISEASE

CHR GENES OF INTEREST WITHIN OR SNP ODDS RATIO (95% CI) ASSOCIATION OF SNP OR PROXY WITH
NEAR ASSOCIATED INTERVAL PER RISK ALLELE OTHER CARDIOVASCULAR PHENOTYPES
1p13 CELSR2, PSRC1, SORT1 rs599839 1.11 (1.08–1.15) eQTL for SORT1, CELSR2, and PSRC1 tran-
script levels in liver
1p32 PCSK9 rs11206510 1.08 (1.05–1.11)
1p32 PPAP2B rs17114036 1.17 (1.13–1.22)
1q41 MIA3 rs17465637 1.14 (1.09–1.20)
2p21 ABCG5, ABCG8 rs4299376 1.07 (1.04–1.11) Serum phytosterols
2q33 WDR12, NBEAL1 rs6725887 1.14 (1.09–1.19) eQTL for NBEAL1 transcript level in aortic
media
3q22 MRAS rs9818870 1.12 (1.07–1.16) eQTL for MRAS transcript level in aortic media
6p21 ANKS1A rs17609940 1.07 (1.05–1.10)
6p24 PHACTR1 rs12526453 1.10 (1.06–1.13) Coronary artery calcification
6p24 c6orf105 rs6903956 1.65 (1.44–1.90)
6q23 TCF21 rs12190287 1.08 (1.06–1.10) eQTL for TCF21 transcript level in liver, fat
6q25 LPA rs3798220 1.51 (1.33–1.70) Lipoprotein(a)
6q26 LPA rs10455872 1.68 (1.43–1.98) Lipoprotein(a)
7q22 BCAP29, DUS4L rs10953541 1.08 (1.05–1.11)
7q32 ZC3HC1 rs11556924 1.09 (1.07–1.12)
8q24 TRIB1 rs17321515 1.06 (1.03–1.10) Triglycerides, HDL cholesterol
9p21 CDKN2A, CDKN2B, ANRIL rs4977574 1.29 (1.23–1.36) Coronary artery calcification, intracranial aneu-
rysm, abdominal aortic aneurysm, among others
9q34 ABO rs579459 1.10 (1.07–1.13) Venous thromboembolism, ACE enzyme activ-
ity, plasma E-selectin level, plasma vWF level,
among others
10p11 KIAA1462 rs2505083 1.07 (1.04–1.09)
10q11 CXCL12 rs1746048 1.09 (1.07–1.13) Plasma CXCL12 level
10q23 LIPA rs1412444 1.09 (1.07–1.12) eQTL for LIPA transcript level in monocytes
10q24 CYP17A1, CNNM2, NT5C2 rs12413409 1.12 (1.08–1.16) Intracranial aneurysm
11q22 PDGFD rs974819 1.07 (1.04–1.09) eQTL for PDGFD transcript level in aortic
media
11q23 ZNF259, APOA5, APOA1 rs964184 1.13 (1.10–1.16) Triglycerides, HDL cholesterol
12q24 SH2B3 rs3184504 1.07 (1.04–1.10) LDL cholesterol, platelet count, plasma eosino-
phil count, among others
13q34 COL4A1, COL4A2 rs4773144 1.07 (1.05–1.09)
14q32 HHIPL1 rs2895811 1.07 (1.05–1.10)
15q25 ADAMTS7 rs3825807 1.08 (1.06–1.10)
17p11 RASD1, PEMT, RAI1 rs12936587 1.07 (1.05–1.09) eQTL for RASD1 and PEMT transcript levels
in monocytes
17p13 SMG6 rs216172 1.07 (1.05–1.09)
17q21 UBE2Z rs46522 1.06 (1.04–1.08) eQTL for UBE2Z transcript level in blood
19p13 LDLR rs6511720 1.18 (1.11–1.25)
19q13 APOE rs2075650 1.14 (1.09–1.19)
21q22 SLC5A3, MRPS6, KCNE2 rs9982601 1.18 (1.12–1.24) eQTL for MRPS6 transcript level in blood
*Adapted from Kathiresan and Srivastava, 2012.

3 2 0   •   G enomics in C linical Practice

for T2D) which were in adjoining LD blocks (Helgadottir
et  al., 2008). While the SNP associated with CAD also
showed association with five other arterial diseases, with
the strongest association being AAA (OR  =  1.31, 95%
CI 1.22–1.42) and intracranial aneurysms (OR  =  1.29,
95% CI 1.16–1.43), rs10811661, the T2D SNP, showed
no significant association with either CAD or the other
arterial diseases. This result has since been replicated in
several studies (Biros et al., 2010; Bown et al., 2008; Gori
et al., 2010) and highlights the complexity of the genomic
region and makes the process of going from an associated
locus to a causal gene even more challenging. While the
mechanism by which the risk “region” affects atheroscle-
rotic processes remains to be determined, the presence of
a consistent association in multiple cohorts makes 9p21 a
prominent area of research efforts.
S O RT I L I N1 (SO RT1)
The elucidation of the role of Sortilin1, a multi-ligand
sorting receptor in cholesterol metabolism, highlights the
power of GWAS findings when put in the context of appro-
priate functional studies.
Samani and colleagues first implicated the 1p31 region
in a GWAS for CAD in 2007 (OR 1.29, 95% CI 1.10–
1.21) (Samani et al., 2007), which has since been replicated
in a number of large-scale GWAS for both MI and CAD
(Kathiresan et al., 2009; Schunkert et al., 2011). The SNPs
tagged in these studies lie in a noncoding region, between
genes PSRC1 and CELSR1, and in the same LD block as
SORT1, making identification of the causal variant and the
gene mediating the association challenging.
SORT1 was first implicated as the putative gene of inter-
est by Linsel-Nitschke and colleagues (Linsel-Nitschke
et  al., 2010). This work identified an expression quan-
titative trait locus (eQTL), uncovering an association
between one of the GWAS SNPs, rs599839, and SORT1
mRNA levels. In addition, they demonstrated a significant
increase in LDL cholesterol uptake in HEK293 cells with
over-expression of SORT1 and laid the foundation for the
hypothesis that increased SORT1 expression could poten-
tially have a protective role in CAD. The role of SORT1
in LDL cholesterol metabolism was further investigated
by Musunuru and colleagues in a series of experiments
that showed the direct effect of Sort1 knockdown and
over-expression on plasma LDL cholesterol and lipopro-
tein levels in murine models of atherosclerosis. They also
presented evidence to show that a common polymor-
phism, rs12740374, previously shown to be associated
with CAD, creates a novel CCAAT-enhancer-binding
G enomics of C om p le x C ardio vascular D isease   •   3 2 1
protein (C/EBP) binding site and hence mediates an
increased LDL cholesterol level via a hepatic secretory
pathway. However, in direct contrast, Kjolby et al. showed
that in a Sort1, Ldlr double knockout mouse model,
plasma LDL cholesterol levels, and thus the atheroscle-
rotic burden, are reduced when compared to a LDLR-null
mouse (Kjolby et al., 2010). Considering the difference in
mouse manipulations and timing of over-expression, the
differences are not altogether surprising, but they defi-
nitely raise questions about the role of SORT1 in the biol-
ogy of cholesterol metabolism and its subsequent effect on
disease processes.
The success of GWAS is demonstrated by the fact that,
in addition to identifying genes that are implicated as puta-
tive candidates because of their biological relevance in the
disease process, we are now also able to identify a long
list of genes that show a strong, definite association with
CVD through yet-unknown mechanisms. Indeed, in a
recent review, Kathiresan and Srivastava (Kathiresan and
Srivastava, 2012) divided the list of loci mapped by GWAS
into those known to be associated with an established risk
factor like LDL-cholesterol or blood pressure, and those
where the association is established but the mechanism is
unknown.
Herein lies the challenge before us: to improve our abil-
ity to identify the causal variant, and thereby the underly-
ing mechanism of action. Over the next couple of years, we
will need to make a strong effort to determine the complete
list of associated loci, and more importantly, to use this list
to direct research efforts to better understand the biology
of normal cardiovascular processes and myocardial infarc-
tion. Aside from identifying novel molecular pathways, the
importance of these genetic variants from a clinical stand-
point is often hard to interpret, given the small effect size
of most GWAS hits and the ability to easily measure tra-
ditional cardiovascular risk factors. The argument in favor
of using variants that have shown repeated association in
several studies for either risk prediction has been shaky at
best, with different groups reporting conflicting levels of
success in using genetic data for risk prediction (Drenos
et al., 2007; Kathiresan et al., 2008; Kathiresan et al., 2009;
Paynter et al., 2010). In a large literature-survey-based test
of a “genetic score” combining the effects of 101 SNPs with
traditional cardiovascular risk factors, Paynter and col-
leagues showed no significant improvement in risk deter-
mination or classification in a cohort of 19,213 women
(Paynter et al., 2010). However, it is important to mention
that while these data suggest that genetic variants, at this
point in time, have little value in cross-sectional measures
of risk, there is strong evidence to suggest that genetic data
might provide a better measure of lifetime risk than indi-
vidual cross-sectional measures of risk factors. This has been
successfully shown by Cohen and colleagues, who observed
significantly reduced levels of plasma LDL cholesterol
(~15%) and incidence of cardiovascular disease in carriers
of the Arg47Leu allele of PCSK9 (hazard ratio 0.50, 95%
CI 0.32–0.79) (Cohen et  al., 2006). While the expected
reduction of the CVD risk corresponding to the decrease
in LDL-c levels that they observed is ~23%, analysis of the
Atherosclerosis Risk in Communities (ARIC) prospective
cohort showed a 47% decrease in R47L carriers, suggest-
ing that there is an accumulated lifetime burden of reduced
LDL-c levels that is not accurately captured by a single mea-
sure of lipid levels.
STROK E
Stroke is one of the leading causes of death and disability
in the developed world, with annual incidence of 795,000
(Roger et  al., 2012). With limited treatment options
available, focus has been on primary prevention, largely
through modification of acquired risk factors (diabetes
mellitus, smoking, high blood pressure, and atrial fibril-
lation) (Goldstein et  al., 2001). However, as with many
common diseases, rare monogenic conditions that cause
stroke have been identified, and a great deal of progress has
been made in identifying the underlying genetic defects.
Indeed, the gene for cerebral autosomal dominant arteri-
opathy with subcortical infarcts and leukoencephalopathy
(CADASIL), which has served as a model for inherited
ischemic stroke, was identified as NOTCH3 in 1996 ( Joutel
et al., 1996). More recently, substantial progress has been
made in cerebral cavernous malformations (CCM), with
genes for all three types of CCM now identified. CCM1
is caused by mutations in KRIT (Laberge-le Couteulx
et al., 1999); CCM2 is caused by mutations in MGC4607
(malcavernin) (Denier et al., 2004); and CCM3 is caused
by Programmed cell-death protein 10 (Bergametti et  al.,
2005). While these genetic defects are rare in the general
population, they have high penetrance; therefore, being
able to identify carriers has a significant clinical impact.
Whether any of these genes has a role in common forms
of stroke remains to be determined, though a study by
Dong and colleagues, in which they screened individuals
with lacunar stroke for coding mutations in NOTCH3, did
not find any association (Dong et al., 2003). This does not,
however, rule out the involvement of common variants,
more likely to be found in noncoding conserved regions
and involved in gene regulation.
3 2 2   •   G enomics in C linical Practice
While rare, cerebral venous thrombosis (CVT), which
accounts for less than 1% of all strokes, is a complex dis-
ease, with numerous etiological risk factors (for review, see
Agostoni et al., 2009), including genetic factors. Marjot and
colleagues (Marjot et al., 2011) conducted a comprehensive
meta-analysis of candidate gene studies for CVT, identify-
ing 26 case-control studies covering six polymorphisms in
six genes. With a sample size of 1183 CVT cases and 5189
controls, they demonstrated significant associations for
three genes (discussed below), including factor V Leiden/
G1691A (OR 2.40, 95% CI 1.75–3.30), prothrombin/
G20210A (OR 5.48, 95% CI 3.88–7.74), and MTHFR/
C677T (OR 2.30, 95% CI 1.20–4.42).
FAC TO R V L E I D E N
Bertina and colleagues (Bertina et al., 1994) identified an
arginine-to-glycine (R506Q) mutation (termed “Leiden
allele”) in factor V in a family with activated protein C
(APC) resistance and prone to thrombosis. APC limits
clot formation by inactivation of factors Va and VIIIa,
and the Leiden mutation is predicted to alter the amino
acid at the APC cleavage site in factor V, causing factor
V to be less efficiently degraded. Thus, individuals who
carry factor V Leiden have increased thrombin genera-
tion and a hypercoaguable state, which could explain
the increased risk for stroke associated with this allele
(Dahlback, 1995).
P ROT H RO M B I N
Poort and colleagues (Poort et  al., 1996)  first identi-
fied the G20210A single-base-pair substitution in the 3′
untranslated region of the parent gene, coagulation factor
II (F2), and demonstrated its association with elevated
plasma prothrombin levels and increased risk for venous
thrombosis. Subsequent studies have demonstrated
that prothrombin levels were probably due to increased
thrombin generation (Franco et al., 1999). More recently,
the mechanism by which G20210A alters prothrombin
levels has been established. Ceelie and colleagues (Ceelie
et al., 2004) have demonstrated that the G20210A muta-
tion results in a more effective poly (A) site (a poly [A]‌
tail is required for mRNA to be efficiently exported from
the nucleus and translated into protein in the cytoplasm),
leading to elevated mRNA levels, resulting in increased
prothrombin production and thrombin formation. Thus,
like factor V Leiden, the G20210A mutation probably
leads to a pro-coagulant state, thereby increasing risk of
stroke.
 M ET H Y L E N ET ET R A H Y D RO F O L AT E
R E D U C TA S E (MTHFR)
MTHFR catalyzes the conversion of 5,10-
methylene­­­­tetrahydro­folate to 5-methyltetrahydrofolate,
a cosubstrate for homocysteine remethylation to methio-
nine. Frosst and colleagues (Frosst et al., 1995) identified
a cytosine-to-thymine single-base-pair substitution at
position 677 (C677T) that converts an alanine to a valine
residue, and produces a thermolabile form of the protein.
They demonstrated that this variant was associated with
reduced enzyme activity and increased levels of homo-
cysteine. Elevated levels of homocysteine are associated
with increased risk for stroke (Wald et  al., 2002), thus,
the C677T variant is likely to contribute to increased risk
of stroke directly due to its reduced ability to metabolize
homocysteine.
More common forms of stroke can be divided into two
major varieties:  ischemic and hemorrhagic. The majority
of strokes are ischemic (80–90%), and can be further sub-
divided into:  1)  large-vessel occlusive disease, usually due
to atherosclerosis and plaque formation; 2)  small-vessel
occlusive disease, which have involvement of small, perfo-
rating end-arteries in the brain; and 3) cardiogenic stroke,
which is secondary to blood clots from a diseased heart.
Traditionally, occlusive disease has been considered to be
due to atherosclerosis, and cardiogenic stroke due to atrial
fibrillation secondary to mitral-valve stenosis; however,
arguments have been put forward that all three forms prob-
ably have a significant atherosclerotic component (Gulcher
et al., 2005). Given the heterogeneous nature of the stroke
phenotype, assumptions about the underlying etiology of
disease can have a significant impact on the ability to iden-
tify genetic determinants of stroke susceptibility.
The leap from the existence of genes for monogenic and
rare forms of stroke to the likely existence of genes contrib-
uting to risk for common forms of stroke is bolstered by
numerous studies that have shown that a genetic compo-
nent to susceptibility to common forms of stroke probably
exists. A comprehensive analysis of these studies, in which
all genetic epidemiology studies of ischemic stroke from
1966 to 2003 were systematically reviewed, was conducted
by Flossmann and colleagues (Flossmann et  al., 2004).
Based on twin (OR 1.65, 95% CI 1.2–2.3), case-control
(OR 1.76, 95% CI 1.7–1.9), and cohort (OR 1.30, 95%
CI 1.2–1.5) studies, they concluded that there is a modest
but significant genetic component to the risk for ischemic
stroke in the general population. Most genetic studies have
focused on candidate genes under case-control designs,
and Casas and colleagues (Casas et al., 2004) performed a
G enomics of C om p le x C ardio vascular D isease   •   3 2 3
meta-analysis of these studies on ischemic stroke, incorpo-
rating 32 genes studies across ~18,000 cases and ~58,000
controls. They identified significant associations for factor
V Leiden (OR 1.33, 95% CI 1.12–1.58), methylenetetra-
hydrofolatereductase (MTHFR) C677T (OR = 1.24, 95%
CI 1.08–1.42), prothrombin G20210A (OR = 1.44, 95%
CI 1.11–1.86) (discussed above, in the context of CVT),
as well angiotensin-converting enzyme (ACE) insertion/
deletion (OR 1.21, 95% CI 1.08–1.35). A  more recent
meta-analysis conducted by Hamzi et  al. (Hamzi et  al.,
2011)  reviewed 300 manuscripts for five candidate genes
among 152,797 individuals (45,433 cases and 107,634
controls) and confirmed the associations for prothrombin
(OR 1.57, 95% CI 1.23–2.89) and ACE (OR 1.11, 95% CI
1.02–1.25), but did not find significant results for factor
V Leiden or MTHFR for ischemic stroke. Therefore, they
concluded that there are common variants in several genes
involved in common forms of stroke, each with a modest
effect. Meta-analyses of the ACE insertion/deletion vari-
ant (see below) in non–European descent individuals also
revealed a significant association with ischemic stroke, indi-
cating the importance of this variant across different ethnic
groups (Ariyaratnam et al., 2007; Wang et al., 2012).
A N G I OT E N S I N- C O N VE RT I N G
E N ZY M E (ACE)
ACE plays an important role in blood pressure regulation
and electrolyte balance, and ACE inhibitors have been
at the forefront of therapy for treating hypertension and
reducing risk for CVD. Indeed, ACE is an important regu-
lator of the renin-angiotensin-aldosterone system through
both its ability to hydrolyze angiotensin I  into angioten-
sin II, a potent vasopressor, and its ability to inactivate
bradykinin, a potent vasodilator that may stimulate nitric
oxide production (Kim and Iwao, 2000). ACE is found on
the surface of vascular endothelial cells and in circulating
plasma, and animal studies have shown the importance
of ACE in regulating blood pressure (Esther et  al., 1997;
Krege et  al., 1995). In 1990, Rigat and colleagues (Rigat
et al., 1990) identified an insertion/deletion polymorphism
(I/D) that was responsible for up to 50% of the variation
in circulating levels of ACE. While the molecular basis of
how the I/D polymorphism affects circulating ACE levels
is not entirely clear, a study using nearby polymorphisms to
measure specific expression of the I and D alleles indicates
that the D allele leads to higher expression of ACE mRNA
(Suehiro et al., 2004).
Given the widespread use of ACE inhibitors in clinical
treatment, and the high frequency of the I/D polymorphism
in the general population (~30% in Caucasian popula-
tions), the ACE I/D polymorphism provides a prime tar-
get for testing the potential impact of a genetic variant on
choice of drug therapy (pharmacogenetics). Arnett and
colleagues (Arnett et al., 2005) have reported the results of
a double-blind, active-controlled randomized trial of anti-
hypertensive treatment in which they examined the impact
of the ACE I/D polymorphism on response to four differ-
ent medications (chlorthalidone, amlodipine, lisinopril,
and doxazosin). The study included 37,939 participants
≥55 years of age with ≥1 risk factor for CVD. These individ-
uals were followed up for four to eight years, with primary
outcomes including fatal coronary heart disease (CHD)
and/or nonfatal MI, and secondary outcomes including
stroke, all-cause mortality, combined CHD, and combined
cardiovascular disease. ACE I/D genotype was not predic-
tive for CHD (though the risk for stroke was consistent
with the meta-analysis of Casas et  al., 2004), nor did it
modify the response to treatment with the different anti-
hypertensive medications. These results were surprising, as
one would predict that those with the D/D genotype, and
therefore higher ACE levels, would be more responsive to
the ACE inhibitor therapy (lisinopril). Indeed, these results
provide a warning for making the leap between genetics and
treatment. Despite a functional variant in a gene whose
product is a direct target of one of the therapies, the choice
of therapy did not affect the outcome. There were however,
some differences in outcome according to gender and dia-
betes status, but given the number of hypotheses tested, fur-
ther follow-up is needed to verify any of these observations.
In addition to candidate-gene studies that led to the rev-
elation of the association of the above-discussed genes with
stroke, traditional genetic approaches have also focused on
using large pedigrees in family-based linkage studies. While
these studies have been at the forefront of identifying genes
for Mendelian diseases, they are often difficult to implement
for phenotypes that occur late in life, such as stroke, due to
the difficulty in obtaining informative pedigrees. However,
deCODE Genetics has leveraged the combination of exten-
sive genealogical records and medical records in Iceland to
be able to perform these types of studies.
P H O S P H O D I E S T E R A S E 4D (PD E4D)
Gretarsdottir and colleagues (Gretarsdottir et al., 2002) ini-
tially performed a genome-wide linkage scan in 476
patients with stroke within 179 extended pedigrees from
Iceland, and identified a locus on 5q12 (Log of odds [LOD]
score  =  4.40), which they designated as “STRK1.” They
employed a broad definition of stroke, including individuals
3 2 4   •   G enomics in C linical Practice
with either ischemic or hemorrhagic stroke, as well as TIA
(transient ischemic attack), which they considered an isch-
emic event, arguing that the same pathophysiological mech-
anisms are responsible for both. In a subsequent study, the
same group fine-mapped this locus using a population-based
case-control study composed of 864 Icelandic affected indi-
viduals and 908 controls, implicating PDE4D, a regulator
of intracellular levels of cyclic adenosine monophosphate
(cAMP) (Gretarsdottir et  al., 2003). PDE4D mRNA is
expressed in cardiac myocytes and may be involved in excita-
tion–contraction coupling (Lehnart et al., 2005). However,
this association was limited to ischemic stroke, and specifi-
cally to the combined cardiogenic and carotid forms (using
Trial of Org 10172 in Acute Stroke Treatment [TOAST]
subcategories). Subsequent studies have yielded ambiguous
replication results, and two large meta-analyses, with the
most recent containing >10,000 cases and >10,000 con-
trols, have not demonstrated an association with ischemic
stroke (Bevan et al., 2008; Lovkvist et al., 2012).
A R AC H I D O NAT E
5-L I P O OX YG E NA S E -AC T I VAT I N G P ROT E I N
(A LOX5A P)
Helgadottir and colleagues (Helgadottir et  al., 2004)
reported a finding of linkage and association with
ALOX5AP and both stroke and MI in an Icelandic popu-
lation. They identified a specific haplotype (HapA) that
is relatively common and carried in 27% of patients with
stroke (Relative risk [RR] 1.7, P <0.0001). The associa-
tion with both MI and stroke was particularly intriguing,
and in the same paper Helgadottir and colleagues identi-
fied another haploytpe (HapB) that was associated with
MI in an independent British cohort. They went on to
demonstrate that the synthesis of leukotriene B4 (LTB4)
in ionomycin-stimulated neutrophils from patients with
a history of MI is greater than from controls without MI
and that this difference is largely accounted for by carriers
of the HapA haplotype. LTB4, which is a biolipid inflam-
matory and vasoactive mediator, is produced from LTA4,
which is produced from arachidonic acid by ALOX5AP.
Thus, the association study is backed by functional evi-
dence that points to a specific mechanism by which vari-
ants in ALOX5AP may increase risk for both stroke and
MI: elevated levels of LTB4 might contribute to increased
inflammation, a known risk factor for CVD events through
the development and atherosclerosis and/or plaque insta-
bility. As with PDE4D, replication studies have yielded
ambiguous results, and the largest meta-analysis to date,
with >5,000 cases and >4,500 controls, does not show a
significant association of the HapA haplotype and ischemic Table 21.5  GENETIC LOCI ASSOCIATED WITH
stroke (Zintzaras et al., 2009). STROKE
A.  LOCI/GENES FROM CANDIDATE-GENE APPROACH

GENE CHR ALLELE/SNP

PIT X2 A N D Z FH X3
ACE 17 . . .
As noted above for MI, the advent of GWAS has begun to ALOX5AP 13 rs17216473
change the landscape of stroke genetics (see Table 21.5). The Angiopoietin-1 8 rs2507800
first successful GWAS for ischemic stroke (Gretarsdottir
APOA (LPA) 6 . . .
et  al., 2008)  identified a signal near PITX2 (paired-like
APOE 19 . . .
homeodomain transcription factor 2), which had previously
CRP 1 rs2794521
been associated with atrial fibrillation (AF) (Gudbjartsson
CYP4AII 1 . . .
et al., 2007). AF, one of the most common forms of electri-
cal instability, is characterized by chaotic electrical activity CYP4F2 19 rs2108622
of the atria, and plays a major role in cardioembolic stroke CYPIIB2 8 rs1799998
(Lip and Tse, 2007). While the association was initially DDAH1 1 . . .
also reported for non-cardiogenic stroke (Gretarsdottir eNOS (NOS3) 7 rs1799983
et al., 2008), subsequent studies have validated the associa- Factor V Leiden 1 . . .
tion with cardiogenic stroke (International Stroke Genetics Fibrinogen 4 . . .
Consortium [ISGC] et al., 2012), but not overall ischemic GP1BA 17 . . .
stroke (Carty et  al., 2012). The utility of using AF as an GPIIIa 17 . . .
endophenotype to identify cardiogenic stroke–associ- IL-6 7 . . .
ated variants is further validated by the observation of an
LTC4S2 5 rs730012
association between stroke and SNPs at the ZFHX3 locus
MTHFR 1 rs2274976
(Gudbjartsson et al., 2007; ISGC et al., 2012), which had
NPY 7 rs16147
previously been associated with AF (Benjamin et al., 2009;
PAI-1 (Serpine) 7 . . .
ISGC et al., 2012).
Paroxonase-1 7 rs662
PDE4D 5 rs12188950
HDAC9 Prothrombin 11 . . .
In the first study of prospectively identified stroke in the SGK1 6 rs1057293
general population, Ikram and colleagues identified a single TNF-Alpha 10 . . .
locus, NINJ2 (Ikram et  al., 2009). NINJ2 encodes nin- VKORC1 16 rs9923231
jurin2, an adhesion molecule that is upregulated in response
B.  LOCI FROM GWAS FOR ISCHEMIC STROKE
to nerve injury. However, this locus was not validated by
subsequent studies (ISGC et al., 2012; ISGC and Wellcome GENES CHR SNP
Trust Case-Control Consortium 2, 2010). A more recent CDKN2A/B 9 rs4977574
GWAS for ischemic stroke, conducted in >9,000 cases and NINJ2 12 rs11833579
>11,000 controls, replicated associations for cardioembolic PCSK9 1 rs11206510
stroke near PITX2 and ZFHX3, as well for the 9p21 locus PITX2 4 rs1906591
(previously implicated in MI; see above) with large-vessel PRKCH 14 rs2230500
disease (Gschwendtner et  al., 2009). They also reported a
ZFHX3 16 rs7193343
novel finding with the HDAC9 locus and large vessel stroke
*Adapted from Bevan et al., 2012.
(OR 1.42, 95% CI 1.28–1.57). HDAC9 encodes histone
deacetylase 9, an enzyme involved in regulating chromatin
structure and gene transcription (Haberland et al., 2009). stroke subtypes. The PITX2 and ZFHX3 variants appear
Using a Bayesian statistical framework to formally test to only affect risk for cardioembolic stroke, whereas the
whether different variants were associated with all subtypes 9p21 variants appear to broadly influence ischemic stroke,
of ischemic stroke or specific subtypes, they clearly demon- and HDAC9 is specific for large-vessel stroke. Based on
strate the importance of subtype classification and provide this specificity, the authors postulate that a mechanism for
strong evidence of heterogeneity of genetic effects across association of HDAC9 with stroke through accelerated

G enomics of C om p le x C ardio vascular D isease   •   3 2 5

atherosclerosis is possible, but note that this hypothesis is
highly speculative.
In an example of population-specific findings, a
gene-based association study using ~50,000 tag SNPs was
conducted in ~1100 Japanese cerebral infarct cases, and
identified a nonsynonymous SNP in PRKCH (V374I),
which encodes protein kinase C eta (Kubo et al., 2007). The
authors go on to show that the identified variant appears
to be functional, resulting in higher autophosphorylation
and kinase activity, which activates its downstream signal-
ing pathway. The variant has a frequency of about 20% in
Asian populations, but less than 1% in European ances-
try individuals, and is not observed in the Yoruba from
Ibadan (International HapMap Consortium, 2005). This
finding has been validated in a Chinese population (Wu
et al., 2009), as well as in a meta-analysis of >3,600 cases of
ischemic stroke and >4,500 controls drawn from Chinese
and Japanese populations (Li et  al., 2012). PRKCH is a
serine-theronine kinase that is mainly expressed in vascular
endothelial cells and foamy macrophages (which play an
important role in atherosclerosis), and is involved in regu-
lation of cell differentiation, proliferation, and apoptosis.
Indeed, increased expression of PRKCH was correlated
with progression of coronary atherosclerotic lesion type,
providing strong evidence for a mechanism by which altered
PRKCH levels influence risk of stroke (Kubo et al., 2007).
I N T R AC R A N I A L A N EU RY S M
A series of GWAS have focused on intracranial aneurysm,
a major cause of hemorrhagic stroke, and identified six
loci using discovery and replication cohorts from Europe
and Japan comprising >5,800 cases and >14,000 con-
trols (Bilguvar et  al., 2008; Yasuno et  al., 2010; Yasuno
et al., 2011). These loci include the 9p21 locus, which is
implicated in CAD and large-vessel ischemic stroke (see
above). Additional loci include SOX17, CNNM2, KL/
STARD13, RBBP8, and EDNRA. The author note that a
common pathway tying these genes together is cell cycle
progression, and these genes may affect proliferation and
senescence of progenitor-cell populations (Yasuno et al.,
2010). Indeed, SOX17 is required for endothelial forma-
tion and maintenance, and Sox17-/- mice show vascular
abnormalities (Sakamoto et  al., 2007). EDNRA, which
encodes a G protein-coupled receptor for endothelin,
is a particularly intriguing candidate, as it mediates the
vasoconstriction and mitogenic effects of EDN1 (Alberts
et al., 1994; Suzuki et al., 1999). Yasuno and colleagues
(Yasuno et  al., 2011)  note that the effects of EDNRA
variants on IA risk may occur in two distinct ways,
3 2 6   •   G enomics in C linical Practice
depending on whether the variant increases or decreases
EDNRA-mediated signaling. An increase in signal might
promote development of atherosclerosis, whereas a
decreased signal might lead to an inability to adequately
repair the vasculature after vascular injury. Understanding
the specific nature of the risk variant may have impor-
tant pharmacogenetic implications, as selective EDNRA
antagonists are in clinical trials for treatment of subarach-
noid hemorrhage (clazosentan) (Kramer and Fletcher,
2009; Macdonald et  al., 2008; Macdonald et  al., 2013;
Vergouwen et al., 2012).
S U D D E N C A R D I AC D E AT H ( S C D )
Sudden cardiac death continues to be one of the lead-
ing causes of death in the United States. According to the
U.S. Centers for Disease Control and Prevention, about
462,000 of the 2,400,000 (19.3%) U.S. deaths in 1999 were
classified as “sudden cardiac deaths” (SCDs) using their
definition of SCD as including all deaths “due to cardiac
disease that occurred out of hospital (~341,000) or in an
emergency department, or one in which the decedent was
reported ‘dead on arrival’ ” (Zheng et al., 2002). From the
standpoint of preventive care, SCD poses a huge burden,
since fewer than 10% of SCD victims survive, and approxi-
mately one-third of all victims manifest SCD as their first
clinical event. Approximately two-thirds of SCD victims do
not have clinical symptoms that would warrant preventive
intervention. Therefore, the ability to identify individuals
who are at high risk for SCD is crucial, and advances in
genetics may fill this gap.
As for stroke and MI, a great deal of progress has been
made identifying the genes involved in Mendelian forms of
disease that contribute to increased risk of SCD. Mutations
in coding sequences in at least seven cardiac sarcolemmal,
sodium, potassium, and calcium ion channels subunit genes
(i.e., KVLQT1 [KCNQ1], HERG [KCNH2], SCN5A,
minK [KCNE1], RYR2, MiRP1 [KCNE2], and Kir2.1
[KCNJ2]) result in increased susceptibility to SCD (Priori
and Napolitano, 2004). Electrophysiological dysfunc-
tion, manifested as delayed myocardial cell depolarization
and repolarization, is caused by mutations in the proteins
encoded by these genes, as was originally discovered by
Keating, Schwartz, Moss, Priori, and others during the
1990s, and it is now known to underlie a whole family of
related pro-arrhythmic conditions, exemplified by the Long
QT syndrome (LQTS) and Brugada’s syndrome (Keating
and Sanguinetti, 2001; Splawski et  al., 2000). Mutations
in these same genes also result in converse disorders, such
as “short QT syndrome,” which also enhances SCD risk
(Brugada et al., 2004; Gaita et al., 2003).
Multiple etiologies probably contribute to increased
SCD risk, including susceptibilities that arise from genetic
variations in sarcomeric proteins, such as beta-myosin heavy
chain (MyHC), cardiac troponin T (cTnT), and myosin
binding protein-C (MyBP-C), which underlie SCDs that
occur in patients with inherited hypertrophic cardiomyopa-
thies (Marian and Roberts, 2001). Another important factor
is genetic changes that impact early patterning events during
embryogenesis, and subsequently cause disturbances in car-
diac electrical function from development through matura-
tion. Chien and collaborators were among the first to report
that genetic alterations affecting early transcription-factor
expression may lead to enhanced arrhythmia susceptibility
(Nguyen-Tran et al., 2000). Similar alterations in develop-
mental factors were also implicated in rare familial vascular
defects that appear to result in enhanced susceptibility to
myocardial infarction (Wang et al., 2003).
The existence of rare inherited monogenic disorders,
such as the LQTS and Brugada syndromes, demonstrate
that mutations in ion-channel genes, structural proteins, and
calcium-handling genes can increase susceptibility to lethal
arrhythmias (Keating and Sanguinetti, 2001; Splawski et al.,
2000). It is thus a small leap to propose that subtler varia-
tions in these same genes may predispose to more common
forms of SCD. Two population-based studies have demon-
strated that a family history of SCD increases the risk for
SCD approximately 1.8-fold, independent of other tradi-
tional cardiovascular disease risk factors (Friedlander et al.,
1998; Jouven et al., 1999). The first study, conducted in the
United States by Friedlander and colleagues (Friedlander
et  al., 1998), analyzed associations with “primary cardiac
events” in a cohort of men and women attended by first
responders in King County, Washington (235 cases, 374
controls). The second, done in Paris by Jouven and col-
leagues ( Jouven et al., 1999), analyzed deaths in a cohort of
7,746 asymptomatic middle-aged males, using retrospective
autopsy and clinical data analyses to ascribe cardiac deaths
to either SCD or MI. Multifactorial statistical analyses indi-
cated that the occurrence of SCD in a parent results in a
1.6–1.8-fold increase in SCD susceptibility, despite con-
trolling for conventional risk factors indicative of coronary
disease (e.g., cholesterol sub-fractions, blood pressure, obe-
sity, tobacco use, etc.). In a very limited number of cases in
the Parisian study, where there was a history of both mater-
nal and paternal SCD events (n = 19), the relative risk in
offspring was ~9 (P = 0.01), indicating an additive genetic
model. Elevated incidence of SCD in the Paris study seg-
regated independently of elevated familial incidence of
G enomics of C om p le x C ardio vascular D isease   •   3 2 7
myocardial infarction, suggesting genetic factors that spe-
cifically associate with risk for SCD, rather than factors that
may underlie overall CVD risk (e.g., atherosclerosis).
α NAV1.5 S U BU N I T (SCN5A)
The idea that ion-channel sequence variations that alter
cardiac de- or re-polarizing currents in patients with rare
inherited syndromes, like LQTS, may also contribute
to enhanced SCD susceptibility seen in more common
forms of cardiac disease represents an attractive hypoth-
esis. Splawski and colleagues (Splawski et  al., 2002)  have
given some insight into this paradigm. In their study, a
single-nucleotide sequence variant in the SCN5A Na chan-
nel gene found in African Americans, S1102Y, was associ-
ated with a modest enhancement of arrhythmia risk. The
aberrant allele was estimated to be present in up to 4.6 mil-
lion African Americans: a level of prevalence far beyond all
previously established SCD susceptibility alleles combined,
and it was not identified in other ethnic populations sam-
pled. In in vitro transfection experiments, the Y1102 allele
accelerated channel activation, providing a plausible mecha-
nism by which this variant may increase the likelihood of
altered cardiac repolarization and arrhythmia. This finding
was followed up in a series of 289 sudden deaths in blacks
by Burke and colleagues (Burke et  al., 2005). Individuals
were classified into four categories: 1) controls, 2) cardiac
deaths with clear anatomical substrate, 3)  cardiac deaths
with no anatomical substrate except mild to moderate car-
diac hypertrophy, and 4) unexplained cardiac arrhythmias.
The frequency of the Y1102 allele was significantly higher
in those with SCD in the absence of a clear morphological
abnormality (categories 3 and 4, combined n = 65). These
findings strongly suggest that this allele is a risk factor for
SCD in African Americans, but require confirmation in a
larger cohort.
SCN5A variants have also been studied in non–African
ancestry populations, yielding mixed results. An examina-
tion of all coding exons in 67 SCD cases with known cor-
onary artery disease and 91 CAD controls in the Oregon
Sudden Unexpected Death Study (primarily of European
descent) found no association between coding variants and
SCD risk (Stecker et al., 2006). In a Han Chinese popula-
tion, the A1673G variant in SCN5A has been observed to
modify risk for SCD (Chen et al., 2004; Fang et al., 2008),
though this SNP has not been found to be associated in
other populations with SCD (Doolan et al., 2008; Stecker
et  al., 2006). Finally, in a sequencing study of SCN5A
and four potassium-channel genes (KCNQ1, KCNH2,
KCNE1, and KCNE2), Albert and colleagues (Albert et al.,
2008)  identified a significantly higher proportion of vari-
ants in SCN5A in women who died suddenly. An exami-
nation of common variants in these same genes in a nested
case-control analysis of 516 cases and 1,522 matched con-
trols of European ancestry demonstrated two intronic SNPs
significantly associated with SCD, one in SCN5A and one
in KCNQ1 (Albert et  al., 2010). While clearly requiring
additional replication, together these results suggest that
both common and rare variants in SCN5A may contribute
to altered risk of SCD.
Thus, while several marginal associations have been
reported for ion-channel genes, none of these results has
been convincingly replicated in independent studies, and
therefore they remain unproven. Indeed, SCD is likely to be
the result of multiple pathways that contribute to increased
susceptibility to arrhythmias, including atherosclerosis and
thrombosis, electrogenesis and propagation, and initiat-
ing influences and triggers (Figure 21.1) (Spooner et  al.,
2001a; Spooner et al., 2001b). These pathways are likely to
involve different genes, and thus extensive phenotyping of
samples becomes important. In the absence of being able
to distinguish SCD due to different underlying etiologies
(e.g., structural defects vs. ion-channel defects), all sam-
ples will fall under the rubric of “SCD,” and the power to
find genetic determinants is greatly reduced (Arking et al.,
2004). Several studies are attempting to address this issue,
including the Oregon Sudden Unexpected Death Study,
which recruits samples through the emergency medi-
cal system and attempts to get electrocardiogram data on
all SCDs and autopsy data when available (Chugh et  al.,
2003). There is also the need to obtain prospective data
in order to assess attributable risk for SCD-susceptibility
variants, and that effort is ongoing in the ARIC and CHS
cohorts (ARIC-Investigators, 1989; Fried et al., 1991). In
the absence of large, well-phenotyped SCD cohorts, a great
deal of focus has been on subclinical phenotypes, including
the QT interval.
QT I N T E RVA L
The QT interval is a measure of cardiac repolarization and
is subject to the joint control of the depolarizing Na+ (INa)
currents, Ca2+ (ICa) currents, and the repolarizing slow (IKs)
and rapid (IKr) K+ currents. QT interval (corrected for heart
rate) is a moderately heritable trait (25% to 52% heritabil-
ity) (Busjahn et al., 1999; Carter et al., 2000; Newton-Cheh
et al., 2005), and extremes of QT interval have been associ-
ated with increased risk for SCD in both Mendelian forms
(LQTS and short QT syndrome [SQTS]), as well as in
population-based settings (de Bruyne et al., 1999; Dekker
3 2 8   •   G enomics in C linical Practice
et al., 1994; Dekker et al., 2004; Elming et al., 1998; Okin
et al., 2000; Schouten et al., 1991; Sharp et al., 1998). Taken
together, these observations suggest that QT interval is
likely to have a genetic component (moderate heritability),
and that genetic variants that modify QT interval may also
modify risk of SCD.
NOS1A P
Testing the utility of endophenotypes to identify
disease-related genes was a major motivation behind one of
the first successful GWAS, in which the investigators iden-
tified a common variant in the 5′ region of the NOS1AP
gene associated with a 2–3 ms increase in QT interval per
minor allele. NOS1AP encodes an adapter protein that
physically bridges neuronal nitric oxide synthase with its
targets and modulator proteins. Specifically how NOS1AP
variants modulate QT interval is currently unknown, but
over-expression of NOS1AP in guinea pig ventricular myo-
cytes results in shortening of the cardiac action potential,
a decrease in the L-type Ca2+ (ICa) current, and a smaller
increase in the rapid delayed rectifier K+ current (IKr), with
a resultant shortening of the QT interval (Chang et  al.,
2008). The genetic association has been extensively repli-
cated (Arking et al., 2009; Eijgelsheim et al., 2009; Lehtinen
et al., 2008; Post et al., 2007; Raitakari et al., 2008; Tobin
et al., 2008). In a follow-up study of NOS1AP with SCD
in the combined Atherosclerosis Risk in Communities
Study and Cardiovascular Health Study cohorts (498
cases, 19,295 controls), Kao and colleagues (Kao et  al.,
2009) demonstrated that the NOS1AP SNP most strongly
associated with QT interval, rs16847548, was associated
with risk for SCD in white American adults, with the
QT-prolonging variant associated with increased SCD risk
(p = 0.002). Individuals homozygous for the risk allele were
approximately 72% more likely to die of SCD than indi-
viduals homozygous for the non-risk genotype, even after
adjusting for age, sex, and heart rate. It is important to note
the risk allele is common, with 39% of the general white
American population carrying one copy and 5% carrying
two copies. These findings have been confirmed by a second
independent study (Eijgelsheim et al., 2009).
EC G -A S S O C I AT E D S N P S
In addition to QT interval, QRS (cardiac ventricular con-
duction) and RR (inverse heart rate) intervals have also
been associated with cardiovascular mortality and SCD
(Desai et al., 2006; Jouven et al., 2001). The publication of
GWAS identifying numerous variants associated with these
 Allelic Variation Among Multiple Inter-linked
Pathways

1 2 3

Atherosclerosis Electrogenesis Initiating Influences

& & &
Thrombosis Propagation Triggers

Cholesterol Metabolism Na & K Channels Central Neural Modulation

Plaque Formation & Stability Ca Channels & Cycling Sympathetic & Para-Symps
Clotting Cascade Proteins Connexins & Gap Junctions Receptor & Signaling Pathways
Inflammatory Mediators Energetics & Redox Factors Ischemia & Ionic Imbalances
Vascular Factors Scarring, Fibrosis & Disarray Vascular Control

SCD

Figure 21.1 
Potential genetic contributors to sudden cardiac death (SCD). Potential and documented elements of susceptibility are suggested in
three broad pathways: 1) those that lead to progressive atherosclerosis and frank coronary disease and the likelihood of an occlusive infarction and
ischemic arrhythmias; 2) those involved in electrogenesis and intromyocardial conduction pathways; and 3) those that may influence the initiation
of aberrant triggering events and the perpetuation of an arrhythmia. Adapted from Spooner et al., 2001, and Arking et al., 2004.

traits (Eijgelsheim et al., 2010; Sotoodehnia et al., 2010) has
allowed a more comprehensive assessment of the role of
electrocardiogram (ECG)-associated SNPs in SCD risk. In
one such study, using data generated from an SCD GWAS
composed of 1,283 SCD cases and >20,000 controls, Arking
and colleagues (Arking et  al., 2011)  examined 49 SNPs
associated (p < 5 × 10–8) with QRS, QT, and RR inter-
vals. In a test looking at direction of the genetic effect, the
ECG-trait-prolonging allele was significantly more often asso-
ciated with increased risk of SCD (31 of 49, p = 0.03), with
this effect almost entirely due to the QRS/QT-associated
SNPs (28 of 40, p  =  0.006). Three loci, including PLN,
KCNQ1, and NOS1AP, showed nominal association with
SCD, while a fourth locus, TKT/CACNA1D/PRKCD, was
significant even after Bonferroni correction to account for the
number of loci tested. The TKT/CACNA1D/PRKCD asso-
ciation is particularly intriguing due to the observation that
the QRS-prolonging allele was protective for risk of SCD,
which is counter to the effect observed with the measured
trait (longer QRS duration is associated with increased risk
of SCD). This result raises the possibility that the effect of the
SNP variant on risk of SCD may not be mediated through
its effect on QRS interval. A similar result was also seen in
NOS1AP, where one of the alleles that were associated with
SCD had no effect on QT interval (Kao et al., 2009).
21q21
While focusing on candidate genes and endophenotypes
has yielded several compelling candidates, there is no doubt
that additional genes play a role in SCD, and they are not
likely to be identified through these approaches. Several
GWAS with SCD as the phenotype of interest have been
published, two of which have reported genome-wide sig-
nificant findings. The AGNES (Arrhythmia Genetics in
the Netherlands) cohort, which is composed of individuals
with a first myocardial infarction and ventricular fibrillation
(VF) who survived to hospital admission (n = 515) com-
pared with individuals with myocardial infarction alone
(n = 457), reported a SNP, rs2824292 in the 21q21 locus,
with an OR of 1.78 (95% CI 1.47–2.13, p = 3.36 × 10–10),
and an OR of 1.49 (95% CI 1.14–1.95, p  =  0.004) in a
replication sample of 146 out-of-hospital SCD cases and
391 controls (Bezzina et al., 2010). This SNP is a common
variant (allele frequency of 47%) located in an intergenic
region. The nearest gene, CXADR (~100 kb away), encodes
a viral receptor implicated in viral myocarditis (Bowles
et al., 1986), but it is not directly implicated by this study.
BA Z2B
A second GWAS, comprising a total of 4,400 SCD cases and
>30,000 controls, all of European ancestry, reported a sig-
nificant signal at the 2q24.2 locus, with the strongest SNP,
rs4665058 (p = 1.8 × 10–10), mapped to an intron of the
BAZ2B gene. This locus contains three genes expressed in
the heart but not previously known to play a role in cardiac
biology (BAZ2B, WDSUB1, and TANC1). The risk allele
had a study-size-weighted frequency of 1.4% and increased
the risk for SCD by 1.92-fold per allele (95% CI 1.57–2.34).

G enomics of C om p le x C ardio vascular D isease   •   3 2 9

Based on non-human primate sequence data, the risk allele
is ancestral; thus its low frequency in European-ancestry
populations suggests strong negative selection, as fewer than
0.8% of ancestral alleles have reached a frequency of 1.4%
or lower. A search for missense/splice mutations correlated
with rs4665058 (r2 > 0.8) using pilot 1 data from the 1000
Genomes Project (November 2010 release) (Durbin et al.,
2010) was unsuccessful, indicating that the functional vari-
ant is probably regulatory in nature. The authors note that
the meta-analysis consisted of both population-based and
case-control studies, with some of the case-control studies
using CAD controls as opposed to population-based con-
trols. They thus suggest that the consistent results in stud-
ies with both CAD controls and population-based controls
(see supplementary ­figure 2 from Arking et al., 2011) pro-
vide evidence that the risk associated with rs4665058 may
be specific to SCD rather than a generic CAD risk factor.
Somewhat surprisingly, this study did not replicate the
21q21 association seen in the AGNES cohort, despite ade-
quate power. This lack of replication may reflect an associa-
tion specific to the underlying population from which the
AGNES cohort is drawn, or may be limited to the highly
specific AGNES phenotype of individuals with a first myo-
cardial infarction who survived a VF event, as opposed to a
more broadly defined class of SCD observed in the general
population.
C AU T I O N A RY TA L E S
M A D S B OX T R A NS C R I P T I O N
E N H A N C E R FAC TO R 2, P O LY P E P T I D E
A (M E F2A)
One approach to identifying genetic determinants of com-
plex traits has been to identify families that exhibit mono-
genic forms of the phenotype of interest, with the idea that
any genes identified are likely to be involved in complex
forms of the phenotype as well. This approach has several
merits, including less phenotypical heterogeneity, since
presumably all affected individuals in the pedigree are
manifesting the same phenotype. Additionally, one can use
traditional linkage analysis, which does not require recruit-
ing a control population: however, if not done carefully, this
can lead to both false positives and false negative results.
This approach was adopted by Wang and colleagues (Wang
et  al., 2003), who studied a large family that displayed
an autosomal dominant form of CAD. They performed
genome-wide linkage analysis, and identified a significant
association on chromosome 15q26. With 93 genes in the
3 3 0   •   G enomics in C linical Practice
associated region, they focused on the transcription fac-
tor MEF2A, largely due to the overall expression pattern;
MEF2A was expressed in blood vessels during mouse early
embryogenesis (Edmondson et  al., 1994), and expression
was similar to vascular endothelial growth factor receptor
2 (VEGFR2) and the Von Willebrand factor (Subramanian
and Nadal-Ginard, 1996). These data led Wang and col-
leagues to speculate that MEF2A can be an early marker
for vasculogenesis. They sequenced the gene in the affected
individuals, and identified a 21-bp deletion in exon 11
(termed “Δ7aa”) that resulted in the loss of seven conserved
amino acids and was present in all affected members in the
family. They further went on to demonstrate altered cellu-
lar trafficking for the mutant protein, and that its ability to
activate atrial natriuretic factor is reduced. In a subsequent
study, the same group examined 207 CAD/MI patients for
mutations in MEF2A, identifying three novel mutations in
exon 7 in four patients, and no mutations in 191 controls
(Bhagavatula et  al., 2004). They demonstrated that these
mutations significantly reduce the transcriptional activity
of MEF2A, and suggest that “a significant percent of the
CAD/MI population may carry mutations in MEF2A.”
The combination of both family- and population-based
evidence would seem to present strong evidence for the
involvement of MEF2A with CAD/MI. However, a
subsequent study by Weng and colleagues (Weng et  al.,
2005) raised some doubts about the strength of the effect
of MEF2A variants on susceptibility to CAD/MI. They
sequenced MEF2A exons in 300 white individuals with
documented CAD with onset before age 55 years (men) or
65  years (women), and in 300 elderly controls (men >60
yrs, women >70 yrs) who did not have signs or symptoms
of CAD. Of five missense mutations identified, one was
unique to the CAD individuals, one was unique to the con-
trols, and three were common to both groups. They further
observed the 21-bp deletion in three unrelated, unaffected
individuals, and demonstrated that the deletion does not
segregate with early CAD. They conclude that “these stud-
ies support that MEF2A mutations are not a common cause
of CAD in white people and [we] argue strongly against
a role for the MEF2A 21-bp deletion in autosomal domi-
nant CAD.” A  similar negative association was reported
by Kajimoto and colleagues in a Japanese population
(Kajimoto et  al., 2005)  and by Gonzalez and colleagues
(Gonzalez et  al., 2006)  in a Spanish cohort for the Δ7aa
allele, though Gonzalez and colleagues reported a positive
association for a rare Pro279Leu mutation (OR 3.06, 95%
CI 1.17–8.06).
In an accompanying commentary to the Weng and col-
leagues paper, Altshuler and Hirschhorn (Altshuler and
Hirschhorn, 2005) concluded that the role of MEF2A vari-
ants in CAD has not been established, and used the MEF2A
studies to illustrate a number of criteria that should be imposed
when performing similar studies. They note that in a complex
disease like CAD, it may not be particularly uncommon to
find large pedigrees that appear to display monogenic forms
of the disease. Thus, they propose looking for an unusual phe-
notype shared by affected individuals, such as early-onset or
a syndromic form of disease. The phenotype also needs to be
consistent across family members, which can be difficult with
CAD, in which some individuals are “affected” by virtue of
having had prior coronary events, while others may have angi-
ographically defined disease in the absence of events. Given
these concerns, linkage across multiple families with identical
ascertainment criteria should be observed.
In the event that a linkage signal in a region is detected,
similar concerns arise when trying to identify the underlying
functional variant. The observation of rare, potentially func-
tional variants segregating within a family is not uncommon,
and any such variation observed in the linkage region will, by
virtue of being in a disease-linked region, segregate with the
disease. Thus, Altshuler and Hirschhorn propose that for a
specific gene in a linked region to be associated with disease,
it must meet one or more of the following criteria:
Multiple different mutations exist, each of which
co-segregates with disease;
There is confirmatory evidence for a particular allele in a
case-control study;
There are multiple rare variants that have been well
ascertained in controls;
There is observation of a de novo mutation in an affected
child (but not in his parent);
There is strong evidence of the effects of a human mutation
in a model organism that recapitulates the human
disease phenotype.
With the cost of genotyping decreasing and
sample-collection increasing, the number of such studies
being performed is on the rise; therefore, adopting rigorous
criteria, as outlined above, for deeming a gene to be involved
in complex disease is warranted.
C O N C LU S I O N S
Our understanding of the genetic landscape of cardiovascu-
lar disease has significantly changed in the last five years. We
G enomics of C om p le x C ardio vascular D isease   •   3 3 1
now have a long list of associated loci that have been robustly
replicated in a number of different cohorts. However, the
daunting challenge before us now is to uncover the underly-
ing causal variant. There are many challenges that impede
our ability to do so.
First, common genetic variants are of modest effect
(OR <2), which has been a significant factor in the fail-
ure to identify and reproducibly demonstrate an effect for
a given variant. Most studies have simply been underpow-
ered, greatly contributing to the confusion in the litera-
ture. Indeed, as large studies with participant size soaring
to over 200,000 become the norm, the list of reproducibly
associated loci climbs steadily. Second, our ability to pick
out genes involved in disease based on our limited under-
standing of biology is insufficient, especially given that half
the genes in the genome are of largely unknown function.
Third, most genes that have been studied have not been
studied in a comprehensive way, limiting our ability to com-
pare across studies and to truly exclude genes (as opposed
to a specific variant) from association with disease. Fourth,
functional variants are likely to be influenced by sex and/or
environment (e.g., smoking, diet); thus, comprehensively
collecting data on study populations is essential. With
genotyping no longer an impediment to progress, the
focus going forward is likely to shift from collecting and
assessing larger populations to identify additional genetic
variants, to finding ways to interrogate the function of the
linked loci, identify the underlying causal genes, and use
this information to better predict general population risk
for cardiovascular disease.
REFERENCES
Agostoni, E., Aliprandi, A., and Longoni, M. (2009). Cerebral venous
thrombosis. Expert Rev Neurother 9, 553–564.
Albert, C.M., MacRae, C.A., Chasman, D.I., et  al. (2010). Common
variants in cardiac ion channel genes are associated with sudden car-
diac death. Circ Arrhythm Electrophysiol 3, 222–229.
Albert, C.M., Nam, E.G., Rimm, E.B., et al. (2008). Cardiac sodium chan-
nel gene variants and sudden cardiac death in women. Circulation
117, 16–23.
Alberts, G.F., Peifley, K.A., Johns, A., Kleha, J.F., and Winkles, J.A.
(1994). Constitutive endothelin-1 overexpression promotes smooth
muscle cell proliferation via an external autocrine loop. J Biol Chem.
269, 10112–10118.
Altshuler, D., and Hirschhorn, J.N. (2005). MEF2A sequence variants
and coronary artery disease: a change of heart? J Clin Invest 115,
831–833.
Amos, C.I., Wang, L.E., Lee, J.E., et al. (2011). Genome-wide association
study identifies novel loci predisposing to cutaneous melanoma.
Hum Mol Genet 20, 5012–5023.
Antoniou, A.C., Kuchenbaecker, K.B., Soucy, P., et al. (2012). Common
variants at 12p11, 12q24, 9p21, 9q31.2 and in ZNF365 are associ-
ated with breast cancer risk for BRCA1 and/or BRCA2 mutation
carriers. Breast Cancer Res 14, R33.
ARIC-Investigators. (1989). The Atherosclerosis Risk in Communities
(ARIC) Study:  design and objectives. Am J Epidemiol 129,
687–702.
Ariyaratnam, R., Casas, J.P., Whittaker, J., Smeeth, L., Hingorani, A.D.,
and Sharma, P. (2007). Genetics of ischaemic stroke among persons
of non-European descent: a meta-analysis of eight genes involving
approximately 32,500 individuals. PLoS Med 4, e131.
Arking, D.E., Chugh, S.S., Chakravarti, A., and Spooner, P.M. (2004).
Genomics in sudden cardiac death. Circ Res 94, 712–723.
Arking, D.E., Junttila, M.J., Goyette, P., et  al. (2011). Identification
of a sudden cardiac death susceptibility locus at 2q24.2 through
genome-wide association in European ancestry individuals. PLoS
Genet 7, e1002158.
Arking, D.E., Khera, A., Xing, C., et  al. (2009). Multiple indepen-
dent genetic factors at NOS1AP modulate the QT interval in a
multi-ethnic population. PLoS One 4, e4333.
Arnett, D.K., Davis, B.R., Ford, C.E., et  al. (2005). Pharmacogenetic
association of the angiotensin-converting enzyme insertion/
deletion polymorphism on blood pressure and cardiovascular
risk in relation to antihypertensive treatment:  the Genetics of
Hypertension-Associated Treatment (GenHAT) study. Circulation
111, 3374–3383.
Beckie, T.M., Beckstead, J.W., and Groer, M.W. (2011). The association
between variants on chromosome 9p21 and inflammatory biomark-
ers in ethnically diverse women with coronary heart disease: a pilot
study. Biol Res Nurs 13, 306–319.
Benjamin, E.J., Rice, K.M., Arking, D.E., et  al. (2009). Variants in
ZFHX3 are associated with atrial fibrillation in individuals of
European ancestry. Nat Genet 41, 879–881.
Bennet, A.M., Di Angelantonio, E., Ye, Z., et  al. (2007). Association
of apolipoprotein E genotypes with lipid levels and coronary risk.
JAMA 298, 1300–1311.
Bergametti, F., Denier, C., Labauge, P., et al. (2005). Mutations within
the programmed cell death 10 gene cause cerebral cavernous malfor-
mations. Am J Hum Genet 76, 42–51.
Bertina, R.M., Koeleman, B.P., Koster, T., et  al. (1994). Mutation in
blood coagulation factor V associated with resistance to activated
protein C. Nature 369, 64–67.
Bevan, S., Dichgans, M., Gschwendtner, A., Kuhlenbaumer, G.,
Ringelstein, E.B., and Markus, H.S. (2008). Variation in the PDE4D
gene and ischemic stroke risk: a systematic review and meta-analysis
on 5200 cases and 6600 controls. Stroke 39, 1966–1971.
Bevan, S., Traylor, M., Adib-Samii, P., et al. (2012). Genetic heritability
of ischemic stroke and the contribution of previously reported can-
didate gene and genomewide associations. Stroke 43, 3161–3167.
Bezzina, C.R., Pazoki, R., Bardai, A., et al. (2010). Genome-wide associa-
tion study identifies a susceptibility locus at 21q21 for ventricular
fibrillation in acute myocardial infarction. Nat Genet 42, 688–691.
Bhagavatula, M.R., Fan, C., Shen, G.Q., et  al. (2004). Transcription
factor MEF2A mutations in patients with coronary artery disease.
Hum Mol Genet 13, 3181–3188.
Bilguvar, K., Yasuno, K., Niemela, M., et al. (2008). Susceptibility loci for
intracranial aneurysm in European and Japanese populations. Nat
Genet 40, 1472–1477.
Biros, E., Cooper, M., Palmer, L.J., Walker, P.J., Norman, P.E., and
Golledge, J. (2010). Association of an allele on chromosome 9 and
abdominal aortic aneurysm. Atherosclerosis 212, 539–542.
Boekholdt, S.M., Bijsterveld, N.R., Moons, A.H., Levi, M., Buller, H.R.,
and Peters, R.J. (2001). Genetic variation in coagulation and fibri-
nolytic proteins and their relation with acute myocardial infarc-
tion: a systematic review. Circulation 104, 3063–3068.
Bowles, N.E., Richardson, P.J., Olsen, E.G., and Archard, L.C. (1986).
Detection of Coxsackie-B-virus-specific RNA sequences in myo-
cardial biopsy samples from patients with myocarditis and dilated
cardiomyopathy. Lancet 1, 1120–1123.
Bown, M.J., Braund, P.S., Thompson, J., London, N.J., Samani, N.J.,
and Sayers, R.D. (2008). Association between the coronary artery
3 3 2   •   G enomics in C linical Practice
disease risk locus on chromosome 9p21.3 and abdominal aortic
aneurysm. Circ Cardiovasc Genet 1, 39–42.
Brugada, R., Hong, K., Dumaine, R., et al. (2004). Sudden death asso-
ciated with short-QT syndrome linked to mutations in HERG.
Circulation 109, 30–35.
Burke, A., Creighton, W., Mont, E., et al. (2005). Role of SCN5A Y1102
polymorphism in sudden cardiac death in blacks. Circulation 112,
798–802.
Busjahn, A., Knoblauch, H., Faulhaber, H.D., et al. (1999). QT interval is
linked to 2 long-QT syndrome loci in normal subjects. Circulation
99, 3161–3164.
Carter, N., Snieder, H., Jeffery, S., et al. (2000). QT interval in twins. J
Hum Hypertens 14, 389–390.
Carty, C.L., Buzková, P., Fornage, M., et  al. (2012). Associations
between incident ischemic stroke events and stroke and cardio-
vascular disease-related genome-wide association studies single
nucleotide polymorphisms in the Population Architecture Using
Genomics and Epidemiology study. Circ Cardiovasc Genet 5,
210–216.
Casas, J.P., Hingorani, A.D., Bautista, L.E., and Sharma, P. (2004).
Meta-analysis of genetic studies in ischemic stroke: thirty-two genes
involving approximately 18,000 cases and 58,000 controls. Arch
Neurol 61, 1652–1661.
Ceelie, H., Spaargaren-van Riel, C.C., Bertina, R.M., and Vos, H.L.
(2004). G20210A is a functional mutation in the prothrombin
gene; effect on protein levels and 3′-end formation. J Thromb
Haemost 2, 119–127.
Chang, K.C., Barth, A.S., Sasano, T., et al. (2008). CAPON modulates
cardiac repolarization via neuronal nitric oxide synthase signaling in
the heart. Proc Natl Acad Sci U S A 105, 4477–4482.
Chen, J., Xie, X., Zhu, J., Tao, Q., and Wang, X. (2004).
Single-nucleotide polymorphisms in SCN5A gene in Chinese
Han population and their correlation with cardiac arrhythmias.
Genet Med 6, 159.
Cheng, X., Shi, L., Nie, S., et al. (2011). The same chromosome 9p21.3
locus is associated with type 2 diabetes and coronary artery disease
in a Chinese Han population. Diabetes 60, 680–684.
Chugh, S.S., Dogra, V., Thompson, B., et al. (2003). Annual incidence
of sudden cardiac death based on operative definition:  a prospec-
tive, population-based, countywide study using multiple sources.
Circulation 108, IV-1040.
Cohen, J.C., Boerwinkle, E., H., M.T., Jr., and Hobbs, H.H. (2006).
Sequence variations in PCSK9, low LDL, and protection against
coronary heart disease. N Engl J Med 354, 1264–1272.
Cunnington, M.S., Santibanez Koref, M., Mayosi, B.M., Burn, J., and
Keavney, B. (2010). Chromosome 9p21 SNPs associated with mul-
tiple disease phenotypes correlate with ANRIL expression. PLoS
Genet 6, e1000899.
Dahlback, B. (1995). New molecular insights into the genetics of throm-
bophilia. Resistance to activated protein C caused by Arg506 to Gln
mutation in factor V as a pathogenic risk factor for venous thrombo-
sis. Thromb Haemost 74, 139–148.
de Bruyne, M.C., Hoes, A.W., Kors, J.A., Hofman, A., van Bemmel, J.H.,
and Grobbee, D.E. (1999). Prolonged QT interval predicts cardiac
and all-cause mortality in the elderly. The Rotterdam Study. Eur
Heart J 20, 278–284.
Dekker, J.M., Crow, R.S., Hannan, P.J., Schouten, E.G., and Folsom, A.R.
(2004). Heart rate-corrected QT interval prolongation predicts risk
of coronary heart disease in black and white middle-aged men and
women: the ARIC study. J Am Coll Cardiol 43, 565–571.
Dekker, J.M., Schouten, E.G., Klootwijk, P., Pool, J., and Kromhout, D.
(1994). Association between QT interval and coronary heart disease
in middle-aged and elderly men. The Zutphen Study. Circulation
90, 779–785.
Denier, C., Goutagny, S., Labauge, P., et  al. (2004). Mutations within
the MGC4607 gene cause cerebral cavernous malformations. Am J
Hum Genet 74, 326–337.
Desai, A.D., Yaw, T.S., Yamazaki, T., Kaykha, A., Chun, S., and Froelicher,
V.F. (2006). Prognostic significance of quantitative QRS duration.
Am J Med 119, 600–606.
Diabetes Genetics Initiative of Broad Institute of Harvard and MIT,
Lund University, and Novartis Institutes of BioMedical Research,
Saxena, R., Voight, B.F., Lyssenko, V., et al. (2007). Genome-wide
association analysis identifies loci for type 2 diabetes and triglyceride
levels. Science 316, 1331–1336.
Dong, Y., Hassan, A., Zhang, Z., Huber, D., Dalageorgou, C., and
Markus, H.S. (2003). Yield of screening for CADASIL mutations
in lacunar stroke and leukoaraiosis. Stroke 34, 203–205.
Doolan, A., Langlois, N., Chiu, C., Ingles, J., Lind, J.M., and Semsarian,
C. (2008). Postmortem molecular analysis of KCNQ1 and SCN5A
genes in sudden unexplained death in young Australians. Int J
Cardiol 127, 138–141.
Drenos, F., Whittaker, J.C., and Humphries, S.E. (2007). The use of
meta-analysis risk estimates for candidate genes in combination to
predict coronary heart disease risk. Ann Hum Genet 71, 611–619.
Durbin, R.M., Abecasis, G.R., Altshuler, D.L., et al. (2010). A map of
human genome variation from population-scale sequencing. Nature
467, 1061–1073.
Edmondson, D.G., Lyons, G.E., Martin, J.F., and Olson, E.N. (1994).
MEF2 gene expression marks the cardiac and skeletal muscle lin-
eages during mouse embryogenesis. Development 120, 1251–1263.
Eijgelsheim, M., Newton-Cheh, C., Aarnoudse, A.L., et  al. (2009).
Genetic variation in NOS1AP is associated with sudden cardiac
death:  evidence from the Rotterdam Study. Hum Mol Genet 18,
4213–4218.
Eijgelsheim, M., Newton-Cheh, C., Sotoodehnia, N., et  al. (2010).
Genome-wide association analysis identifies multiple loci related to
resting heart rate. Hum Mol Genet 19, 3885–3894.
Elming, H., Holm, E., Jun, L., et al. (1998). The prognostic value of the
QT interval and QT interval dispersion in all-cause and cardiac
mortality and morbidity in a population of Danish citizens. Eur
Heart J 19, 1391–1400.
Enciso-Mora, V., Hosking, F.J., Sheridan, E., et  al. (2012). Common
genetic variation contributes significantly to the risk of childhood
B-cell precursor acute lymphoblastic leukemia. Leukemia 26,
2212–2215.
Esther, C.R., Marino, E.M., Howard, T.E., et al. (1997). The critical role
of tissue angiotensin-converting enzyme as revealed by gene target-
ing in mice. J Clin Invest 99, 2375–2385.
Fang, D.H., Wu, L.Q., Lu, L., et  al. (2008). Association of human
SCN5A polymorphisms with idiopathic ventricular arrhythmia in
a Chinese Han cohort. Circ J 72, 592–597.
Flossmann, E., Schulz, U.G., and Rothwell, P.M. (2004). Systematic
review of methods and results of studies of the genetic epidemiol-
ogy of ischemic stroke. Stroke 35, 212–227.
Folkersen, L., Kyriakou, T., Goel, A., et al., and PROCARDIS Consortia.
(2009). Relationship between CAD risk genotype in the chromo-
some 9p21 locus and gene expression. Identification of eight new
ANRIL splice variants. PLoS One 4, e7677.
Franco, R.F., Trip, M.D., ten Cate, H., et al. (1999). The 20210 G-->A
mutation in the 3′-untranslated region of the prothrombin gene and
the risk for arterial thrombotic disease. Br J Haematol 104, 50–54.
Fried, L.P., Borhani, N.O., Enright, P., et al. (1991). The Cardiovascular
Health Study: design and rationale. Ann Epidemiol 1, 263–276.
Friedlander, Y., Siscovick, D.S., Weinmann, S., et  al. (1998). Family
history as a risk factor for primary cardiac arrest. Circulation 97,
155–160.
Frosst, P., Blom, H.J., Milos, R., et al. (1995). A candidate genetic risk
factor for vascular disease: a common mutation in methylenetetra-
hydrofolate reductase. Nat Genet 10, 111–113.
Gaita, F., Giustetto, C., Bianchi, F., et al. (2003). Short QT syndrome: a
familial cause of sudden death. Circulation 108, 965–970.
Goldstein, L.B., Adams, R., Becker, K., et al. (2001). Primary prevention
of ischemic stroke:  a statement for healthcare professionals from
G enomics of C om p le x C ardio vascular D isease   •   3 3 3
the Stroke Council of the American Heart Association. Circulation
103, 163–182.
Gonzalez, P., Garcia-Castro, M., Reguero, J.R., et  al. (2006). The
Pro279Leu variant in the transcription factor MEF2A is associated
with myocardial infarction. J Med Genet 43, 167–169.
Gori, F., Specchia, C., Pietri, S., et al., GISSI Prevenzione Investigators,
and SIBioC-GISSI Prevenzione Group. (2010). Common genetic
variants on chromosome 9p21 are associated with myocardial
infarction and type 2 diabetes in an Italian population. BMC Med
Genet 11, 60.
Gretarsdottir, S., Sveinbjornsdottir, S., Jonsson, H.H., et  al. (2002).
Localization of a susceptibility gene for common forms of stroke to
5q12. Am J Hum Genet 70, 593–603.
Gretarsdottir, S., Thorleifsson, G., Manolescu, A., et al. (2008). Risk vari-
ants for atrial fibrillation on chromosome 4q25 associate with isch-
emic stroke. Ann Neurol 64, 402–409.
Gretarsdottir, S., Thorleifsson, G., Reynisdottir, S.T., et al. (2003). The
gene encoding phosphodiesterase 4D confers risk of ischemic
stroke. Nat Genet 35, 131–138.
Gschwendtner, A., Bevan, S., Cole, J.W., et al. (2009). Sequence variants
on chromosome 9p21.3 confer risk for atherosclerotic stroke. Ann
Neurol 65, 531–539.
Gudbjartsson, D.F., Arnar, D.O., Helgadottir, A., et al. (2007). Variants
conferring risk of atrial fibrillation on chromosome 4q25. Nature
448, 353–357.
Gulcher, J.R., Gretarsdottir, S., Helgadottir, A., and Stefansson, K.
(2005). Genes contributing to risk for common forms of stroke.
Trends Mol Med 11, 217–224.
Haberland, M., Montgomery, R.L., and Olson, E.N. (2009). The many
roles of histone deacetylases in development and physiology: impli-
cations for disease and therapy. Nat Rev Genet 10, 32–42.
Hamzi, K., Tazzite, A., and Nadifi, S. (2011). Large-scale meta-analysis
of genetic studies in ischemic stroke: five genes involving 152,797
individuals. Indian J Hum Genet 17, 212–217.
Harismendy, O., Notani, D., Song, X., et al. (2011). 9p21 DNA variants
associated with coronary artery disease impair interferon-gamma
signalling response. Nature 470, 264–268.
Helgadottir, A., Manolescu, A., Thorleifsson, G., et al. (2004). The gene
encoding 5-lipoxygenase activating protein confers risk of myocar-
dial infarction and stroke. Nat Genet 36, 233–239.
Helgadottir, A., Thorleifsson, G., Magnusson, K.P., et  al. (2008). The
same sequence variant on 9p21 associates with myocardial infarc-
tion, abdominal aortic aneurysm and intracranial aneurysm. Nat
Genet 40, 217–224.
Helgadottir, A., Thorleifsson, G., Manolescu, A., et al. (2007). A com-
mon variant on chromosome 9p21 affects the risk of myocardial
infarction. Science 316, 1491–1493.
Humphries, S.E., Talmud, P.J., Hawe, E., Bolla, M., Day, I.N., and
Miller, G.J. (2001). Apolipoprotein E4 and coronary heart disease
in middle-aged men who smoke: a prospective study. Lancet 358,
115–119.
IBC 50K CAD Consortium. (2011). Large-scale gene-centric analysis
identifies novel variants for coronary artery disease. PLoS Genet 7,
e1002260.
Ikram, M.A., Seshadri, S., Bis, J.C., et al. (2009). Genomewide associa-
tion studies of stroke. N Engl J Med 360, 1718–1728.
International HapMap Consortium. (2005). A haplotype map of the
human genome. Nature 437, 1299–1320.
International Stroke Genetics Consortium (ISGC), Wellcome Trust
Case Control Consortium 2 (WTCCC2), Bellenguez, C., Bevan,
S., Gschwendtner, A., et al. (2012). Genome-wide association study
identifies a variant in HDAC9 associated with large vessel ischemic
stroke. Nat Genet 44, 328–333.
International Stroke Genetics Consortium, and Wellcome Trust
Case-Control Consortium 2. (2010). Failure to validate association
between 12p13 variants and ischemic stroke. N Engl J Med 362,
1547–1550.
Jarinova, O., Stewart, A.F., Roberts, R., et al. (2009). Functional analy-
sis of the chromosome 9p21.3 coronary artery disease risk locus.
Arterioscler Thromb Vasc Biol 29, 1671–1677.
Joutel, A., Corpechot, C., Ducros, A., et  al. (1996). NOTCH3 muta-
tions in CADASIL, a hereditary adult-onset condition causing
stroke and dementia. Nature 383, 707–710.
Jouven, X., Desnos, M., Guerot, C., and Ducimetiere, P. (1999).
Predicting sudden death in the population:  the Paris Prospective
Study I. Circulation 99, 1978–1983.
Jouven, X., Zureik, M., Desnos, M., Guerot, C., and Ducimetiere, P.
(2001). Resting heart rate as a predictive risk factor for sudden death
in middle-aged men. Cardiovasc Res 50, 373–378.
Kajimoto, K., Shioji, K., Tago, N., et al. (2005). Assessment of MEF2A
mutations in myocardial infarction in Japanese patients. Circ J 69,
1192–1195.
Kao, W.H., Arking, D.E., Post, W., et al. (2009). Genetic variations in
nitric oxide synthase 1 adaptor protein are associated with sud-
den cardiac death in US white community-based populations.
Circulation 119, 940–951.
Kathiresan, S., Melander, O., Anevski, D., et al. (2008). Polymorphisms
associated with cholesterol and risk of cardiovascular events. N Engl
J Med 358, 1240–1249.
Kathiresan, S., and Srivastava, D. (2012). Genetics of human cardiovas-
cular disease. Cell 148, 1242–1257.
Kathiresan, S., Voight, B.,F., Purcell, S., et al. (2009). Genome-wide asso-
ciation of early-onset myocardial infarction with single nucleotide
polymorphisms and copy number variants. Nat Genet 41, 334–341.
Keating, M.T., and Sanguinetti, M.C. (2001). Molecular and cellular
mechanisms of cardiac arrhythmias. Cell 104, 569–580.
Kim, S., and Iwao, H. (2000). Molecular and cellular mechanisms
of angiotensin II-mediated cardiovascular and renal diseases.
Pharmacol Rev 52, 11–34.
Kjolby, M., Andersen, O.M., Breiderhoff, T., et  al. (2010). SORT1,
encoded by the cardiovascular risk locus 1p13.3, is a regulator of
hepatic lipoprotein export. Cell Metab 12, 213–223.
Kral, B.G., Mathias, R.A., Suktitipat, B., et al. (2011). A common variant
in the CDKN2B gene on chromosome 9p21 protects against coro-
nary artery disease in Americans of African ancestry. J Hum Genet
56, 224–229.
Kramer, A., and Fletcher, J. (2009). Do endothelin-receptor antago-
nists prevent delayed neurological deficits and poor outcomes after
aneurysmal subarachnoid hemorrhage? A meta-analysis. Stroke 40,
3403–3406.
Krege, J.H., John, S.W., Langenbach, L.L., et  al. (1995). Male-female
differences in fertility and blood pressure in ACE-deficient mice.
Nature 375, 146–148.
Kubo, M., Hata, J., Ninomiya, T., et al. (2007). A nonsynonymous SNP
in PRKCH (protein kinase C eta) increases the risk of cerebral
infarction. Nat Genet 39, 212–217.
Laberge-le Couteulx, S., Jung, H.H., Labauge, P., et al. (1999). Truncating
mutations in CCM1, encoding KRIT1, cause hereditary cavernous
angiomas. Nat Genet 23, 189–193.
Lahoz, C., Schaefer, E.J., Cupples, L.A., et  al. (2001). Apolipoprotein
E genotype and cardiovascular disease in the Framingham Heart
Study. Atherosclerosis 154, 529–537.
Lehnart, S.E., Wehrens, X.H., Reiken, S., et al. (2005). Phosphodiesterase
4D deficiency in the ryanodine-receptor complex promotes heart
failure and arrhythmias. Cell 123, 25–35.
Lehtinen, A.B., Newton-Cheh, C., Ziegler, J.T., et al. (2008). Association
of NOS1AP genetic variants with QT interval duration in families
from the Diabetes Heart Study. Diabetes 57, 1108–1114.
Li, J., Luo, M., Xu, X., and Sheng, W. (2012). Association between
1425G/A SNP in PRKCH and ischemic stroke among Chinese
and Japanese populations: a meta-analysis including 3686 cases and
4589 controls. Neurosci Lett 506, 55–58.
Linsel-Nitschke, P., Heeren, J., Aherrahrou, Z., et  al. (2010). Genetic
variation at chromosome 1p13.3 affects sortilin mRNA expression,
3 3 4   •   G enomics in C linical Practice
cellular LDL-uptake and serum LDL levels which translates to the
risk of coronary artery disease. Atherosclerosis 208, 183–189.
Lip, G.Y., and Tse, H.F. (2007). Management of atrial fibrillation. Lancet
370, 604–618.
Liu, Y., Sanoff, H.K., Cho, H., et  al. (2009). INK4/ARF transcript
expression is associated with chromosome 9p21 variants linked to
atherosclerosis. PLoS One 4, e5027.
Lovkvist, H., Olsson, S., Hoglund, P., et  al. (2012). A large-sample
assessment of possible association between ischaemic stroke and
rs12188950 in the PDE4D gene. Eur J Hum Genet 20, 783–789.
Lusis, A.J., Mar, R., Pajukanta, P. (2004). Genetics of atherosclerosis.
Annu Rev Genomics Hum Genet 5, 189–218.
Macdonald, R.L., Higashida, R.T., Keller, E., et al. (2013). Randomised
trial of Clazosentan, an endothelin receptor antagonist, in patients
with aneurysmal subarachnoid hemorrhage undergoing surgical
clipping (CONSCIOUS-2). Acta Neurochir Suppl 115, 27–31.
Macdonald, R.L., Kassell, N.F., Mayer, S., et  al., and CONSCIOUS-1
Investigators. (2008). Clazosentan to overcome neurological isch-
emia and infarction occurring after subarachnoid hemorrhage
(CONSCIOUS-1): randomized, double-blind, placebo-controlled
phase 2 dose-finding trial. Stroke 39, 3015–3021.
Marian, A.J., and Roberts, R. (2001). The molecular genetic basis for
hypertrophic cardiomyopathy. J Mol Cell Cardiol 33, 655–670.
Marjot, T., Yadav, S., Hasan, N., Bentley, P., and Sharma, P. (2011).
Genes associated with adult cerebral venous thrombosis. Stroke 42,
913–918.
McPherson, R., Pertsemlidis, A., Kavaslar, N., et al. (2007). A common
allele on chromosome 9 associated with coronary heart disease.
Science 316, 1488–1491.
Motulsky, A.G., and Brunzell, J.D. (2002). Genetics of coronary ath-
erosclerosis. In The Genetic Basis of Common Diseases (pp 105–
126). King, R.A., Rotter, J.I. and Motulsky, A.G., eds. (Oxford,
UK: Oxford University Press).
Newton-Cheh, C., Larson, M.G., Corey, D.C., et al. (2005). QT interval
is a heritable quantitative trait with evidence of linkage to chromo-
some 3 in a genome-wide linkage analysis: the Framingham Heart
Study. Heart Rhythm 2, 277–284.
Nguyen-Tran, V.T., Kubalak, S.W., Minamisawa, S., et al. (2000). A novel
genetic pathway for sudden cardiac death via defects in the transi-
tion between ventricular and conduction system cell lineages. Cell
102, 671–682.
Okin, P.M., Devereux, R.B., Howard, B.V., Fabsitz, R.R., Lee, E.T., and
Welty, T.K. (2000). Assessment of QT interval and QT dispersion
for prediction of all-cause and cardiovascular mortality in American
Indians: The Strong Heart Study. Circulation 101, 61–66.
Palomaki, G.E., Melillo, S., and Bradley, L.A. (2010). Association
between 9p21 genomic markers and heart disease: a meta-analysis.
JAMA 303, 648–656.
Pasmant, E., Laurendeau, I., Héron, D., Vidaud, M., Vidaud, D., and
Bièche, I. (2007). Characterization of a germ-line deletion, includ-
ing the entire INK4/ARF locus, in a melanoma-neural system
tumor family:  identification of ANRIL, an antisense noncoding
RNA whose expression coclusters with ARF. Cancer Research 67,
3963–3969.
Patel, R.S., Eapen, D.J., Zafari, A.M., Vaccarino, V., and Quyyumi,
A.A. (2011). Letter by Patel et al. regarding article “Chromosome
9p21 haplotypes and prognosis in white and black patients with
coronary artery disease.” Circ Cardiovasc Genet 4, e11; author
reply e12.
Paynter, N.P., Chasman, D.I., Pare, G., et al. (2010). Association between
a literature-based genetic risk score and cardiovascular events in
women. JAMA 303, 631–637.
Poort, S.R., Rosendaal, F.R., Reitsma, P.H., and Bertina, R.M. (1996).
A common genetic variation in the 3′-untranslated region of
the prothrombin gene is associated with elevated plasma pro-
thrombin levels and an increase in venous thrombosis. Blood 88,
3698–3703.
Post, W., Shen, H., Damcott, C., et  al. (2007). Associations between
genetic variants in the NOS1AP (CAPON) gene and cardiac repo-
larization in the Old Order Amish. Hum Hered 64, 214–219.
Priori, S.G., and Napolitano, C. (2004). Genetics of cardiac arrhythmias
and sudden cardiac death. Ann N Y Acad Sci 1015, 96–110.
Raitakari, O.T., Blom-Nyholm, J., Koskinen, T.A., Kahonen, M., Viikari,
J.S., and Lehtimaki, T. (2009). Common variation in NOS1AP
and KCNH2 genes and QT interval duration in young adults. The
Cardiovascular Risk in Young Finns Study. Ann Med 41, 144–151.
Rajaraman, P., Melin, B.S., Wang, Z., et  al. (2012). Genome-wide
association study of glioma and meta-analysis. Hum Genet 131,
1877–1888.
Rigat, B., Hubert, C., Alhenc-Gelas, F., Cambien, F., Corvol, P., and
Soubrier, F. (1990). An insertion/deletion polymorphism in the
angiotensin I-converting enzyme gene accounting for half the vari-
ance of serum enzyme levels. J Clin Invest 86, 1343–1346.
Roger, V.L., Go, A.S., Lloyd-Jones, D.M., et al. (2012). Heart disease and
stroke statistics—2012 update: a report from the American Heart
Association. Circulation 125, e2–e220.
Sakamoto, Y., Hara, K., Kanai-Azuma, M., et al. (2007). Redundant roles
of Sox17 and Sox18 in early cardiovascular development of mouse
embryos. Biochem Biophys Res Commun 360, 539–544.
Samani, N.J., Erdmann, J., Hall, A.S., et al. (2007). Genomewide associa-
tion analysis of coronary artery disease. N Engl J Med 357, 443–453.
Schouten, E.G., Dekker, J.M., Meppelink, P., Kok, F.J., Vandenbroucke,
J.P., and Pool, J. (1991). QT interval prolongation predicts cardio-
vascular mortality in an apparently healthy population. Circulation
84, 1516–1523.
Schunkert, H., König, I.  R., Kathiresan, S., et  al. (2011). Large-scale
association analysis identifies 13 new susceptibility loci for coronary
artery disease. Nat Genet 43, 333–338.
Scott, L.J., Mohlke, K.L., Bonnycastle, L.L., et al. (2007). A genome-wide
association study of type 2 diabetes in Finns detects multiple suscep-
tibility variants. Science 316, 1341–1345.
Sharp, D.S., Masaki, K., Burchfiel, C.M., Yano, K., and Schatz, I.J.
(1998). Prolonged QTc interval, impaired pulmonary function, and
a very lean body mass jointly predict all-cause mortality in elderly
men. Ann Epidemiol 8, 99–106.
Shen, G.Q., Rao, S., Martinelli, N., et  al. (2008). Association between
four SNPs on chromosome 9p21 and myocardial infarction is repli-
cated in an Italian population. J Hum Genet 53, 144–150.
Sherborne, A.L., Hosking, F.J., Prasad, R.B., et al. (2010). Variation in
CDKN2A at 9p21.3 influences childhood acute lymphoblastic leu-
kemia risk. Nat Genet 42, 492–494.
Shete, S., Hosking, F.J., Robertson, L.B., et  al. (2009). Genome-wide
association study identifies five susceptibility loci for glioma. Nat
Genet 41, 899–904.
Sotoodehnia, N., Isaacs, A., de Bakker, P.I., et al. (2010). Common vari-
ants in 22 loci are associated with QRS duration and cardiac ven-
tricular conduction. Nat Genet 42, 1068–1076.
Splawski, I., Shen, J., Timothy, K.W., et  al. (2000). Spectrum of muta-
tions in long-QT syndrome genes, KVLQT1, HERG, SCN5A,
KCNE1, and KCNE2. Circulation 102, 1178–1185.
Splawski, I., Timothy, K.W., Tateyama, M., Clancy, C.E., and Malhotra,
A. (2002). Variant of SCN5A sodium channel implicated in risk of
cardiac arrhythmia. Science 297, 1333.
Spooner, P.M., Albert, C., Benjamin, E.J., et al. (2001a). Sudden cardiac
death, genes, and arrhythmogenesis: consideration of new popula-
tion and mechanistic approaches from a National Heart, Lung, and
Blood Institute workshop, Part I. Circulation 103, 2361–2364.
Spooner, P.M., Albert, C., Benjamin, E.J., et al. (2001b). Sudden cardiac
death, genes, and arrhythmogenesis: consideration of new popula-
tion and mechanistic approaches from a National Heart, Lung, and
Blood Institute workshop, Part II. Circulation 103, 2447–2452.
Stecker, E.C., Sono, M., Wallace, E., Gunson, K., Jui, J., and Chugh, S.S.
(2006). Allelic variants of SCN5A and risk of sudden cardiac arrest
in patients with coronary artery disease. Heart Rhythm 3, 697–700.
G enomics of C om p le x C ardio vascular D isease   •   3 3 5
Subramanian, S.V., and Nadal-Ginard, B. (1996). Early expression of the
different isoforms of the myocyte enhancer factor-2 (MEF2) pro-
tein in myogenic as well as non-myogenic cell lineages during mouse
embryogenesis. Mech Dev 57, 103–112.
Suehiro, T., Morita, T., Inoue, M., Kumon, Y., Ikeda, Y., and Hashimoto,
K. (2004). Increased amount of the angiotensin-converting enzyme
(ACE) mRNA originating from the ACE allele with deletion. Hum
Genet 115, 91–96.
Suzuki, E., Nagata, D., Kakoki, M., et al. (1999). Molecular mechanisms
of endothelin-1-induced cell-cycle progression:  involvement of
extracellular signal-regulated kinase, protein kinase C, and phos-
phatidylinositol 3-kinase at distinct points. Circ Res 84, 611–619.
Takeuchi, F., Yokota, M., Yamamoto, K., et  al. (2012). Genome-wide
association study of coronary artery disease in the Japanese. Eur J
Hum Genet 20, 333–340.
Tobin, M.D., Kahonen, M., Braund, P., et al. (2008). Gender and effects
of a common genetic variant in the NOS1 regulator NOS1AP on
cardiac repolarization in 3761 individuals from two independent
populations. Int J Epidemiol 37, 1132–1141.
Utermann, G., Hees, M., and Steinmetz, A. (1977). Polymorphism of
apolipoprotein E and occurrence of dysbetalipoproteinaemia in
man. Nature 269, 604–607.
Vergouwen, M.D., Algra, A., and Rinkel, G.J. (2012). Endothelin recep-
tor antagonists for aneurysmal subarachnoid hemorrhage: a system-
atic review and meta-analysis update. Stroke 43, 2671–2676.
Visel, A., Zhu, Y., May, D., et al. (2010). Targeted deletion of the 9p21
non-coding coronary artery disease risk interval in mice. Nature
464, 409–412.
Volcik, K.A., Barkley, R.A., Hutchinson, R.G., et  al. (2006).
Apolipoprotein E polymorphisms predict low density lipoprotein
cholesterol levels and carotid artery wall thickness but not incident
coronary heart disease in 12,491 ARIC study participants. Am J
Epidemiol 164, 342–348.
Wald, D.S., Law, M., and Morris, J.K. (2002). Homocysteine and cardio-
vascular disease:  evidence on causality from a meta-analysis. BMJ
325, 1202.
Wang, B., Guo, Q., Peng, Y., Lu, J., Singh, B., and Hua, B. (2012).
Association of AGT M235T and ACE I/D polymorphisms with
the risk of ischemic stroke: meta-analysis in Han Chinese popula-
tion. J Neurol Sci 320, 79–84.
Wang, L., Fan, C., Topol, S.E., Topol, E.J., and Wang, Q. (2003).
Mutation of MEF2A in an inherited disorder with features of coro-
nary artery disease. Science 302, 1578–1581.
Ward, H., Mitrou, P.N., Bowman, R., et al. (2009). APOE genotype, lip-
ids, and coronary heart disease risk: a prospective population study.
Arch Intern Med 169, 1424–1429.
Watkins, H., Farrall, M. (2006). Genetic susceptibility to coronary artery
disease: from promise to progress. Nat Rev Genet 7, 163–173.
Weng, L., Kavaslar, N., Ustaszewska, A., et al. (2005). Lack of MEF2A
mutations in coronary artery disease. J Clin Invest 115, 1016–1020.
Wu, L., Shen, Y., Liu, X., et al. (2009). The 1425G/A SNP in PRKCH
is associated with ischemic stroke and cerebral hemorrhage in a
Chinese population. Stroke 40, 2973–2976.
Xie, F., Chu, X., Wu, H., et al. (2011). Replication of putative suscep-
tibility loci from genome-wide association studies associated with
coronary atherosclerosis in Chinese Han population. PLoS One 6,
e20833.
Yamagishi, K., Folsom, A.R., Rosamond, W.D., Boerwinkle, E., and
ARIC Investigators. (2009). A genetic variant on chromosome
9p21 and incident heart failure in the ARIC study. Eur Heart J 30,
1222–1228.
Yang, X.R., Liang, X., Pfeiffer, R.M., et al. (2010). Associations of 9p21
variants with cutaneous malignant melanoma, nevi, and pigmen-
tation phenotypes in melanoma-prone families with and without
CDKN2A mutations. Fam Cancer 9, 625–633.
Yasuno, K., Bakircioglu, M., Low, S.K., et al. (2011). Common variant
near the endothelin receptor type A (EDNRA) gene is associated
 with intracranial aneurysm risk. Proc Natl Acad Sci U S A 108,
19707–19712.
Yasuno, K., Bilguvar, K., Bijlenga, P., et al. (2010). Genome-wide asso-
ciation study of intracranial aneurysm identifies three new risk loci.
Nat Genet 42, 420–425.
Ye, Z., Liu, E.H., Higgins, J.P., et al. (2006). Seven haemostatic gene poly-
morphisms in coronary disease: meta-analysis of 66,155 cases and
91,307 controls. Lancet 367, 651–658.
Zdravkovic, S., Wienke, A., Pedersen, N.L., Marenberg, M.E., Yashin,
A.I., and De Faire, U. (2002). Heritability of death from coronary
heart disease: a 36-year follow-up of 20,966 Swedish twins. J Intern
Med 252, 247–254.
Zeggini, E., Scott, L.J., Saxena, R., et  al. (2008). Meta-analysis of
genome-wide association data and large-scale replication identifies
3 3 6   •   G enomics in C linical Practice
additional susceptibility loci for type 2 diabetes. Nat Genet 40,
638–645.
Zeggini, E., Weedon, M.N., Lindgren, C.M., et al. (2007). Replication of
genome-wide association signals in UK samples reveals risk loci for
type 2 diabetes. Science 316, 1336–1341.
Zheng, Z.J., Croft, J.B., Giles, W.H., et al. (2002). State-specific mortal-
ity from sudden cardiac death—United States, 1999. MMWR 51,
123–126.
Zintzaras, E., Rodopoulou, P., and Sakellaridis, N. (2009). Variants of
the arachidonate 5-lipoxygenase-activating protein (ALOX5AP)
gene and risk of stroke: a HuGE gene-disease association review and
meta-analysis. Am J Epidemiol 169, 523–532.
 22.
GENOMICS OF T YPE 2 DIABETES MELLITUS
AND OBESIT Y
Venkatesan Radha and Viswanathan Mohan
INTRODUCTION
Type 2 diabetes mellitus (T2DM) is a heterogeneous dis-
ease resulting from defects of both insulin secretion and
insulin action and characterized by persistent hyperglyce-
mia (Eriksson et al., 1989). In addition to the consequences
of abnormal metabolism of glucose (e.g., hyperlipidemia,
glycosylation of proteins, etc.), there are many long-term
complications associated with the disease. These include
cardiovascular, peripheral vascular, ocular, neurological,
and renal abnormalities, which are responsible for mor-
bidity, disability, and premature death in young adults.
According to the International Diabetes Federation (IDF)
Diabetes Atlas 2012, more than 371  million people have
diabetes worldwide (IDF, 2012)  and this number is pro-
jected to increase rapidly.
Obesity is clinically defined as having a body mass
index (BMI) above 30kg/m2. Approximately 500  million
people worldwide are estimated to have obesity, and 1.4
billion are estimated to be overweight. This is projected to
rise to 700  million by 2015 (World Health Organization
[WHO], 2012). The rise in prevalence of both diabetes and
obesity has been staggering. Over the course of less than a
generation, the total number is projected to double. The
prevalence rates of the burden of diabetes and obesity are
climbing fastest in Asia, Africa, and South America.
The global epidemic of type 2 diabetes is tied to rising
rates of overweight and obesity in adults as well as in youth.
The prevalence of overweight (BMI 25–30kg/m2) and
obesity (BMI 30kg/m2 or more) in the adult population
is predicted to rise from 33% in 2005 to 57.8% in 2030,
if recent secular trends of obesity continue (Kelly et  al.,
2008). Overweight and obesity are the single most impor-
tant predictors of diabetes (Hu et al., 2001), and the impact
of obesity on lifetime risk of diabetes is stronger in younger
adults (Narayan et al., 2007).
337
As with other common diseases, type 2 diabetes and
obesity are best described as complex, multifactorial traits.
Individual susceptibility to these conditions reflects an
interplay of multiple genetic and environmental factors,
each of which has a modest effect on risk (Figure 22.1).
T Y P E 2 D I A B ET E S A N D O B E S I T Y
A S C O M P L E X T R A I T S :  T H E R O L E
O F E N VI R O N M E N T A N D G E N E S
Evidence for an environmental component of susceptibil-
ity is strong. The rapid changes in prevalence rates provide
the most compelling evidence, and also identify the likely
culprits:  the increasing availability of cheap sources of
energy-dense foods, and a shift to more sedentary lifestyles
(Alberti and Zimmet, 2013). Studies over the past decade
have suggested that there is an important role for environ-
ment in precipitating type 2 diabetes and other components
of metabolic syndromes.
While environmental factors certainly play a major role
in the diabetes and obesity epidemic, there are multiple
lines of evidence to support the view that genetic factors are
very important in the susceptibility of these two conditions.
These are the concordant rates in twin studies and familial
clustering and ethnic variation in prevalence (Diamond,
2003).
The rates of concordance are much higher for monozy-
gotic than for dizygotic twins (Poulsen et al., 1999). Twin
studies have been the much-used model to assess the genetic
component of T2D and obesity because monozygotic
(M2) twins are generally identical, while dizygotic (D2)
twins share 50% of their genetic material; the concordance
for T2D among M2 twins is reported to be about 70%,
while for D2 twins, it is in the range of 20–30%. Similarly,
the concordance for fat mass among M2 twins has been
 Genetic factors Genetic factors
Genes influencing Genes influencing
β-cell mass Obesity
β-cell development Insulin action
β-cell function
β-cell immunogenicity
Insulin
secretory
Type 2 Insulin
DM resistance
defect
Environmental factors
Obesity
Environmental factors Age
Pregnancy
Perinatal malnutrition Sedentary lifestyle
Diabetic mother Diabetogenic drugs
Figure 22.1  Gene–environment interaction in type 2 diabetes.
reported to range from 70–90%, while in D2 twins, it is
35–45% (Bell et al., 2005).
Evidence from adoption and family studies has shown
the role of genes in T2D and obesity compared to environ-
mental factors. For example, in the case of obesity, it has
been shown that, while there is no association between the
BMI of non-identical twins separated at birth, there seems
to be a significant relationship for identical twins raised
apart (Stunkard et al., 1986).
The heritability estimates of obesity seems as high as 0.8
(Stunkard et al., 1990), while that of diabetes and related
traits is more variable, but derived measures of B-cell func-
tion show a consistently high heritability, 0.5–0.8, across
family and twin studies (Stumvoll et al., 2005; McCarthy,
2008). The sibling relative risk (RR) in the European popu-
lation is between 3 and 4. The equivalent figures for severe
obesity are on a similar scale (Allison, Faith, et al., 1996).
Recently, heritability of quantitative traits associated with
type 2 diabetes in large multiplex families from South India
was estimated (Mathias et al., 2009). Heritability estimates
were calculated for all quantitative traits at the univariate
level, and bivariate analyses were done to determine the cor-
relation in genetic and environmental control across these
quantitative traits. The study revealed strong familial aggre-
gation of quantitative traits that are typically associated
with type 2 diabetes, lending credence to the role of genetic
factors in type 2 diabetes.
In addition, differences in the prevalence of T2D and
obesity among racial/ethnic groups also provide genetic
underpinning for these two conditions. For example,
the prevalence of obesity is less than 35% in Caucasian
and Asian populations, but reaches 50% or more in Pima
Indians (Knowler et al., 1990). Although the prevalence of
3 3 8   •   G enomics in C linical P ractice
overweight or obesity (as a measure of BMI) is much lower
in most Asian populations compared to Europeans, Asians
tend to develop diabetes at a lower BMI level, and the risk
of T2D tends to be higher in Asian populations compared
to Europeans (Raji et al., 2001; Radha, et al., 2007). These
are clear indicators that ethnically specific variations in phe-
notypes exist amongst various populations in the world.
All these findings strongly support the view that genes
play a central role in the development of T2D, BMI, and,
consequently, obesity. However, the etiological pathways
responsible for the development of diabetes and obesity are
not clear, and understanding of these pathways involved in
development and progression of disease will aid in thera-
peutic developments. In this regard, research in genetics
and genomics offers promise in identifying the etiological
pathways that are common to people with diabetes and obe-
sity (McCarthy et al., 2008).
C U R R E N T U N D E R S TA N D I N G O F
G E N ET I C S O F T Y P E 2 D I A B ET E S
AND OBESIT Y
MO N O G E N I C F O R M S O F T H E
DISEASE
A small proportion of the incidence of T2D and obesity
is attributable to monogenic and/or syndromic forms.
Disease susceptibility in such individuals is determined by
the segregation of mutation at a single, highly penetrant
gene (Tattersall, 1974). In case of T2D and obesity, some of
these genes are also associated with the multifactorial form
of type 2 diabetes and obesity; therefore, studies of such
rare monogenic and syndromic subtypes provide powerful
insights into the network and pathways that are critical for
normal homeostasis.
Monogenic Diabetes
Monogenic diabetes consists of different subtypes of
single-gene disorders comprising a large spectrum of phe-
notypes, such as neonatal diabetes (NDM), monogenic
diabetes of infancy, dominantly inherited familial forms of
early onset of diabetes called MODY (maturity onset dia-
betes of the young), and rare diabetes associated syndromic
disorders (Tattersall et al., 1975). All of these are unrelated
to autoimmunity and are diagnosed at a very young age.
 N E O N ATA L D I A B ET E S
Neonatal diabetes mellitus (NDM) is defined as diabetes
diagnosed within the first six months of life. It is a relatively
rare entity that includes many clinically and genetically het-
erogeneous disorders that affect 1:100,000–260,000 live
births (Slingerland et al., 2009). NDM can be either per-
manent (PNDM), requiring lifelong treatment, or transient
(TNDM), with insulin dependence in the first months
only and a spontaneous remission of diabetes usually by
18 months of age (Hattersley et al., 2009). The severe hyper-
glycemia and minimal ketosis appearing in the first days of
life may have dramatic complications in the neonate, such
as failure to thrive, acidosis, dehydration, and neurological
alterations. Neonatal diabetes is a monogenic disorder and
is mostly unrelated to auto-immunity, and it is conferred by
mutations in genes that play a key role in beta cell function
or development, including genes for glucokinase, the potas-
sium sensitive ATP (KATP) channel (Gloyn et al., 2003), and
the insulin (Greeley et al., 2011).
Mutations in the gene encoding the Kir6.2 and SUR1
(ABCC8) subunits of the KATP channel are the most com-
mon cause of PNDM, accounting for around 50% of the
cases (Flanagan et al., 2006). If the channel is open, the insu-
lin is not secreted. The mutations prevent channel closure
and thus also prevent insulin secretion. The specific muta-
tions determine the phenotype, and for KIR mutations,
there seems to be a striking correlation with functional
severity of the mutation. Mutations in KCNJ11 or ABCC8
can also cause TNDM. In addition to diabetes, around 20%
of the patients with KATP channel mutations have neuro-
logical symptoms. These features occasionally constitute
severe syndrome of developmental delay, epilepsy, and neo-
natal diabetes (DEND), or more commonly, intermediate
DEND, which is characterized by diabetes and less severe
developmental delay without epilepsy. The identification
of KATP channel mutations can have a dramatic impact
on the type of diabetic therapy. Most patients with KATP
channel mutations are switched over from insulin to an oral
sulfonylurea drug (Hattersley et  al., 2005; Pearson et  al.,
2006).
S T U D I E S O N N EO NATA L D I A B ET E S
IN INDIA
The molecular basis of neonatal diabetes has not been sys-
tematically studied in India, except in four isolated case
reports (Letha et  al., 2007; Ahamed et  al., 2008; Kochar
et  al., 2010). The genes implicated in the pathology of
neonatal diabetes include KCNJ11, ABCC8, and INS.
G enomics of T ype 2 D iabetes M ellitus and O besity   •   3 3 9
Very recently, in our group, we studied the molecular and
clinical aspects of neonatal diabetes in Indian children
( Jahnavi et  al., 2013). We also studied the children with
infantile-onset diabetes and monogenic syndromic dia-
betes. We identified mutations in genes such as KCNJ11,
ABCC8, INS, AGPAT2, SLC2A2, and EIF2AK3 in 10
children ( Jahnavi et  al., 2013). The most interesting out-
come of this genetic work was that the identification of
KCNJ11 and ABCC8 gene mutations in five children made
it possible for us to convert them to oral sulfonylurea ther-
apy from insulin injections. From the clinical point of view,
the most dramatic consequence of this discovery has been
the capacity to convert a large proportion of children suffer-
ing from PNDM from insulin therapy to oral agents such as
sulfonylurea (Pearson, 2006).
M AT U R I T Y- O N S ET D I A B ET E S O F
T H E YO U N G ( M O DY )
MODY is defined as a dominantly inherited young-onset
non-autoimmune diabetes that occurs in adolescence or
young adulthood (usually <25 years of age) due to a primary
defect in pancreatic β-cell function (Fajans et  al., 2011).
MODY is a subtype of type 2 diabetes and does not gener-
ally require insulin at the time of diagnosis. This is because
a residual insulin secretion is still maintained for some years
after diagnosis. MODY accounts for 1–3% of all T2D cases
(Hattersley et al., 1998).
Heterozygous mutations or partial/whole gene dele-
tions in 11 susceptibility genes so far explain the clini-
cal heterogeneity of the MODY subtypes. The MODY
genes encode the enzyme glucokinase (GCK) (Froguel,
1993); the transcription factors HNF1α (Yamagata et  al.,
1996), HNF4α (Ellard et  al., 2006), HNF1β, PDX1, and
NEUROD1 (Vaxillaire et al., 2006); the preproinsulin INS
(Meur et al., 2010), KLF11 (Scohy, 2010), CEL (Torsvik
et al., 2010), PAX4 (Inoue et al., 1998), and BLK (Drebin,
1995); each having a crucial role in the development and/
or function of the pancreatic β-cells. Amongst these genes,
mutations in GCK, HNF1α and HNF4α are the most com-
mon causes of MODY, although the mutation prevalence
may vary across populations (Lehto et al., 1999).
MODY 1 subtype is caused by mutations in HNF4α
gene, and this gene plays a role in development of the
liver, kidney, and intestines. Heterozygous mutation in the
human HNF4α gene results in progressive decrease in insu-
lin production (Yamagata et al., 1996).
Since the first reports of the GCK-related MODY
subtype (MODY 2)  (Osbak et  al., 2009), more than
600 GCK mutations distributed throughout the gene
have been reported worldwide in more than 1,400 fami-
lies. Heterozygous gain-of function GCK mutations are
shown to cause hypoglycemia, whereas heterozygous
loss-of function GCK mutations result in alterations of
both glucose-stimulated insulin secretion and hepatic gly-
cogen synthesis, which cause mild fasting hyperglycemia
(5.5–8.0 mmol/l) (Velho et al., 1992; Fajans et al., 2001).
Interestingly, homozygous GCK mutations causing per-
manent neonatal diabetes mellitus have also been reported
(Osbak et al., 2009).
Mutations in the HNFIα gene result in the MODY3
subtype. More than 200 different HNF1α gene mutations
located in the promoter and coding regions have been
described in some 370 families of various ethnic origins
(Velho et al., 1992), and MODY3 is the commonest form
of MODY worldwide (Fajans et al., 2001). Clinically, the
patients are generally non-obese, often presenting with
severe hyperglycemia, and may develop micro-vascular
complications of diabetes. The prevalence of MODY3 in
patients with early-onset T2DM varies from 2.5% to 36%
(Aguilar-Salinas et al., 2001; Owen, 2003) The penetrance
of MODY3 is high, and this subtype is a more severe form
of diabetes, often evolving towards insulin dependency and
micro-vascular complications of diabetes (Vaxillaire and
Froguel, 2006).
HNF1α and HNF4α mutations are associated with dia-
betes onset in early adulthood and cause a progressive and
severe deterioration in glucose tolerance, requiring hypo-
glycemic drugs or even insulin therapy at a young age. This
shared phenotype is consistent with the interdependence
between HNF1α and HNF4α forming part of a regulatory
network in the pancreatic β-cell (Maestro et al., 2007).
Two other transcription factor genes, PDX1 and
NEUROD1, have an important role in the development
of the endocrine pancreas, although they only rarely cause
MODY (Stoffers et al., 1997).
Mutations in HNF1β/TCF2 were at first associated
with MODY5 in a few families (Vaxillaire et  al., 2006).
In addition to having an important role in early pancreas
development, HNFI B protein function is also crucial
in kidney development and nephron differentiation, and
HNF1β mutations were shown to be a more common cause
of renal cystic diseases and multiple renal malformations
(Thomas et al., 2008).
The basic helix-loop helix (bHLH) transcription fac-
tor NEUROD1 (or BETA2) with other factors (like
NEUROG3) specifies the pancreatic endocrine lineage. This
is also known as MODY6. Heterozygous loss-of-function
mutations in NEUROD1 were reported in a few MODY
3 4 0   •   G enomics in C linical P ractice
families (Stoffers et  al., 1997), and very rare homozygous
mutations are associated with PNDM, cerebellar hypopla-
sia, learning difficulties, and visual and hearing impairment
(Rubio-Cabezas, 2010). MODY 6 highlights the critical
role of NEUROD1 in the development of both endocrine
pancreas and central nervous system.
Krueppel-like factor 11 is a protein that in humans is
encoded by the KLF11 gene. MODY7 is caused by muta-
tions in the KLF11 gene. KLF 11 regulates exocrine cell
growth and behaves like a tumor suppressor for pancreatic
malignancy (Neve et al., 2005).
The carboxyl-ester lipase gene (CEL) controls both exo-
crine and endocrine functions of the pancreas. MODY8 is
caused by mutation in the CEL gene on chromosome 9. It
is caused by frameshift deletions in the variable number of
tandem repeats (VNTR) of the carboxyl-ester lipase gene
(Torsvik et al., 2010).
Paired box gene 4 encodes for PAX4 protein, and
mutations in this gene result in MODY9 (Inoue et  al.,
1998). The paired box gene 4 is involved in pancreatic islet
development.
The proinsulin precursor of insulin is encoded by
the INS gene (Bell et  al., 1980). MODY10 is caused by
mutation in the INS (PROINSULIN) gene. MODY11
is caused by mutation in the B lymphocyte kinase (BLK)
gene (Drebin et  al., 1995). This gene preferentially
expresses in B-lymphoid cells, and it probably functions in a
signal-transduction pathway specific to this lineage.
MO DY I N S O U T H E A S T A S I A
Earlier studies reported on the high prevalence of MODY
(per the clinical criteria used at that time) in South Indians
(4.8%) (Mohan et  al., 1985). The insulin responses in
MODY and the beta-cell response in the offspring of
MODY indicated an increased insulin resistance compared
to classical Indian type 2 diabetic subjects.
Recently, our work has revealed important insights
into the genetics of MODY in south India. We identi-
fied nine novel variants comprising seven mutations (one
novel mutation,  –538G→C at promoter region, and six
novel coding region mutations) and two polymorphisms
in the HNF1A gene (Radha et  al., 2009). Functional
studies revealed reduced transcriptional activity of the
HNF1A promoter for two of the promoter variants. One
of the novel mutations, Arg263His, was identified in a
family of 30 individuals. The mutation co-segregated
with diabetes in this family, and this mutation was not
seen in nondiabetic members in the family, thus pro-
viding evidence for the mutation to be involved in
causing MODY (Radha et al., 2009). Furthermore, three
novel MODY1 mutations have also been identified in
HNF4A gene by us in the same population (Anuradha
et al., 2011). Thus, about 9% of the clinically diagnosed
MODY subjects are MODY3, and 3% are MODY1, in
south India.
A molecular genetic diagnosis in patients suspected of
MODY is important (Shields et al., 2010) because it con-
firms a diagnosis of MODY, classifies the subtype, predicts
the likely clinical course, defines risk for relatives, and may
change the patient’s treatment.
MO N O G E N I C A N D SY N D RO M I C
F O R M S O F O B E S IT Y
As in the case of diabetes, there are several rare mono-
genic and syndromic forms of obesity now established.
“Monogenic obesity” refers to number of rare forms of
severe obesity that result from mutations of large effect size
in an individual gene or chromosomal region. Studies on
monogenic obesity have been good models for getting the
first insights into the genetics of the trait, with initial studies
in rodent models being fruitful. The initial evidence came
from mouse studies. The ob/ob mouse strain was the first to
be studied, and they exhibited excess adiposity along with
impairment in reproductive behavior and weighing three
times more than the normal mice. The discovery of muta-
tions in the leptin gene in these mice and mutations in the
leptin receptor gene in another strain of obese diabetic mice,
db/db mice (Chen et al., 1996), provided novel pathways in
studying diabetes in man (Zhang et  al., 1994). Following
this, leptin mutations were also found in human patients
(Echwald et al., 1997). Also, homozygotes for a mutation
in leptin receptor gene showed impaired growth-hormone
secretion and early-onset morbid obesity (Clement et  al.,
1998). Rare homozygous mutation in leptin genes have
been identified in patients with severe early-onset obesity.
This showed the importance of the leptin melanocortin
pathway in hyperphagia and obesity susceptibility ( Julia
et al., 2013).
The discovery of mutations within the leptin gene
resulting in severe obesity induced by leptin deficiency
represents one of the key successes of obesity research,
whereby exogenous administration of leptin to patients
carrying a homozygous mutation resulted in the rever-
sal of the obesity phenotype. This is akin to shifting
PNDM patients from insulin to oral drugs in the case
of diabetes. These two instances have clinical relevance
and implication. These results also led to the understand-
ing of the leptin melanocortin signalling pathway and
G enomics of T ype 2 D iabetes M ellitus and O besity   •   3 4 1
its role in energy homeostasis and body-weight regula-
tion. Mutations in single genes of this pathway have been
shown to have a monogenic effect on the development
of obesity. Of particular interest is the mutation in the
MC4R gene, which accounts up to 6% of and 2% of adult
obesity cases (den Hoed., 2010; Xia et al., 2013). While
studies on genetic syndromes of obesity in rodents have
provided some insights into the mechanisms involved in
energy homeostasis, research in recent years has begun
to directly identify disorders of energy balance due to
defects in other loci, as can be observed in syndromes
such as Prader-Willi, Alstronis, and Bardet-Biedl (Beales
et al., 2003).
MU LT I FAC TO R I A L F O R M S O F
D I A B ET E S A N D O B E S I T Y
Compared to the success in pinpointing the causation of
monogenic and syndromic forms of diabetes, progress in
identifying genes influencing the development of common,
polygenic T2D and related traits has been slow (McCarthy,
2004).
Effect sizes associated with any single susceptibility vari-
ant are generally modest, in addition to conditionality due
to gene–gene and gene–environment interactions in the
case of polygenic or multifactorial forms of diabetes. It is
generally assumed that, for polygenic or quantitative traits,
each allele has a small additive effect. Susceptibility-gene
discovery in obesity recapitulates much of that in T2D
diabetes.
Until recently, the same strategies—genome-wide link-
age following positional cloning and the candidate-gene
approach—have been employed extensively. However, the
development of high-density single-nucleotide polymor-
phism (SNP) genotyping arrays has expanded association
analysis of common variants to a genome-wide scale, with
the high-throughput nature of these technologies permit-
ting genotyping of millions of SNPs in thousands of indi-
viduals (Loos et al., 2009).
Numerous genomic loci have been shown to be asso-
ciated with type 2 diabetes and its associated traits and
similarly BMI/waist–hip ratios as measures of obesity, with
many of them having been replicated in various studies and
populations ( Julia et al., 2013).
It is important to recognize the development of tech-
nologies and strategies that led to the discovery of gene loci
associated with these complex diseases, for better under-
standing of and perspective on genomic studies. These have
been outlined in the following section.
 L I N K AG E A N A LYS E S
Linkage analysis is a method of mapping disease genes in
affected families by genotyping about 400–500 genetic
markers, whereby disease loci can be mapped on a
genome-wide level. It attempts to identify chromosomal
regions co-segregating with a disease phenotype in individ-
uals who are related. Finding that affected family members
share a certain marker that is identical by descent—that is,
identical because it was inherited from the same parent—
more often than expected by chance, is evidence that a
disease-causing variant is in linkage disequilibrium with the
genotyped marker.
This method has been very useful in identifying familial
genetic variants with large effects such as those giving rise to
MODY (Vaxillaire et al., 2006). However, it has been less
successful in identifying genes that cause complex diseases
such as T2D. Although great efforts have been put into
linkage studies of T2D, only two genes have been reported
to have been identified by linkage: calpain 10 (CAPN10)
and transcription factor 7–like 2 (TCF7L2). In the case of
TCF7L2, a T2D locus was mapped to chromosome 10q
in both an Icelandic and a Mexican-American population
(Duggirala et al., 1999). This region was later fine-mapped
in the Icelandic population by the use of 228 microsatel-
lite markers covering a 10.5-Mbp region, pinpointing the
locus to intron 3 of the TCF7L2 gene (Grant et al., 2006).
The association between T2D and a number of SNPs in the
TCF7L2 gene has since been confirmed in numerous studies
in different ethnic groups (Scott et al., 2007; Cauchi et al.,
2007; Chandak, 2007; Bodhini et  al., 2007). All of these
studies demonstrate robust and convincing statistical evi-
dence of association with diabetic risk and consistent effect
sizes. The risk allele confers a relative risk of approximately
1.4 compared to homozygous carriers of the non-risk allele,
making this the strongest association with T2D by far.
The other gene mapped by linkage analysis is a locus
on chromosome 2 that was first mapped in 1996 by Hanis
and colleagues. This locus was fine-mapped and the proxi-
mal gene shown to be CAPN10, the gene for calpain 10,
a cysteine protease with largely unknown functions in
glucose metabolism (Horikawa et  al., 2000; Weedon,
2003). Despite a number of negative replication studies,
several meta-analyses have shown consistent association
of CAPN10 with T2D. Nevertheless, none of the large
genome-wide association studies (GWAS) has identified
CAPN10 as being associated with T2D.
Genome-wide linkage scans in families with more
common forms of obesity than the monogenic forms
have resulted in several loci, but the genes within these
3 4 2   •   G enomics in C linical P ractice
loci have remained elusive. In order to find yet-unknown
genes involved in obesity, the entire genome scanning
was employed to search for linkage between polymorphic
markers and disease-related phenotypes. This method had
been successful in mapping monogenic obesity; while
some quantitative trait loci such as body weight have been
mapped, including the genes for glutamic acid decarbox-
ylase enzyme (GAD2), ectonucleotide pyrophosphatase/
phosphodiesterase1 (ENPP1), and solute carrier family 6
member 14 (SLC6A14) (Sladek et al., 2007). Validation in
some of these is yet to be completed.
C A N D I DAT E G E N E A P P R OAC H
Identification of disease genes can also be made on the basis
of association testing in populations rather than in families.
Candidate-gene studies detect associations between genetic
markers in specific genes considered to be candidates for
the phenotype of interest. These may be functional candi-
dates, based on biological evidence in pathways relevant to
disease, or positional candidates mapping to regions impli-
cated in animal or linkage studies. Candidate-gene studies
are hypothesis-driven.
In diabetes genetics, based on hypotheses, the
candidate-gene approach focuses on the search for an asso-
ciation between T2D and sequence variants in or near bio-
logically defined candidate genes, which have been chosen
based on their known physiological function. The impor-
tance of these variants or other nearby variants is tested
by comparing the frequency in T2D patients and normal
glucose-tolerant subjects. Table 22.1 highlights the genetic
studies in type 2 diabetes in different populations using can-
didate gene approach.
Extending the analysis of genes implicated in mono-
genic forms of diabetes has also proved successful for T2D,
as exemplified by HNF4A, HNF1A, and KCNJ11 genes.
Common variants of HNF4A (MODY1) have been associ-
ated with T2D in both Finnish and Ashkenazi Jewish pop-
ulations (Silander et al., 2004; Love-Gregory et al., 2004).
One of the main candidate genes that is implicated in
adipogenesis, insulin resistance, and T2D is the peroxi-
some proliferator activated receptor-γ(PPAR-γ) gene. This
is a transcription factor that is involved in adipogenesis and
in the regulation of adipocyte gene expression and glucose
metabolism. Within a unique domain of PPAR-γ2 gene
that enhances ligand-independent activation, a common
Pro12Ala polymorphism has been identified (Altshuler
et al., 2000). Deeb and colleagues (1998) reported that
the Ala allele of this polymorphism was associated with
Table 22.1  SUMMARY OF GENETIC STUDIES IN TYPE 2 DIABETES IN DIFFERENT POPULATIONS USING
CANDIDATE-GENE APPROACH

1 PPAR g gene (Pro12Ala) Deeb et al., 1998; Lyssenko et al., 2005 Pro12Ala polymorphism protective in Caucasians
2 PPAR g gene (Pro12Ala) Radha et al., 2006 Pro12Ala polymorphism not protective in South Asians
3 PGC-1 α gene (Gly482Ser) Ek et al., 2001 Associated with relative risk of type 2 diabetes in a
European population
4 PGC-1 α gene (Thr394Thr) Vimaleswaran et al., 2005, 2006 Associated with type 2 diabetes and with body fat
5 PGC-1 α gene (Gly482Ser) Liang et al., 2006 Associated with the development of insulin resistance and
type 2 diabetes
6 PGC-1 α gene (Thr394Thr, Bhat et al., 2007 Associated with type 2 diabetes
Gly482Ser)
7 PC-1 gene (K121Q) Bacci et al., 2005 Associated with insulin resistance/atherogenic phenotypes
8 PC-1 gene (K121Q) Abate et al., 2005 Associated with type 2 diabetes
9 TCF7L2 gene(rs7903146) Grant et al., 2006; Scott et al.; T allele of rs7903146 associated with an increased risk of
Cauchi et al. type 2 diabetes
10 TCF7L2 gene(rs7903146) Chandak et al., 2007 Associated with type 2 diabetes
11 TCF7L2 gene(rs12255372; Bodhini et al., 2007 Associated with type 2 diabetes in Asian Indians
rs7903146)
12 TCF7L2 gene(rs7903146) Sanghera et al., 2008 Associated with type 2 diabetes in Asian Indians
13 Adiponectin gene (+45T/G; Hara et al., 2001; Menzaghi et al., 2002; Significantly associated with type 2 diabetes and adiponec-
+276G/T) Vasseur et al., 2002 tin level in Japanese population and with insulin resistance
in some Caucasian populations
14 Adiponectin gene Stumvoll et al., 2002 SNP 45 is associated with obesity in a German population
15 Adiponectin gene Vimaleswaran et al., 2008 +10211T→G Associated with type 2 diabetes
16 FTO gene Frayling et al., 2007 Associated with body mass index
17 FTO gene Dina Associated with obesity-related traits
18 FTO gene Yajnik et al., 2009 Associated with type 2 diabetes in south Asian Indians
19 FTO gene Ramya et al., 2010 Associated with type 2 diabetes and obesity in South Asian
Indians

increased insulin sensitivity and decreased risk of T2D in
a Finnish and a second-generation Japanese cohort. Since
this initial work, the preponderance of evidence has sup-
ported PPARG’s association with T2D, with an odds ratio
(OR) of ~1.2. The risk of T2D conferred by this SNP
has been studied prospectively in the Finnish Diabetes
Prevention Study and the larger Botnia Prevention Study.
In the Finnish study (Uusitupa et al., 2001), 500 subjects
with impaired glucose tolerance, the relative risk of devel-
oping diabetes was doubled in alanine carriers, contradict-
ing the prior evidence that the alanine allele was protective.
In the larger Botnia study, comprising >2000 subjects, pro-
line homozygotes were 1.7 times more likely to develop
diabetes than alanine carriers. In contrast in our group we
found that the Pro12Ala polymorphism of the PPAR–G
gene, which is protective against diabetes in Caucasians,
does not offer protection in two cohorts of South Asians
studied at Chennai, India, and Dallas in the United States
(Radha et al., 2006).
The SNP E23K of KCNJ11 has now been convincingly
associated with T2D. Although initial smaller studies failed
to replicate the association of the E23K polymorphism
with T2D, large-scale studies and meta-analyses have con-
sistently associated the lysine variant with T2D, with an
OR of 1.15 (Moore et al., 2008).
Adiponectin, encoded by the ADIPOQ gene, is one
of the adipocyte-expressed proteins that enhances insulin
sensitivity and functions in regulating the homeostatic con-
trol of glucose, lipid, and energy metabolism (Diez et  al.,
2003). Genome-wide scans have mapped a susceptibility
locus for type 2 diabetes and obesity/metabolic syndrome
to chromosome 3q27, where the ADIPOQ gene is located
(Kissebah et  al., 2000; Vionnet et  al., 2000; Comuzzie
et  al., 2001; Lindsay et  al., 2003)  SNPs of the ADIPOQ
gene have been genotyped in large datasets from various
ethnic groups, and several SNPs associated with hypoadi-
ponectinemia, obesity, and type 2 diabetes have been iden-
tified (Menzaghi et al., 2002; Vasseur et al., 2003; Gibson

G enomics of T ype 2 D iabetes M ellitus and O besity   •   3 4 3

et al., 2004; Berthier et al., 2005; Heid et al., 2006). Two
SNPs in the adiponectin gene, a silent T to G substitution
in exon 2 (+45T/G) and a G to T substitution in intron 2
(+276G/T), were significantly associated with type 2 diabe-
tes and adiponectin level in a Japanese population and with
insulin resistance in some Caucasian populations (Italy,
Germany) (Hara et al., 2001;Ge et al., 2004); and SNP 45 is
associated with obesity in a German population (Stumvoll
et al., 2002). In the proximal promoter region of the APM1
gene, SNP-11426A/G and -11391A/-11377G haplo-
type predicted the associations with fasting plasma glu-
cose, type 2 diabetes, and adiponectin levels. Adiponectin
has been associated with low diabetes risk. The metabolic
effects of adiponectin are mediated by adiponectin recep-
tors 1 (ADIPOR1) and 2 (ADIPOR2). A  study on sicx
polymorphisms in ADIPOR1 and 16 polymorphisms in
ADIPOR2 was carried out and a significant association
between ADIPOR1 haplotypes and diabeted risk was
observed. Adiponectin is an adipose tissue–specific protein
that is decreased in subjects with obesity and type 2 diabe-
tes (Lindsay et al., 2003; Qi et al., 2007). Our study showed
for the first time that the +10211T→G polymorphism in
the first intron of the adiponectin gene is associated with
type 2 diabetes, obesity, and hypoadiponectinemia (odds
ratio [OR] 1.28; 95% CI 1.07–1.54; P = 0.008) in an Asian
Indian population (Vimaleswaran et  al., 2008), thereby
suggesting adiponectin to be a very important gene for obe-
sity and type 2 diabetes.
Most candidate gene studies of common obesity looked
at variants in genes already implicated in rarer, mono-
genic forms of the disease. Since the pathophysiological
basis of obesity is largely unknown, a hypothesis-driven
candidate-gene approach could identify only a small frac-
tion of the genetic risk factors for the disease.
Common variants in the genes leptin and leptin recep-
tor were associated with BMI and obesity in several popu-
lations (Loos, 2012). Adiponectin, a hormone that plays
a key role in regulation of glucose and fatty acids and has
reduced levels in obesity and type 2 diabetes, has been
found to be genetically associated with these diseases
around the world (Diaz, Iglesias, 2003). Association stud-
ies of various candidate genes have also implicated the gene
encoding such factors as cannabinoid receptor 1 (CNR1),
dopamine receptor 2 (DRD2), serotonin receptor 2C
(htr2c), and SLC6A4 (McCarthy et  al., 2005); but the
most replicated of them is Pro12 Ala substitution in the
PPARG gene, which has been extensively associated with
both obesity and type 2 diabetes. All these studies support
the view that obesity is a complex disease influenced by
many genes with small effect size.
3 4 4   •   G enomics in C linical P ractice
Both the linkage analysis and candidate-gene approaches
had limited success and were unable to explain the genetics
of complex diseases satisfactorily. Scientists hence embarked
on looking at the genome completely, to discover multiple
gene variants with individually small effects. That led to the
era of the genome-wide association studies.
G E N O M E -W I D E A S S O C I AT I O N
S T U D I E S ( GWA S )
The ability to interrogate the entire genome was made pos-
sible by two key advances: the Human Genome Project and
the International HapMap project. Thus the novel approach
of searching for genetic association in a “genome-wide”
fashion came into practice. Scientists embarked on GWAS,
which allowed them to discover multiple gene variants with
individually small effects. This is a high-throughput meth-
odology that allows the scanning of a dense set of SNPs
spanning across the entire human genome in an unbiased
manner, using powerful statistical methods to study asso-
ciations between a given disease phenotype and a repre-
sentation of all common variations in the genome. Once
a specific polymorphism is associated with a disease, it is
usually annotated by naming the gene in closest proximity
to it. These variants are not causal for the phenotype but
operate as tag-SNPs that capture the common haplotype
variation in a given region of the human genome. The tag
SNP flags a genomic region that harbors the causal variant,
which may itself be acting at a certain distance; for instance,
by modulating expression of a faraway gene. Therefore,
while association signals are often identified by gene names,
deep-sequencing efforts, fine-mapping, and functional
approaches only demonstrate a causal relationship between
gene locus and the phenotype.
Since GWAS is a non–hypothesis-driven approach, this
strategy can be used to uncover new insights into the biol-
ogy of a given phenotype without any prior knowledge of
function (Shu et al., 2010). GWAS have an advantage over
genome-wide linkage studies in that they do not require the
study subjects to be related, which allows for studies with
larger sample sizes, thus increasing their power to detect
true associations. Larger samples enable the discovery of
new genes and smaller effects, and also provide more accu-
rate effect-size estimates.
GWAS have been a very successful approach, with many
new loci implicated in complex traits (Groop et al., 2013).
Type 2 diabetes and obesity have been beneficiaries of this
strategy, since substantial progress in our knowledge has
been elucidated by GWAS studies.
 The first GWAS for T2D was conducted in a French dis-
covery cohort composed of 661 cases of T2D and 614 non-
diabetic controls who were genotyped on two genotyping
platforms. In total, 392,935 SNPs from two different geno-
typing platforms were analyzed for association with T2D.
This study identified novel and reproducible association sig-
nals at SLC30A8 and HHEX and validated the well-known
association at TCF7L2 (Sladak et al., 2007). Investigators
from the Icelandic company deCODE and their collabora-
tors confirmed the association of loci SLC30A8 and HHEX
with T2D and identified an additional signal in CDKAL1
(Frayling et al., 2007). Three other collaborating groups, the
Wellcome Trust Case Control Consortium (WTCCC), the
Finland–United States Investigation of non-insulin depen-
dent diabetes mellitus (NIDDM), Genetics (FUSION)
group, and the Diabetes Genetics Initiative (DGI), pub-
lished their findings replicating SLC30A8 and HHEX, and
independently discovering novel associations at CDKAL1,
IGF2BP2, and CDKN2A/B (Steinthorsdottir et al., 2007;
Scott, 2007). These discoveries led to a plethora of studies
that replicated the top signals in various ethnic popula-
tions (Zeggini et al., 2007; Yajnik et al., 2009; Ramya et al.,
2011). A number of loci were replicated, but many of them
were unable to be replicated.
The first GWAS studies on T2D in non-European popu-
lations were published in 2008 using a multi-stage approach
(Yasuda et al., 2008) and genotyping 100,000 SNPs in 187
T2D cases and two different control populations, each
including 752 individuals, thereby identifying 2800 candi-
date SNPs for follow-up in replication cohorts. Both stud-
ies identified the KCNQ1 gene locus. The second wave of
discoveries comprised meta-analyses of more than 50,000
individuals. Many research groups worked together in con-
sortia like DIAGRAM (DIAbetes Genetics Replication
and Metaanalysis Consortium), MAGIC (Meta-Analyses
of Glucose and Insulin related traits Consortium). These
have resulted in more than 50 loci for T2D.
The identification of novel genes by GWAS is the dis-
covery phase of the work, while the replication of these loci
forms the validation phase. Both these phases are essential
for the discovery of novel genes/gene loci. Our own repli-
cation study inChennai Urban Rural Epidemiology Study
(CURES) resulted in the identification of genes/gene vari-
ants such as rs7756992, rs7754840, and rs6931514 of the
CDKAL1, rs7020996 of the CDKN2A/B gene, rs7923837
of the HHEX gene, and rs12056034 of the BAZ1B genes
as associated with T2D in our population (Chidambaram
et al., 2010).
The coming together of big study groups in collabora-
tion resulted in meta-analysis of the GWAS that identified
G enomics of T ype 2 D iabetes M ellitus and O besity   •   3 4 5
genes, and this gave us a large sample size, increasing the
power and credibility of the study and enabling us to dis-
cover common variants with frequency >5%. In a collabora-
tive effort to study the population of South Asian ancestry,
we performed a GWAS followed by replication of top
SNPs, and identified novel common genetic variants at six
loci (GRB14, ST6GAL1, VPS26A, HMG20A, AP3S2 and
HNF4A) newly associated with T2D (P  =  4.1  × 10−8 to
P = 1.9 × 10−11). SNPs at GRB14 were also associated with
insulin sensitivity (P = 5.0 × 10−4), and SNPs at ST6GAL1
and HNF4A were also associated with pancreatic beta-cell
function (P = 0.02 and P = 0.001, respectively). Our find-
ings clearly provide additional insight into mechanisms
underlying T2D and show the potential for new discovery
from genetic association studies in South Asians, a popu-
lation with increased susceptibility to T2D (Kooner et al.,
2011).
Recently, in a first Indian GWAS study, we identified
a novel gene locus rs998451 (OR = 1.56, P = 6.3×10–12)
within TMEM163 gene locus, which encodes a probable
vesicular transporter in nerve terminals (Tabassum et al.,
2013). TMEM163 variants also showed association with
decreased fasting plasma insulin and also homeostatic
model assessment of insulin resistance, indicating plau-
sible effect through impaired insulin secretion. Forty-nine
of 56 previously reported signals showed consistency in
direction, with similar effect sizes in Indians and previ-
ous studies; 25 of them were also associated with T2D
(P <0.05). Known loci and the newly identified 2q21
locus altogether explained 7.65% variance in risk of type 2
diabetes in Indians. This study suggests that common sus-
ceptibility variants for type 2 diabetes are largely the same
across populations, with population-specific loci, and it
also provides further insights into the genetic architec-
ture and etiology of type 2 diabetes in Indians. In another
GWAS and multistage meta-analysis study in Punjabi
Sikhs from Northern India (Saxena et al., 2013), a novel
locus at 13q12 in the SGCG gene (rs9552911, P = 1.82 x
10–8) was associated with T2D susceptibility. This finding
demonstrates a population-specific association that could
lead to additional biological insights into T2D pathogen-
esis (Table 22.2).
Like in the case of most GWAS, BMI and obesity
GWAS have implicated multiple loci (Zeggini et al., 2008).
To date there have been four waves of discovery for BMI,
with each wave increasing the number of studies and num-
ber of loci identified. Insulin-induced gene 2 (INSIG2) was
the first locus to be reported in a GWAS of obesity although
replication efforts yielded very inconsistent results (Herbert
et al., 2006).
 Table 22.2  SUMMARY OF GENETIC STUDIES IN TYPE 2 DIABETES IN ASIAN INDIAN
POPULATIONS BY REPLICATION AND GWAS APPROACH
1 CDKAL1 gene Chidambaram et al., 201
CDKN2A/B gene
HHEX gene
BAZ1B gene
2 GRB14 gene Kooner et al., 2011
ST6GAL1 gene
VPS26A gene
AP3S2 gene
HMG20A gene
HNF4A gene
3 TCF7L2 gene Tabassum et al., 2012
TMEM163 gene
TMEM163 gene
MAP3K1 gene
TGFBR3 gene
FLJ35379 gene
The GWAS that discovered the first loci associated
with BMI was part of the Welcome Trust Case Control
Consortium studies examining the genetics of type 2 dia-
betes (Frayling et  al., 2007). Genetic variation in 1,924
individuals with type 2 diabetes was compared with
that in 2938 controls. A  SNP in the FTO (fat mass and
obesity-associated) gene was found to show strong associa-
tion with type 2 diabetes, which was replicated in the sec-
ond stage using a larger sample size. Once the analyses were
adjusted for BMI, the association with T2D was abolished.
The effect of FTO on T2D was on a pathway through BMI.
The association with BMI was subsequently established
by replicating in a sample comprising 19,424 adults from
seven studies and 10,172 children from two different stud-
ies (Grant et al., 2008). Thus came the conclusion that the
FTO locus affects diabetes through its effect on adiposity.
Subsequently, numerous replication studies also confirmed
the association of the FTO locus with BMI and related obe-
sity traits. Earlier studies in Asian Indians on rs9939609
T→A and rs7193144 C→T variants of the intron 1 of the
FTO gene showed an association with type 2 diabetes that
was independent of BMI (Chauhan et al., 2011). Our recent
association study showed that the rs8050136 C→A vari-
ant is associated with both Generalised Obesity is the one
in which BMI is taken as the measure of obesity. Central
Obesity is the one in which waist circumference and waist
hip ratio are the measures of obesity. The rs8050136 C/A
polymorphism was associated with generalized obesity. The
odds ratio for obesity for the CA genotype was significant
even after adjustment for age, sex, and diabetes [OR:  2.0
(95% CI: 1.57 –2.76), p <0.0001]. This study demonstrated
that there is no independent association of rs8050136 C→A
with T2DM, as its ((association with T2DM appears to be
3 4 6   •   G enomics in C linical P ractice
Associated with type 2 diabetes in south Indians
Associated with type 2 diabetes in south Asian
ancestry including south Indians
Associated with type 2 diabetes in Indians
linked through obesity (Ramya et  al., 2011). FTO is still
considered the locus with the largest effect on BMI, which,
together with its high minor allele frequency, implies that
it is a low-hanging fruit that could be easily identified by
moderately powerful studies.
The second wave of discoveries was possible because of
the increase in their sample size. In order to increase the
statistical power in the analysis of gene variants associated
with BMI, a multinational collaboration known as the
Genetic Investigation of ANthropometric Traits (GIANT)
Consortium was established. This study confirmed the
strong association with variation in FTO and identified one
new locus near melanocortin 4 receptor (MC4R), a gene in
which mutations are a common cause of extreme childhood
obesity (Cauchi et al., 2009; Heid et al., 2010). At the same
time, the GWAS in subjects of Asian Indian origin identi-
fied the same locus to be associated with waist circumfer-
ence and related traits. This finding was further replicated in
South Asians, East Asians, various white European popula-
tions, and African Americans (Chambers et al., 2008). The
sample size was doubled in a third wave of discoveries by
the GIANT Consortium, using 32,387 individuals of white
European descent, and identifying six new loci robustly
associated with BMI. Ten loci were near TMEM18, near
KCTD15, near GNPDA2, in SH2B1, in MTCH2, near
NEGR1, near FAIM2, near SEC16B, near ETV5, and in
BDNF (Willer et al., 2009; Thorleifsson et al., 2009).
In the fourth wave of discoveries, the GIANT
Consortium expanded its GWAS stage to comprise 123,865
individuals from 46 populations of white European descent.
By the end of the fourth wave, GWAS had identified 32 loci
unequivocally associated with BMI (Speliotes et al., 2010;
Day et  al., 2011; Loos, 2012). Although the number of
discovered loci increased with each wave of discovery, the
32 loci combined explained only 1.45 % of the phenotypi-
cal variation in BMI, and accounted for only 2–4% of the
heritability.
For the overall adiposity, BMI is a good proxy, but it
does not allow distinguishing between lean mass and fat
mass. Body fat percentage is a better proxy, and a recent
meta-analysis of 15 GWAS of body fat percentage again
confirmed FTO as an obesity-susceptibility locus, and iden-
tified SPRY2 and IRS1 loci.
Two waves of GWAS meta-analyses for waist hip ratio
(WHR) have been performed. Because waist circumference
and BMI are highly correlated, all loci identified in GWAS
for waist circumference were also identified in a GWAS
for BMI, suggesting that these are associated with overall
obesity, rather than with abdominal obesity in particular
(Saunders et al., 2007; Lindgren et al., 2009).
Yet another meta-analysis by the GIANT Consortium
on abdominal obesity was performed by testing for associa-
tion with WHR adjusted for BMI. Fourteen loci reached
genome-wide significance, but none of the WHR loci over-
lapped with any of the BMI loci, suggesting that they are
likely to influence abdominal obesity, rather than overall
obesity (Heid et al., 2010). Furthermore, five GWAS have
been done on extreme and/or early-onset obesity. Three
new loci were identified besides FTO and MC4R loci.
Considering these results together, much remains to
be uncovered in the search for genetic variants influencing
BMI and obesity susceptibility (Walley et al., 2009). GWAS
have identified at least 52 genetic loci associated with obe-
sity and related traits in a span of just five years (Walters
et al., 2010; den Hoed et al., 2010).
A promising area of research involves the effect of
obesity-associated variants on bariatric surgery outcomes.
Sarzynski and colleagues (2011) studied the role of 11
common BMI-associated variants on weight loss after bar-
iatric surgery. They found a SNP within the FTO gene to
be significantly associated with postsurgical weight loss. In
another study, Still et al.(2011) stratified the patients who
underwent bariatric surgery into “low,” “intermediate,” and
“high” risk groups based on their genetic profiles at four
obesity-associated SNPs and found significant differences
in postoperative weight loss between each of the three
groups of patients.
The GWAS have successfully identified and replicated
nearly 75 susceptibility loci associated with T2D and
related metabolic traits so far, mostly in Europeans, and
some in African and South Asian populations (Sanghera
et  al., 2012). In the meantime, new loci are being identi-
fied and validated, and these findings reveal new molecular
G enomics of T ype 2 D iabetes M ellitus and O besity   •   3 4 7
pathways underlying diabetes etiology, gene–environment
interactions, epigenetic modifications, and gene functions.
While most of the studies have focused on SNPs, the roles
of copy number variations and rare structural variants
remain to be understood for complex diseases such as dia-
betes and obesity.
C O N C LU S I O N
The use and application of genetic information to clinical
settings has been limited to prediction of monogenic forms
of T2D and obesity. Noteworthy examples are the cases of
neonatal diabetes and extreme childhood obesity (Bradfield
et  al., 2012). The predictive value of established diabetic
susceptibility and obesity susceptibility loci (Day et  al.,
2011) and their ability to discriminate between individuals
who are at higher risk of T2D or obesity in adult life versus
those who are at lower risk, is still not good enough to be
translated to clinical settings and personalized medicine.
A substantial portion of the predictive heritability
of type 2 diabetes and obesity and inter-individual vari-
ability in BMI remain elusive. This missing heritability
raises the question of whether the heritability of these
diseases is overestimated, or whether hitherto unex-
plained genetic variations are still present that remain
unexplored by the current methodologies for large-scale
association analyses. Deep-sequencing, fine-mapping
of the regions identified by GWAS, and the functional
studies of specific genetic variations are underway and
would bring to light the role of the GWAS-established
genetic loci. Thus the primary outputs of these discover-
ies are in gaining new insights into the biological path-
ways that underlie these diseases. Each new discovery
brings a number of functional and clinical/epidemiolog-
ical follow-up studies that will help increase our under-
standing of the mechanisms involved.
REFERENCES
Abate N et al. ENPP1/ Pc–1 K121Q polymorphism and genetic suscep-
tibility to type 2 diabetes. Diabetes 54: 1207–1213, 2005.
Aguilar–Salinas CA et  al. Early-onset type 2 diabetes:  Metabolic
and genetic characterization in the Mexican population. J Clin
Endocrinol Metab 86: 220–226, 2001.
Ahamed A et al. Permanent neonatal diabetes mellitus due to a C96Y
heterozygous mutation in the insulin gene. A case report. J Pancreas
9: 715–718, 2008.
Alberti K G et al. Global burden of disease—where does diabetes mel-
litus fit in? Nature Rev Endocrinol 95: 258–260, 2013.
Allison et al. Risch’s lambda values for human obesity. Int J Obes Relat
Metab Disord, 20: 990–999, 1996.
Altshuler D et al. The common PPAR–gamma Pro12Ala polymorphism
is associated with decreased risk of Type 2 diabetes. Nat Genet
26: 76–80, 2000.
Anuradha S et al. A prevalent amino acid polymorphism at Codon 98
(Ala98Val) of the hepatocyte nuclear factor–1 is associated with
maturity onset diabetes of the young and younger age at onset of
type 2 diabetes in Asian Indians. Diabetes Care 28:  2430–2435,
2005.
Anuradha S et al. Association of novel variants in the hepatocyte nuclear
factor 4A gene with maturity onset diabetes of the young and early
onset type 2 diabetes. Clin Genet 80 (6): 541–549, 2011.
Beales PL et  al. Genetic interaction of BBS1 mutations with alleles at
other BBS loci can result in non-Mendelian Bardet-Biedl syndrome.
Am J Hum Genet 72: 1187–1199, 2003.
Bell CG et al. The genetics of human obesity. Nat Rev Genet 6: 221–234,
2005.
Bell GI et al. Sequence of the human insulin gene. Nature 284: 26–32,
1980.
Bellanne-Chantelot C et al. Clinical spectrum associated with hepato-
cyte nuclear factor-1beta mutations. Ann Intern Med 140:  510–
517, 2004.
Bellanne-Chantelot C et al. Large genomic rearrangements in the hepa-
tocyte nuclear factor-1{beta} (TCF2) gene are the most frequent
cause of maturity-onset diabetes of the young type 5. Diabetes
54: 3126–3132, 2005.
Berthier MT et al. Impact of adiponectin gene polymorphisms on plasma
lipoprotein and adiponectin concentrations of viscerally obese men.
J Lipid Res 46: 237–244, 2005.
Bodhini D et al. The rs12255372 (G/T) and rs7903146 (C/t) polymor-
phisms of the TCF7L2 gene are associated with type 2 diabetes mel-
litus in Asian Indians. Metabolism 56 (9): 1174–1178, 2007.
Bradfield JP et  al. A genome-wide association meta-analysis identifies
new childhood obesity loci. Nat Genet 44: 526–531, 2012.
Cauchi S et al. Combined effects of MC4R and FTO common genetic
variants on obesity in European general populations. J Mol Med
87: 537–546, 2009.
Cauchi S et al. TCF7L2 is reproducibly associated with type 2 diabetes
in various ethnic groups: a global meta-analysis. J Mol Med (Berl).
85 (7): 777–782, 2007.
Chambers JC et  al. Common genetic variation near MC4R is associ-
ated with waist circumference and insulin resistance. Nat Genet
40: 716–718, 2008.
Chandak GR et al. Common variants in the TCF7L2 gene are strongly
associated with type 2 diabetes mellitus in the Indian population.
Diabetologia 50: 63–67, 2007.
Chang YC et al. Common variation in the fat mass and obesity-associated
(FTO) gene confers risk of obesity and modulates BMI in the
Chinese population. Diabetes 57: 2245–2252, 2008.
Chauhan G et al. Common variants of FTO and the risk of obesity and
type 2 diabetes in Indians. J Hum Genet 56: 720–726, 2011.
Chen H et al. Evidence that the diabetes gene encodes the leptin recep-
tor: identification of a mutation in the leptin receptor gene in db/db
mice. Cell 84: 491–495, 1996.
Chidambaram M et  al. Replication of recently described type 2 gene
variants in a South Indian population. Metabolism 59 (12): 1760–
1766, 2010.
Clement K et al. A mutation in the human leptin receptor gene causes
obesity and pituitary dysfunction. Nature 392: 398–401, 1998.
Comuzzie AG et al. The genetic basis of plasma variation in adiponectin,
a global endophenotype for obesity and the metabolic syndrome. J
Clin Endocrinol Metab 86: 4321–4325, 2001.
Day FR et al. Developments in obesity genetics in the era of genome–
wide association studies. J Nutrigenet Nutrigenomics 4: 222–238,
2011.
Deeb SS et  al. A Pro12Ala substitution in PPARγ2 associated with
decreased receptor activity, lower body mass index and improved
insulin sensitivity. Nat Genet 20: 284–287, 1998.
3 4 8   •   G enomics in C linical P ractice
den Hoed M et al. Genetic susceptibility to obesity and related traits in
childhood and adolescence. Diabetes 59: 2980–2988, 2010.
Diabetes Genetics Initiative of Broad Institute of Harvard and MIT,
Lund University, and Novartis Institutes of BioMedical Research.
Genome-wide association analysis identifies loci for type 2 diabetes
and triglyceride levels. Science 316: 1331–1336, 2007.
Diamond J. The double puzzle of diabetes. Nature 423: 599–602, 2003.
Diez JJ et al. The role of the novel adipocyte-derived hormone adiponec-
tin in human disease. Eur J Endocrinol 148: 293–300, 2003.
Dina C et  al. Variation in FTO contributes to childhood obesity and
severe adult obesity. Nat Genet 39: 724–726, 2007.
Drebin JA et  al. Molecular cloning and chromosomal localization of
the human homologue of a B-lymphocyte specific protein tyrosine
kinase (blk). Oncogene 10: 477–486, 1995.
Dubern B et  al. Mutational analysis of melanocortin-4 receptor,
agouti-related protein, and alpha-melanocyte-stimulating hormone
genes in severely obese children. J Pediatr 139: 204–209, 2001.
Duggirala R et al. Linkage of type 2 diabetes mellitus and of age at onset
to a genetic location on chromosome 10q in Mexican Americans.
Am J Hum Genet 64: 1127–1140, 1999.
Echwald SM et al. Amino acid variants in the human leptin receptor: Lack
of association to juvenile onset obesity. BBRC 52 (233): 248–252,
1997.
El-Gharbawy AH et  al. Serum brain-derived neurotrophic factor con-
centrations in lean and overweight children and adolescents. J Clin
Endocrinol Metab 9: 3548–3552, 2006.
Ellard S et al. Mutations in the genes encoding the transcription factors
hepatocyte nuclear factor 1 alpha (HNF1A) and 4 alpha (HNF4A)
in maturity-onset diabetes of the young. Hum Mutat 27: 854–869,
2006.
Eriksson J et al. Early metabolic defects in persons at increased risk for
non-insulin-dependent diabetes mellitus. N Engl J Med 321: 337–
343, 1989.
Fajans SS et al. Molecular mechanisms and clinical pathophysiology of
maturity-onset diabetes of the young. N Engl J Med 345: 971–980,
2001.
Fajans SS et al. MODY: history, genetics, pathophysiology, and clinical
decision making. Diabetes Care 34: 1878–1884, 2011.
Flanagan SE et  al. Mutations in KCNJ11, which encodes Kir6.2, are
a common cause of diabetes diagnosed in the first 6  months of
life, with the phenotype determined by genotype. Diabetologia
49: 1190–1197, 2006.
Frayling TM et al. A common variant in the FTO gene is associated with
body mass index and predisposes to childhood and adult obesity.
Science 316: 889–894, 2007.
Froguel P et  al. Familial hyperglycemia due to mutations in glucoki-
nase. Definition of a subtype of diabetes mellitus. N Engl J Med
328: 697–702, 1993.
Gibson F et  al. Genetics of the ADIPOQ locus and its contribution
to Type 2 diabetes susceptibility in French Caucasians. Diabetes
53: 2977–2983, 2004.
Gloyn AL et  al. Large scale association studies of variants in genes
encoding the pancreatic beta-cell K-ATP channel subunits Kir6.2
(KCNJ11) and SUR1 (ABCC8) confirm that the KCNJ11
E23K variant is associated with increased risk of Type 2. Diabetes
52: 568–572, 2003.
Grant SF et  al. Variant of transcription factor 7–like 2 (TCF7L2)
gene confers risk of type 2 diabetes. Nat Genet 38:  320–323,
2006.
Grant SF et al. Association analysis of the FTO gene with obesity in chil-
dren of Caucasian and African ancestry reveals a common tagging
SNP. PLoS One 233: 248–252, 2008.
Greeley SA et al. Neonatal diabetes: An expanding list of genes allows for
improved diagnosis and treatment. Curr Diabetic Rep 11: 519–532,
2011.
Groop L et  al. Genetics of diabetes—Are we missing the genes or the
disease? Mol Cell Endocrinol. 382: 726–239, 2014.
Gu HF et al. Single nucleotide polymorphisms in the proximal promoter
region of the adiponectin (APM1) gene are associated with type 2
diabetes in Swedish Caucasians. Diabetes 53 (Suppl. 1): S31–S35,
2004.
Hanis CL et al. A genome-wide search for human non-insulin-dependent
(type 2) diabetes genes reveals a major susceptibility locus on chro-
mosome 2. Nat Genet 13: 161–166, 1996.
Hara K et  al. Genetic variation in the gene encoding adiponectin is
associated with increased risk of type 2 diabetes. Diabetes 50
(Suppl. 1): A244, 2001.
Hattersley A et al. The diagnosis and management of monogenic diabe-
tes in children and adolescents. Pediatr Diabetes 10: 33–42, 2009.
Hattersley AT et al. Activating mutations in Kir6.2 and neonatal diabe-
tes: new clinical syndromes, new scientific insights, and new ther-
apy. Diabetes 54 (9): 2503–2513, 2005.
Hattersley AT et al. Mutations in the glucokinase gene of the fetus result
in reduced birth weight. Nat Genet 19: 268–270, 1998.
Heid IM et al. Association of the 103I MC4R allele with decreased body
mass in 7937 participants of two population based surveys. J Med
Genet 42: e21, 2005.
Heid IM et  al. Genetic architecture of the APM1 gene and its inX-
uence on adiponectin plasma levels and parameters of the
metabolic syndrome in 1,727 healthy Caucasians. Diabetes
55: 375–384, 2006.
Heid IM et al. Meta-analysis identifies 13 new loci associated with waist–
hip ratio and reveals sexual dimorphism in the genetic basis of fat
distribution. Nat Genet 42: 949–960, 2010.
Herbert A et al. A common genetic variant is associated with adult and
childhood obesity. Science 312 (5771): 279–283, 2006.
Horikawa Y et al. Genetic variation in the gene encoding calpain-10 is
associated with type 2 diabetes mellitus. Nat Genet 26: 163–175,
2000.
Hu, FB et al. Diet, lifestyle, and the risk of type 2 diabetes mellitus in
women. N Engl J Med 345: 790–797, 2001.
Inoue H et al. Isolation of full-length cDNA of mouse PAX4 gene and
identification of its human homologue. Biochem Biophys Res
Commun 243: 628–633, 1998.
International Diabetes Federation. IDF Diabetes Atlas,
Update: International Diabetes Federation, 2012. Brussels.
Jahnavi S et al. Clinical and molecular characterization of neonatal dia-
betes and monogenic syndromic diabetes in Asian Indian children.
Clin Genet 83 (5): 439–445, 2013.
Julia S et al. From obesity genetics to the future of personalized obesity
therapy. Nat Rev Endocrinol 9: 402–413, 2013.
Kelly T et al. Global burden of obesity in 2005 and projections to 2030.
Int J Obes (Lond) 32: 1431–1437, 2008.
Kissebah AH, et al. Quantitative trait loci on chromosomes 3 and 17 in
Xuence phenotypes of the metabolic syndrome. Proc Natl Acad Sci
97: 14478–14483, 2000.
Kochar IPS et al. Transient neonatal diabetes due to KCNJ11 mutation.
Indian Pediatr 47: 359–360, 2010.
Kooner JS et  al. Genome-wide association study in people of South
Asian ancestry identifies six novel susceptibility loci for type 2 dia-
betes. Nat Genet 43: 984–989, 2011.
Lehto M et al. High frequency of mutations in MODY and mitochon-
drial genes in Scandinavian patients with familial early-onset diabe-
tes. Diabetologia 42: 1131–1137, 1999.
Letha S et al. Permanent neonatal diabetes mellitus due to KCNJ11 gene
mutation. Indian J Pediatr 74: 947–949, 2007.
Lindgren CM et al. Genome-wide association scan meta-analysis identi-
fies three loci influencing adiposity and fat distribution. PLoS Genet
5: e1000508, 2009.
Lindsay RS et al. Genome-wide linkage analysis of serum adiponectin in
the Pima Indian population. Diabetes 52: 2419–2425, 2003.
Liu G et  al. Common variants near melanocortin 4 receptor are asso-
ciated with general and visceral adiposity in European and
African-American youth. J Pediatr 156: 598–605, 2010.
G enomics of T ype 2 D iabetes M ellitus and O besity   •   3 4 9
Liu G et al. FTO variant rs9939609 is associated with body mass index
and waist circumference, but not with energy intake or physi-
cal activity in European and African-American youth. BMC Med
Genet 11: 57, 2010.
Loos RJ. Genetic determinants of common obesity and their value in
prediction. Best Pract Res Clin Endocrinol Metab 26 (2): 211–226,
2012.
Loos RJ. Recent progress in the genetics of common obesity. Br J Clin
Pharmacol 68: 811–829, 2009.
Loos RJF. The genetic determinants of common obesity-susceptibility.
In: Adipose Tissue Biology. Symonds ME, ed. New York, Springer
Science. 317–378, 2012.
Love-Gregory LD et  al. A common polymorphism in the upstream
promoter region of the hepatocyte nuclear factor-4α gene on
chromosome 20q is associated with type 2 diabetes and appears to
contribute to the evidence for linkage in an Ashkenazi Jewish popu-
lation. Diabetes 53: 1134–1140, 2004.
Lyssenko V et al. Genetic prediction of future type 2 diabetes. PLoS Med
2 (12): e345, 2005.
Maestro MA et  al. Distinct roles of HNF1 beta, HNF1 alpha, and
HNF4 alpha in regulating pancreas development, beta-cell function
and growth. Endocrine Dev 12: 33–45, 2007.
Mathias RA et al. Heritability of quantitative traits associated with type
2 diabetes mellitus in large multiplex families from South India.
Metabolism 58 (10): 1439–1445, 2009.
McCarthy MI. Progress in defining the molecular basis of type 2 diabe-
tes mellitus through susceptibility-gene identification. Hum Mol
Genet. 13: Spec No 1: R33–41, 2004.
McCarthy MI et  al. Making the right associations. Diabetologia.
48(7): 1241–1243, 2005.
McCarthy, MI. Diabetes mellitus and obesity. Chapter 12 in: Genetics
and Medicine, 2008.Oxford Uuniversity Press, Oxford.
McCarthy MI. Progress in defining the molecular basis of type 2 diabe-
tes through susceptibility gene identification. Hum Mol Genet 13
(Suppl. 1): R33–R41, 2004. Epub 2004 Jan 13.
Menzaghi C et  al. A haplotype at the adiponectin locus is associated
with obesity and other features of the insulin resistance syndrome.
Diabetes 51: 2306–2312, 2002.
Meur G et  al. Insulin gene mutations resulting in early-onset diabe-
tes:  Marked differences in clinical presentation, metabolic status,
and pathogenic effect through endoplasmic reticulum retention.
Diabetes 59: 653–661, 2010.
Mohan V et al. High prevalence of maturity onset diabetes of the young
(MODY). Diabetes Care 8: 371–374, 1985.
Moore A F et al. Genetic susceptibility to type 2 diabetes and implica-
tions for antidiabetic therapy. Annu Rev Med 59: 95–111, 2008.
Narayan KM et al. Effect of BMI on lifetime risk for diabetes in the U.S.
Diabetes Care 30: 1562–1566, 2007.
Neve B et al. Role of transcription factor KLF11 and its diabetes-associated
gene variants in pancreatic beta-cell function. Proc Natl Acad Sci U
S A 102: 4807–4812, 2005.
Ng MC et al. Implication of genetic variants near TCF7L2, SLC30A8,
HHEX, CDKAL1, CDKN2A/B, IGF2BP2, and FTO in type 2
diabetes and obesity in 6,719 Asians. Diabetes 57: 2226–2233, 2008.
Osbak KK et  al. Update on mutations in glucokinase (GCK), which
cause maturity-onset diabetes of the young, permanent neona-
tal diabetes, and hyperinsulinemic hypoglycemia. Hum Mutat
30: 1512–1526, 2009.
Owen KR et  al. Etiological investigation of diabetes in young adults
presenting with apparent type 2 diabetes. Diabetes Care 26: 2088–
2093, 2003.
Pearson ER et al. Switching from insulin to oral sulfonylureas in patients
with diabetes due to Kir6.2 mutations. N Engl J Med 355: 467–477,
2006.
Poulsen P et al. Heritability of type II (non-insulin-dependent) diabetes
mellitus and abnormal glucose tolerance—a population-based twin
study. Diabetologia 42: 139–145, 1999.
Qi L et al. Variations in adiponectin receptor genes and susceptibility to
type 2 diabetes in women:  A  tagging–single nucleotide polymor-
phism haplotype analysis. Diabetes 56 (6): 1586–1591, 2007.
Radha V et al. Genetic predisposition to type 2 diabetes among Asian
Indians. Indian J Med Res: 125: 259–274, 2007.
Radha V et al. Identification of novel variants in the hepatocyte nuclear
factor 1 alpha gene in South Indian patients with maturity onset dia-
betes of young. J Clin Endocrinol Metab 94 (6): 1959–1965, 2009.
Radha V et al. Role of genetic polymorphism peroxisome proliferator—
Activated Receptor-2 Pro 12Ala on ethnic susceptibility to diabetes
in south Asian and Caucasian subjects. Diabetes Care 29:  1046–
1051, 2006.
Raji A et al. Body fat distribution and insulin resistance in healthy Asian
Indians and Caucasians. JCEM: 86 (11): 5366–5371, 2001.
Ramya K et al. Genetic variations in the FTO gene are associated with
type 2 diabetes and obesity in south Indians. Diabetes Technol Ther
13 (1): 33–42, 2011.
Rubio-Cabezas O et  al. Homozygous mutations in NEUROD1 are
responsible for a novel syndrome of permanent neonatal diabetes
and neurological abnormalities. Diabetes 59: 2326–2331, 2010.
Sanghera DK et al. Type 2 diabetes genetics: Beyond GWAS. J Diabetes
Metab. 3 (198), 2012.
Sanghera DK et al. TCF7L2 polymorphisms are associated with type 2
diabetes in Khatri Sikhs from North India: Genetic variation affects
lipid levels. Ann Hum Genet 72 (4): 499–509, 2008.
Sarzynski MA et  al. Associations of markers in 11 obesity candidate
genes with maximal weight loss and weight regain in the SOS bar-
iatric surgery cases. Int J Obes (Lond.) 35: 676–683, 2011.
Saunders CL et al. Meta-analysis of genome-wide linkage studies in BMI
and obesity. Obesity (Silver Spring) 15: 2263–2275, 2007.
Saxena R et  al. Genome-Wide Association Study Identifies a Novel
Locus Contributing to Type 2 Diabetes Susceptibility in Sikhs
of Punjabi Origin From India. Diabetes Diabetes 62(5):  1746–
1755, 2013.
Scohy S et al. Identification of KLF13 and KLF14 (SP6), novel members
of the SP/XKLF transcription factor family. Genomics 70: 93–101,
2010.
Scott LJ et  al. A genome-wide association study of type 2 diabetes in
Finns detects multiple susceptibility variants. Science 316:  1341–
1345, 2007.
Shields BM et al. Maturity-onset diabetes of the young (MODY): How
many cases are we missing? Diabetologia 53: 2504–2508, 2010.
Shu XO et al. Identification of new genetic risk variants for Type 2 diabe-
tes. PLoS Genet 6: e1001127, 2010.
Silander K et  al. Genetic variation near the hepatocyte nuclear fac-
tor–4 alpha gene predicts susceptibility to type 2 diabetes. Diabetes
53: 1141–1149, 2004.
Sladek R et al. A genome-wide association study identifies novel risk loci
for type 2 diabetes. Nature 445: 881–885, 2007.
Slingerland AS et al. Referral rates for diagnostic testing support an inci-
dence of permanent neonatal diabetes in three European countries
of at least 1 in 260, 000 live births. Diabetologia 52: 1683–1685,
2009.
Speliotes EK et al. Association analyses of 249,796 individuals reveal 18
new loci associated with body mass index. Nat Genet 42: 937–948,
2010.
Steinthorsdottir V et  al. A variant in CDKAL1 influences insulin
response and risk of type 2 diabetes. Nat Genet 39:  770–775,
2007.
Still CD et  al. High allelic burden of four obesity SNPs is associated
with poorer weight loss outcomes following gastric bypass surgery.
Obesity (Silver Spring) 19: 1676–1683, 2011.
Stoffers DA et al. Early-onset type-II diabetes mellitus (MODY4) linked
to IPF1. Nat Genet 17: 138–139, 1997.
Stumvoll M et al. Association of the T–G polymorphism in adiponectin
(exon 2) with obesity and insulin sensitivity: interaction with family
history of type 2 diabetes. Diabetes 51: 37–41, 2002.
3 5 0   •   G enomics in C linical P ractice
Stumvoll M, et al., Type 2 diabetes: principles of pathogenesis and ther-
apy. Lancet 365: 1333–1346, 2005.
Stunkard et  al. An adoption study of human obesity. N Engl J Med
314: 193–198, 1986.
Stunkard et al. The body-mass index of twins who have been reared apart.
N Engl J Med 322: 1483–1487, 1990.
Tabassum R et al. Genome wide association study for type 2 diabetes in
Indians identifies a new susceptibility locus at 2q21. Diabetes 62
(3): 977–986, 2013.
Tattersall RB et  al. A difference between the inheritance of classical
juvenile-onset and maturity-onset type diabetes of young people.
Diabetes 24: 44–53, 1975.
Tattersall RB. Mild familial diabetes with dominant inheritance. QJM
43 (2): 339–357, 1974.
Thomas CP et al. A genetic syndrome of chronic renal failure with mul-
tiple renal cysts and early onset diabetes. Kidney Int 74 (8): 1094–
1099, 2008.
Thorleifsson G et al. Genome-wide association yields new sequence vari-
ants at seven loci that associate with measures of obesity. Nat Genet
41, 18–24, 2009.
Torsvik J et al. Mutations in the VNTR of the carboxyl-ester lipase gene
(CEL) are a rare cause of monogenic diabetes. Hum Genet 127
(1): 55–64, 2010.
Uusitupa M et  al. Impact of Pro12Ala polymorphism of the PPAR-γ
gene on body weight and diabetes incidence in the Finnish diabetes
prevention study. Int J Obesity 25: S5, 2001.
Vasseur F et al. The genetics of adiponectin. Curr Diab Rep 3: 151–158,
2003.
Vaxillaire M et al. Genetic basis of maturity-onset diabetes of the young.
Endocrinol Metab Clin North Am 35 (2): 371–384, 2006.
Velho G et  al. Primary pancreatic beta-cell secretory defect caused by
mutations in glucokinase gene in kindreds of maturity onset diabe-
tes of the young. Lancet 340: 444–448, 1992.
Vimaleswaran KS et  al. A novel association of a polymorphism in the
first intron of adiponectin gene with type 2 diabetes, obesity and
hypoadiponectinemia in Asian Indians. Hum Genet 123 (6): 599–
605, 2008.
Vimaleswaran KS et al. Effect of polymorphisms in the PPARGC1A
gene on body fat in Asian Indians. Int J Obes 30:  884–891,
2006.
Vimaleswaran KS et  al. Peroxisome proliferators-activated receptor—
co-activator-1 (PGC-1) gene polymorphisms and their relationship
to Type 2 diabetes in Asian Indians. Diabetic Med 22: 1516–1521,
2005.
Vionnet N et  al. Genomewide search for type 2 diabetessusceptibility
genes in French Whites: evidence for a novel susceptibility locus for
early-onset diabetes on chromosome 3q27–qter and independent
replication of a type 2-diabetes locus on chromosome 1q21-q24.
Am J Hum Genet 67: 1470–1480, 2000.
Walley A J et al. The genetic contribution to non-syndromic human obe-
sity. Nature Rev Genet 10: 431–442, 2009.
Walters RG et al. A new highly penetrant form of obesity due to dele-
tions on chromosome 16p11.2 Nature 463: 671–675, 2010.
Weedon MN et al. Meta-analysis and a large association study confirm a
role for calpain-10 variation in type 2 diabetes susceptibility. Am J
Hum Genet 73: 1208–1212, 2003.
Wellcome Trust Case Control Consortium. Genome-wide association
study of 14,000 cases of seven common diseases and 3,000 shared
controls. Nature. 447: 661–678, 2007.
Willer CJ et al. Six new loci associated with body mass index highlight
a neuronal influence on body weight regulation. Nat Genet 41: 25–
34, 2009.
World Health Organization: Obesity and overweight http://www.who.
int/mediacentre/factsheets/fs311/en/online, 2012.
Xia Q et al. The genetics of human obesityXia Ann N Y Acad Sci 2013
April; 1281(1):  178–190. Published online 2013 January 29.
doi:10.1111/nyas.12020
Yajnik CS et al. FTO gene variants are strongly associated with type 2
diabetes in South Asian Indians. Diabetologia 52 (2):  247–252,
2009.
Yamagata K et  al. Mutations in the hepatocyte nuclear factor-1 alpha
gene in maturity-onset diabetes of the young (MODY3). Nature
5: 384 (6608): 455–458, 1996.
Yasuda K et al. Variants in KCNQ1 are associated with susceptibility to
type 2 diabetes mellitus. Nat Genet 40: 1092–1097, 2008.
Zeggini E et  al. Meta-analysis of genome-wide association data and
large-scale replication identifies additional susceptibility loci for
type 2 diabetes. Nat Genet 40: 638–645, 2008.
Zeggini E et al. Replication of genome-wide association signals in UK
samples reveals risk loci for type 2 diabetes. Science 316:  1336–
1341, 2007.
Zhang Y et al. Positional cloning of the mouse obese gene and its human
homologue. Nature 372: 425–432, 1994.

G enomics of T ype 2 D iabetes M ellitus and O besity   •   3 5 1

 23.
GENETICS AND GENOMICS OF OSTEOPOROSIS
AND RELATED DISORDER S
Yoshiji Yamada
INTRODUCTION
Osteoporosis is a major health problem of the elderly that
is evenly distributed across the world. It is characterized
by a reduction in bone mineral density (BMD) and dete-
rioration in the microarchitecture of bone, both of which
result in a predisposition to fractures[1]‌. The clinical impor-
tance of osteoporosis lies in its association with fractures[2].
Osteoporotic fractures are common, expensive to treat,
and a substantial cause of morbidity and mortality in most
developed countries. The most common sites for osteopo-
rotic fractures are the hip, spine, and distal forearm. The
incidence of such fractures increases with age and is greater
for women than for men as a result of the increase in the
rate of bone loss after menopause[3]. BMD is one of the
most important determinants of the risk for osteoporotic
fracture[4]. Other factors include susceptibility to falling[5],
bone quality, and skeletal geometry[6].
Bone mass is determined by various mechanisms that
are affected by environmental factors as well as by multiple
genes (Figure 23.1). The environmental factors that con-
tribute to the determination of bone mass include calcium
intake[7]‌, smoking[8,9], alcohol consumption[8], and physical
activity[10]. Despite the importance of these environmental
factors, genetic factors play the predominant role in regu-
lating bone mass and also contribute substantially to other
determinants of osteoporotic fracture risk[2]. Twin studies
have suggested that most of the variance in BMD is geneti-
cally determined, with the actual contribution depending
on the site examined. For example, the heritability of BMD
for the spine and hip has been estimated at between 70%
and 85%, compared to 50% and 60% for the wrist[11–14].
Segregation studies in families also support a strong genetic
contribution to the regulation of bone mass, with a pat-
tern of inheritance that is most consistent with the additive
action of several genes, each with modest effects, rather than
with that of a few genes with large effects[15].
352
Other determinants of the risk for osteoporotic fracture
probably have a genetic component, as reflected by mater-
nal or grandmaternal history of hip fracture independent of
the bone mass[16,17]. This observation could be explained by
genetic effects on hip axis length, ultrasound properties of
bone[11], or bone turnover[18]. The heritability estimates for
quantitative ultrasound properties of bone, hip axis length,
and other aspects of femoral neck geometry range from
60% to 85%[11,18,19]. Although a positive family history is a
risk factor for fracture[20], twin and family studies indicate
that the heritability of fracture itself is only on the order
of 25% to 35%[21,22]. This heritability is much lower than
that of the skeletal phenotypes that predispose to fracture,
probably because of the importance of fall-related factors in
determining the fracture risk[23,24].
Personalized prevention and treatment of osteoporo-
sis and osteoporotic fractures are important public health
goals. One approach is to identify disease-susceptibility
genes. Genetic linkage analyses and candidate-gene asso-
ciation studies have implicated various loci and genes in
the regulation of BMD and the pathogenesis of osteo-
porosis or fractures. Recent genome-wide association
studies (GWASs) have also been successful in identify-
ing genes that predispose to osteoporosis or fractures.
Although there has been extensive progress in the genet-
ics of osteoporosis over the past decades, the genes that
confer susceptibility to osteoporosis remain to be identi-
fied definitively.
This chapter describes the genetic and genomic
aspects of osteoporosis, including an overview of related
monogenic bone diseases, and the chromosomal loci and
genes identified by linkage analyses, the candidate-gene
approach, or GWAS for BMD or predisposition to osteo-
porosis or fractures. Such studies may provide insight into
the function of these genes as well as into the role of genet-
ics in the development of osteoporosis or osteoporotic
fractures.
 Physical risk factors Environmental risk factors

Aging Low calcium intake

Women Low vitamin D intake
Early menopause Low vitamin K intake
Late menarche Physical inactivity
Smoking
Excessive alcohol intake
Genetic risk factors Lack of sunlight
Primary
osteoporosis

Bone mineral density Osteoporosis

Secondary
osteoporosis
Hyperthyroidism Hyperparathyroidism
Cushing’s syndrome Osteomalacia
Glucocorticoid therapy Bone metastasis of cancer
Rheumatoid arthritis Multiple myeloma
Diabetes mellitus

Figure 23.1 
Pathogenesis of osteoporosis. Primary osteoporosis is a bone disorder of relatively unknown origin that occurs with aging and
accelerates after menopause. There is no direct or singular cause for primary osteoporosis. There are two types of primary osteoporosis. Type
I (postmenopausal) osteoporosis develops when estrogen levels decline in postmenopausal women. Type II (senile) osteoporosis occurs in elderly
people when bone formation cannot keep up with the natural process of bone loss. Secondary osteoporosis is caused by the use of corticosteroids
or by various medical conditions, including hormonal imbalances, certain cancers, diabetes mellitus, and rheumatoid arthritis. In addition to
environmental and physical risk factors, genetic factors and interactions between multiple genes and environmental factors are important in the
development of primary osteoporosis.

G E N ET I C S O F O S T E O P O R O S I S vitreous opacities, and fragility fractures. Positional cloning

G E N E S F O R MO N O G E N I C B O N E D I S E A S E
In rare inherited bone diseases, mutations in individual genes
have profound effects on bone mass or the risk of fragility
fractures[24,25]. The classic example of such diseases is osteo-
genesis imperfecta, which is characterized by severe bone
fragility, reduced bone mass, and other connective tissue
manifestations[26]. Most cases of osteogenesis imperfecta are
caused by mutations in collagen, type I genes (COL1A1 or
COL1A2), although clinical features of this condition also
occur in individuals from families with no abnormalities of
collagen genes[27,28]. Recent studies have shown that muta-
tions in the genes for cartilage associated protein (CRTAP),
leucine proline-enriched proteoglycan 1 (LEPRE1), and
peptidylprolyl isomerase B (PPIB), which form a protein
complex necessary for prolyl-3-hydroxylation of collagen,
type I, cause recessive forms of osteogenesis imperfecta[29–31].
Osteoporosis is also a prominent manifestation in patients
with inactivating mutations of the genes for cytochrome
P450, family 19, subfamily A, polypeptide 1 (CYP19A1)
or estrogen receptor 1 (ESR1), as a result of the absence of
estrogen action in bone in these individuals[32,33].
Osteoporosis-pseudoglioma syndrome is a rare autoso-
mal recessive disease characterized by reduced bone mass,
identified homozygous inactivating mutations in the low
density lipoprotein receptor–related protein 5 gene (LRP5)
as the cause of this syndrome[34]. A  monogenic syndrome
characterized by unusually high BMD (high bone mass syn-
drome) was also mapped to the osteoporosis-pseudoglioma
syndrome locus on chromosome 11q12-13 and shown to be
caused by an activating mutation of LRP5[35,36].
Mutations in the region of the transforming growth fac-
tor, beta 1 gene (TGFB1) that encodes the latency-activating
peptide domain have been shown to be responsible for
Camurati-Engelmann disease[37,38], a condition character-
ized by osteosclerosis of the diaphysis of long bones. These
mutations prevent or inhibit binding of latency-associated
peptide to the mature TGFB1 molecule[39] and increase
levels of bioactive TGFB1, which cause the increased bone
turnover that is characteristic of the disease[40].
Mutations in the coding or regulatory regions of the
sclerostin gene (SOST) have been identified as the cause of
the sclerosing bone dysplasias known as sclerosteosis and
van Buchem disease[41, 42].
Osteopetrosis is a group of syndromes characterized
by failure of osteoclastic bone resorption. Osteopetrosis
most often occurs as the result of defects in osteoclast
function (osteoclast-rich osteopetrosis), but it is occa-
sionally caused by defects in osteoclast differentiation

G e n etic s a nd G e no m ic s o f O s t e o p oro s i s a nd R e l at e d D i s ord e r s   •   3 5 3

(osteoclast-poor osteopetrosis)[43]. Many gene mutations
have been identified in osteoclast-rich osteopetrosis, all of
which impair the ability of osteoclasts to resorb bone[43, 44].
Inactivating mutations of the T cell immune regulator 1
gene (TCIRG1), which encodes a subunit of the vacuolar
proton pump of osteoclasts, are responsible for autosomal
recessive osteopetrosis[45], whereas inactivating mutations
in the chloride channel 7 gene (CLCN7), which encodes
a chloride channel expressed in osteoclasts, cause severe
infantile osteopetrosis. Although haploinsufficiency of
CLCN7 does not result in an obvious bone phenotype,
specific heterozygous mutations in the gene give rise to
autosomal dominant osteopetrosis (Albers-Schönberg
disease), presumably by exerting a dominant negative
effect on chloride channel function[46]. A missense muta-
tion of the carbonic anhydrase II gene (CA2) was shown
to be responsible for an autosomal recessive syndrome
of osteopetrosis with renal tubular acidosis and cerebral
calcification[47,48]. Osteoclast-poor osteopetrosis is caused
by inactivating mutations in the genes for tumor necrosis
factor receptor superfamily, member 11a (TNFRSF11A)
or tumor necrosis factor superfamily, member 11
(TNFSF11)[49,50].
Finally, pyknodysostosis, an autosomal recessive osteo-
chondrodysplasia characterized by osteosclerosis and short
stature, has been found to be caused by nonsense or mis-
sense mutations in the cathepsin K gene (CTSK)[51].
Some of these genes for monogenic bone diseases may
also contribute to regulation of BMD in the general pop-
ulation. For example, polymorphisms of COL1A1[52,53],
ESR1[54,55], LRP5[56], and SOST[57] have been shown to be
associated with BMD or the prevalence of osteoporosis or
fractures.
Strategies for Genetic Analysis of Osteoporosis
and Related Phenotypes
There are two basic strategies for identifying genes that influ-
ence BMD or the predisposition to osteoporosis or frac-
tures[58]:  linkage analyses and association studies. Linkage
analysis involves the proposition of a model to account for
the pattern of inheritance of a phenotype observed in a
pedigree. It determines whether the phenotypical locus is
transmitted together with genetic markers of known chro-
mosomal position. Association studies determine whether a
certain allele occurs at a frequency higher than that expected
by chance in individuals with a particular phenotype. Such
an association is thus suggested by a statistically significant
difference in the prevalence of alleles with respect to the
phenotype[58]. Association studies consisted of two strate-
gies:  the candidate-gene approach and the genome-wide
approach (Figure 23.2). The candidate-gene approach
involves the direct examination of whether an individual
gene or genes might contribute to the trait of interest. This
strategy has been widely applied to analysis of the possible
association between genetic variants and disease outcome,
with genes selected on the basis of a priori hypotheses
regarding their potential etiological role. It is characterized
as a hypothesis-testing approach because of the biological
observation supporting the proposed candidate gene. The

Genome-wide approach Candidate gene approach

Linkage analysis Association study Linkage analysis Association study

(microsatellite markers) (SNPs, CNVs): GWAS (microsatellite markers) (polymorphisms)

Candidate locus Candidate gene Association study

(SNPs)

Association study Disease susceptibility

(SNPs, CNVs) genes

Functional analysis

Figure 23.2 
Strategies for identifying disease susceptibility genes. There are two basic strategies for identifying genes that influence common diseases or
other complex traits, the genome-wide approach and the candidate-gene approach, both of which rely on linkage analyses and association studies. In
the genome-wide association study (GWAS), single-nucleotide polymorphisms (SNPs) or copy number variations (CNVs) distributed throughout
the entire genome are used to identify genomic regions that harbor genes that influence the trait of interest with a detectable effect size. The
candidate-gene approach involves the direct examination of whether an individual gene or genes might contribute to the trait of interest.

3 5 4   •   G e no m ic s in C l inic a l Pr actic e
candidate-gene approach is not, however, able to identify
disease-associated polymorphisms in unknown genes. In
the genome-wide scan, single-nucleotide polymorphisms
(SNPs) or copy number variations (CNVs) distributed
throughout the entire genome are used to identify genomic
regions that harbor genes that influence the trait of interest
with a detectable effect size. This is a hypothesis-generating
approach, allowing the detection of previously unknown
potential trait loci.
L I N K AG E A NA LYS I S O F B M D O R
O S T EO P O RO S I S
The published results of whole-genome and partial-genome
linkage analyses for BMD or osteoporosis are summa-
rized in Table 23.1. Initial linkage analyses found that a
locus linked to low BMD mapped to chromosomal region
1p36.3-36.2[59–61], whereas a locus linked to high BMD
mapped to 11q12-13[62]. Genes responsible for other
BMD-related phenotypes, such as autosomal recessive
osteoporosis-pseudoglioma syndrome[63] and autosomal
dominant osteopetrosis[64], are also located at 11q12-13.
Osteoporosis-pseudoglioma syndrome and high bone
mass syndrome are caused by loss-of-function and activat-
ing mutations, respectively, of LRP5 at this locus[34–36], as
described in detail in a later section below.
Additional studies have indicated that BMD is linked
to multiple chromosomal loci and that such loci differ
for BMD at different sites (Table 23.1). Osteoporosis
has also been linked to chromosomal region 20p12.3[65].
Family-based genome-wide scans have revealed that the
loci that affect BMD are largely gender, age, and site
specific[66].
C A N D I DAT E - G E N E A S S O C I AT I O N
S T U D I E S O F B M D, O S T EO P O RO S I S ,
O R FR AC T U R E S
Polymorphisms of a variety of candidate genes have been
associated with BMD or with susceptibility to osteoporosis
or fractures (Table 23.2). It is also possible that polymor-
phisms in these genes are in linkage disequilibrium with
other polymorphisms in nearby genes that are the actual
determinants of BMD or susceptibility to osteoporosis or
fracture. Despite the identification of a variety of candidate
genes related to osteoporosis or BMD, the replicability of
such findings is relatively low, mainly because of the lim-
ited population size of the studies, the ethnic diversity of
gene polymorphisms, gene–gene interactions, complicating
environmental factors, gene–environment interactions, and
G e n etic s a nd G e no m ic s o f O s t e o p oro s i s a nd R e l at e d D i s ord e r s   •   3 5 5
linkage disequilibrium with nearby genes. In the following
sections, two candidate genes (VDR and COL1A1) that
are of particular interest in the genetics of osteoporosis are
reviewed.
Vitamin D Receptor (VDR)
Vitamin D is a potent regulator of bone and calcium homeo-
stasis as well as of cellular differentiation and proliferation
in many tissues. Its active form, 1,25-dihydroxyvitamin D3
(calcitriol), interacts with the highly specific vitamin D
receptor (VDR) and thereby affects the expression of target
genes. A Bsm I restriction fragment length polymorphism
of VDR was associated with BMD in Australian women[67].
Of the many studies performed subsequently, some[68–70]
have supported this association, whereas others[71–73] have
not. These apparently contradictory results are possibly
attributable to differences in several factors among the
studies, including size, age, and ethnic background of the
study populations, as well as environmental aspects such as
diet, especially vitamin D and calcium intake. In addition,
the effects of VDR genotype on BMD appear to be rela-
tively small. A meta-analysis of 16 studies concluded that
BMD was 2.5% lower for the spine and 2.4% lower for the
femoral neck in individuals with the BB genotype (B allele,
absence of the Bsm I restriction site) than in those with the
bb genotype (b allele, presence of the restriction site)[74].
Another meta-analysis of 75 studies, including those exam-
ining four polymorphisms (Bsm I, Apa I, Taq I, and Fok I),
confirmed the association of VDR polymorphisms with
BMD[75]. Furthermore, a more recent meta-analysis of the
relation between VDR Bsm I  genotype and either BMD
or osteoporosis in women revealed that individuals with
the BB genotype had a lower BMD than did those with
the Bb or bb genotypes at baseline, which led to a greater
proportional loss in BMD in those with the BB genotype
over time[76].
Determination of the nucleotide sequence of human
VDR[77] revealed two potential translation initiation sites, the
most 5′ of which is affected by a T→C SNP (ATG→ACG).
Individuals with the T allele of this SNP thus have two start
sites and are able to initiate translation from the first ATG
codon, whereas those with the C allele initiate translation at
the second ATG. The predicted protein produced by initia-
tion at the first ATG is three amino acids longer than is that
generated by initiation at the second start site. The T→C SNP
of VDR has been associated with BMD in postmenopausal
Mexican-American women[78], premenopausal American
white women[79] and Japanese women[80], with the TT geno-
type implicated as a risk factor for reduced BMD. However,
Table 23.1  Whole-Genome and Partial-Genome Linkage Analyses of BMD or Osteoporosis

CHROMOSOMAL LOCUS MARKER PHENOTYPE REFERENCE

1p36.3 D1S2694 BMD (femoral neck) 61
1p36.3-36.2 D1S214 Low BMD (femoral neck) 60
1p36 D1S450 Low BMD (hip) 59
1p36 BMD (whole body)
1q21-23 D1S484 BMD (lumbar spine)
1q22-23 D1S445 BMD (lumbar spine)
2p25 D2S1780 BMD (femoral neck)
2p24-23 D2S149 Low BMD (spine) 59
2p24-21.1 D2S2976, D2S1400, D2S405 BMD (radius)
3p26 D3S1297 BMD (wrist)
3q21 D3S1298, D3S1285 BMD (lumbar spine)
3q25 D3S1279, D3S1565 BMD (lumbar spine) 66
4p D4S2639 BMD (radius)
4q25 D4S1572, D4S406 BMD (femoral neck) 66
4q32 D4S413 BMD (spine, wrist)
4q32-34 D4S1539, D4S1554 Low BMD (hip) 59
5q33-35 D5S422 BMD (femoral neck)
6p21.2 D6S2427 BMD (femoral neck, lumbar spine)
6p12-11 D6S257, D6S462 BMD (lumbar spine)
6q27 D6S446 BMD (trochanter) 61
7p15 D7S493 BMD (lumbar spine) 61
7p14 D7S516, D7S510 BMD (femoral neck) 66
7q22 D7S531 BMD (spine)
8q24.3 D8S373 BMD (Ward’s triangle)
10q21 D10S196, D10S537 BMD (femoral neck) 66
10q26 D10S1651 BMD (hip)
11q12-13 D11S987 High BMD (lumbar spine) 62
11q12-13 D11S1313, D11S935 BMD (lumbar spine, femoral neck)
12q24 D12S1723 BMD (spine)
12q24 D12S2070 BMD (radius)
12q24.2 D12S395 BMD (lumbar spine)
13q14-22 D13S788 BMD (femoral neck, trochanter)
13q21-34 D13S788, D13S800 BMD (radius)
14q D14S592, D14S588 BMD (trochanter)
14q31 D14S587 BMD (lumbar spine)
15q D15S1507 BMD (femoral neck)
16p13 D16S3075, D16S261 BMD (femoral neck) 66
16q23 D16S3091, D16S520 BMD (lumbar spine) 66
17p13 D17S1852 BMD (wrist)
18p11 D18S53, D18S478 BMD (lumbar spine) 66
20p12.3 D20S882, D20S900 Osteoporosis 65
20q13 D20S196, D20S173 BMD (lumbar spine) 66
21q22.2 D21S2055 BMD (trochanter)
21qter D21S1446 BMD (trochanter)

3 5 6   •   G e no m ic s in C l inic a l Pr actic e
Table 23.2  Candidate Gene Association Studies of BMD, Osteoporosis, or Fractures

CANDIDATE GENE CHROMOSOMAL LOCUS PHENOTYPE REFERENCE

Methylenetetrahydrofolate reductase (MTHFR) 1p36.3 BMD
Tumor necrosis factor receptor 2 (TNFRSF1B) 1p36.3-36.2 Low BMD (spine)
Retinoblastoma protein–binding zinc finger protein RIZ (PRDM2) 1p36 BMD
Alkaline phosphatase, liver (ALPL) 1p36.1-34 BMD
Gamma-carboxyglutamic acid protein, bone (BGLAP) 1q25-31 BMD, osteopenia
Glutaminyl-peptide cyclotransferase (QPCT) 2p22.2 BMD
Interleukin 1, beta (IL1B) 2q14 BMD
Interleukin-1 receptor antagonist (IL1RN) 2q14.2 Bone loss (spine)
Osteoporotic fracture
Lactase (LCT) 2q21 BMD
Tumor necrosis factor receptor–associated factor–interacting 2q24-31 BMD
protein (TANK)
Peroxisome proliferator–activated receptor gamma (PPARG) 3p25 BMD
Chemokine receptor 2 (CCR2) 3p21 BMD
Calcium-sensing receptor (CASR) 3q13.3-21 BMD
Microsomal triglyceride transfer protein (MTP) 4q22-24 BMD
Phosphodiesterase 4D, cAMP-specific (PDE4D) 5q12 BMD
PDZ and LIM domain protein 4 (PDLIM4) 5q31.1 BMD
Runt-related transcription factor 2 (RUNX2) 6p21 BMD, osteoporotic
fracture
Estrogen receptor 1 (ESR1) 6q25.1 BMD 54, 55
Interleukin 6 (IL6) 7p21 BMD
Calcitonin receptor (CALCR) 7q21.3 BMD, osteoporotic
fracture
Paraoxonase 1 and 2 (PON1, PON2) 7q21.3 BMD
Collagen, type I, alpha 2 (COL1A2) 7q22.1 BMD
Gonadotropin-releasing hormone 1 (GNRH1) 8p21-11.2 BMD
Tumor necrosis factor receptor superfamily, member 11b 8q24 Vertebral fracture 117
(TNFRSF11B)
BMD 119
Very low density lipoprotein receptor (VLDLR) 9p24 BMD
Cytochrome P450, family 17, subfamily A, polypeptide 1 10q24.3 BMD
(CYP17A1)
Cyclin-dependent kinase inhibitor 1C (CDKN1C) 11p15.5 BMD
Dopamine receptor D4 (DRD4) 11p15.5 BMD
Parathyroid hormone (PTH) 11p15.3-15.1 BMD
Calcitonin (CALCA) 11p15.2-15.1 BMD
Low density lipoprotein receptor–related protein 5 (LRP5) 11q13.4 BMD 56
T cell immune regulator 1 (TCIRG1) 11q13.4-13.5 BMD
Matrix metalloproteinase 1 (MMP1) 11q22-23 BMD
Vitamin D receptor (VDR) 12q12-14 BMD 67, 85
Insulin-like growth factor 1 (IGF1) 12q22-24.1 BMD, osteoporosis
Klotho (KL) 13q12 BMD
Estrogen receptor 2 (ESR2) 14q BMD
Vitamin D–binding protein (GC) 14q12 BMD
(continued)
Table 23.2 CONTINUED
CANDIDATE GENE CHROMOSOMAL LOCUS PHENOTYPE REFERENCE
Bone morphogenetic protein 4 (BMP4) 14q22-23 BMD
Cytochrome P450, family 19, subfamily A, polypeptide 1 15q21.1 BMD, osteoporosis,
(CYP19A1) spinal fracture
Cytochrome P450, family 1, subfamily A, polypeptide 1 (CYP1A1) 15q22-24 BMD
Sclerostin (SOST) 17q12-21 BMD 57
Collagen, type I, alpha 1 (COL1A1) 17q21.31-22 Osteoporosis 52
Osteoporotic fracture 53
BMD 94
Growth hormone (GH1) 17q22-24 BMD
Transforming growth factor, beta 1 (TGFB1) 19q13.1 BMD
Osteoporosis
Vertebral fracture
Apolipoprotein E (APOE) 19q13.2 BMD
Catechol-O-methyltransferase (COMT) 22q11.2 BMD
Androgen receptor (AR) Xq11-12 BMD

no association of this SNP with BMD was detected in pre- Collagen, Type I, Alpha 1 (COL1A1)
menopausal American black[79] or premenopausal French[81]
Collagen, type I  is the most abundant protein of bone
women. The T→C SNP of VDR was found to affect not
matrix. Mutations in the coding regions of the genes for
only the molecular mass of the encoded protein (T allele, 50
the two collagen, type I  chains (COL1A1 and COL1A2)
kDa; C allele, 49.5 kDa), but also the transcriptional activa-
result in a severe autosomal dominant pediatric condition
tion of the gene by vitamin D (T allele < C allele)[80]. This
known as osteogenesis imperfecta[87]. A G→T SNP at the
latter observation was not independently confirmed, how-
first base of a consensus binding site for the transcription
ever[82]. The functional impact of this SNP thus remains to
factor Sp1 in the first intron of COL1A1 was associated not
be determined.
only with BMD in white women[52] but also with osteopo-
A large meta-analysis for 26,242 subjects failed to
rotic fracture in postmenopausal women[53, 88]. Other stud-
demonstrate any relation of Bsm I, Apa I, Taq I, or Fok
ies, however, showed only a weak association of this SNP
I  polymorphism to BMD or fracture[83]. In addition, the
with BMD or osteoporotic fracture in premenopausal
candidate-gene meta-analysis in 19,195 subjects showed
French women[89], or a lack of association in postmeno-
no significant association of VDR genotypes with BMD or
pausal Swedish women[90], in American women[91], or in
fracture[84].
postmenopausal Danish women[92]. The T allele of the Sp1
A –3731A→G SNP that affects the binding site of the
binding site polymorphism affects collagen gene regulation
caudal-related homeodomain protein Cdx-2 in the VDR
in such a manner that it increases the production of the
promoter was associated both with transcriptional activ-
α1(I) collagen chain relative to that of the α2(I) chain and
ity of the promoter and with BMD for the lumbar spine
leads to reduced bone strength by a mechanism that is partly
in Japanese women, with the G allele corresponding to
independent of bone mass[93]. A  –1997G→T SNP in the
reduced transcriptional activity and low BMD[85]. The A
COL1A1 promoter was also associated with BMD for the
allele of this polymorphism exhibited a protective effect
lumbar spine[94]. The –1997G→T SNP and the G→T SNP
on fracture, consistent with the association of the G allele
of the Sp1 binding site were in linkage disequilibrium[94].
with reduced BMD[86]. An association of the –3731A→G
A meta-analysis of the effect of the Spl binding site poly-
polymorphism with BMD and fracture was also detected
morphism of COLIA1 on the prevalence of fractures in
in a large meta-analysis[83]. These various observations
3641 subjects revealed that the risk was 1.7-fold greater in ss
thus suggest that VDR may be a susceptibility gene for
(TT) homozygotes versus SS (GG) homozygotes, 1.4-fold
osteoporosis or fractures, although a VDR locus has not
greater in ss homozygotes versus Ss (GT) heterozygotes, and
attained a significant association with these conditions by
1.3-fold greater in Ss heterozygotes versus SS homozygotes.
GWASs.

3 5 8   •   G e no m ic s in C l inic a l Pr actic e
The effects of this polymorphism were slightly increased
when the analysis was limited to vertebral fractures
(odds ratios, 2.1, 1.5, and 1.3, respectively)[95]. Another
meta-analysis of an association of this polymorphism
with BMD or osteoporotic fracture in 7849 participants
revealed that BMD for the lumbar spine was significantly
lower in individuals with the Ss genotype than in those
with the SS genotype. BMD for the femoral neck was also
lower in individuals with the Ss genotype or those with the
ss genotype than in those with the SS genotype. Analysis of
the prevalence of fractures showed an increased odds ratio
for any fracture in subjects with the Ss genotype (1.3) and
an even greater increase in those with the ss genotype (1.8).
Increased risk was largely attributable to vertebral fracture,
for which the odds ratio was 1.4 for individuals with the
Ss genotype and 2.5 for those with the ss genotype[96]. In
the study of 20,786 subjects, the Sp1 polymorphism of
COL1A1 was associated with BMD for the spine and hip
and vertebral fractures in women[97]. These observations
thus suggest that genetic variants that affect collagen, type
I  metabolism are important determinants of the develop-
ment of osteoporosis and fractures, although COL1A1 has
not attained a significant association with these conditions
by GWASs.
G E N O M E -WI D E A S S O C I AT I O N
S T U D I E S O F B M D, O S T EO P O RO S I S ,
O R FR AC T U R E S
Candidate-gene association studies have substantial limita-
tions for detecting the genetic basis of osteoporosis because
this approach relies on the selection of the genes for associa-
tion studies based on either a biological hypothesis or the
location of a particular gene in implicated linkage regions.
In addition, most candidate-gene association studies have
generated inconsistent or inconclusive results[98]. The
recent development of high-density genotyping arrays has
improved the resolution of unbiased genome-wide scans for
common variants associated with multifactorial diseases.
Currently, the GWAS makes use of high-throughput geno-
typing technologies that include up to 4.3 million markers
for SNPs and CNVs to examine their relation to clinical
conditions or measurable traits. As of January 20, 2012, the
Catalog of Published Genome-Wide Association Studies
(National Human Genome Research Institute, National
Institutes of Health [NIH]; http://www.genome.gov/
gwastudies/) included 1152 publications and 5637 SNPs
associated with various diseases or traits, many in genes
not previously suspected of having a role in the condition
studied, and some in genomic regions containing no known
G e n etic s a nd G e no m ic s o f O s t e o p oro s i s a nd R e l at e d D i s ord e r s   •   3 5 9
genes. GWASs represent a substantial advance in the search
for genetic variants that confer susceptibility to multifacto-
rial polygenic diseases. GWASs, however, have the disad-
vantages that currently available marker sets are designed to
identify common alleles and are not well suited to study-
ing the effects of rare polymorphisms (< 1% to 5% popula-
tion frequency) within a gene of interest[25]. The published
results of GWASs for BMD, osteoporosis, and fractures are
summarized in Table 23.3.
In the following sections, six genes (TNFRSF11B,
TNFSF11, TNFRSF11A, LRP5, ESR1, and SOST) that
were identified by GWASs and are of particular interest in
the genetics of osteoporosis are reviewed.
Tumor Necrosis Factor Receptor Superfamily,
Member 11b (TNFRSF11B)
Tumor necrosis factor (TNF) receptor superfamily, mem-
ber 11b (TNFRSF11B, or osteoprotegerin) is a secreted
glycoprotein that was independently identified by three
groups of research workers[99–101]. In vitro studies suggest
that TNFRSF11B inhibits osteoclastogenesis by inter-
rupting intercellular signaling between osteoblastic stromal
cells and osteoclast progenitors[99]. TNFRSF11B-deficient
mice exhibit a condition similar to juvenile Paget’s disease
that is characterized by a marked decrease in trabecular and
cortical bone density, pronounced thinning of the parietal
bone of the skull, and a high incidence of fractures[102],
whereas hepatic expression of TNFRSF11B in transgenic
mice results in osteopetrosis and a coincident decrease in
the proportion of osteoclasts in the later stages of differ-
entiation[99]. The systemic administration of recombinant
TNFRSF11B also results in a marked increase in BMD in
normal rats as well as in prevention of bone loss in ovari-
ectomized rats[99,103]. Furthermore, a single subcutaneous
injection of TNFRSF11B reduced bone resorption in post-
menopausal women[104]. Similar treatment with a recombi-
nant TNFRSF11B construct suppressed bone resorption in
patients with multiple myeloma or breast cancer with bone
metastases[105].
Osteoclastogenesis is regulated by three TNF– or TNF
receptor–related proteins:  tumor necrosis factor recep-
tor superfamily, member 11a (TNFRSF11A, or recep-
tor activator of nuclear factor–κB [RANK])[106,107], tumor
necrosis factor superfamily, member 11 (TNFSF11, or
RANK ligand [RANKL])[108,109], and TNFRSF11B[99–101].
TNFSF11 expressed on the surface of bone marrow stromal
cells induces the differentiation of osteoclasts, enhances the
activity of mature osteoclasts, and inhibits osteoclast apop-
tosis by binding to its functional receptor, TNFRSF11A,
Table 23.3  Genome-Wide Association Studies of BMD, Osteoporosis, or Fractures

CHROMOSOMAL DBSNP NUCLEOTIDE GENE (NEARBY GENE) PHENOTYPE REFERENCE

LOCUS SUBSTITUTION
1p36 rs7524102 G→A, ZBTB40 BMD, fracture 122
rs6696981 G→T
rs6426749 C→G BMD 124
1p33-p32 rs17131547 A→G TGFBR3 BMD 156
1p31.3 rs1430742 A→G, WLS BMD 124
rs2566755 A→G
1p13.2 rs494453 G→A RAP1A BMD 123
2p21 rs11898505 G→A SPTBN1 BMD 124
2q11.2 rs2278729 G→A TBC1D8 BMD 123
2q24 rs1863196 A→G, GALNT3 BMD 157
rs6710518 C→T
2q33.1 rs7605378 A→C FONG Osteoporosis 158
3p22 rs87938 G→A CTNNB1 BMD 124
4q21.1 rs1471403 C→T MEPE BMD 124
4q22 rs1054627 G→A IBSP BMD 157
5q14 rs1366594 G→T MEF2C BMD 124
5q31 rs13182402 A→G ALDH7A1 BMD, fracture 159
6p22 rs9466056 G→A SOX4 BMD 157
rs17563605 T→C, RSPO3 BMD 157
rs13204965 A→C
6p21 rs3130340 C→T MHC BMD, fracture 122
6q25 rs9479055 C→A, C6orf97 BMD 122
rs4870044 T→C,
rs1038304 G→A,
rs6929137 G→A,
rs1999805 C→T ESR1 BMD 122
rs2504063 A→G, 124
rs2941740 C→T
7p14 rs1721400 C→T SFRP4 BMD 160
rs1524058 C→T STARD3NL BMD 124
7q21.3 rs4729260 C→G, FLJ42280 BMD 124
rs7781370 C→T
7q31 rs7776725 T→C FAM3C BMD 160
8q24 rs4355801 A→G TNFRSF11B BMD, osteoporosis 121
rs6993813 T→C, BMD 122
rs6469804 A→G
rs2062377 T→A, 124
rs11995824 G→C
rs2062375 C→G 123
11p15.3 rs297325 T→C, SOX6 BMD 161
rs4756846 C→T
rs7117858 A→G 124
11p15 rs9630182 T→C, PTH BMD 162
rs2036417 A→G,
rs7125774 C→T
11p14.1 rs16921914 G→A DCDC5 BMD 124
11p13 rs1152620 A→G LTBP3 BMD 157
11p11.2 rs7932354 T→C ZNF408 BMD 124
(continued)

3 6 0   •   G e no m ic s in C l inic a l Pr actic e
Table 23.3 CONTINUED
CHROMOSOMAL DBSNP NUCLEOTIDE GENE (NEARBY GENE) PHENOTYPE REFERENCE
LOCUS SUBSTITUTION
11q13.4 rs3736228 C→T (Ala1330Val) LRP5 BMD, osteoporo- 121
sis, fracture
rs599083 T→G BMD 124
12q13 rs10876432 A→G SP7 BMD 130
rs2016266 G→A 124
13q14 rs9594738 C→T, TNFSF11 BMD 122
rs9594759 C→T
rs1021188 T→C 127
rs9533090 C→T AKAP11 BMD 124
14q32 rs2010281 A→G MARK3 BMD, fracture 130
14q32.12 rs1298989 T→C, CATSPERB BMD 163
rs1285635 A→G
16p13 rs13336428 A→G CLCN7 BMD 157
16p11 rs8057551 A→G, IL21R BMD 162
rs8061992 C→A,
rs7199138 C→G
16q23 rs16945612 T→C, ADAMTS18 BMD 156
rs11859065 G→A,
rs11864477 T→C,
rs11860781 A→T
16q24 rs10048146 A→G FOXL1 BMD 124
17q12 rs9303521 G→T CRHR1 BMD 124
17q21 rs1513670 G→A, SOST BMD, fracture 130
rs7220711 A→G,
rs1107748 C→T
rs228769 C→G HDAC5 BMD 124
18q11.2 rs7227401 G→T OSBPL1A BMD 123
18q21 rs3018362 G→A TNFRSF11A BMD, fracture 130, 122
rs884205 G→T BMD 124
20p12.1-p11.23 rs2273061 A→G JAG1 BMD 164

expressed on osteoclasts or their progenitors[108–111]. The
interaction between TNFSF11 and TNFRSF11A is antag-
onized by TNFRSF11B, which acts as a decoy receptor for
TNFSF11. The biological effects of TNFRSF11B include
inhibition of the later stages of osteoclastogenesis[99,103,112],
suppression of the activation of mature osteoclasts[108,110],
and induction of osteoclast apoptosis[113]. The balance
between TNFRSF11B and TNFSF11 may thus represent
an important determinant of bone resorption[112,114]. The
importance of TNFRSF11B in the regulation of bone
remodeling in humans has been indicated by the occurrence
of juvenile Paget’s disease, characterized by rapid remodel-
ing of woven bone, osteopenia, fractures, and progressive
skeletal deformity, in Navajo individuals homozygous for a
deletion of ~100 kb in TNFRSF11B[115].
Several SNPs have been detected in TNFRSF11B, some
of which have been shown to be associated with BMD or
fractures. Both 209G→A and 245T→G SNPs of the
TNFRSF11B promoter were found to be associated with
BMD for the lumbar spine in postmenopausal Slovenian
women, with the 209GA/245TG genotype representing
a risk factor for reduced BMD[116]. Both 163A→G and
245T→G SNPs were associated with vertebral fractures
in Danish women and men, with the G allele of each poly-
morphism representing a risk factor for fracture[117]. The
245T→G SNP was associated with BMD at various sites
in postmenopausal Japanese women, with the GG genotype
representing a risk factor for reduced BMD[118]. A 950T→C
SNP was also associated with BMD in postmenopausal
Japanese women[119]. In contrast, SNPs of TNFRSF11B

G e n etic s a nd G e no m ic s o f O s t e o p oro s i s a nd R e l at e d D i s ord e r s   •   3 6 1

were not related to BMD or the prevalence of fractures in
elderly Swedish women[120].
Several SNPs at the TNFRSF11B locus were significantly
associated with BMD in the GWASs[121–123]. Furthermore,
TNFRSF11B SNPs were associated with an increased risk
of fractures[122]. The association of TNFRSF11B with BMD
was confirmed by the meta-analysis of GWASs[124].
The functional mechanisms by which TNFRSF11B
alleles predispose to osteoporosis have not been fully deter-
mined. The expression of the risk allele at the rs4355801 SNP
was associated with reduced expression of TNFRSF11B in
lymphoblastoid cell lines[121]. This observation suggests that
the variants of TNFRSF11B associated with osteoporosis
result in reduced gene expression, thereby increasing bone
resorption and bone loss[25].
Several lines of evidence thus indicate that TNFRSF11B
plays an important role in bone remodeling. The fact that
SNPs in TNFRSF11B have been associated with BMD or
fractures in different populations and ethnic groups further
suggests that the encoded protein is a prominent regulator
of bone mass and a determinant of the predisposition to
osteoporosis or fractures.
Tumor Necrosis Factor Superfamily, Member 11
(TNFSF11)
A previous study showed that an intronic SNP
(rs2277438) of TNFSF11 interacted with the 1181G→C
polymorphism of TNFRSF11B to affect BMD in Korean
postmenopausal women, although rs2277438 alone was
not related to BMD[125]. A  significant association was
observed between rs9594782 SNP in the TNFSF11
promoter and low BMD, whereas the intronic SNP
(rs2277438) was not related to BMD in Chinese men
and women[126].
TNFSF11 was confirmed to be a susceptibility gene
for osteoporosis by the GWASs, which showed that sev-
eral SNPs in TNFSF11 were associated with BMD[122,127].
The functional mechanisms by which polymorphisms of
TNFSF11 regulate BMD remain unclear[25].
Tumor Necrosis Factor Receptor Superfamily,
Member 11a (TNFRSF11A)
A significant association was observed between a C→T
(Ala192Val) polymorphism of TNFRSF11A and BMD in
Korean postmenopausal women[128]. This polymorphism
was associated with hip BMD in Chinese men, but not
women[126]. Two intronic SNPs were also related to BMD
in Korean postmenopausal women[129].
3 6 2   •   G e no m ic s in C l inic a l Pr actic e
The GWASs showed an association of TNFRSF11A
polymorphisms with BMD[130] and fracture[122]. This asso-
ciation has been confirmed by the GWAS meta-analysis[124].
The functional relevance of polymorphisms of the
TNFRSF11A locus with osteoporosis or fracture remains
to be elucidated[25].
Low-Density Lipoprotein Receptor-Related Protein
5 (LRP5)
LRP5 is expressed in osteoblasts and regulates the prolif-
eration, survival, and activity of these cells. Population
genetics and in vitro studies have shown that LRP5 is
a key regulator of bone metabolism and acts via the
Wnt signaling pathway[131,132]. Mutations in LRP5 cause
osteoporosis-pseudoglioma syndrome[34], an autosomal
recessive condition of juvenile onset that is characterized
both by blindness, resulting from aberrant vitreo-retinal
vascular growth, and by osteoporosis, resulting in frac-
tures and deformation. Analysis of bone biopsy specimens
from affected individuals has revealed that the volume
of trabecular bone is reduced but that the number and
appearance of osteoblasts and osteoclasts are normal. Six
disease-causing frameshift or nonsense mutations in the
regions of LRP5 encoding the signal peptide, the epidermal
growth factor–like domain, the YWTD domains, and the
low density lipoprotein receptor–like domains have been
identified, suggesting that osteoporosis-pseudoglioma syn-
drome results from loss of LRP5 function. The consequent
down-regulation of the Wnt signaling pathway may impair
bone formation and thereby lead to osteoporosis.
In contrast to osteoporosis-pseudoglioma syndrome,
activating mutations in LRP5 cause an autosomal dominant
syndrome characterized by increased bone mass. Affected
individuals show enhanced bone synthesis as assessed by
serum markers, whereas bone resorption and bone archi-
tecture appear normal. Affected members of families that
manifest this condition, but not unaffected relatives or
unrelated controls, harbor a G→T transversion in exon 3
of LRP5[35,36]. This mutation results in the substitution of
a conserved glycine residue at position 171 by valine. The
mutant LRP5 protein constitutively activates the Wnt sig-
naling pathway and renders it resistant to Dkk1-mediated
inhibition. The activation of LRP5 function thus results in
excessive bone accumulation.
LRP5 mutations have also been detected in individu-
als with a heterogeneous group of conditions characterized
by autosomal recessive[133] or autosomal dominant[64] forms
of osteopetrosis. The increased BMD apparent in affected
individuals results from disruption of bone turnover and
remodeling. Some forms of osteopetrosis, especially those
attributable to osteoclast dysfunction, are associated with
increased fracture risk. However, this is not seen in the high
bone mass syndrome. Four mutations of LRP5 (331G→T
[Asp111Tyr], 511G→C [Gly171Arg], 724G→A
[Ala242Thr], and 758C→T [Thr253Ile]) have been identi-
fied in five families affected with osteopetrosis[134].
LRP5 polymorphisms have also been associated with
endosteal hyperostosis and its variant, van Buchem dis-
ease. Like osteopetrosis, these conditions are marked by
excessive accrual of bone, although, in this instance, such
accrual is limited to the inner (endosteal) surface, leading to
obliteration of the medullary space. Two coding polymor-
phisms of LRP5 (640G→A [Ala214Thr] and 724G→A
[Ala242Thr]) have been identified in individuals with these
conditions[134].
Several studies have implicated common polymor-
phisms of LRP5 in BMD variation in the general popula-
tion. A 2047G→A (Val667Met) SNP in exon 9 was thus
significantly associated with bone mineral content of the
lumbar spine, with bone area, and with stature in a cohort
of 889 healthy white men and women, with the association
being most pronounced in adult men[56]. Haplotype analy-
sis of five SNPs at the LRP5 locus suggested that additional
genetic variation at this locus might also contribute to
determination of bone mass and size. SNPs of LRP5 were
associated with BMD for the lumbar spine, total hip, or
femoral neck in a cohort of 909 British Caucasian adults[135].
Family studies identified one SNP (171346C→A in intron
21) with a significant relation to BMD, with the association
being stronger in men than in women. The association of
LRP5 polymorphisms with BMD was also observed as the
over-representation of variants of three SNPs in osteopo-
rotic probands compared with unrelated postmenopausal
women with increased BMD. In addition, an association
between haplotypes of SNPs and BMD was detected by
comparison of the osteoporotic subjects with the indi-
viduals with a high BMD. Association of SNPs in LRP5
with BMD has also been observed in Japanese[136] and
Australian[137] women. Although a significant association
was observed between LRP5 SNPs and BMD for the hip
or spine in a study of healthy premenopausal white women,
only a small proportion of the total variation in BMD was
accounted for by these SNPs[138]. The genotyped SNPs thus
accounted for ~0.8% of the variation in BMD for the femo-
ral neck and 1.1% of that for the lumbar spine, suggesting
that natural variation in and around LRP5 is not a major
contributing factor to the observed variability in peak
BMD at either of these sites in white women. In a study of
37,534 subjects, nonsynonymous variants (Val667Met and
G e n etic s a nd G e no m ic s o f O s t e o p oro s i s a nd R e l at e d D i s ord e r s   •   3 6 3
Ala1330Val) of LRP5 were significantly associated with
BMD and fracture[139]. The LRP5 locus was also identified
as a significant determinant of BMD in the GWAS[121] and
in the GWAS meta-analysis[124].
The most likely functional candidates among LRP5
variants are a Val667Met (rs4988321) and an Ala1330Val
(rs3736228). Promoter reporter assays showed that differ-
ent haplotypes for the Val667Met and Ala1330Val poly-
morphisms differently activated reporter-gene transcription,
suggesting that they may be functional[140]. The various stud-
ies showed that rare mutations in LRP5 have major effects
on BMD, whereas common polymorphisms have smaller
effects on BMD and fracture risk in the general popula-
tion[25]. LRP5 may thus be an important regulator of bone
mass and determinant of predisposition to osteoporosis.
Estrogen Receptor 1 (ESR1)
The importance of estrogen receptor 1 in the regulation of
bone mass was indicated by the occurrence of osteoporo-
sis in a man with a nonsense mutation in ESR1[33] as well
as by the observation that BMD in mice lacking a func-
tional ESR1 allele is 20 to 25% less than that in wild-type
mice[141]. Two SNPs have been identified in the first intron
of ESR1: a T→C polymorphism that is recognized by the
restriction endonuclease Pvu II (T and C alleles correspond
to the presence [p allele] and absence [P allele] of the restric-
tion site, respectively), and an A→G polymorphism that
is recognized by Xba I (A and G alleles correspond to the
presence [x allele] and absence [X allele] of the restriction
site, respectively). These SNPs, alone or in combination,
have been associated with BMD in postmenopausal[54,55] or
premenopausal[142,143] women or with the response to hor-
mone replacement therapy[144]. However, other studies did
not confirm these observations[145–150]. Although no asso-
ciation was observed between BMD in women and either
of these two ESR1 polymorphisms alone, an association
was detected between BMD for the lumbar spine or femo-
ral neck and the haplotype of the SNPs[55]. In addition, a
microsatellite (TA repeat) polymorphism in the promoter
region of ESR1 was associated with BMD and with the
prevalence of fractures in studies in which such an associa-
tion with the T→C or A→G SNPs in the first intron was
not apparent[55,145,149]. No association was detected between
estrogen responsiveness of BMD and ESR1 genotype in
postmenopausal Korean women who had undergone hor-
mone replacement therapy[148]. In contrast, women with
the TT genotype (Pvu II SNP) have been suggested to be
relatively estrogen insensitive; those with the C allele thus
appeared to benefit more from the protective effect of
hormone replacement therapy with regard to fracture risk
than did women with the TT genotype[144].
A meta-analysis of the relations of the A→G (Xba I)
and T→C (Pvu II) SNPs of ESR1 to BMD and fracture
risk in 5834 women from 30 study groups showed that
homozygotes for the G allele of the A→G SNP had a higher
BMD and a lower risk of fractures compared with carriers
of the A allele, whereas the T→C SNP was not associated
with either BMD or fracture risk[151]. A large meta-analysis
of the relations of three polymorphisms (A→G and T→C
SNPs in intron 1 and the TA repeat polymorphism in the
promoter) of ESR1 and their haplotypes to BMD and
fractures in 18,917 individuals in eight European centers
showed that none of the polymorphisms or haplotypes had
any significant effect on BMD. However, in women homo-
zygous for the G allele of the A→G SNP, the risk for all
fractures was reduced by 19% (odds ratio, 0.81) and that for
vertebral fractures by 35% (odds ratio, 0.65). No significant
effects on fracture risk were apparent for the T→C SNP or
TA repeat polymorphism[152].
The association of ESR1 polymorphisms with
BMD was confirmed by the GWAS[122] and the GWAS
meta-analysis[124]. Functional studies suggested that the
intronic polymorphisms may affect gene transcription. The
Pvu II polymorphism locates within consensus recogni-
tion sites for the AP4 and Myb transcription factors[55,153],
and promoter reporter assays showed that the Pvu II poly-
morphism influences Myb-driven transcription in vitro[153].
These observations suggest that ESR1 is a susceptibility
gene for osteoporosis, although the effect of ESR1 poly-
morphisms on BMD or fracture risk may be small. The
molecular mechanisms that underlie the association of
ESR1 polymorphisms with BMD or with susceptibility to
osteoporosis or fractures remain unclear[25].
Sclerostin (SOST)
SOST is produced almost exclusively by osteocytes and
inhibits bone formation, by preventing members of the Wnt
family from binding to the LRP5 receptor[154]. Inactivating
mutations of SOST cause syndromes of sclerosteosis and
van Buchem disease, indicating that SOST is a genetic deter-
minant of BMD. A previous study showed an association
of SOST polymorphisms with BMD in men and women,
with effects that increased with age[57]. The SOST locus
was associated with BMD and fracture in a candidate-gene
meta-analysis[84]. Three SNPs near SOST were also signifi-
cantly associated with total hip BMD in the GWAS[130].
The SOST locus, however, did not reach a genome-wide
significance level in the GWAS meta-analysis[124]. The
3 6 4   •   G e no m ic s in C l inic a l Pr actic e
various studies suggest that polymorphisms at the SOST
locus may contribute to the regulation of BMD, although
the mechanisms responsible for the association remain to
be determined[25].
C L I N I C AL I MPL I C AT I O N S
The increasing body of information garnered from studies
on the genetics of osteoporosis has resulted in the emer-
gence of a greater understanding of the biology of bone in
health and disease. Such knowledge has clinical implica-
tions for the prediction, diagnosis, prognosis, and treatment
of osteoporosis. The genes responsible for the regulation of
BMD and bone fragility as well as their encoded proteins
are potentially important therapeutic targets in the design
of new treatments for bone disease. Genetic markers are
potential diagnostic tools for the assessment of individuals
at risk of developing osteoporotic fractures. Genetic mark-
ers of bone fragility or bone loss, together with measure-
ments of BMD, might also form the basis for promotion
of preventive therapies in individuals at risk of fracture.
It should be remembered, however, that gene–gene and
gene–environment interactions make the interpretation
of information based on genetic markers of osteoporosis
more complex than is that of information based on mark-
ers for monogenic bone diseases. Another use of genetic
markers might be to distinguish treatment responders from
non-responders and to identify patients who are at risk
of developing unfavorable side effects. It is likely that the
use of gene polymorphisms to predict the response to, and
adverse effects of, therapies for osteoporosis will increase
in the future and will give rise to major advances in patient
care[155]. Genetic analysis of osteoporosis or fractures is thus
likely to have important direct clinical applications.
S U MMA RY
This chapter has summarized monogenic bone diseases and
the candidate loci and genetic variants related to bone mass
or to predisposition to osteoporosis or fractures. There has
been a growing effort to find genetic variants that confer risk
for osteoporosis and fractures as a means to understand the
underlying biological events. Clarification of the functional
relevance of SNPs at various loci to osteoporosis and fractures
may provide insight into the pathogenesis of these conditions
as well as into the role of genetic factors in their development.
Such studies may ultimately lead to the personalized preven-
tion of osteoporosis and fractures. It may thus become possible
to predict the future risk for osteoporosis in each individual
on the basis of measurement of BMD and genetic analyses.
Furthermore, it may be possible to prevent an individual from
having osteoporotic fractures by medical intervention based
on his or her genotypes for specific SNPs. In the future, we may
have the ability to use specific agents particularized for certain
genetic susceptibility factors, thereby increasing their efficacy
and limiting any side effects of treatment. Identification of sus-
ceptibility genes for osteoporosis and clarification of the func-
tional relevance of genetic variants to this condition will thus
contribute to the personalized prevention, early diagnosis, and
treatment of osteoporosis.
R EFE R E N C ES
1. Kanis JA et  al.:  The diagnosis of osteoporosis. J Bone Miner Res
9:1137–1141, 1994.
2. Ralston SH:  Genetics of osteoporosis. Rev Endocr Metab Disord
2:13–21, 2001.
3. Melton LJ III: How many women have osteoporosis now? J Bone
Miner Res 10:175–177, 1995.
4. Cummings SR et al.: Bone density at various sites for prediction of
hip fractures. The Study of Osteoporotic Fractures Research Group.
Lancet 341:72–75, 1993.
5. Dargent-Molina P et  al.:  Fall-related factors and risk of hip frac-
ture: The EPIDOS prospective study. Lancet 348:145–149, 1996.
6. Faulkner KG et al.: Simple measurement of femoral geometry pre-
dicts hip fracture: the study of osteoporotic fractures. J Bone Miner
Res 8:1211–1217, 1993.
7. Holbrook TL et al.: Dietary calcium and risk of hip fracture: 14 year
prospective population study. Lancet ii:1046–1049, 1988.
8. Hannan MT et al.: Risk factors for longitudinal bone loss in elderly
men and women:  The Framingham Osteoporosis Study. J Bone
Miner Res 15:710–720, 2000.
9. Krall EA et al.: Smoking increases bone loss and decreases intestinal
calcium absorption. J Bone Miner Res 14:215–220, 1999.
10. Cooper C et  al.:  Physical activity, muscle strength, and calcium
intake in fracture of the proximal femur in Britain. Br Med J
297:1443–1446, 1988.
11. Arden NK et  al.:  The heritability of bone mineral density, ultra-
sound of the calcaneus and hip axis length:  a study of postmeno-
pausal twins. J Bone Miner Res 11:530–534, 1996.
12. Flicker L et al.: Bone density in elderly women: a twin study. J Bone
Miner Res 10:1607–1613, 1995.
13. Howard GM et al.: Genetic and environmental contributions to the
association between quantitative ultrasound and bone mineral density
measurements: a twin study. J Bone Miner Res 13:1318–1327, 1998.
14. Pocock NA et al.: Genetic determinants of bone mass in adults: a
twin study. J Clin Invest 80:706–710, 1987.
15. Gueguen R et  al.:  Segregation analysis and variance components
analysis of bone mineral density in healthy families. J Bone Miner
Res 10:2017–2022, 1995.
16. Cummings SR et al.: Risk factors for hip fracture in white women.
Study of Osteoporotic Fractures Research Group. N Engl J Med
332:767–773, 1995.
17. Stewart A et al.: Prediction of fractures in perimenopausal women: a
comparison of dual energy X-ray absorptiometry and broadband
ultrasound attenuation. Ann Rheum Dis 55:140–142, 1996.
18. Garnero P et al.: Genetic influence on bone turnover in postmeno-
pausal twins. J Clin Endocrinol Metab 81:140–146, 1996.
G e n etic s a nd G e no m ic s o f O s t e o p oro s i s a nd R e l at e d D i s ord e r s   •   3 6 5
19. Slemenda CW et  al.:  The genetics of proximal femur geometry,
distribution of bone mass and bone mineral density. Osteoporos Int
6:178–182, 1996.
20. Keen RW et al.: Family history of appendicular fracture and risk of
osteoporosis: a population-based study. Osteoporos Int 10:161–166,
1999.
21. Deng HW et al.: Genetic determination of Colles’ fracture and dif-
ferential bone mass in women with and without Colles’ fracture.
J Bone Miner Res 15:1243–1252, 2000.
22. MacGregor AJ et al.: Genetic factors and osteoporotic fractures in
elderly people. Br Med J 320:1669–1670, 2000.
23. Kannus P et al.: Genetic factors and osteoporotic fractures in elderly
people:  prospective 25  year follow up of a nationwide cohort of
elderly Finnish twins. Br Med J 319:1334–1337, 1999.
24. Ralston SH: Genetic control of susceptibility to osteoporosis. J Clin
Endocrinol Metab 87:2460–2466, 2002.
25. Ralston SH et  al.:  Genetics of osteoporosis. Endocr Rev 31:629–
662, 2010.
26. Rauch F et  al.:  Osteogenesis imperfecta. Lancet 363:1377–1385,
2004.
27. Glorieux FH et  al.:  Osteogenesis imperfecta type VI:  a form of
brittle bone disease with a mineralization defect. J Bone Miner Res
17:30–38, 2002.
28. Marini JC et al.: Consortium for osteogenesis imperfecta mutations
in the helical domain of type I collagen: regions rich in lethal muta-
tions align with collagen binding sites for integrins and proteogly-
cans. Hum Mutat 28:209–221, 2007.
29. Barnes AM et al.: Deficiency of cartilage-associated protein in reces-
sive lethal osteogenesis imperfecta. N Engl J Med 355:2757–2764,
2006.
30. Cabral WA et al.: Prolyl 3-hydroxylase 1 deficiency causes a reces-
sive metabolic bone disorder resembling lethal/severe osteogenesis
imperfecta. Nat Genet 39:359–365, 2007.
31. Barnes AM et al.: Lack of cyclophilin B in osteogenesis imperfecta
with normal collagen folding. N Engl J Med 362:521–528, 2010.
32. Morishima A et al.: Aromatase deficiency in male and female sib-
lings caused by a novel mutation and the physiological role of estro-
gens. J Clin Endocrinol Metab 80:3689–3698, 1995.
33. Smith EP et  al.:  Estrogen resistance caused by a mutation in the
estrogen-receptor gene in a man. N Engl J Med 331:1056–1061,
1994.
34. Gong Y et al.: LDL receptor-related protein 5 (LRP5) affects bone
accrual and eye development. Cell 107:513–523, 2001.
35. Boyden LM et  al.:  High bone density due to a mutation in
LDL-receptor-related protein 5. N Engl J Med 346:1513–1521, 2002.
36. Little RD et al.: A mutation in the LDL receptor-related protein 5
gene results in the autosomal dominant high-bone-mass trait. Am J
Hum Genet 70:11–19, 2002.
37. Janssens K et  al.:  Mutations in the latency-associated peptide of
TGF-β1 cause Camurati-Engelmann disease. Nat Genet 26:273–
275, 2000.
38. Kinoshita A et al.: Domain-specific mutations in TGFB1 result in
Camurati-Engelmann disease. Nat Genet 26:19–20, 2000.
39. Janssens K et  al.:  Transforming growth factor-β1 mutations in
Camurati-Engelmann disease lead to increased signaling by alter-
ing either activation or secretion of the mutant protein. J Biol Chem
278:7718–7724, 2003.
40. McGowan NW et al.: A mutation affecting the latency-associated
peptide of TGFβ1 in Camurati-Engelmann disease enhances osteo-
clast formation in vitro. J Clin Endocrinol Metab 88:3321–3326,
2003.
41. Balemans W et al.: Increased bone density in sclerosteosis is due to
the deficiency of a novel secreted protein (SOST). Hum Mol Genet
10:537–543, 2001.
42. Balemans W et al.: Identification of a 52 kb deletion downstream of
the SOST gene in patients with van Buchem disease. J Med Genet
39:91–97, 2002.
 43. Villa A et al.: Infantile malignant, autosomal recessive osteopetro-
sis: the rich and the poor. Calcif Tissue Int 84:1–12, 2009.
44. Balemans W et al.: A clinical and molecular overview of the human
osteopetroses. Calcif Tissue Int 77:263–274, 2005.
45. Frattini A et al.: Defects in TCIRG1 subunit of the vacuolar pro-
ton pump are responsible for a subset of human autosomal recessive
osteopetrosis. Nat Genet 25:343–346, 2000.
46. Cleiren E et  al.:  Albers-Schonberg disease (autosomal dominant
osteopetrosis, type II) results from mutations in the ClCN7 chlo-
ride channel gene. Hum Mol Genet 10:2861–2867, 2001.
47. Roth DE et  al.:  Molecular basis of human carbonic anhydrase II
deficiency. Proc Natl Acad Sci U S A 89:1804–1808, 1992.
48. Sly WS et al.: Carbonic anhydrase II deficiency identified as the pri-
mary defect in the autosomal recessive syndrome of osteopetrosis
with renal tubular acidosis and cerebral calcification. Proc Natl Acad
Sci U S A 80:2752–2756, 1983.
49. Guerrini MM et  al.:  Human osteoclast-poor osteopetrosis with
hypogammaglobulinemia due to TNFRSF11A (RANK) muta-
tions. Am J Hum Genet 83:64–76, 2008.
50. Sobacchi C et  al.:  Osteoclast-poor human osteopetrosis due to
mutations in the gene encoding RANKL. Nat Genet 39:960–962,
2007.
51. Gelb BD et  al.:  Pycnodysostosis, a lysosomal disease caused by
cathepsin K deficiency. Science 273:1236–1238, 1996.
52. Grant SFA et al.: Reduced bone density and osteoporosis associated
with a polymorphic Sp1 binding site in the collagen type Iα 1 gene.
Nat Genet 14:203–205, 1996.
53. Uitterlinden AG et al.: Relation of alleles of the collagen type Iα 1
gene to bone density and the risk of osteoporotic fractures in post-
menopausal women. N Engl J Med 338:1016–1021, 1998.
54. Kobayashi S et al.: Association of bone mineral density with poly-
morphism of the estrogen receptor gene. J Bone Miner Res 11:306–
311, 1996.
55. Albagha OME et al.: Estrogen receptor α gene polymorphisms and
bone mineral density: haplotype analysis in women from the United
Kingdom. J Bone Miner Res 16:128–134, 2001.
56. Ferrari SL et  al.:  Polymorphisms in the low-density lipoprotein
receptor-related protein 5 (LRP5) gene are associated with variation
in vertebral bone mass, vertebral bone size, and stature in whites. Am
J Hum Genet 74:866–875, 2004.
57. Uitterlinden AG et  al.:  Polymorphisms in the sclerosteosis/
van Buchem disease gene (SOST) region are associated with
bone-mineral density in elderly whites. Am J Hum Genet 75:1032–
1045, 2004.
58. Nguyen TV et al.: Genetic epidemiological approaches to the search
for osteoporosis genes. J Bone Miner Res 15:392–401, 2000.
59. Devoto M et al.: First-stage autosomal genome screen in extended
pedigrees suggests genes predisposing to low bone mineral density
on chromosomes 1p, 2p and 4q. Eur J Hum Genet 6:151–157, 1998.
60. Devoto M et  al.:  Variance component linkage analysis indicates a
QTL for femoral neck bone mineral density on chromosome 1p36.
Hum Mol Genet 10:2447–2452, 2001.
61. Devoto M et al.: Univariate and bivariate variance component link-
age analysis of a whole-genome scan for loci contributing to bone
mineral density. Eur J Hum Genet 13:781–788, 2005.
62. Johnson ML et  al.:  Linkage of a gene causing high bone mass to
human chromosome 11 (11q12–13). Am J Hum Genet 60:1326–
1332, 1997.
63. Gong Y et  al.:  Osteoporosis-pseudoglioma syndrome, a disorder
affecting skeletal strength and vision, is assigned to chromosome
region 11q12–13. Am J Hum Genet 59:146–151, 1996.
64. Van Hul E et al.: Localization of the gene causing autosomal domi-
nant osteopetrosis type I to chromosome 11q12–13. J Bone Miner
Res 17:1111–1117, 2002.
65. Styrkarsdottir U et  al.:  Linkage of osteoporosis to chromosome
20p12 and association to BMP2. PLoS Biol 1:351–360, 2003.
3 6 6   •   G e no m ic s in C l inic a l Pr actic e
66. Ralston SH et  al.:  Loci for regulation of bone mineral density in
men and women identified by genome wide linkage scan:  the
FAMOS study. Hum Mol Genet 14:943–951, 2005.
67. Morrison NA et  al.:  Prediction of bone density from vitamin D
receptor alleles. Nature 367:284–287, 1994.
68. Sainz J et  al.:  Vitamin D-receptor gene polymorphisms and bone
density in prepubertal American girls of Mexican descent. N Engl J
Med 337:77–82, 1997.
69. Tokita A et  al.:  Vitamin D receptor alleles, bone mineral density
and turnover in premenopausal Japanese women. J Bone Miner Res
11:1003–1009, 1996.
70. Uitterlinden AG et al.: A large-scale population-based study of the
association of vitamin D receptor gene polymorphisms with bone
mineral density. J Bone Miner Res 11:1241–1248, 1996.
71. Garnero P et al.: Vitamin D receptor gene polymorphisms are not
related to bone turnover, rate of bone loss, and bone mass in post-
menopausal women: The OFELY Study. J Bone Miner Res 11:827–
834, 1996.
72. Hustmyer FG et al.: Bone mineral density in relation to polymor-
phism at the vitamin D receptor gene locus. J Clin Invest 94:2130–
2134, 1994.
73. Melhus H et  al.:  Vitamin D receptor genotypes in osteoporosis.
Lancet 344:949–950, 1994.
74. Cooper GS et  al.:  Are vitamin D receptor polymorphisms associ-
ated with bone mineral density? A meta-analysis. J Bone Miner Res
11:1841–1849, 1996.
75. Gong G et al.: The association of bone mineral density with vitamin
D receptor gene polymorphisms. Osteoporos Int 9:55–64, 1999.
76. Thakkinstian A et al.: Meta-analysis of molecular association stud-
ies:  vitamin D receptor gene polymorphisms and BMD as a case
study. J Bone Miner Res 19:419–428, 2004.
77. Baker AR et al.: Cloning and expression of full-length cDNA encod-
ing human vitamin D receptor. Proc Natl Acad Sci U S A 85:3294–
3298, 1988.
78. Gross C et al.: The presence of a polymorphism at the translation
initiation site of the vitamin D receptor gene is associated with
low bone mineral density in postmenopausal Mexican-American
women. J Bone Miner Res 11:1850–1855, 1996.
79. Harris SS et al.: The vitamin D receptor start codon polymorphism
(Fok I) and bone mineral density in premenopausal American black
and white women. J Bone Miner Res 12:1043–1048, 1997.
80. Arai H et al.: A vitamin D receptor gene polymorphism in the trans-
lation initiation codon:  effect on protein activity and relation to
bone mineral density in Japanese women. J Bone Miner Res 12:915–
921, 1997.
81. Eccleshall TR et al.: Lack of correlation between start codon poly-
morphism of the vitamin D receptor gene and bone mineral density
in premenopausal French women: The OFELY Study. J Bone Miner
Res 13:31–35, 1998.
82. Gross C et al.: The vitamin D receptor gene start codon polymor-
phism:  a functional analysis of Fok I  variants. J Bone Miner Res
13:1691–1699, 1998.
83. Uitterlinden AG et al.: The association between common vitamin
D receptor gene variations and osteoporosis:  a participant-level
meta-analysis. Ann Intern Med 145:255–264, 2006.
84. Richards JB et al.: Collaborative meta-analysis: associations of 150
candidate genes with osteoporosis and osteoporotic fracture. Ann
Intern Med 151:528–537, 2009.
85. Arai H et  al.:  The polymorphism in the caudal-related homeodo-
main protein Cdx-2 binding element in the human vitamin D recep-
tor gene. J Bone Miner Res 16:1256–1264, 2001.
86. Fang Y et al.: Cdx-2 polymorphism in the promoter region of the
human vitamin D receptor gene determines susceptibility to frac-
ture in the elderly. J Bone Miner Res 18:1632–1641, 2003.
87. Sykes B:  Human genetics. Bone disease cracks genetics. Nature
348:18–20, 1990.
 88. Langdahl BL et  al.:  An Sp1 binding site polymorphism in the
COLIA1 gene predicts osteoporotic fractures in both men and
women. J Bone Miner Res 13:1384–1389, 1998.
89. Garnero P et  al.:  Collagen Iα 1 Sp1 polymorphism, bone mass,
and bone turnover in healthy French premenopausal women: The
OFELY Study. J Bone Miner Res 13:813–817, 1998.
90. Liden M et al.: Polymorphism at the Sp1 binding site in the col-
lagen type Iα 1 gene does not predict bone mineral density in
postmenopausal women in Sweden. Calcif Tissue Int 63:293–295,
1998.
91. Hustmyer FG et  al.:  Polymorphism at an Sp1 binding site of
COL1A1 and bone mineral density in premenopausal female
twins and elderly fracture patients. Osteoporos Int 9:346–350,
1999.
92. Heegaard A et al.: Lack of influence of collagen type Iα 1 Sp1 bind-
ing site polymorphism on the rate of bone loss in a cohort of post-
menopausal Danish women followed for 18 years. Calcif Tissue Int
66:409–413, 2000.
93. Mann V et al.: A COL1A1 Sp1 binding site polymorphism predis-
poses to osteoporotic fracture by affecting bone density and qual-
ity. J Clin Invest 107:899–907, 2001.
94. Garcia-Giralt N et al.: Two new single-nucleotide polymorphisms
in the COL1A1 upstream regulatory region and their relationship
to bone mineral density. J Bone Miner Res 17:384–393, 2002.
95. Efstathiadou Z et  al.:  Association of collagen Iα 1 Sp1 polymor-
phism with the risk of prevalent fractures: a meta-analysis. J Bone
Miner Res 16:1586–1592, 2001.
96. Mann V et  al.:  Meta-analysis of COL1A1 Sp1 polymorphism in
relation to bone mineral density and osteoporotic fracture. Bone
32:711–717, 2003.
97. Ralston SH et  al.:  Large-scale evidence for the effect of the
COLIA1 Sp1 polymorphism on osteoporosis outcomes:  the
GENOMOS Study. PLoS Med 3:e90, 2006.
98. Xu XH et al.: Molecular genetic studies of gene identification for
osteoporosis: the 2009 update. Endocr Rev 31:447–505, 2010.
99. Simonet WS et  al.:  Osteoprotegerin:  a novel secreted protein
involved in the regulation of bone density. Cell 89:309–319, 1997.
100. Tan KB et  al.:  Characterization of a novel TNF-like ligand and
recently described TNF ligand and TNF receptor superfamily
genes and their constitutive and inducible expression in hemato-
poietic and non-hematopoietic cells. Gene 204:35–46, 1997.
101. Tsuda E et al.: Isolation of a novel cytokine from human fibroblasts
that specifically inhibits osteoclastogenesis. Biochem Biophys Res
Commun 234:137–142, 1997.
102. Bucay N et al.: Osteoprotegerin-deficient mice develop early onset
osteoporosis and arterial calcification. Genes Dev 12:1260–1268,
1998.
103. Yasuda H et  al.:  Identity of osteoclastogenesis inhibitory fac-
tor (OCIF) and osteoprotegerin (OPG): a mechanism by which
OPG/OCIF inhibits osteoclastogenesis in vitro. Endocrinology
139:1329–1337, 1998.
104. Bekker PJ et al.: The effect of a single dose of osteoprotegerin in
postmenopausal women. J Bone Miner Res 16:348–360, 2001.
105. Body JJ et  al.:  A phase I  study of AMGN-0007, a recombinant
osteoprotegerin construct, in patients with multiple myeloma or
breast carcinoma related bone metastases. Cancer 97:887–892,
2003.
106. Anderson DM et al.: A homologue of the TNF receptor and its
ligand enhance T-cell growth and dendritic-cell function. Nature
390:175–179, 1997.
107. Hsu H et  al.:  Tumor necrosis factor receptor family member
RANK mediates osteoclast differentiation and activation induced
by osteoprotegerin ligand. Proc Natl Acad Sci U S A 96:3540–
3545, 1999.
108. Lacey DL et al.: Osteoprotegerin ligand is a cytokine that regulates
osteoclast differentiation and activation. Cell 93:165–176, 1998.
G e n etic s a nd G e no m ic s o f O s t e o p oro s i s a nd R e l at e d D i s ord e r s   •   3 6 7
109. Yasuda H et  al.:  Osteoclast differentiation factor is a ligand for
osteoprotegerin/osteoclastogenesis-inhibitory factor and is identi-
cal to TRANCE/RANKL. Proc Natl Acad Sci U S A 95:3597–
3602, 1998.
110. Burgess TL et al.: The ligand for osteoprotegerin (OPGL) directly
activates mature osteoclasts. J Cell Biol 145:527–538, 1999.
111. Fuller K et  al.:  TRANCE is necessary and sufficient for
osteoblast-mediated activation of bone resorption in osteoclasts. J
Exp Med 188:997–1001, 1998.
112. Takai H et al.: Transforming growth factor-β stimulates the pro-
duction of osteoprotegerin/osteoclastogenesis inhibitory factor by
bone marrow stromal cells. J Biol Chem 273:27091–27096, 1998.
113. Akatsu T et  al.:  Osteoclastogenesis inhibitory factor suppresses
osteoclast survival by interfering in the interaction of stromal cells
with osteoclast. Biochem Biophys Res Commun 250:229–234, 1998.
114. Hofbauer LC et al.: Clinical implications of the osteoprotegerin/
RANKL/RANK system for bone and vascular diseases. JAMA
292:490–495, 2004.
115. Whyte MP et al.: Osteoprotegerin deficiency and juvenile Paget’s
disease. N Engl J Med 347:175–184, 2002.
116. Arko B et  al.:  Sequence variations in the osteoprotegerin gene
promoter in patients with postmenopausal osteoporosis. J Clin
Endocrinol Metab 87:4080–4084, 2002.
117. Langdahl BL et al.: Polymorphisms in the osteoprotegerin gene are
associated with osteoporotic fractures. J Bone Miner Res 17:1245–
1255, 2002.
118. Yamada Y et al.: Association of polymorphisms of the osteoprote-
gerin gene with bone mineral density in Japanese women but not
men. Mol Genet Metab 80:344–349, 2003.
119. Ohmori H et al.: Linkage and association analysis of the osteopro-
tegerin gene locus with human osteoporosis. J Hum Genet 47:400–
406, 2002.
120. Brändström H et al.: Single nucleotide polymorphisms in the human
gene for osteoprotegerin are not related to bone mineral density or
fracture in elderly women. Calcif Tissue Int 74:18–24, 2004.
121. Richards JB et  al.:  Bone mineral density, osteoporosis, and
osteoporotic fractures:  a genome-wide association study. Lancet
371:1505–1512, 2008.
122. Styrkarsdottir U et al.: Multiple genetic loci for bone mineral den-
sity and fractures. N Engl J Med 358:2355–2365, 2008.
123. Hsu YH et al.: An integration of genome-wide association study
and gene expression profiling to prioritize the discovery of novel
susceptibility loci for osteoporosis-related traits. PLoS Genet
6:e1000977, 2010.
124. Rivadeneira F et al.: Twenty bone-mineral-density loci identified
by large-scale meta-analysis of genome-wide association studies.
Nat Genet 41:1199–1206, 2009.
125. Kim JG et al.: Association between osteoprotegerin (OPG), recep-
tor activator of nuclear factor-κB (RANK), and RANK ligand
(RANKL) gene polymorphisms and circulating OPG, soluble
RANKL levels, and bone mineral density in Korean postmeno-
pausal women. Menopause 14:913–918, 2007.
126. Hsu YH et  al.:  Variations in genes involved in the RANKL/
RANK/OPG bone remodeling pathway are associated with bone
mineral density at different skeletal sites in men. Hum Genet
118:568–577, 2006.
127. Paternoster L et  al.:  Genome-wide association meta-analysis of
cortical bone mineral density unravels allelic heterogeneity at the
RANKL locus and potential pleiotropic effects on bone. PLoS
Genet 6:e1001217, 2010.
128. Choi JY et al.: Genetic polymorphisms of OPG, RANK, and ESR1
and bone mineral density in Korean postmenopausal women.
Calcif Tissue Int 77:152–159, 2005.
129. Koh JM et al.: Identification of novel RANK polymorphisms and
their putative association with low BMD among postmenopausal
women. Osteoporos Int 18:323–331, 2007.
 130. Styrkarsdottir U et al.: New sequence variants associated with bone
mineral density. Nat Genet 41:15–17, 2009.
131. Koay MA et  al.:  Genetic disorders of the LRP5-Wnt signalling
pathway affecting the skeleton. Trends Mol Med 11:129–137,
2005.
132. Westendorf JJ et  al.:  Wnt signaling in osteoblasts and bone dis-
eases. Gene 341:19–39, 2004.
133. Heaney C et al.: Human autosomal recessive osteopetrosis maps to
11q13, a position predicted by comparative mapping of the murine
osteosclerosis (oc) mutation. Hum Mol Genet 7:1407–1410, 1998.
134. Van Wesenbeeck L et al.: Six novel missense mutations in the LDL
receptor-related protein 5 (LRP5) gene in different conditions
with an increased bone density. Am J Hum Genet 72:763–771,
2003.
135. Koay MA et  al.:  Influence of LRP5 polymorphisms on normal
variation in BMD. J Bone Miner Res 19:1619–1627, 2004.
136. Mizuguchi T et al.: LRP5, low-density-lipoprotein-receptor-related
protein 5, is a determinant for bone mineral density. J Hum Genet
49:80–86, 2004.
137. Bollerslev J et  al.:  LRP5 gene polymorphisms predict bone mass
and incident fractures in elderly Australian women. Bone 36:599–
606, 2005.
138. Koller DL et al.: Contribution of the LRP5 gene to normal varia-
tion in peak BMD in women. J Bone Miner Res 20:75–80, 2005.
139. van Meurs JB et al.: Large-scale analysis of association between LRP5
and LRP6 variants and osteoporosis. JAMA 299:1277–1290, 2008.
140. Kiel DP et  al.:  Genetic variation at the low-density lipoprotein
receptor-related protein 5 (LRP5) locus modulates Wnt signaling
and the relationship of physical activity with bone mineral density
in men. Bone 40:587–596, 2007.
141. Korach KS: Insights from the study of animals lacking functional
estrogen receptor. Science 266:1524–1527, 1994.
142. Patel MS et al.: Alleles of the estrogen receptor α-gene and an estro-
gen receptor cotranscriptional activator gene, amplified in breast
cancer-1 (AIBI1), are associated with quantitative calcaneal ultra-
sound. J Bone Miner Res 15:2231–2239, 2000.
143. Willing M et  al.:  Bone mineral density and its change in white
women:  estrogen and vitamin D receptor genotypes and their
interaction. J Bone Miner Res 13:695–705, 1998.
144. Salmén T et al.: The protective effect of hormone-replacement ther-
apy on fracture risk is modulated by estrogen receptor α g­ enotype
in early postmenopausal women. J Bone Miner Res 15:2479–2486,
2000.
145. Becherini L et  al.:  Evidence of a linkage disequilibrium between
polymorphisms in the human estrogen receptor α gene and their
relationship to bone mass variation in postmenopausal Italian
women. Hum Mol Genet 9:2043–2050, 2000.
146. Brown MA et  al.:  Genetic control of bone density and turn-
over: role of the collagen 1α 1, estrogen receptor, and vitamin D
receptor genes. J Bone Miner Res 16:758–764, 2001.
147. Gennari L et  al.:  Vitamin D and estrogen receptor allelic vari-
ants in Italian postmenopausal women: evidence of multiple gene
­contribution to bone mineral density. J Clin Endocrinol Metab
83:939–944, 1998.
3 6 8   •   G e no m ic s in C l inic a l Pr actic e
148. Han KO et  al.:  Nonassociation of estrogen receptor genotypes
with bone mineral density and estrogen responsiveness to hor-
mone replacement therapy in Korean postmenopausal women.
J Clin Endocrinol Metab 82:991–995, 1997.
149. Langdahl BL et  al.:  A TA repeat polymorphism in the estrogen
receptor gene is associated with osteoporotic fractures but poly-
morphisms in the first exon and intron are not. J Bone Miner Res
15:2222–2230, 2000.
150. Vandevyver C et al.: Lack of association between estrogen recep-
tor genotypes and bone mineral density, fracture history, or muscle
strength in elderly women. J Bone Miner Res 14:1576–1582, 1999.
151. Ioannidis JP et al.: Association of polymorphisms of the estrogen
receptor α gene with bone mineral density and fracture risk in
women: a meta-analysis. J Bone Miner Res 17:2048–2060, 2002.
152. Ioannidis JP et al.: Differential genetic effects of ESR1 gene poly-
morphisms on osteoporosis outcomes. JAMA 292:2105–2114,
2004.
153. Herrington DM et al.: Common estrogen receptor polymorphism
augments effects of hormone replacement therapy on E-selectin
but not C-reactive protein. Circulation 105:1879–1882, 2002.
154. Balemans W et  al.:  The binding between sclerostin and LRP5 is
altered by DKK1 and by high-bone mass LRP5 mutations. Calcif
Tissue Int 82:445–453, 2008.
155. Hobson EE et al.: Role of genetic factors in the pathophysiology
and management of osteoporosis. Clin Endocrinol 54:1–9, 2001.
156. Xiong DH et  al.:  Genome-wide association and follow-up rep-
lication studies identified ADAMTS18 and TGFBR3 as bone
mass candidate genes in different ethnic groups. Am J Hum Genet
84:388–398, 2009.
157. Duncan EL et al.: Genome-wide association study using extreme
truncate selection identifies novel genes affecting bone mineral
density and fracture risk. PLoS Genet 7:e1001372, 2011.
158. Kou I et al.: Common variants in a novel gene, FONG on chro-
mosome 2q33.1 confer risk of osteoporosis in Japanese. PLoS One
6:e19641, 2011.
159. Guo Y et al.: Genome-wide association study identifies ALDH7A1
as a novel susceptibility gene for osteoporosis. PLoS Genet
6:e1000806, 2010.
160. Cho YS et  al.:  A large-scale genome-wide association study of
Asian populations uncovers genetic factors influencing eight quan-
titative traits. Nat Genet 41:527–534, 2009.
161. Liu YZ et al.: Powerful bivariate genome-wide association analyses
suggest the SOX6 gene influencing both obesity and osteoporosis
phenotypes in males. PLoS One 4:e6827, 2009.
162. Guo Y et al.: IL21R and PTH may underlie variation of femoral
neck bone mineral density as revealed by a genome-wide associa-
tion study. J Bone Miner Res 25:1042–1048, 2010.
163. Koller DL et al.: Genome-wide association study of bone mineral
density in premenopausal European-American women and rep-
lication in African-American women. J Clin Endocrinol Metab
95:1802–1809, 2010.
164. Kung AW et al.: Association of JAG1 with bone mineral density
and osteoporotic fractures: a genome-wide association study and
follow-up replication studies. Am J Hum Genet 86:229–239, 2010.
 24.
GENETICS AND GENOMICS OF CHRONIC KIDNEY
DISEASE
Albert C. M. Ong and Alexander P. Maxwell

INTRODUCTION Two Genes Cause ADPKD

The prevalence of end-stage renal disease (ESRD) is steadily
increasing, contributing to growth in the provision of renal
replacement therapy (dialysis and transplantation) for over
one million individuals worldwide (Collins et  al., 2003).
Living with ESRD is challenging for individuals, and the
economic costs of renal replacement therapy for society are
enormous (Meguid El Nahas and Bello, 2005).
Figure 24.1 shows the relative prevalence of different
primary renal diseases in the English and Welsh popula-
tion on renal replacement therapy (RRT) in 2010 and in
the American population in 2009. Diabetic nephropathy
is the most common cause of ESRD in the United States
and is increasing in prevalence worldwide. There is increas-
ing evidence for genetic (polygenic) susceptibility for dia-
betic nephropathy. Autosomal dominant polycystic kidney
disease (ADPKD) accounts for 5–10% of prevalent ESRD
and is the single most common monogenic cause of ESRD
worldwide. In this chapter, we review the evidence of how
genetic factors can modify the phenotypical expression
of these two common diseases, and illustrate the differ-
ent experimental approaches that are being used to iden-
tify modifying genes that are important in chronic kidney
disease.
Mutation to PKD1 (chromosome region 16p13.3) is the
most common cause of ADPKD (~86% cases), with most of
the remainder due to changes to PKD2 (4q22). PKD1 and
PKD2 patients have indistinguishable renal and extra-renal
phenotypes: they were only recognized as distinct diseases by
genetic linkage analysis in the late 1980s (Parfrey et al., 1990).
PKD1 is a complex gene with 46 exons that generates a large
transcript (~14kb) containing a long open-reading frame
predicted to encode a 4302aa protein named polycystin-1
(Hughes et al., 1995). PKD2 has 15 exons, generates a ~5kb
transcript, and encodes polycystin-2, a protein of 968aa
(Mochizuki et al., 1996) (Figure 24.2).
Gene Locus Effect
PKD2 is a significantly milder disease in terms of the mean
age at diagnosis, a lower prevalence of hypertension, and a
later age at onset of ESRD (PKD1, 54.3 yrs; PKD2, 74.0 yrs)
(Hateboer et al., 1999). Consistent with this milder pheno-
type, there appears to be an enrichment of PKD2 patients
among ADPKD patients who present late (after 63 yrs) with
ESRD (Torra et al., 2000). Conversely the vast majority of
very early onset (VEO) ADPKD manifesting in utero or
infancy is related to mutations in PKD1 (Peral et al., 1996).

The Effect of Gender

G E N ET I C H ET E R O G E N E I T Y O F A D P K D
While there is no clear gender difference in PKD1, PKD2
females have a significantly better prognosis (age at onset of
ADPKD AS A COMPLEX TRAIT DISEASE
ESRD: males, 68.1 yrs; females, 76.0 yrs): the reason for this
Autosomal dominant polycystic kidney disease (ADPKD) is unclear (Magistroni et al., 2003; Rossetti et al., 2002a).
is one of the most common human monogenic diseases,
with an incidence of 1:400–1:1000. It is characterized by
Allelic Heterogeneity
the progressive development and enlargement of focal cysts
in both kidneys, typically resulting in end-stage renal dis- Both PKD1 and PKD2 exhibit marked allelic heterogene-
ease (ESRD) by the fifth decade. ity, with over 400 different PKD1 and 100 different PKD2

369
 (A)

Renal vascular Other

disease 4% 15% Aetiology
unknown 21%
Hypertension
6%

GN (biopsy proven)
PKD 10% 17%

Diabetes Pyelonephritis
15% 12%

(B) Aetiology
unknown 4%

0% Other
13% GN 15%

0%
Hypertension
25%

Diabetes 38%

PKD 5%

Figure 24.1 
Primary diagnosis in prevalent patients on renal replacement therapy (RRT) in England and Wales (E&W) in 2010 and in the United
States in 2009. Diabetes was the second most common cause in the U.K. (A) and the most common cause of ESRD in prevalent RRT patients in the
U.S. (B). In the U.K., the major change in the over-65 age group was an increase in renovascular disease (from 1.1% to 7.8%). The male:female ratio
was greater than unity for all groups except for PKD. Source: UK Renal Registry data, 2010, US Renal Data System 2011.

mutations reported to date (ADPKD mutation database,
http://pkdb.mayo.edu). Unlike PKD2, PKD1 is highly poly-
morphic, with on average 10–13 changes described per indi-
vidual (Garcia-Gonzalez et al., 2007; Rossetti et al., 2007).
The majority of mutations are private (i.e., unique to a single
family) and indicate that a significant level of new mutation
is occurring (Rossetti et al., 2001). For PKD1, mutations 5′
to the median are associated with more severe disease (aver-
age age at onset of ESRD: 5′, 53 yrs; 3′, 56 yrs) and a sig-
nificantly greater risk of developing intracranial aneurysms
(Rossetti et al., 2002a; Rossetti et al., 2003).This association
is not related to mutation type and may be due to the pro-
posed cleavage of polycystin-1 via a G-protein coupled recep-
tor proteolytic site (GPS) into two different proteins (Figure
24.2) with mutations to each half having potentially different
phenotypical consequences (Qian et al., 2002; Rossetti et al.,
2003).As yet, no clear phenotype–genotype correlations
have been reported for PKD2 (Magistroni et al., 2003).
Gene Syndromes
Patients with a contiguous gene deletion of both PKD1
and the neighboring tuberous sclerosis gene (TSC2) allele
have more severe early-onset PKD than those with TSC2
mutations alone (Brook-Carter et al., 1994). Most probably
this indicates a synergistic role between polycystin-1 and
tuberin (the TSC2 protein) in cyst development:  tuberin
may play a role in trafficking polycystin-1 to the lateral cell
membrane in kidney epithelial cells (Kleymenova et  al.,
2001; Ong et al., 1999).
A family with bilineal inheritance of germline PKD1
and PKD2 mutations has been described and recently

3 7 0   •   G eno m ic s in C l inica l P ractice

 Structure and function of the polycystin proteins
NH2

Key

Signal sequence PLAT—lipid binding?

Leucine rich repeats REJ module
WSC domain GPS domain

PKD repeat G-protein binding

C-type lectin Coiled coil

LDL-A related EF-hand

Transmembrane region

Voltage activated/TRP channel homology

GPS

NH2 COOH
COOH
Polycystin-1 Polycystin-2

Figure 24.2 
Predicted structure and topology of the ADPKD proteins, polycystin-1 and polycystin-2. Several functional protein, carbohydrate, and
lipid-binding functional domains of both proteins are depicted (Ong and Harris, 2005). Both proteins interact via a coiled coil domain in their
C-termini. A G-protein coupled receptor proteolytic site (GPS) may allow polycystin-1 to be cloven into two halves that remain tethered by
non-covalent bonds.

endoplasmic reticulum (ER) located polycystin-2 to release

further characterized (Pei et al., 2012; Pei et al., 2001). The
PKD1 mutation is predicted to be a hypomorphic change
(Y528C), while the PKD2 mutation is a truncating muta-
tion (L736X). Although these patients have more severe
disease than cases with either mutation alone, the difference
is not dramatic (i.e., not every renal tubular cell gives rise to
a cyst). The effect of a trans-heterozygous mutation in either
gene appears to act as a modifying factor for the other in
terms of the risk of cyst development or hastening its pro-
gression, rather than an effect on cyst initiation itself (Wu
et al., 2002).
MO D E L S O F P O LYC YS T I N
FUNCTION
Polycystin-1 and polycystin-2 form a receptor-ion com-
plex that regulates tubular morphogenesis (Ong and
Harris, 2005). However, both proteins have been local-
ized in multiple subcellular compartments that may differ
between cell types and developmental status (Ong, 2000).
Several models of polycystin function have thus been pro-
posed (Giamarchi et  al., 2010). It is likely that they form
mechano-sensitive and ligand-gated channel complexes
in primary cilia and at the basolateral surface of epithelial
cells. Plasma membrane polycystin-1 may couple to activate
intracellular Ca2+. Finally, both proteins have been found
in urinary exosomes and may represent a form of urocrine
signaling (Hogan et al., 2009).
MU TAT I O NA L M EC H A N I S M
I N A D P K D
Somatic Mutations
A “two-hit” mechanism of cyst formation has been pro-
posed for ADPKD (consisting of a germline mutation
to one allele and a somatic mutation to the other). There
is evidence for epithelial cell clonality and a high rate of
somatic mutations in cells isolated from individual kid-
ney and liver cysts (Brasier and Henske, 1997; Qian et al.,
1996; Watnick et al., 1998). Experimentally, a unique Pkd2
knockout mouse (Pkd2WS25 mutant) which has an unstable
allele (prone to inactivation by recombination) develops
progressive cystic disease in a manner consistent with a
two-hit model (Lu et al., 1997; Wu et al., 1998a). Certainly,
a two-hit model could help explain both the focal nature
of cyst development and the striking intrafamilial pheno-
typical variability seen in most families. However, there
remain questions as to whether a two-hit mechanism is the
only means to generate a cyst, and indeed, whether these

G enetic s and G eno m ic s of C hronic K idney D i s ea s e   •   3 7 1

somatic events may be later events more important for cyst
expansion and progression rather than initiation (Ong and
Harris, 1997).
Evidence for Gene Dosage
A decrease (loss of a single allele or haploinsufficiency) or
increase (transgenic expression) in polycystin-1 dosage may
itself result in cyst formation (Lantinga-van Leeuwen et al.,
2004; Pritchard et  al., 2000; Kurbegovic et  al., 2010). In
addition, haploinsufficient Pkd2 mice have reduced survival
(not due to renal failure) indicating that a dosage reduction
of polycystin-2 itself can exert a phenotypical load (Wu
et  al., 2000). Recent studies have shown that Pkd2+/- vas-
cular smooth muscle cells (which express a lower level of
polycystin-2) have altered intracellular Ca2+ homeostasis
and are hyperproliferative (Qian et al., 2003). Similar find-
ings have been shown for Pkd1+/- cyst epithelia (Yamaguchi
et al., 2004) and mouse collecting duct cells (Ahrabi et al.,
2007). Recent work has pinpointed the influence of hypo-
morphic alleles in PKD1 and PKD2 in very early onset
PKD, proving the importance of gene dosage.
In summary, there appear to be multiple mechanisms
that could underlie cyst formation. Cyst development
requires a germline mutation, but beyond this, the likeli-
hood of cyst formation is influenced by a number of differ-
ent factors, which could include somatic genetic events at
the other (normal) allele, mutations at the other ADPKD
gene, and possibly a wide array of other genetic loci (e.g.
TSC2). In effect, these loci act as modifiers of disease pre-
sentation in ADPKD. Environmental or genetic factors
that modulate the rate of somatic mutation or DNA repair
could modify the disease phenotype (Peters and Breuning,
2001). Beyond the genetic events, stochastic factors prob-
ably also influence whether a cell, haploinsufficient for an
ADPKD mutation, is diverted into a cystogenic pathway.
Another factor that could modify the cystic phenotype is
the presence of hypomorphic alleles giving rise to partially
functional (mutant) polycystin-1 protein, such as in Pkd1
del34 mice (Lu et  al., 2001); alternatively, nonfunctional
mutant protein could act in a dominant negative manner.
I N T R A FA M I L I A L VA R I A B I L I T Y
The wide range of intrafamilial phenotypical variability
in large ADPKD families and more recently for PKD1
and PKD2 has been described (Magistroni et  al., 2003;
Milutinovic et al., 1992; Rossetti et al., 2003). The phenom-
enon of genetic anticipation in ADPKD families was first
3 7 2   •   G eno m ic s in C l inica l P ractice
postulated in 1925 with the description of very early onset
(VEO) cases, often born to relatively asymptomatic parents
(Cairns, 1925). However, no evidence for dynamic unstable
mutations in PKD1 or PKD2 has been found. Nevertheless,
there is a high recurrence risk (45%) of a severely affected
sibling in subsequent pregnancies (Zerres et al., 1993). For
VEO cases, it is difficult to envisage (though impossible
to exclude) that somatic mutations to PKD1 could be the
underlying reason for increased disease severity in utero
(Ong et al., 1999).
VE RY E A R LY O N S ET P K D
The genetic characterization of very early onset families has
identified hypomorphic alleles in PKD1 that are incom-
pletely penetrant but, if co-inherited with a truncating
or hypomorphic PKD1 mutation or another cystic gene
(HNF-1b), give rise to a severe presentation (Rossetti et al.,
2009). The combination of a hypomorphic PKD2 mutation
and frameshifting PKHD1 mutation was reported in a single
early-onset family (Bergmann et  al., 2011). These findings
raise several important points. First is the question as to how
many unclassified variants (UCV) reported for PKD1 could
represent hypomorphic alleles. Second, mutations in other
cystic genes can potentially determine phenotype in an addi-
tive and dosage-dependent manner. Third, they imply that
there may be a critical polycystin threshold for cyst forma-
tion and raise questions about the role of somatic mutation
in cyst initiation. At present, we do not know the relative
contribution that each makes to the phenotype in either very
early onset disease or more typical adult-onset PKD.
MO S A I C I S M
A family with germline and somatic mosaicism in PKD1
has been reported (Connor et  al., 2008). This could be a
rare cause of intrafamilial variability.
Non-Allelic Factors
Other non-allelic causes of variability include modifying
genes, environmental factors, or stochastic events. Many
studies have attempted to define the possible role of modi-
fying genes (heritability) in determining renal outcome.
An informative study of 74 parent–offspring pairs where
the age of renal death of both was known showed a wide
Gaussian distribution of values around zero (Geberth et al.,
1995). Furthermore, there was no difference in the median
age of renal death between two decades (1950–1971, 1975–
1985) thus excluding the effect of secular trend as a major
confounding factor. A second study utilizing sib-pairs com- Coexistent Diseases—Cystic Fibrosis
pared the age of ESRD between 56 sibships and 9 pairs of
A significant body of evidence suggests that ADPKD
monozygotic twins and calculated the intraclass correlation
cyst epithelial cells undergo a phenotypical switch to
coefficient (ICC) within both groups (Persu et al., 2004).
cAMP responsiveness as measured by cell proliferation
The ICC, which compares the similarity within sibships
and fluid secretion (Grantham, 2003). The cystic fibro-
to that between sibships, was significantly higher in the
sis transmembrane conductance regulator (CFTR) is a
twins (0.92) than in other sibs (0.49). Discordance in the
cAMP-dependent chloride channel protein that is mutated
renal phenotype between a pair of dizygotic twins carrying
in patients with the autosomal recessive disease cystic fibro-
the same PKD1 mutation has been reported in one family
sis (CF). CFTR is expressed in the kidney and in cyst lin-
(Peral et al., 1996). Later reanalysis of this family revealed
ing cells. A family with two individuals with coexistent CF
that the more severely affected twin had inherited both the
and ADPKD has been reported with normal kidney func-
germline PKD1 mutation and a potential hypomorphic
tion, suggesting a possible protective effect of the absence
PKD1 mutation (from the clinically unaffected parent)
of CFTR (O’Sullivan et al., 1998). However, this was not
(Rossetti et al., 2009).
substantiated in another family where neither heterozygous
The possible influence of modifying genes (heritability) in
nor homozygous CFTR mutations (DeltaF508) had a pro-
PKD1 families has also been addressed in two large studies.
tective effect on renal outcome (Persu et al., 2000)
Heritability (h2) for the age of ESRD was estimated at between
43–78% (Fain et al., 2005; Paterson et al., 2005). However,
the number of potential modifiers, their allele frequency, and A S S O C I AT I O N S T U D I E S O F
their relative effects in relation to determining loss of renal C A N D I DAT E - G E N E P O LY MO R P H I S M S
function are not clear. Paterson and colleagues calculated that
Numerous studies examining the association between
a genome-wide scan to look for a single modifier locus will
candidate-gene polymorphisms (e.g., Angiotensin Converting
require very large cohorts to be powered adequately and thus
Enzyme [ACE], Endothelial Nitric Oxide Synthase
necessitate international collaboration (Paterson et al., 2005).
[ENOS]) and kidney function have been performed.
These studies also point to the likely significance of
Overall, these finding have proven to be inconclusive and
(unknown) environmental factors as contributing to the
contradictory (Persu et  al., 2002; Walker et  al., 2003).
ESRD phenotype. These factors could include urinary tract
A recent single-nucleotide polymorphism (SNP) genotyping
infections in men and the number of pregnancies in women
association study of 173 candidate genes in a cohort of 794
(Gabow et al., 1992), but require verification in larger scale
white patients initially identified 12 SNPs associated with
populations. Figure 24.3 summarizes the potential causes of
Estimated Glomerular Filtration Rate (eGFR) or renal sur-
phenotypical variability in this disease.
vival (time to ESRD). However, only three SNPs remained
associated with eGFR in a second replicative cohort of 472
Locus-specific
white patients (Liu et al., 2010): these were all located at the
Genotype
Allele-specific DKK3 locus, DKK3 being a secreted antagonist of Wnt sig-
naling. It is estimated that GWAS studies of PKD1 disease
variance will require 3000–4000 patients to detect multiple
Somatic Gender
mutations loci with similar effect size.
Environmental
Modifying genes Factors PAT H O G E N E S I S O F C YS T I C
50–70% (30–50%)
CHANGES IN ADPKD
Phenotype
MO D I FI E R S I N RO D E N T P K D MO D E L S
Figure 24.3 
Genetic and environmental factors determining the renal The apparent similarities in the process of cyst formation
phenotype in ADPKD. There is a strong locus effect, but allelic-specific
factors are weak. Co-inheritance of two hypomorphic mutations in between different rodent models and human disease have led
PKD1 as well as mutations in other cystic genes (PKHD1, HNF1b) some investigators to exploit the approach of cross-breeding
have been recently reported in some very early onset cases. An effect a PKD allele onto different genetic backgrounds to map
of gender has been reported in PKD2 (but not PKD1) patients. The
precise contributions of different non-allelic factors (somatic mutation, modifier loci. In theory, this approach allows environ-
genetic modifiers, and environmental agents) are uncertain. mental factors to be controlled for, thus permitting the

G enetic s and G eno m ic s of C hronic K idney D i s ea s e   •   3 7 3

contribution of genetic factors to be paramount. In one P H E N OT Y P I C A L D I FFE R E N C E S
of the most-studied murine PKD models, the cpk mouse, B ET WE E N PK D1 K N O C KO U T M I C E
it is apparent that disease expression and the severity of
Numerous Pkd1 deficient mice have been reported, all
both renal and extra-renal phenotypes can be modified by
of which are associated with a cystic kidney phenotype.
genetic background. For instance, the biliary abnormality
Almost all are embryonically lethal in the recessive state
in cpk is not penetrant in the original C57BL/6 (B6) back-
apart from a hypomorphic allele (Pkd1nl) (Lantinga-van
ground but is variably penetrant when expressed in other
Leeuwen et al., 2004). Of interest, some, but not all, mutants
strain backgrounds, such as CAST/Ei, DBA/2J, BALB/c
develop skeletal or cardiovascular defects, unrelated to the
or CD1 (Guay-Woodford, 2003).
site of the gene disrupted. There is a high prevalence of aor-
PKD due to the pcy allele is more severe in a DBA/2
tic aneurysms in homozygous Pkd1nl mice that has not been
than a C57BL/6 background. A  good example of the
described in other models (Lantinga-van Leeuwen et  al.,
approach of whole-genome quantitative trait loci (QTL)
2004). The expression of these phenotypes could therefore
mapping is one that led to the identification of two major
be related to genetic background; that is, other modifying
modifiers (MOP1, MOP2) regulating renal disease pro-
genes. This may reflect the situation in human ADPKD
gression in the pcy mouse (Woo et al., 1997). The pcy gene
where not all patients (i.e., ~8–10%) develop vascular com-
was subsequently identified as the murine homologue of
plications (aneurysms), and, more rarely, patients associated
the nephronopthisis Type 3 (NPHP3) gene, which gives
with a Marfanoid phenotype have been reported (Rossetti
rise to adolescent-onset nephronopthisis, a phenotype that
et  al., 2003; Somlo et  al., 1993). Several other models
can include cysts (Olbrich et al., 2003). It would therefore
have been reported, including (1) conditional (inducible)
be of interest to investigate whether MOP1 and MOP2
tissue-specific Cre-Lox deletion mutants; (2) hypomorphic
can also act as modifiers in the human disease. Similar
alleles; and (3) knockdown alleles. These will be valuable as
approaches have been taken in other rodent strains, includ-
models to test new drugs in a preclinical setting.
ing the recessive cpk mouse and the dominant Han:SPRD
(cy/+) rat, genes that have not yet been associated with
P O LYC YS T I N H O MO L O GU E S
human PKD (Bihoreau et  al., 2002; Mrug et  al., 2005).
Table 24.1 summarizes the most common rodent PKD Numerous polycystin-like homologues have been identified
models that have been characterized and those in which from whole-genomic and Expressed Sequence Tag (EST)
QTL have been mapped. It is hoped that a comparative databases. In essence, they divide into “PC1-like” and
genome analysis will enable any modifying loci identified “PC2-like” proteins (Table 24.2). In theory, these homo-
in a rodent model to be mapped back to a human suscepti- logues could modulate the PKD phenotype caused by
bility locus in ADPKD. mutations in PKD1 or PKD2. However, the restricted tissue

Table 24.1  QTL MAPPING IN RODENT MODELS OF PKD

MODEL TRANSMISSION GENE PROTEIN HUMAN DISEASE ORIGINAL QTL

STRAIN
Mouse
bpk AR BiccI Bicaudal C BALB/c +
cpk AR CysI Cystin C57BL/6J (B6) +
inv AR Invs Inversin Yes (NPHP2) OVE210
jck AR Nek8 Nek8 Yes (NPHP9) MMTV/c-myc +
jcpk AR BiccI Bicaudal C
kat AR Nek1 Nek1 RBF/Dn +
kat2J C57BL/6J
orpk AR TgN737Rpw Polaris FBV/N +
pcy AR NPHP3 Nephrocystin-3 Yes KKDBA/2J +
Rat
Han:SPRD-cy AD Pkdr1 SamCystin SPRD +
pck AR Pkhd1 Fibrocystin Yes (ARPKD) Crj:CD/SD
wpk AR Mks3 Meckelin Yes (MKS3) Wistar

3 7 4   •   G eno m ic s in C l inica l P ractice

Table 24.2  POLYCYSTIN-1 AND POLYCYSTIN-2 HUMAN HOMOLOGUES

PROTEIN GENE CHROMOSOME FUNCTION HUMAN DISEASE TISSUE DISTRIBUTION REFERENCE

Polycystin-1 PKD1 16p13.3 ?Adhesion ADPKD Widespread, higher in foetal tissue Hughes et al., 1995; Nauli et al., 2003; Streets
?Channel regulator et al., 2002
?Mechanosensor
Polycystin-REJ PKDREJ 22q13 Acrosome reaction Unknown Testis Hughes et al., 1999; Sutton et al., 2008
Polycystin-1L1 PKD1L1 7p12-13 LR determination Unknown Heart, brain, placenta, testis, mam- Yuasa et al., 2002; Kamura et al., 2011; Field
mary tissue et al., 2011
Polycystin-1L2 PKD1L2 16q 22-23 ?Neuromuscular function Unknown Heart, brain, placenta, testis, lung, Li et al., 2003; Mackenzie et al., 2009
pancreas, liver, skeletal muscle
Polycystin-1L3 PKD1L3 16q 22-23 Sour taste receptor Unknown Placenta, liver, testis Li et al., 2003; Ishimaru et al., 2006
Polycystin-2 PKD2 4q21-23 Ion channel ADPKD Widespread McGrath et al., 2003; Mochizuki et al., 1996;
LR determination Neill et al., 2004
?Acrosome reaction
Polycystin-L PKDL 10q24-25 Ion channel Unknown Heart, brain, testis, retina, spleen, Chen et al., 1999; Wu et al., 1998b; Ishimaru
(Polycystin-2L; (PKD2L; Sour taste receptor kidney, skeletal muscle et al., 2006
Polycystin-2L1) PKD2L1)
Polycystin-2L2 PKD2L2 5q31 ?Spermatogenesis Unknown Testis Guo et al., 2000; Chen et al., 2008
distribution of each homologue (most are not expressed in
the kidney) suggests that they are more likely to play a role
in expression of the extra-renal phenotype. Three of these
complexes have been shown to play functional roles in
unexpected phenotypes:  the PKD1L1-PKD2 complex in
left-right axis determination; PKD1L3-PKD2L1 in sour
taste sensation; PKDREJ-PKD2 in the acrosome reaction
(Hughes et al., 1999). It is likely that other combinations of
PC1 and PC2 homologues could fulfill other yet undiscov-
ered biological functions.
P O S S I B L E PAT H WAYS D ET E R M I N I N G
D I S E A S E P RO G R E S S I O N A N D
T H E R A P EU T I C S
Our discussion so far has focused largely on factors that
could modulate the rate of cyst initiation. However, factors
that could regulate the rate of cyst growth (proliferation,
apoptosis, fluid secretion) may be more appropriate thera-
peutic targets than attempts to replace the gene (Qian et al.,
2001)  (Figure 24.4). Also, the diseased ADPKD kidney
is characterized by non-cystic features such as interstitial
fibrosis, inflammation, and vessel wall thickening (leading
to ischemia) (Ong and Harris, 2005). Although it is unclear
whether these are the direct downstream consequences of
cystic transformation, alternative pharmacological strate-
gies targeting these pathways might be effective ways to
prevent or delay progression to ESRD. A large number of
published studies have tested potential drugs, treatments, or
Cyst expansion and disease progression
Germline
Somatic mutations
hypomorphic alleles
PKD1/2 Tubular cell other cystic gene
mutation mutations
stoichastic factors
Cystic cell
Fluid Proliferation Matrix
transport Apoptosis accumulation
Figure 24.4 
Pathophysiology of disease progression in the cystic kidney.
The rate of cyst formation as well as downstream phenotypical changes
in the cystic cell (cell turnover, fluid secretion, matrix accumulation)
could be modified by non-allelic and environmental factors (Ong and
Harris, 2005). This may account for the phenotypical variability arising
between individuals carrying the same germline mutation. Potential
triggers to a cystic cell phenotype could include somatic mutations or
stochastic factors.
3 7 6   •   G eno m ic s in C l inica l P ractice
diets in rodent PKD models, especially in the Han:SPRD
rat (Guay-Woodford, 2003). However, it should be noted
that this model displays some unusual features, such as
gender dimorphism, which could limit its applicability
to human ADPKD. Moreover, it is clear that what works
in one model may not always be reproduced in another
(Guay-Woodford, 2003). Therefore, future studies should
ideally concentrate on orthologous ADPKD models.
Figure 24.5 summarizes the major signaling pathways in
ADPKD that have been identified and targeted based on
a mechanism-based therapeutic strategy (Chang and Ong,
2011). The main approaches can be divided into those that
target cell proliferation, fluid secretion, or cAMP turnover.
Agents that block cAMP production, such as the vasopres-
sin type 2 receptor (VP2R) antagonists, are likely to affect
both proliferation and fluid secretion (Torres et al., 2004).
Genetic variants in key components of these pathways
could act as modifiers of the disease phenotype or deter-
mine response to treatment.
D I A B ET I C N E P H R O PAT H Y:
A   MU LT I FAC TO R I A L K I D N EY D I S E A S E
W I T H A G E N ET I C P R E D I S P O S I T I O N
Diabetic nephropathy is the most common identified cause
of chronic kidney disease in the United States and the sec-
ond most common cause of ESRD in the United Kingdom
(U.S. Renal Data System 2010; U.K. Renal Registry 2010;
Levey and Coresh, 2012). The prevalence of diabetic kid-
ney disease continues to rise, reflecting the epidemic of
type 2 diabetes, the improving survival of persons with dia-
betes, and the increased number of diabetic patients with
end-stage renal disease (ESRD) accepted for renal replace-
ment therapy (chronic dialysis or renal transplantation)
( Jones et al., 2005).
The clinical diagnosis of diabetic nephropathy is based
on the presence of persistent proteinuria (albuminuria) in
an individual with diabetes in the absence of other renal
diseases, urinary tract infection, or heart failure. It is usually
associated with rising blood pressure and eventual decline
in glomerular filtration rate (GFR). The fall in GFR is nor-
mally identified once the serum creatinine rises outside the
normal range.
The early changes in kidney function and structure
include development of glomerular capillary hyperten-
sion, epithelial cell (podocyte) dysfunction, and glomerular
basement membrane thickening with associated mesan-
gial matrix expansion (Adler, 2004; Fioretto and Mauer,
2007; Fioretto et  al., 1992). These pathological changes
(A) (B) K+

NKCC1 TRAM-34

Sorafenib KCa3.1
EGFR SR
CFTR inhibitors Cl– + Metformin
RAS B-Raf MAPK – + – AMPK
Bosutinib + Somatostatin CFTR
Src cAMP Cl– cAMP SR
Tolvaptan +
cAMP + – + Somatostatin
+ V2R
Fluid secretion –
Cilia – Tolvaptan
Proliferation +
2+ Cilia V2R
Ca ATP ATP
– Roscovitine Ca2+
+
PC1 PC2 Genz-123346 HDAC inhibitors
PC1 PC2 PC2
+ – AMPK Ca2+
mTOR Calcimimetics + ER
GlcCer – + + +
TSC1/TSC2 CaSR
Sirolimus Metformin Triptolide
Everolimus
PC1 PC2 PC1 PC2

Figure 24.5 
Mechanism-based therapeutics in ADPKD. The major signaling pathways involved in the pathogenesis of ADPKD and the targets of
candidate drugs (Chang and Ong, 2011). (A) The “secretory” pathways; (B) the “proliferative” pathways. Abbreviations: AMPK, AMP-activated protein kinase; CaSR,
calcium-sensing receptor; CFTR, cystic fibrosis transmembrane regulator; ER, endoplasmic reticulum; GlcCer, glucosylceramide; HDAC, histone deacetylase; KCa3.1, endothelial Ca2+-activated K+-channel; MAPK,
mitogen-activated protein kinase; mTOR, the mammalian target of rapamycin; NKCC1, Na-K-Cl cotransporter; PC, polycystin; SR, somatostatin receptor; TSC, tuberous sclerosis;V2R, vasopressin V2 receptor.

are associated with the loss of the normal barrier function
of the glomerular basement membrane; consequently,
albumin can pass from the glomerular capillaries through
“leaky” glomerular basement membranes and disrupted
slit diaphragms into the urinary space (Gross et al., 2005a)
(Figures 24.6a and 24.6b ). Initially, this abnormal urinary
albumin excretion is low in quantity and not detected on
dipstick urinalysis. Microalbuminuria, a term that describes
this small amount of urinary albumin excretion, is important
since it is often the first detectable clinical sign of diabetic
kidney disease (Shields and Maxwell, 2010). Persistent glo-
merular injury is associated with renal tubular dysfunction,
clinical proteinuria (dipstick urinalysis positive), and pro-
gressive renal failure (Figure 24.6c).
Diabetic nephropathy is arguably the most important
microvascular complication of diabetes in terms of health-
care costs, morbidity, and premature mortality (Alicic
and Tuttle, 2010; Groop et  al., 2009; Hex et  al., 2012).
Individuals with diabetes have an increased mortality rate
due to cardiovascular disease, such as coronary heart disease
and stroke (Morgan et al., 2000; Williams and Airey, 2002;
Ninomiya et  al., 2009; Drury et  al., 2011). The develop-
ment of diabetic nephropathy further amplifies this risk of
macrovascular disease which is increased two- to four-fold

(B)
(A) Ang II
Afferent Efferent
arteriole arteriole Angiotensin II induced
P vasoconstriction of the
efferent arteriole
Ang II
P

Thickened but “leaky”

glomerular basement
Glomerular membrane Microalbuminuria:
Glomerular basement membrane Filtrate with a small amount
filtrate of albumin crossing the glomerular
basement membrane
Figure 24.6a 
Normal glomerular filtration. Normal glomerular filtration
depends on glomerular capillary pressure (P) providing a hydraulic Figure 24.6b 
Early structural and functional changes in the pathogenesis
force for filtration across the semipermeable glomerular basement of diabetic nephropathy. Glomerular capillary pressure (P) is elevated
membrane. Constant glomerular capillary pressure is maintained by due to vasoconstriction of the efferent arteriole mediated by higher
autoregulation of glomerular blood flow by adjustment of the vascular renal tissue levels of angiotensin II. The glomerular basement membrane
resistance provided by the afferent (pre-glomerular) and efferent is widened but has decreased charge and size selectivity permitting an
(post-glomerular) arterioles. increased flux of albumin from blood to urinary space.

G enetic s and G eno m ic s of C hronic K idney D i s ea s e   •   3 7 7

 (C)
Progressive Renal Disease
Glomerulus
Renal tubules
Glomerulus
Glomerulus
Glomerulus
Figure 24.6c 
Progressive glomerular sclerosis and tubulointerstitial fibrosis are associated with loss of renal function. An inflammatory interstitial
infiltrate is seen in association with tubular atrophy.
in early renal disease (microalbuminuria), nine times with
established nephropathy (proteinuria) and up to 50-fold
in persons with ESRD (Deckert et  al., 1996; Dinneen
and Gerstein, 1997; Tuomilehto et al., 1998; Fuller et al.,
2001). Diabetic nephropathy is associated with prolonged
duration of diabetes, poor glycemic control, and raised
blood pressure, and is more common in diabetic individu-
als of Asian or African ancestry (The Diabetes Control
and Complications Trial (DCCT) Research Group, 1993;
UK Prospective Diabetes Study [UKPDS] Group  1998;
Nelson et al., 1997).
Primary prevention of diabetic nephropathy requires
excellent glycemic control that is often difficult to achieve
in practice (The Diabetes Control and Complications Trial
(DCCT) Research Group,1993; UK Prospective Diabetes
Study (UKPDS) Group  1998; Barnett, 2004). Intensive
control of diabetes reduces the incidence of diabetic
nephropathy in type 1 diabetes (de Boer et al., 2011) and
type 2 diabetes (Stratton et al., 2000). Nevertheless, efforts
to intensively manage glycemic control in patients with
both type 2 diabetes and cardiovascular disease have been
reported recently to be associated with an increased overall
patient mortality despite improvements in renal outcomes
(Dluhy and McMahon, 2008). Progression of renal disease
can be modified by inhibition of the renin-angiotensin
3 7 8   •   G eno m ic s in C l inica l P ractice
Tubulointerstitial
fibrosis
Glomerular
sclerosis
Neutrophil
Interstitial
inflammatory Lymphocyte
infiltrate
Macrophage
system with angiotensin-converting enzyme inhibitors
and/or angiotensin II receptor blockers (Brenner et  al.,
2001; Lewis et al., 1993; Lewis et al., 2001). These agents
have beneficial effects beyond blood pressure lowering,
including lowering glomerular capillary pressure and reduc-
ing proteinuria (Chiurchiu et  al., 2005). Smoking cessa-
tion, aspirin therapy, and lipid-lowering drugs are also part
of the multiple-risk-factor management to reduce cardio-
vascular events in persons with nephropathy (Fioretto
and Solini, 2005; Gaede et  al., 2003; Gaede et  al., 2008).
Dietary protein restriction (to 1g/kg body weight) can also
slow progression in proteinuric diabetic patients, although
the clinical utility of this strategy is limited (Pedrini et al.,
1996; Robertson et al., 2007).
Unfortunately, despite these interventions, kidney dis-
ease still develops in up to 30% of individuals with type 1
diabetes and up to 50% of persons with type 2 diabetes, and
progresses in approximately 30% of diabetics (Krolewski
et al., 1996; Pambianco et al., 2006; Parving et al., 2006).
Additional novel strategies are therefore urgently required
to fight the epidemic of diabetic renal disease (Wolf and
Ritz, 2003). Of interest, diabetic nephropathy is not an
inevitable complication of a prolonged duration of dia-
betes (Bain et al., 2003). Also, whilst the cumulative inci-
dence of nephropathy is greatest in those persons with the
worst glycemic control, the majority of persons with poorly
controlled diabetes still do not develop clinically apparent
renal disease (Krolewski et  al., 1996). This illustrates that
hyperglycemia is necessary but not sufficient for the devel-
opment of nephropathy. These clinical paradoxes have led
to renewed interest in determining patients’ genetic suscep-
tibility to diabetic nephropathy.
PAT H O P H Y S I O L O GY
Mesangial expansion and tubulointerstitial fibrosis are two
important renal structural changes associated with the pro-
gression of diabetic nephropathy. The degree of tubuloint-
erstitial fibrosis also correlates with prognosis (Gilbert and
Cooper, 1999). There is clinical and experimental evidence
for a wide range of mechanisms mediating the effects of
hyperglycemia and hypertension on the diabetic kidney
(Conway et al., 2012).
Upregulation of the renin-angiotensin system
(RAS) plays a prominent role in the pathophysiology of
nephropathy, and experimental evidence demonstrates
that treatment with angiotensin-converting enzyme
(ACE) inhibitors can reduce proteinuria, tubulointersti-
tial injury, and glomerulosclerosis (Gilbert and Cooper,
1999; Zatz et  al., 1986). Local activation of intra-renal
RAS leads to high levels of angiotensin II in diabetic kid-
neys (Anderson et al., 1993). Whilst angiotensin II medi-
ates its effects in part through alteration in systemic blood
pressure and glomerular hemodynamics, angiotensin II is
also recognized to be a potent inducer of transforming
growth factor-β(TGF-β). Overexpression of TGF-β is
associated with tissue fibrosis (Ziyadeh and Wolf, 2008).
Accumulation of extracellular matrix is a prominent
histological feature of diabetic nephropathy. This extra-
cellular matrix expansion is associated with elevated
TGF-β mRNA and protein levels in both renal glomeruli
and the tubulointerstitium in diabetic patients and ani-
mal models of diabetes (Rocco et  al., 1992; Yamamoto
et al., 1993; Yamamoto et al., 1996). ACE inhibitors can
reduce renal TGF-β production in diabetic rodent mod-
els (Benigni et al., 2006).
Persistent hyperglycemia can induce formation
of polyol compounds, as glucose is reduced to sorbi-
tol by the nicotinamide adenine dinucleotide phos-
phate (NADPH)-dependent enzyme, aldose reductase
(Dunlop, 2000). The role of this pathway in development
of microvascular diabetic complications is supported by
the efficacy of aldose reductase inhibitors such as tolres-
tat and sorbinol in animal models (Greene et  al., 1999;
G enetic s and G eno m ic s of C hronic K idney D i s ea s e   •   3 7 9
Oates and Mylari, 1999). However, toxicity of these
agents has largely prevented their therapeutic use in dia-
betic nephropathy.
Advanced glycosylation end products (AGEs) are
non-enzymatic modifications of amino acids and proteins
formed in states of chronic hyperglycemia (Brownlee,
2001). Serum levels of AGEs are increased in diabetic
nephropathy and are also found in glomeruli and tubules
in diabetic animals (Chen et  al., 2000). Receptors for
AGEs (RAGE) are upregulated in diabetic nephropa-
thy and may play a role in the epithelial to mesenchymal
transdifferentiation of renal tubular cells in diabetes.
Infusion of AGEs causes albuminuria and glomeruloscle-
rosis (Vlassara et al., 1994).
Hyperglycemia may also play a direct role in upregulat-
ing the protein kinase C (PKC) family of serine threonine
kinases involved in multiple cell signaling pathways that
regulate gene transcription (Murphy et al., 1998). The PKC
enzymes are implicated in regulation of vascular permeabil-
ity, blood flow, smooth muscle contractility, angiogenesis,
and fibrinolysis (Koya et  al., 1997). Thus hyperglycemia,
by stimulation of PKC-regulated pathways, can induce
increased production of cytokines and growth factors criti-
cal to the progression of nephropathy. Experimental models
of diabetes have shown that pharmacological inhibitors of
PKC can reduce severity of renal injury (Koya et al., 2000).
P R E D I S P O S IT I O N TO D I A B ET I C
N E P H RO PAT H Y
Several environmental risk factors are next described that
contribute to the pathogenesis of diabetic nephropathy.
Hypertension
Raised blood pressure is arguably the most important
modifiable risk factor affecting progression of nephropa-
thy (Cooper, 1998; Marshall, 2004; Ritz et al., 2001; James
et al., 2010). Patients with type 1 diabetes usually develop a
rise in blood pressure after the onset of microalbuminuria,
whereas hypertension often predates the onset of clini-
cal renal disease in persons with type 2 diabetes (Remuzzi
et  al., 2002). Familial clustering of hypertension has been
reported where there are diabetic offspring affected by
nephropathy (De Cosmo et al., 1997; Roglic et al., 1998).
Hyperglycemia
The risk of developing diabetic nephropathy is increased
by poor glycemic control (Chase et  al., 1989; Molitch
et  al., 1993). The Diabetes Control and Complications
Trial (DCCT) conclusively demonstrated that intensive
glycemic control in patients with type 1 diabetes reduced
the risk of developing nephropathy (microalbuminuria)
and the progression to early nephropathy (proteinuria) (de
Boer et  al., 2011). The effect of improved glycemic con-
trol remained after four years of follow-up, despite the fact
that the difference in glycemic control between the inten-
sive and conventional groups had begun to converge (The
Diabetes Control and Complications Trial/Epidemiology
of Diabetes Interventions and Complications [DCCT/
EDIC] Research Group, 2000). Recently the DCCT/
EDIC Research Group reported that the long-term risk of
an impaired GFR in individuals with type 1 diabetes was
significantly lower with intensive diabetes therapy early
in the course of the disease, compared to those treated
with conventional diabetes therapy (The DCCT/EDIC
Research Group, 2011). The evidence for a beneficial effect
of intensive glycemic control in retarding progression in
patients who already have established nephropathy is less
convincing. Reversal of diabetic nephropathy pathology
has been reported following successful pancreatic trans-
plantation for type 1 diabetes (Fioretto and Mauer, 2012;
Fioretto et al., 1998).
Multifactorial intervention, including improved gly-
cemic control, can also achieve impressive reductions in
the risk of nephropathy in patients with type 2 diabe-
tes. In the Steno study, the intensively treated group had
a target HbA1c of <6.5% and a target blood pressure of
<130/80 mmHg (Gaede et al., 1999; Gaede et al., 2008).
In addition, the intensively treated group received low-dose
aspirin, ACE inhibition, and lipid-lowering therapy. During
eight years of follow up, the intensive treatment achieved
a 61% reduction in nephropathy and 58% reduction in
retinopathy compared to the conventionally treated group
(Gaede et al., 2003).
Birth Weight
Low birth weight is associated with an increased risk of
cardiovascular disease and type 2 diabetes (Barker and
Bagby, 2005; Barker et al., 1993; Hales and Barker, 2001;
Phillips et al., 1994). It has been suggested that intrauter-
ine growth retardation is associated with a reduction in
nephron number in man, a hypothesis supported by animal
models (Brenner and Chertow, 1994; Luyckx and Brenner,
2005, 2010; Schreuder et  al., 2005). Some clinical stud-
ies have challenged this concept, reporting no association
between birth weight and progression of diabetic nephrop-
athy ( Jacobsen et  al., 2003). Careful autopsy studies have
3 8 0   •   G eno m ic s in C l inica l P ractice
correlated the reduced nephron number with the presence
of essential hypertension prior to death (Gross et al., 2005b;
Keller et al., 2003). It has also been reported that fetal expo-
sure to maternal type 1 diabetes is associated with renal
dysfunction in adulthood, possibly as a result of reduced
nephron numbers in offspring of diabetic mothers (Abi
Khalil et al., 2010).
Diet
Modification of dietary protein intake has been studied in
an effort to retard the progression of renal failure (Klahr
et al., 1994; Levey et al., 1999). Animal studies have shown
dietary protein-restriction reduces glomerular hyperfiltra-
tion, intraglomerular capillary hypertension, and the pro-
gression of renal disease (Molitch et al., 2003). In clinical
studies, protein-restricted diets have been associated with a
fall in proteinuria and reduction in the rate of decline of
GFR (Hansen et al., 2002; Pedrini et al., 1996). In routine
practice, however, the effects of dietary protein restriction
have been difficult to replicate. Interest has also been focused
on the influence of dietary fat intake on the development
of nephropathy and atherosclerosis. Hypercholesterolemia
has been reported to be associated with more rapid fall in
GFR (Breyer et  al., 1996). Treatment with statin therapy
reduced cardiovascular events in type 2 diabetics with
microalbuminuria and proteinuria (Colhoun et al., 2004),
but paradoxically, statin therapy did not reduce cardio-
vascular mortality in dialysis-dependent diabetic patients
(Wanner et al., 2005). Potentially atherogenic lipoprotein
profiles are associated with renal dysfunction in type 1 dia-
betes ( Jenkins et al., 2003). Individuals with type 2 diabe-
tes, chronic kidney disease, and the most “atherogenic” lipid
profiles appear to derive clinical benefits (in terms of fewer
cardiovascular events, less retinopathy, and reduction in
albuminuria) from combined statin and fibrate treatment
(Rosenblit, 2012).
Smoking
Cigarette smoking is associated with development of
nephropathy in patients with either type 1 or type 2 diabe-
tes (Nilsson et al., 2004; Olivarius Nde et al., 1993; Telmer
et  al., 1984). Smoking increases the risk of cardiovascular
death, particularly in diabetic patients with proteinuria
(Borch-Johnsen et  al., 1987; Moy et  al., 1990). There is
also evidence that smoking is associated with more rapid
progression of renal disease (Orth et al., 1997; Ritz et al.,
2000). This may be induced by smoking-related changes
to glomerular structure and function (Baggio et al., 2002).
Smoking-cessation is still an important strategy in reducing
the risk of developing nephropathy in persons with diabetes
(Tonstad, 2009).
The Impact of Diabetes Duration on the Risk
of Nephropathy
If renal disease were directly due to persistent hypergly-
cemia, then a linear relationship between cumulative
incidence and duration of diabetes would be expected.
Background diabetic retinopathy is present in almost all
diabetic individuals after 50  years’ duration, but in con-
trast, the cumulative incidence of nephropathy in persons
of European ancestry plateaus at 30% after approximately
25 years’ duration (Doria et al., 1995).
Ethnicity
The prevalence of diabetic nephropathy varies among
ethnic groups, being highest amongst Asian, African
American, and Native American populations (Cowie et al.,
1989; Satko et al., 2002). After 25 years’ diabetes duration,
the cumulative risk for diabetic nephropathy in persons of
European ancestry is approximately 30%, compared to 80%
in Pima Indians (Ballard et al., 1988; Nelson et al., 1993).
FA M I L I A L C LUS T E R I N G O F R E NA L
DISEASE
Familial aggregation of renal disease is not explained by
the known environmental risk factors, and this finding
provides strong support for an inherited genetic suscepti-
bility to nephropathy. The study of familial clustering can
be achieved by comparing the incidence of renal disease
in families where a proband has both diabetes and renal
disease, to the incidence of renal disease in families with
diabetic probands without obvious renal disease. Such
families, with multiple diabetic offspring, are more difficult
to recruit than individual diabetic patients, with or with-
out nephropathy. The likelihood of developing diabetic
nephropathy is higher when the individual with diabetes
has a parental history of cardiovascular disease (De Cosmo
et al., 1997). In family-based studies, estimates of the herita-
bility of albuminuria and glomerular filtration rate (GFR)
have ranged from 35–75% (Placha et al., 2005; O’Seaghdha
and Fox, 2011).
Familial clustering of nephropathy in white individu-
als with type 1 diabetes was first reported by Seaquist
and colleagues (Seaquist et  al., 1989)  (Figure 24.7). They
found that renal disease was present in 83% of the diabetic
G enetic s and G eno m ic s of C hronic K idney D i s ea s e   •   3 8 1
90
Prevalance of nephropathy in siblings (%)
80 Proband with
nephropathy
70
Proband without
60
nephropathy
50
40
30
20
10
0
Minnesota Steno Joslin
Figure 24.7 
Familial clustering of diabetic nephropathy in type 1 diabetes.
Data are from Minnesota, USA, (Seaquist et al., 1989); Steno Clinic,
Copenhagen, Denmark (Borch-Johnsen et al., 1992); and Joslin
Diabetes Center, Boston, USA (Quinn et al., 1996).
siblings of probands who had received kidney transplants.
In contrast, renal disease affected only 17% of diabetic sib-
lings of probands without nephropathy (p < 0.001). Both
groups had similar glycemic control. Confirmation of this
familial clustering phenomenon in type 1 diabetes melli-
tus in other studies demonstrates that, despite similar gly-
cemic control, the prevalence of nephropathy was greater
in diabetic siblings of a proband with diabetic renal disease
(Borch-Johnsen et al., 1992; Quinn et al., 1996; Harjutsalo
et al., 2004).
These observations on familial clustering of diabetic
nephropathy have been extended to type 2 diabetes mellitus
and are evident in different ethnic groups, including Native
Americans, African Americans, and Asians (Pettitt et  al.,
1990; Satko and Freedman, 2005). Of interest, the risk of
developing renal failure is also increased in first-degree rela-
tives of probands with ESRD due to etiologies other than
diabetes, suggesting common genetic risk factors for pro-
gressive kidney disease (Freedman et al., 2005; O’Dea et al.,
1998).
Careful analysis of renal biopsy material has demon-
strated apparent inherited differences in diabetes-induced
glomerular pathology (Fioretto et  al., 1999). These
insights into inherited differences in glomerular structure
are reinforced by assessments of the heritability of albu-
minuria and determinants of renal function such as GFR
(Forsblom et  al., 1999; Hunter et  al., 2002; Langefeld
et  al., 2004). An inherited predisposition to progressive
renal disease may exist; for example, in individuals with
reduced nephron number or subtle defects in glomeru-
lar function resulting in albuminuria. The phenotype of
progressive renal failure only becomes apparent if the sus-
ceptible individual is exposed to additional injury from
hypertension or hyperglycemia.
 FI N D I N G T H E G E N E S R E S P O NS I B L E
F O R D I A B ET I C N E P H RO PAT H Y
The ultimate goal of genetic studies of diabetic nephropa-
thy is the identification and characterization of the gene
variants conferring susceptibility to progressive renal fail-
ure. Investigators need to bear in mind the multifactorial
etiology of nephropathy and the likelihood that multiple
common gene variants, each conferring a small individual
relative risk of this complication, may be responsible (Rich,
2006). Alternatively, multiple rare variants (minor allele fre-
quency 1–5%) with large effect sizes may be important in
individual susceptibility to diabetic nephropathy (Bodmer
and Bonilla, 2008; Schork et al., 2009; Marian, 2012).
The molecular methods for detecting variants associ-
ated with complex disease are candidate-gene analyses and
whole-genome scans, and both approaches have been uti-
lized for diabetic nephropathy. The studies are designed to
test either association between the frequency of a specific
genetic marker in different populations, or linkage to assess
the inheritance of a particular genetic locus within families.
C A N D I DAT E - G E N E A NA LYS I S F O R
D I A B ET I C N E P H RO PAT H Y
In candidate-gene analysis, candidate genes for diabetic
nephropathy are selected on the basis of an a priori hypoth-
esis that the protein products of such candidate genes are
involved in the pathogenesis of the disease (Conway and
Maxwell, 2009; McKnight et al., 2010). It is possible to pri-
oritize the search for candidate genes in a rational manner
based on knowledge of the physiological and biochemical
pathways implicated in diabetic nephropathy. For instance,
genes involved in the renin-angiotensin system regulating
blood pressure or enzymes regulating glucose metabolism
are studied. Problems with this approach include the rela-
tively the small number of genes studied to date (<1% of the
known genome) and the fact that candidates can only be
selected based on current, and therefore incomplete, under-
standing of disease pathogenesis (Thomas et al., 2012).
Rational selection of candidate genes can also be aug-
mented by data derived from subtractive hybridization
experiments and DNA microarrays that may detect upreg-
ulated or downregulated genes that had not been previ-
ously implicated in disease pathogenesis (Clarkson et  al.,
2002; Connolly et al., 2003; Holmes et al., 1997; Murphy
et  al., 1999). These gene-expression profiles may be taken
from direct analysis of mRNA derived from microdis-
sected renal tissue from patients with diabetic nephropa-
thy (Baelde et al., 2004). Once a candidate gene has been
selected, the sequence variation within the gene can be
3 8 2   •   G eno m ic s in C l inica l P ractice
tested for association with the disease, using case-control or
family-based study designs.
Case-Control Studies
Case-control studies are arguably the simplest design and
a powerful method to detect association between a gene
variant and disease. The frequency of a marker (sequence
variant) in a population with a disease like nephropathy
(cases) is compared with that of a matched population
without disease (controls). A  2 x 2 contingency table
and an X2 test is used to determine if there is a statisti-
cal difference between the observed and expected marker
frequency.
Case-control studies are relatively easy to undertake
but are bedeviled by numerous potential sources of errors.
Replication of studies where a positive association has been
found has been notoriously difficult (Hirschhorn et  al.,
2002; Conway and Maxwell, 2009; Currie et  al., 2008).
One review of 166 candidate-gene association studies that
had been replicated at least three times found only six asso-
ciations that remained significant (Hirschhorn et al., 2002).
Nevertheless, those that did remain positive after multiple
replication studies were clinically relevant and included the
associations of variants of the ApoE gene with Alzheimer’s
disease, CTLA4 with Grave’s disease, and factor V Leiden
with deep-venous thrombosis (Hirschhorn et  al., 2002).
Further examination of genetic association studies in a series
of meta-analyses has provided grounds for optimism that
positive associations can be replicated if efforts are made
to avoid smaller, underpowered studies (Ioannidis et  al.,
2003; Lohmueller et  al., 2003). A  recent meta-analysis of
671 genetic association studies for diabetic nephropathy
identified 21 replicated genetic variants that remained sig-
nificant (Mooyaart et al., 2011). These genetic variants were
in or near the following genes:  ACE, AKR1B1, APOC1,
APOE, EPO, NOS3, HSPG2, VEGFA, FRMDS, CARS,
UNC13B, CPVL and CHN2, and GREM1, plus four vari-
ants not near genes. The odds ratios of associated genetic
variants ranged from 0.48 to 1.70.
Potential pitfalls in the interpretation of case-control
studies performed to date include incorrect phenotype
definition, small sample size, population stratification from
ethnic admixture, over-interpretation of results including
subgroup analysis, multiple testing, the effect of duration of
exposure to risk factors, and publication bias. Sufficient cases
and controls must be recruited in order to minimize the
risk of identifying false associations that are due to chance
alone (type 1 error); or conversely, of failing to detect a true
association between a variant and a disease (type 2 error).
It is also important to remember that statistically significant
genetic association does not prove causation.
Family-Based Investigations of Genetic Susceptibility
Family-based studies, where family members are used as
controls, have been employed to avoid the problem of pop-
ulation stratification in case-control studies. The relatives
used may be siblings or parents and are ethnically matched,
thereby eliminating the risk of stratification. In the trans-
mission disequilibrium test (TDT), first pioneered by
Spielman and colleagues to study the genetics of type 1
diabetes, affected offspring and their parents are genotyped
(Spielman et al., 1993). The test compares the actual num-
ber of alleles transmitted from heterozygous parents to
offspring to the number expected to be transmitted (using
the MacNemar X2 test). If the allele is transmitted more fre-
quently than the expected 50:50 ratio, then it is likely to be
associated with development of disease. The power of these
studies is reduced, however, since parents may be homozy-
gous for the marker allele, and in that case, no information
regarding transmission can be determined.
Unfortunately, family-based studies using a TDT analy-
sis have been of very limited value in later-onset disease such
as nephropathy in type 2 diabetics (where parents are likely
to be deceased), and this design has also been difficult to use
for nephropathy in type 1 diabetics, as there is familial clus-
tering of cardiovascular disease (and premature death) in
the parents of type 1 diabetic offspring with nephropathy.
One further option, avoiding the requirement to recruit
parents, is the use of sibling pairs in a modification known
as the sib TDT analysis (Spielman and Ewens, 1998).
The criteria for ideal genetic association studies include
large sample sizes, small p values, associations that make
biological sense, alleles (sequence variants) that alter the
gene product in a physiological way, an initial study that
is independently replicated in a separate population and
(where possible) the association is seen in both case-control
and family-based studies (Anonymous, 1999). Efforts have
been made to improve the consistency and quality of pub-
lished genetic association studies by encouraging authors
to address issues such as population stratification, genotyp-
ing errors, haplotype variation, Hardy-Weinberg equilib-
rium, replication, rationale for choice of genes and variants,
and phenotyping of recruited individuals (Little et  al.,
2009). Several publicly accessible databases now collate
hundreds of genetic epidemiology studies relevant to dia-
betic nephropathy, such as the HuGEnavigator (Yu et  al.,
2010) and the Centralized Online Renal Genetics Initiative
(Currie et al., 2008).
G enetic s and G eno m ic s of C hronic K idney D i s ea s e   •   3 8 3
G E N O M E -WI D E S C A N S F O R
D I A B ET I C N E P H RO PAT H Y
A complementary approach to identify candidate genes for
complex disease such as diabetic nephropathy is the use of
genome-wide scans. A major advantage of this approach is that
no detailed understanding of the pathophysiology of the dis-
ease of interest is required, and there is no prior hypothesis as
to which gene or genes are implicated in causation. Previously,
these screens were based on the principle of linkage disequilib-
rium; that is, where a genetic marker and a disease susceptibil-
ity gene are in close proximity, the marker will segregate with
the disease more often than expected by chance alone. Further
detailed mapping of the chromosomal segment adjacent to
the linked marker should in theory reveal the identity of a dis-
ease susceptibility gene. The linkage method employs several
hundred genetic markers (microsatellites consisting of vari-
able sequence lengths of repeated CA base pairs). In practice,
this approach has been much more successful in identifying
monogenic renal diseases such as X-linked Alport syndrome
(Barker et al., 1990) or ADPKD (Reeders et al., 1985) where
the pattern of inheritance is known.
There have been relatively few linkage studies in diabetic
nephropathy, reflecting the practical difficulty in recruiting
large families with multiple affected individual members. For
instance, in a study of familial clustering of nephropathy in
siblings with type 1 diabetes, Seaquist and colleagues identi-
fied a cohort of 696 patients with diabetic nephropathy who
had received a renal transplant. Of these 696 patients, only
113 had a type 1 diabetic sibling, and only 26 of these siblings
could be enrolled in the study (Seaquist et al., 1989).
Genome-wide linkage studies have been performed for
nephropathy in persons with type 2 diabetes. The earliest
study, performed in Pima Indians, employed 98 sib-pairs
concordant for nephropathy and reported four chromo-
somal regions linked to nephropathy:  3q26.9, 7q35, 9q,
and 20q (Imperatore et al., 1998). The strongest linkage was
with chromosome 7q35. One issue in using sib-pairs con-
cordant for both type 2 diabetes and nephropathy is unrav-
elling whether the linkage is with diabetes or with renal
disease. To resolve this problem, a further genome-wide
scan in Pima Indians showed no evidence of linkage to
type 2 diabetes on chromosome 7q35, suggesting that this
region may indeed harbor a susceptibility gene for nephrop-
athy (Hanson et  al., 1998). The Family Investigation of
Nephropathy and Diabetes (FIND) study recruited dia-
betic sibling pairs concordant and discordant for diabetic
nephropathy (Iyengar et al., 2007). The strongest evidence
of linkage to diabetic nephropathy was on chromosomes
7q21.3, 10p15.3, 14q23.1, and 18q22.3. True linkage has
proven hard to find in genome-wide scans of common com-
plex disease (Altmüller et al., 2001). Overall, most linkage
data for diabetic nephropathy does not reach genome-wide
significance, and only a few regions (3q, 7q, 18q) have been
identified and replicated in multiple studies.
An alternative approach to determining the genetic sus-
ceptibility to diabetic nephropathy is to study an intermedi-
ate trait, such as urine protein excretion. Segregation analysis
has been employed in Caucasian families to model inheri-
tance patterns of proteinuria (urinary albumin/creatinine
ratio or ACR) in type 2 diabetes (Fogarty et  al., 2000b).
Sib-pairs, discordant for type 2 diabetes, had urinary ACR
tests performed, and this quantitative trait was modelled
with age, gender, and duration of diabetes as covariables.
Urinary ACR was heritable and genetically correlated to
blood pressure (Fogarty et al., 2000a). A similar segregation
analysis for urine protein excretion has been performed in
Pima Indians with type 2 diabetes (Imperatore et al., 2000).
These studies are both consistent with a Mendelian model
of inheritance, albeit that the heritability was 27% and pro-
teinuria will be influenced by multiple gene variants and
environmental factors (Fogarty et al., 2000b).
Previously, investigators had only these two broad
options (candidate-gene studies or linkage analysis) to
determine genetic susceptibility to common disease such
as diabetic nephropathy. Direct-association studies testing
candidate-gene variants for association with disease were
undertaken based on an a priori hypothesis. The limitations
of this approach included the limited number of candidate
genes that have been systematically examined for any dis-
ease. The alternative approach of genome scans to identify
chromosomal regions linked to common disease has proved
less successful than for monogenic disorders, with some
notable exceptions such as type 1 diabetes (Nistico et  al.,
1996) and inflammatory bowel disease (Hugot et al., 2001).
Several additional approaches are now technically
feasible and potentially affordable to identify genetic sus-
ceptibility to diabetic nephropathy. These are indirect
genome screens for association with nephropathy (using
an extended panel of genetic markers) and direct genome
screens employing panels of SNPs to directly assess thou-
sands of candidate genes simultaneously.
G E N O M E -WI D E A S S O C I AT I O N S T U D I E S
TO I D E N T I F Y C H RO MO S O M A L R EG I O N S
A N D C A N D I DAT E G E N E S F O R D I A B ET I C
N E P H RO PAT H Y
There has been remarkable progress in the development
of resources to enable genome-wide association studies
3 8 4   •   G eno m ic s in C l inica l P ractice
(GWAS), including the identification of more than 40 mil-
lion SNPs and the further genotyping of 3.8  million of
these SNPs in individual DNA samples from different
ethnic groups within the HapMap Project (Consortium,
2003). This allowed more efficient assessment of genetic
architecture based on haplotype blocks with the human
genome (Barrett et  al., 2005). These initiatives, coupled
with technological advances in genotyping and reduction
in DNA-sequencing costs, have enabled more affordable
GWAS approaches for discovery of genetic susceptibility to
common complex diseases. Furthermore, sophisticated ana-
lytical tools have been designed to quality-control and mine
large datasets (Marchini et al., 2005; Marchini et al., 2007;
Purcell et al., 2007).
A number of individual research centers have col-
laborated to develop much larger collections of cases and
controls with rigorous definitions of clinical phenotypes
relevant to diabetic nephropathy. These larger multicenter
collections include FIND, DCCT/EDIC, Genetics of
Kidneys in Diabetes [GoKinD] US, Genetics of Kidneys in
Diabetes [GoKinD] UK, and European rational approach
for the genetics of diabetic complications [EuraGedic]
(Iyengar et  al., 2007; Group, 1999; Mueller et  al., 2006;
Currie et al., 2008; Tarnow et al., 2008).
To identify chromosomal regions that harbor possible
candidate genes for diabetic nephropathy, it has been pos-
sible to perform a low-resolution, genome-wide, micro-
satellite association screen. This method employs several
thousand fluorescently labelled microsatellite markers
(compared to the usual 300–400 microsatellite markers for
a linkage screen) and a pooling strategy to minimize the
costs of typing large numbers of samples with individual
genetic markers. Careful analysis of separate pools of DNA
from cases and controls generates a series of allele frequen-
cies for the microsatellite markers. This strategy was success-
ful in identifying regions of chromosome 10 significantly
associated with diabetic nephropathy in a cohort of Irish
patients with type 1 diabetes (McKnight et al., 2006). The
data from these indirect screens allow investigators to select
candidate genes, identified close to the associated microsat-
ellite markers, for further study.
Significantly larger marker sets of SNPs have been used,
as an alternative to limited microsatellite panels, in an indi-
rect genome-wide association screen. Using information
from the HapMap project (Altshuler et al., 2005; Phimister,
2005), it is possible to reduce the size of the marker set
required by employing haplotype-tagged SNPs (htSNPs).
Early proof of principle for this strategy was reported for
type 1 diabetes ( Johnson et  al., 2001; Lowe et  al., 2004).
A  non-synonymous SNP (nsSNP) scan is an alternative
method of directly detecting gene variants associated with
disease. An nsSNP may influence phenotype by altering
gene expression or protein function (Savage et al., 2008).
D I R E C T G E N O M E -W I D E
A SS O C I AT I O N S T U D I E S
Direct genome-wide association studies employ either
a large panel of unselected SNPs randomly distributed
throughout the genome, or a more focused set of markers
utilizing non-synonymous SNPs, which may potentially
alter gene expression or protein structure.
In any study using commercially available arrays, a very
large number of SNPs will be genotyped, and it is therefore
likely that many putative disease-susceptibility variants will
be detected. Most of the direct associations between SNPs
and disease will be by chance alone (for example, one would
expect 50,000 SNPs in a scan employing 1,000,000 SNPs to
be associated by chance alone at the 5% confidence level).
Therefore, a major challenge is to rationalize the output
from genome-wide association studies. One possibility is
to apply a correction factor for the number of SNPs geno-
typed. For example, in a genome screen with 1,000,000
markers, adoption of a significance threshold of p = 5 x 10–8
for individual SNPs would result in a genome-wide false
positive rate of 1 in 20, the accepted level of confidence for
an individual genotype. However, this correction may be
too rigorous and therefore miss some true associations that
could be tested in a replication study.
One approach for a GWAS is to adopt a two-stage
screening strategy. Markers showing association in an initial
screen (the discovery cohort) could be reassessed in an inde-
pendent, larger sample (the replication cohort). The thresh-
old of significance for associated markers in the first screen
would purposely be set low so that it remained sufficiently
powerful to detect loci with a relatively modest effect on
disease, despite employing only a fraction of the total sam-
ples available. The potential for detecting susceptibility
variants in this initial screen may be enhanced by including
only extreme phenotypes: cases with onset of nephropathy
after a relatively short duration of diabetes or despite good
glycemic control; and conversely, controls that have not
developed nephropathy despite a long duration of diabe-
tes or poor glycemic control. The majority of associations
detected in the first screen will be false positives; therefore,
a much more stringent significance threshold is required
for the replication study. Hence a subset of patient samples
would be screened using the complete marker set, with
only a fraction of these markers being genotyped in a larger
G enetic s and G eno m ic s of C hronic K idney D i s ea s e   •   3 8 5
sample, thereby improving efficiency without a reduction
in power, while reducing the risk of detecting false-positive
associations.
A screen that employed a two-stage strategy similar to
that described has been undertaken in a Japanese popula-
tion with type 2 diabetes (Shimazaki et  al., 2005). Over
80,000 SNP markers were genotyped, and SNPs in two
candidate genes were associated with diabetic nephropathy.
The genes identified were a sodium-chloride co-transporter
known to be mutated in Gitelman syndrome (Tanaka
et al., 2003) and the engulfment and cell motility (ELMO
1) gene, which is upregulated in high extracellular glucose
concentrations and promotes accumulation of matrix pro-
teins (Shimazaki et al., 2005). A GWAS employing 100,000
SNPs in Pima Indians (105 cases and 102 controls) using
pooled DNA samples highlighted association between the
plasmacytoma variant translocation gene (PVT1) and dia-
betic nephropathy (Hanson et al., 2007).
One of the largest published GWAS for diabetic
nephropathy was published by the U.S. Genetics of Kidneys
in Diabetes (GoKinD) group with attempted confirma-
tion of associated SNPs in the DCCT/EDIC collection
(Pezzolesi et al., 2009). The U.S. GoKinD study employed
820 cases with nephropathy and type 1 diabetes (284 with
proteinuria and 536 with end-stage renal disease) and 885
controls. Despite its size, this GWAS only identified 13 SNPs
associated with diabetic nephropathy with P < 1 x 10–5, and
none reached the conventional threshold for genome-wide
significance. The strongest association was at the FRMD3
(4.1 protein ezrin, radixin, moesin [FERM] domain con-
taining 3) locus (odds ratio [OR] = 1.45, P = 5.0 x 10–7).
Arguably, the sample size was simply too small to detect
alleles with effect sizes now being reported in GWAS of
common complex diseases. International efforts are now
being made to integrate available case-control collections
into much larger datasets to optimize the investigation
of the genetics of diabetic nephropathy (Hirschhorn and
Gajdos, 2011; Wheeler and Barroso, 2011).
C O N C LU S I O N
The Human Genome Project and the HapMap initiative
accelerated the technological advances necessary to allow
dissection of the molecular basis of disease. Well-powered
genome-wide association studies with high-density SNP
arrays have recently identified new signals for renal pheno-
types such as IgA nephropathy and membranous nephropa-
thy (Feehally et al., 2010; Stanescu et al., 2011). It has proven
more challenging to determine the genetic architecture of
diabetic nephropathy. The current focus on rare gene vari-
ants and epigenetic modifications of the genome may dis-
cover new insights into the genetic susceptibility to this
common kidney disease.
AC K N OW L E D G E M E N T
We thank Pauline Whittaker for secretarial assistance.
REFERENCES
Abi Khalil, C., Travert, F., Fetita, S., et al. (2010). Fetal exposure to mater-
nal type 1 diabetes is associated with renal dysfunction at adult age.
Diabetes 59, 2631–2636.
Adler, S. (2004). Diabetic nephropathy: Linking histology, cell biology,
and genetics. Kidney Int 66, 2095–2106.
Ahrabi, A.K., Terryn, S., Valenti, G., et al. (2007). PKD1 haploinsuffi-
ciency causes a syndrome of inappropriate antidiuresis in mice. J Am
Soc Nephrol 18, 1740–1753.
Alicic, R.Z., and Tuttle, K.R. (2010). Management of the diabetic patient
with advanced chronic kidney disease. Semin Dial 23, 140–147.
Altmüller, J., Palmer, L.J., Fischer, G., Scherb, H., Wjst, M. (2001).
Genomewide scans of complex human diseases: true linkage is hard
to find. Am J Hum Genet 69, 936–950.
Altshuler, D., Brooks, L.D., Chakravarti, A., Collins, F.S., Daly, M.J.,
and Donnelly, P. (2005). A haplotype map of the human genome.
Nature 437, 1299–1320.
Anderson, S., Jung, F.F., and Ingelfinger, J.R. (1993). Renal
renin-angiotensin system in diabetes:  functional, immunohisto-
chemical, and molecular biological correlations. Am J Physiol 265,
F477–F486.
Anonymous (1999). Freely associating. Nat Genet 22, 1–2.
Baelde, H.J., Eikmans, M., Doran, P.P., Lappin, D.W., de Heer, E., and
Bruijn, J.A. (2004). Gene expression profiling in glomeruli from
human kidneys with diabetic nephropathy. Am J Kidney Dis 43,
636–650.
Baggio, B., Budakovic, A., Dalla Vestra, M., Saller, A., Bruseghin, M.,
and Fioretto, P. (2002). Effects of cigarette smoking on glomerular
structure and function in type 2 diabetic patients. J Am Soc Nephrol
13, 2730–2736.
Ballard, D.J., Humphrey, L.L., Melton, L.J., et al. (1988). Epidemiology of
persistent proteinuria in type II diabetes mellitus. Population-based
study in Rochester, Minnesota. Diabetes 37, 405–412.
Barker, D.F., Hostikka, S.L., Zhou, J., et  al. (1990). Identification of
mutations in the COL4A5 collagen gene in Alport syndrome.
Science 248, 1224–1227.
Barker, D.J., and Bagby, S.P. (2005). Developmental antecedents of car-
diovascular disease: a historical perspective. J Am Soc Nephrol 16,
2537–2544.
Barker, D.J., Gluckman, P.D., Godfrey, K.M., Harding, J.E., Owens, J.A.,
and Robinson, J.S. (1993). Fetal nutrition and cardiovascular dis-
ease in adult life. Lancet 341, 938–941.
Barnett, A.H. (2004). Treating to goal:  challenges of current manage-
ment. Eur J Endocrinol 151 Suppl 2, T3–7; discussion T29–30.
Barrett, J.C., Fry, B., Maller, J., and Daly, M.J. (2005). Haploview: analy-
sis and visualization of LD and haplotype maps. Bioinformatics 21,
263–265.
Benigni, A., Zoja, C., Campana, M., et  al. (2006). Beneficial effect of
TGFbeta antagonism in treating diabetic nephropathy depends on
when treatment is started. Nephron Exp Nephrol 104, e158–168.
3 8 6   •   G eno m ic s in C l inica l P ractice
Bergmann, C., von Bothmer, J., Ortiz Bruchle, N., et  al. (2011).
Mutations in multiple PKD genes may explain early and severe
polycystic kidney disease. J Am Soc Nephrol: JASN 22, 2047–2056.
Bihoreau, M.T., Megel, N., Brown, J.H., et al. (2002). Characterization of
a major modifier locus for polycystic kidney disease (Modpkdr1) in
the Han:SPRD(cy/+) rat in a region conserved with a mouse modi-
fier locus for Alport syndrome. Hum Mol Genet 11, 2165–2173.
Blakemore, A.I., Watson, P.F., Weetman, A.P., and Duff, G.W. (1995).
Association of Graves’ disease with an allele of the interleukin-1
receptor antagonist gene. J Clin Endocrinol Metab 80, 111–115.
Bodmer, W., and Bonilla, C. (2008). Common and rare variants in
multifactorial susceptibility to common diseases. Nat Genet 40,
695–701.
Borch-Johnsen, K., Nissen, H., Henriksen, E., et  al. (1987). The natu-
ral history of insulin-dependent diabetes mellitus in Denmark:
1. Long-term survival with and without late diabetic complications.
Diabet Med 4, 201–210.
Borch-Johnsen, K., Norgaard, K., Hommel, E., et al. (1992). Is diabetic
nephropathy an inherited complication? Kidney Int 41, 719–722.
Brasier, J.L., and Henske, E.P. (1997). Loss of the polycystic kidney dis-
ease (PKD1) region of chromosome 16p13 in renal cyst cells sup-
ports a loss-of-function model for cyst pathogenesis. J Clin Invest
99, 1–6.
Brenner, B.M., and Chertow, G.M. (1994). Congenital oligonephropa-
thy and the etiology of adult hypertension and progressive renal
injury. Am J Kidney Dis 23, 171–175.
Brenner, B.M., Cooper, M.E., de Zeeuw, D., et al. (2001). Effects of losar-
tan on renal and cardiovascular outcomes in patients with type 2
diabetes and nephropathy. N Engl J Med 345, 861–869.
Brenner, B.M., and Chertow, G.M. (1994). Congenital oligonephropa-
thy and the etiology of adult hypertension and progressive renal
injury. Am J Kidney Dis 23, 171–175.
Breyer, J.A., Bain, R.P., Evans, J.K., et al. (1996). Predictors of the pro-
gression of renal insufficiency in patients with insulin-dependent
diabetes and overt diabetic nephropathy. The Collaborative Study
Group. Kidney Int 50, 1651–1658.
Brook-Carter, P.T., Peral, B., Ward, C.J., et  al. (1994). Deletion of the
TSC2 and PKD1 genes associated with severe infantile polycys-
tic kidney disease—a contiguous gene syndrome. Nat Genet 8,
328–332.
Brownlee, M. (2001). Biochemistry and molecular cell biology of dia-
betic complications. Nature 414, 813–820.
Cairns, H.W.B. (1925). Heredity in polycystic disease of the kidneys.
Q J Med 18, 359–392.
Chang, H.R., Cheng, C.H., Shu, K.H., Chen, C.H., Lian, J.D., and
Wu, M.Y. (2003). Study of the polymorphism of angiotensinogen,
angiotensin-converting enzyme and angiotensin receptor in type II
diabetes with end-stage renal disease in Taiwan. J Chin Med Assoc
66, 51–56.
Chang, M.Y., and Ong, A.C. (2011). Mechanism-based therapeutics for
autosomal dominant polycystic kidney disease: recent progress and
future prospects. Nephron Clin Pract 120, c25-c35.
Chase, H.P., Jackson, W.E., Hoops, S.L., Cockerham, R.S., Archer,
P.G., and O’Brien, D. (1989). Glucose control and the renal and
retinal complications of insulin-dependent diabetes. JAMA 261,
1155–1160.
Chen, S., Cohen, M.P., and Ziyadeh, F.N. (2000). Amadori-glycated
albumin in diabetic nephropathy: pathophysiological connections.
Kidney Int Suppl 77, S40–44.
Chen, X.Z., Vassilev, P.M., Basora, N., et  al. (1999). Polycystin-L is a
calcium-regulated cation channel permeable to calcium ions. Nature
401, 383–386.
Chen, Y., Zhang, Z., Lv, X.Y., et al. (2008). Expression of Pkd2l2 in testis
is implicated in spermatogenesis. Biol Pharm Bull 31, 1496–1500.
Chiurchiu, C., Remuzzi, G., and Ruggenenti, P. (2005). Angiotensin-
converting enzyme inhibition and renal protection in nondiabetic
 patients: the data of the meta-analyses. J Am Soc Nephrol 16 Suppl
1, S58–63.
Clarkson, M.R., Murphy, M., Gupta, S., et  al. (2002). High
glucose-altered gene expression in mesangial cells. Actin-regulatory
protein gene expression is triggered by oxidative stress and cytoskel-
etal disassembly. J Biol Chem 277, 9707–9712.
Colhoun, H.M., Betteridge, D.J., Durrington, P.N., et  al. (2004).
Primary prevention of cardiovascular disease with atorvastatin in
type 2 diabetes in the Collaborative Atorvastatin Diabetes Study
(CARDS):  multicentre randomised placebo-controlled trial.
Lancet 364, 685–696.
Collins, A.J., Kasiske, B., Herzog, C., et  al. (2003). Excerpts from the
United States Renal Data System 2003 Annual Data Report: atlas
of end-stage renal disease in the United States. Am J Kidney Dis 42,
A5–7.
Connolly, S.B., Sadlier, D., Kieran, N.E., Doran, P., and Brady, H.R.
(2003). Transcriptome profiling and the pathogenesis of diabetic
complications. J Am Soc Nephrol 14, S279–283.
Connor, A., Lunt, P.W., Dolling, C., et al. (2008). Mosaicism in autoso-
mal dominant polycystic kidney disease revealed by genetic testing
to enable living related renal transplantation. Am J Transplant 8,
232–237.
Consortium, T.I.H. (2003). The International HapMap Project. Nature
426, 789–796.
Conway, B.R., and Maxwell, A.P. (2009). Genetics of diabetic nephropa-
thy:  are there clues to the understanding of common kidney dis-
eases? Nephron Clin Pract 112, c213–221.
Conway, B.R., Rennie, J., Bailey, M.A., et  al. (2012). Hyperglycemia
and renin-dependent hypertension synergize to model diabetic
nephropathy. J Am Soc Nephrol: JASN 23, 405–411.
Cooper, M.E. (1998). Pathogenesis, prevention, and treatment of dia-
betic nephropathy. Lancet 352, 213–219.
Cowie, C.C., Port, F.K., Wolfe, R.A., Savage, P.J., Moll, P.P., and
Hawthorne, V.M. (1989). Disparities in incidence of diabetic
end-stage renal disease according to race and type of diabetes.
N Engl J Med 321, 1074–1079.
Currie, D., Maxwell, A.P., Sadlier, D., and McKnight, A.J. (2008).
Investigation of adducin 2 (beta) DNA polymorphisms in genetic
predisposition to diabetic nephropathy in Type 1 diabetes. Diabet
Med 25, 1001–1005.
de Boer, I.H., Sun, W., Cleary, P.A., et al. (2011). Intensive diabetes ther-
apy and glomerular filtration rate in type 1 diabetes. N Engl J Med
365, 2366–2376.
De Cosmo, S., Bacci, S., Piras, G.P., et al. (1997). High prevalence of risk
factors for cardiovascular disease in parents of IDDM patients with
albuminuria. Diabetologia 40, 1191–1196.
Deckert, T., Yokoyama, H., Mathiesen, E., et al. (1996). Cohort study
of predictive value of urinary albumin excretion for atherosclerotic
vascular disease in patients with insulin dependent diabetes. BMJ
312, 871–874.
Diabetes Control and Complications Trial Research Group. (1993). The
effect of intensive treatment of diabetes on the development and
progression of long-term complications in insulin-dependent diabe-
tes mellitus. N Engl J Med 329, 977–986.
Dinneen, S.F., and Gerstein, H.C. (1997). The association of microal-
buminuria and mortality in non-insulin-dependent diabetes melli-
tus. A systematic overview of the literature. Arch Intern Med 157,
1413–1418.
Dluhy, R.G., and McMahon, G.T. (2008). Intensive glycemic con-
trol in the ACCORD and ADVANCE trials. N Engl J Med 358,
2630–2633.
Doria, A., Warram, J.H., and Krolewski, A.S. (1995). Genetic suscepti-
bility to nephropathy in insulin-dependent diabetes: from epidemi-
ology to molecular genetics. Diabetes Metab Rev 11, 287–314.
Dunlop, M. (2000). Aldose reductase and the role of the polyol pathway
in diabetic nephropathy. Kidney Int Suppl 77, S3–12.
G enetic s and G eno m ic s of C hronic K idney D i s ea s e   •   3 8 7
Fain, P.R., McFann, K.K., Taylor, M.R., et al. (2005). Modifier genes play
a significant role in the phenotypic expression of PKD1. Kidney Int
67, 1256–1267.
Feehally, J., Farrall, M., Boland, A., et al. (2010). HLA has strongest asso-
ciation with IgA nephropathy in genome-wide analysis. J Am Soc
Nephrol: JASN 21, 1791–1797.
Field, S., Riley, K.L., Grimes, D.T., et  al. (2011). Pkd1l1 estab-
lishes left-right asymmetry and physically interacts with Pkd2.
Development 138, 1131–1142.
Fioretto, P., and Mauer, M. (2007). Histopathology of diabetic nephrop-
athy. Sem Nephrol 27, 195–207.
Fioretto, P., and Mauer, M. (2012). Reversal of diabetic nephropathy:
lessons from pancreas transplantation. J Nephrol 25, 13–18.
Fioretto, P., and Solini, A. (2005). Antihypertensive treatment and
multifactorial approach for renal protection in diabetes. J Am Soc
Nephrol 16 Suppl 1, S18–21.
Fioretto, P., Steffes, M.W., Barbosa, J., Rich, S.S., Miller, M.E., and Mauer,
M. (1999). Is diabetic nephropathy inherited? Studies of glomerular
structure in type 1 diabetic sibling pairs. Diabetes 48, 865–869.
Fioretto, P., Steffes, M.W., Brown, D.M., and Mauer, S.M. (1992). An
overview of renal pathology in insulin-dependent diabetes mellitus
in relationship to altered glomerular hemodynamics. Am J Kidney
Dis 20, 549–558.
Fioretto, P., Steffes, M.W., Sutherland, D.E., Goetz, F.C., and Mauer, M.
(1998). Reversal of lesions of diabetic nephropathy after pancreas
transplantation. N Engl J Med 339, 69–75.
Fioretto, P., Steffes, M.W., Brown, D.M., and Mauer, S.M. (1992). An
overview of renal pathology in insulin-dependent diabetes mellitus
in relationship to altered glomerular hemodynamics. Am J Kidney
Dis 20, 549–558.
Fioretto, P., Steffes, M.W., Sutherland, D.E., Goetz, F.C., and Mauer, M.
(1998). Reversal of lesions of diabetic nephropathy after pancreas
transplantation. N Engl J Med 339, 69–75.
Fogarty, D.G., Hanna, L.S., Wantman, M., Warram, J.H., Krolewski,
A.S., and Rich, S.S. (2000a). Segregation analysis of urinary
albumin excretion in families with type 2 diabetes. Diabetes 49,
1057–1063.
Fogarty, D.G., Rich, S.S., Hanna, L., Warram, J.H., and Krolewski, A.S.
(2000b). Urinary albumin excretion in families with type 2 diabetes
is heritable and genetically correlated to blood pressure. Kidney Int
57, 250–257.
Forsblom, C.M., Kanninen, T., Lehtovirta, M., Saloranta, C., and Groop,
L.C. (1999). Heritability of albumin excretion rate in families of
patients with Type II diabetes. Diabetologia 42, 1359–1366.
Freedman, B.I., Volkova, N.V., Satko, S.G., et al. (2005). Population-based
screening for family history of end-stage renal disease among inci-
dent dialysis patients. Am J Nephrol 25, 529–535.
Fuller, J.H., Stevens, L.K., and Wang, S.L. (2001). Risk factors for cardio-
vascular mortality and morbidity: the WHO Multinational Study
of Vascular Disease in Diabetes. Diabetologia 44 Suppl 2, S54–64.
Gabow, P.A., Johnson, A.M., Kaehny, W.D., et al. (1992). Factors affect-
ing the progression of renal disease in autosomal-dominant polycys-
tic kidney disease. Kidney Int 41, 1311–1319.
Gaede, P., Lund-Andersen, H., Parving, H.H., and Pedersen, O. (2008).
Effect of a multifactorial intervention on mortality in type 2 diabe-
tes. N Engl J Med 358, 580–591.
Gaede, P., Vedel, P., Larsen, N., Jensen, G.V., Parving, H.H., and Pedersen,
O. (2003). Multifactorial intervention and cardiovascular disease in
patients with type 2 diabetes. N Engl J Med 348, 383–393.
Gaede, P., Vedel, P., Parving, H.H., and Pedersen, O. (1999). Intensified
multifactorial intervention in patients with type 2 diabetes mellitus
and microalbuminuria: the Steno type 2 randomised study. Lancet
353, 617–622.
Garcia-Gonzalez, M.A., Jones, J.G., Allen, S.K., et al. (2007). Evaluating
the clinical utility of a molecular genetic test for polycystic kidney
disease. Mol Genet Metab 92, 160–167.
Geberth, S., Ritz, E., Zeier, M., and Stier, E. (1995). Anticipation of age
at renal death in autosomal dominant polycystic kidney disease
(ADPKD)? Nephrol Dial Transpl 10, 1603–1606.
Giamarchi, A., Feng, S., Rodat-Despoix, L., et al. (2010). A polycystin-2
(TRPP2) dimerization domain essential for the function of hetero-
meric polycystin complexes. EMBO J 29, 1176–1191.
Gilbert, R.E., and Cooper, M.E. (1999). The tubulointerstitium in pro-
gressive diabetic kidney disease: more than an aftermath of glomeru-
lar injury? Kidney Int 56, 1627–1637.
Grantham, J.J. (2003). Lillian Jean Kaplan International Prize for
advancement in the understanding of polycystic kidney dis-
ease. Understanding polycystic kidney disease:  a systems biology
approach. Kidney Int 64, 1157–1162.
Greene, D.A., Arezzo, J.C., and Brown, M.B. (1999). Effect of aldose
reductase inhibition on nerve conduction and morphometry in dia-
betic neuropathy. Zenarestat Study Group. Neurology 53, 580–591.
Groop, P.H., Thomas, M.C., Moran, J.L., et al. (2009). The presence and
severity of chronic kidney disease predicts all-cause mortality in
type 1 diabetes. Diabetes 58, 1651–1658.
Gross, J.L., de Azevedo, M.J., Silveiro, S.P., Canani, L.H., Caramori,
M.L., and Zelmanovitz, T. (2005a). Diabetic nephropathy: diagno-
sis, prevention, and treatment. Diabetes Care 28, 164–176.
Gross, M.L., Amann, K., and Ritz, E. (2005b). Nephron number and
renal risk in hypertension and diabetes. J Am Soc Nephrol 16 Suppl
1, S27–29.
Diabetes Control and Complications Trial/Epidemiology of Diabetes
Interventions and ComplicationsGroup. (2000). Retinopathy and
nephropathy in patients with type 1 diabetes four years after a trial
of intensive therapy. N Engl J Med 342, 381–389.
Diabetes Control and Complications Trial/Epidemiology of Diabetes
Interventions and ComplicationsGroup. (2003). Sustained effect
of intensive treatment of type 1 diabetes mellitus on development
and progression of diabetic nephropathy:  the Epidemiology of
Diabetes Interventions and Complications (EDIC) study. JAMA
290, 2159–2167.
Diabetes Control and Complications Trial/Epidemiology of Diabetes
Interventions and ComplicationsGroup. (1999). Epidemiology of
Diabetes Interventions and Complications (EDIC). Design, imple-
mentation, and preliminary results of a long-term follow-up of the
Diabetes Control and Complications Trial cohort. Diabetes Care
22, 99–111.
Diabetes Control and Complications Trial Group. (1993). The effect of
intensive treatment of diabetes on the development and progression
of long-term complications in insulin-dependent diabetes mellitus.
N Engl J Med 329, 977–986.
Guay-Woodford, L.M. (2003). Murine models of polycystic kidney dis-
ease: molecular and therapeutic insights. Am J Physiol Renal Physiol
285, F1034–1049.
Guo, L., Schreiber, T.H., Weremowicz, S., Morton, C.C., Lee, C.,
and Zhou, J. (2000). Identification and characterization of a
novel polycystin family member, polycystin-L2, in mouse and
human: sequence, expression, alternative splicing, and chromosomal
localization. Genomics 64, 241–251.
Hales, C.N., and Barker, D.J. (2001). The thrifty phenotype hypothesis.
Br Med Bull 60, 5–20.
Hansen, H.P., Tauber-Lassen, E., Jensen, B.R., and Parving, H.H. (2002).
Effect of dietary protein restriction on prognosis in patients with
diabetic nephropathy. Kidney Int 62, 220–228.
Hanson, R.L., Ehm, M.G., Pettitt, D.J., et  al. (1998). An autoso-
mal genomic scan for loci linked to type II diabetes mellitus
and body-mass index in Pima Indians. Am J Hum Genet 63,
1130–1138.
Hanson, R.L., Craig, D.W., Millis, M.P., et al. (2007). Identification of
PVT1 as a candidate gene for end-stage renal disease in type 2 diabe-
tes using a pooling-based genome-wide single nucleotide polymor-
phism association study. Diabetes 56, 975–983.
3 8 8   •   G eno m ic s in C l inica l P ractice
Harjutsalo, V., Katoh, S., Sarti, C., Tajima, N., and Tuomilehto, J. (2004).
Population-based assessment of familial clustering of diabetic
nephropathy in type 1 diabetes. Diabetes 53, 2449–2454.
Hateboer, N., van Dijk, M.A., Bogdanova, N., et al. (1999). Comparison
of phenotypes of polycystic kidney disease types 1 and 2. European
PKD1-PKD2 Study Group. Lancet 353, 103–107.
Hex, N., Bartlett, C., Wright, D., Taylor, M., and Varley, D. (2012).
Estimating the current and future costs of type 1 and type 2 diabetes
in the UK, including direct health costs and indirect societal and
productivity costs. Diabet Med. 29, 855–862.
Hirschhorn, J.N., Lohmueller, K., Byrne, E., and Hirschhorn, K. (2002).
A comprehensive review of genetic association studies. Genet Med
4, 45–61.
Hirschhorn, J.N., and Gajdos, Z.K. (2011). Genome-wide association
studies:  results from the first few years and potential implications
for clinical medicine. Annu Rev Med 62, 11–24.
Hogan, M.C., Manganelli, L., Woollard, J.R., et  al. (2009).
Characterization of PKD protein-positive exosome-like vesicles.
J Am Soc Nephrol 20, 278–288.
Holmes, D.I., Abdel Wahab, N., and Mason, R.M. (1997). Identification
of glucose-regulated genes in human mesangial cells by mRNA dif-
ferential display. Biochem Biophys Res Commun 238, 179–184.
Hughes, J., Ward, C.J., Aspinwall, R., Butler, R., and Harris, P.C. (1999).
Identification of a human homologue of the sea urchin receptor for
egg jelly: a polycystic kidney disease-like protein. Hum Mol Genet
8, 543–549.
Hughes, J., Ward, C.J., Peral, B., et al. (1995). The polycystic kidney dis-
ease 1 (PKD1) gene encodes a novel protein with multiple cell rec-
ognition domains. Nat Genet 10, 151–160.
Hugot, J.P., Chamaillard, M., Zouali, H., et  al. (2001). Association of
NOD2 leucine-rich repeat variants with susceptibility to Crohn’s
disease. Nature 411, 599–603.
Hunter, D.J., Lange, M., Snieder, H., et al. (2002). Genetic contribution
to renal function and electrolyte balance:  a twin study. Clin Sci
(Lond) 103, 259–265.
Imperatore, G., Hanson, R.L., Pettitt, D.J., Kobes, S., Bennett, P.H.,
and Knowler, W.C. (1998). Sib-pair linkage analysis for suscepti-
bility genes for microvascular complications among Pima Indians
with type 2 diabetes. Pima Diabetes Genes Group. Diabetes 47,
821–830.
Imperatore, G., Knowler, W.C., Pettitt, D.J., Kobes, S., Bennett, P.H.,
and Hanson, R.L. (2000). Segregation analysis of diabetic nephrop-
athy in Pima Indians. Diabetes 49, 1049–1056.
Ioannidis, J.P., Trikalinos, T.A., Ntzani, E.E., and Contopoulos-Ioannidis,
D.G. (2003). Genetic associations in large versus small studies: an
empirical assessment. Lancet 361, 567–571.
Ishimaru, Y., Inada, H., Kubota, M., Zhuang, H., Tominaga, M., and
Matsunami, H. (2006). Transient receptor potential family mem-
bers PKD1L3 and PKD2L1 form a candidate sour taste receptor.
Proc Natl Acad Sci U S A 103, 12569–12574.
Iyengar, S.K., Abboud, H.E., Goddard, K.A., et al. (2007). Genome-wide
scans for diabetic nephropathy and albuminuria in multiethnic
populations: the family investigation of nephropathy and diabetes
(FIND). Diabetes 56, 1577–1585.
Jacobsen, P., Rossing, P., Tarnow, L., Hovind, P., and Parving, H.H.
(2003). Birth weight—a risk factor for progression in diabetic
nephropathy? J Intern Med 253, 343–350.
James, M.T., Hemmelgarn, B.R., and Tonelli, M. (2010). Early rec-
ognition and prevention of chronic kidney disease. Lancet 375,
1296–1309.
Jenkins, A.J., Lyons, T.J., Zheng, D., et  al. (2003). Lipoproteins in the
DCCT/EDIC cohort:  associations with diabetic nephropathy.
Kidney Int 64, 817–828.
Johnson, G.C., Esposito, L., Barratt, B.J., et al. (2001). Haplotype tag-
ging for the identification of common disease genes. Nat Genet 29,
233–237.
Jones, C.A., Krolewski, A.S., Rogus, J., Xue, J.L., Collins, A., and Warram,
J.H. (2005). Epidemic of end-stage renal disease in people with
diabetes in the United States population:  do we know the cause?
Kidney Int 67, 1684–1691.
Kamura, K., Kobayashi, D., Uehara, Y., et al. (2011). PKD1L1 complexes
with PKD2 on motile cilia and functions to establish the left-right
axis. Development 138, 1121–1129.
Keller, G., Zimmer, G., Mall, G., Ritz, E., and Amann, K. (2003).
Nephron number in patients with primary hypertension. N Engl J
Med 348, 101–108.
Klahr, S., Levey, A.S., Beck, G.J., et al. (1994). The effects of dietary pro-
tein restriction and blood-pressure control on the progression of
chronic renal disease. Modification of Diet in Renal Disease Study
Group. N Engl J Med 330, 877–884.
Kleymenova, E., Ibraghimov-Beskrovnaya, O., Kugoh, H., et al. (2001).
Tuberin-dependent membrane localization of polycystin-1: a func-
tional link between polycystic kidney disease and the TSC2 tumor
suppressor gene. Mol Cell 7, 823–832.
Koya, D., Haneda, M., Nakagawa, H., et  al. (2000). Amelioration of
accelerated diabetic mesangial expansion by treatment with a PKC
beta inhibitor in diabetic db/db mice, a rodent model for type 2
diabetes. FASEB J 14, 439–447.
Koya, D., Jirousek, M.R., Lin, Y.W., Ishii, H., Kuboki, K., and King, G.L.
(1997). Characterization of protein kinase C beta isoform activa-
tion on the gene expression of transforming growth factor-beta,
extracellular matrix components, and prostanoids in the glomeruli
of diabetic rats. J Clin Invest 100, 115–126.
Krolewski, M., Eggers, P.W., and Warram, J.H. (1996). Magnitude of
end-stage renal disease in IDDM: a 35 year follow-up study. Kidney
Int 50, 2041–2046.
Kurbegovic, A., Cote, O., Couillard, M., Ward, C.J., Harris, P.C., and
Trudel, M. (2010). Pkd1 transgenic mice: adult model of polycys-
tic kidney disease with extrarenal and renal phenotypes. Hum Mol
Genet 19, 1174–1189.
Langefeld, C.D., Beck, S.R., Bowden, D.W., Rich, S.S., Wagenknecht,
L.E., and Freedman, B.I. (2004). Heritability of GFR and albumin-
uria in Caucasians with type 2 diabetes mellitus. Am J Kidney Dis
43, 796–800.
Lantinga-van Leeuwen, I.S., Dauwerse, J.G., Baelde, H.J., et al. (2004).
Lowering of PKD1 expression is sufficient to cause polycystic kid-
ney disease. Hum Mol Genet 13, 3069–3077.
Levey, A.S., Greene, T., Beck, G.J., et al. (1999). Dietary protein restric-
tion and the progression of chronic renal disease: what have all of
the results of the MDRD study shown? Modification of Diet in
Renal Disease Study group. J Am Soc Nephrol 10, 2426–2439.
Levey, A.S., and Coresh, J. (2012). Chronic kidney disease. Lancet 379,
165–180.
Lewis, E.J., Hunsicker, L.G., Bain, R.P., and Rohde, R.D. (1993). The
effect of angiotensin-converting-enzyme inhibition on diabetic
nephropathy. The Collaborative Study Group. N Engl J Med 329,
1456–1462.
Lewis, E.J., Hunsicker, L.G., Clarke, W.R., et al. (2001). Renoprotective
effect of the angiotensin-receptor antagonist irbesartan in patients
with nephropathy due to type 2 diabetes. N Engl J Med 345,
851–860.
Li, A., Tian, X., Sung, S.W., and Somlo, S. (2003). Identification of two
novel polycystic kidney disease-1-like genes in human and mouse
genomes. Genomics 81, 596–608.
Little, J., Higgins, J.P., Ioannidis, J.P., et  al. (2009). STrengthening the
REporting of Genetic Association Studies (STREGA): an extension
of the STROBE statement. PLoS Med 6, e22.
Liu, M., Shi, S., Senthilnathan, S., et  al. (2010). Genetic variation of
DKK3 may modify renal disease severity in ADPKD. J Am Soc
Nephrol 21, 1510–1520.
Lohmueller, K.E., Pearce, C.L., Pike, M., Lander, E.S., and Hirschhorn,
J.N. (2003). Meta-analysis of genetic association studies supports a
G enetic s and G eno m ic s of C hronic K idney D i s ea s e   •   3 8 9
contribution of common variants to susceptibility to common dis-
ease. Nat Genet 33, 177–182.
Lowe, C.E., Cooper, J.D., Chapman, J.M., et  al. (2004). Cost-effective
analysis of candidate genes using htSNPs: a staged approach. Genes
Immun 5, 301–305.
Lu, W., Peissel, B., Babakhanlou, H., et al. (1997). Perinatal lethality with
kidney and pancreas defects in mice with a targetted Pkd1 mutation.
Nat Genet 17, 179–181.
Lu, W., Shen, X., Pavlova, A., et al. (2001). Comparison of Pkd1-targeted
mutants reveals that loss of polycystin-1 causes cystogenesis and
bone defects. Hum Mol Genet 10, 2385–2396.
Luyckx, V.A., and Brenner, B.M. (2005). Low birth weight, nephron
number, and kidney disease. Kidney Int Suppl, S68–77.
Luyckx, V.A., and Brenner, B.M. (2010). The clinical importance of
nephron mass. J Am Soc Nephrol: JASN 21, 898–910.
Mackenzie, F.E., Romero, R., Williams, D., et al. (2009). Upregulation
of Pkd1l2 provokes a complex neuromuscular disease in the mouse.
Hum Mol Genet 18, 3553–3566.
Magistroni, R., He, N., Wang, K., et al. (2003). Genotype-renal function
correlation in type 2 autosomal dominant polycystic kidney disease.
J Am Soc Nephrol 14, 1164–1174.
Marchini, J., Donnelly, P., and Cardon, L.R. (2005). Genome-wide strat-
egies for detecting multiple loci that influence complex diseases. Nat
Genet 37, 413–417.
Marchini, J., Howie, B., Myers, S., McVean, G., and Donnelly, P. (2007).
A new multipoint method for genome-wide association studies by
imputation of genotypes. Nat Genet 39, 906–913.
Marian, A.J. (2012). Molecular genetic studies of complex phenotypes.
Transl Res 159, 64–79.
Marshall, S.M. (2004). Recent advances in diabetic nephropathy. Clin
Med 4, 277–282.
McGrath, J., Somlo, S., Makova, S., Tian, X., and Brueckner, M. (2003).
Two populations of node monocilia initiate left-right asymmetry in
the mouse. Cell 114, 61–73.
McKnight, A.J., Maxwell, A.P., Sawcer, S., et al. (2006). A genome-wide
DNA microsatellite association screen to identify chromosomal
regions harbouring candidate genes in diabetic nephropathy. J Am
Soc Nephrol 17, 831–836.
McKnight, A.J., Currie, D., Maxwell, A.P. (2010). Unravelling the
genetic basis of renal diseases; from single gene to multifactorial dis-
orders. J Pathol 220, 198–216.
Meguid El Nahas, A., and Bello, A.K. (2005). Chronic kidney dis-
ease: the global challenge. Lancet 365, 331–340.
Milutinovic, J., Rust, P.F., Fialkow, P.J., et al. (1992). Intrafamilial pheno-
typical expression of autosomal dominant polycystic kidney disease.
Am J Kidney Dis 19, 465–472.
Mochizuki, T., Wu, G., Hayashi, T., et al. (1996). PKD2, a gene for poly-
cystic kidney disease that encodes an integral membrane protein.
Science 272, 1339–1342.
Molitch, M.E., DeFronzo, R.A., Franz, M.J., Keane, W.F., Mogensen,
C.E., and Parving, H.H. (2003). Diabetic nephropathy. Diabetes
Care 26 Suppl 1, S94–98.
Molitch, M.E., Steffes, M.W., Cleary, P.A., and Nathan, D.M. (1993).
Baseline analysis of renal function in the Diabetes Control and
Complications Trial. The Diabetes Control and Complications
Trial Research Group [corrected]. Kidney Int 43, 668–674.
Mooyaart, A.L., Valk, E.J., van Es, L.A., et al. (2011). Genetic associations
in diabetic nephropathy: a meta-analysis. Diabetologia 54, 544–553.
Morgan, C.L., Currie, C.J., Stott, N.C., Smithers, M., Butler, C.C., and
Peters, J.R. (2000). The prevalence of multiple diabetes-related com-
plications. Diabet Med 17, 146–151.
Moy, C.S., LaPorte, R.E., Dorman, J.S., et al. (1990). Insulin-dependent
diabetes mellitus mortality. The risk of cigarette smoking.
Circulation 82, 37–43.
Mrug, M., Li, R., Cui, X., Schoeb, T.R., Churchill, G.A., and
Guay-Woodford, L.M. (2005). Kinesin family member 12 is a
 candidate polycystic kidney disease modifier in the cpk mouse. J Am
Soc Nephrol 16, 905–916.
Mueller, P.W., Rogus, J.J., Cleary, P.A., et al. (2006). Genetics of Kidneys
in Diabetes (GoKinD) study:  a genetics collection available for
identifying genetic susceptibility factors for diabetic nephropathy in
type 1 diabetes. J Am Soc Nephrol: JASN 17, 1782–1790.
Murphy, M., Godson, C., Cannon, S., et  al. (1999). Suppression sub-
tractive hybridization identifies high glucose levels as a stimulus for
expression of connective tissue growth factor and other genes in
human mesangial cells. J Biol Chem 274, 5830–5834.
Murphy, M., McGinty, A., and Godson, C. (1998). Protein kinases
C: potential targets for intervention in diabetic nephropathy. Curr
Opin Nephrol Hypertens 7, 563–570.
Nauli, S.M., Alenghat, F.J., Luo, Y., et  al. (2003). Polycystins 1 and 2
mediate mechanosensation in the primary cilium of kidney cells.
Nat Genet 33, 129–137.
Neill, A.T., Moy, G.W., and Vacquier, V.D. (2004). Polycystin-2 associ-
ates with the polycystin-1 homolog, suREJ3, and localizes to the
acrosomal region of sea urchin spermatozoa. Mol Reprod Dev 67,
472–477.
Nelson, R.G., Knowler, W.C., Pettitt, D.J., Saad, M.F., and Bennett, P.H.
(1993). Diabetic kidney disease in Pima Indians. Diabetes Care 16,
335–341.
Nelson, R.G., Meyer, T.W., Myers, B.D., and Bennett, P.H. (1997). Clinical
and pathological course of renal disease in non-insulin-dependent
diabetes mellitus: the Pima Indian experience. Semin Nephrol 17,
124–131.
Nilsson, P.M., Gudbjornsdottir, S., Eliasson, B., and Cederholm, J. (2004).
Smoking is associated with increased HbA1c values and microalbu-
minuria in patients with diabetes—data from the National Diabetes
Register in Sweden. Diabetes Metab 30, 261–268.
Nistico, L., Buzzetti, R., Pritchard, L.E., et al. (1996). The CTLA-4 gene
region of chromosome 2q33 is linked to, and associated with, type 1
diabetes. Belgian Diabetes Registry. Hum Mol Genet 5, 1075–1080.
O’Dea, D.F., Murphy, S.W., Hefferton, D., and Parfrey, P.S. (1998).
Higher risk for renal failure in first-degree relatives of white patients
with end-stage renal disease: a population-based study. Am J Kidney
Dis 32, 794–801.
O’Seaghdha, C.M., and Fox, C.S. (2011). Genetics of chronic kidney dis-
ease. Nephron Clin Pract 118, c55–63.
O’Sullivan, D.A., Torres, V.E., Gabow, P.A., et al. (1998). Cystic fibrosis
and the phenotypical expression of autosomal dominant polycystic
kidney disease. Am J Kidney Dis 32, 976–983.
Oates, P.J., and Mylari, B.L. (1999). Aldose reductase inhibitors: thera-
peutic implications for diabetic complications. Expert Opin Inv
Drugs 8, 2095–2119.
Olbrich, H., Fliegauf, M., Hoefele, J., et al. (2003). Mutations in a novel
gene, NPHP3, cause adolescent nephronophthisis, tapeto-retinal
degeneration and hepatic fibrosis. Nat Genet 34, 455–459.
Olivarius Nde, F., Andreasen, A.H., Keiding, N., and Mogensen, C.E.
(1993). Epidemiology of renal involvement in newly diagnosed
middle-aged and elderly diabetic patients. Cross-sectional data from
the population-based study “Diabetes Care in General Practice,”
Denmark. Diabetologia 36, 1007–1016.
Ong, A.C., and Harris, P.C. (1997). Molecular basis of renal cyst forma-
tion—one hit or two? Lancet 349, 1039–1040.
Ong, A.C., and Harris, P.C. (2005). Molecular pathogenesis of
ADPKD:  The polycystin complex gets complex. Kidney Int 67,
1234–1247.
Ong, A.C., Harris, P.C., Davies, D.R., et al. (1999). Polycystin-1 expres-
sion in PKD1, early-onset PKD1, and TSC2/PKD1 cystic tissue.
Kidney Int 56, 1324–1333.
Ong, A.C. (2000). Polycystin expression in the kidney and other tis-
sues: complexity, consensus and controversy. Exp Nephrol 8, 208–214.
Ong, A.C., and Harris, P.C. (2005). Molecular pathogenesis of
ADPKD:  The polycystin complex gets complex. Kidney Int 67,
1234–1247.
3 9 0   •   G eno m ic s in C l inica l P ractice
Orth, S.R., Ritz, E., and Schrier, R.W. (1997). The renal risks of smoking.
Kidney Int 51, 1669–1677.
Pambianco, G., Costacou, T., Ellis, D., Becker, D.J., Klein, R., and
Orchard, T.J. (2006). The 30-year natural history of type 1 dia-
betes complications:  the Pittsburgh Epidemiology of Diabetes
Complications Study experience. Diabetes 55, 1463–1469.
Parfrey, P.S., Bear, J.C., Morgan, J., et al. (1990). The diagnosis and prog-
nosis of autosomal dominant polycystic kidney disease. N Engl J
Med 323, 1085–1090.
Parving, H.H., Lewis, J.B., Ravid, M., Remuzzi, G., and Hunsicker,
L.G. (2006). Prevalence and risk factors for microalbuminuria in
a referred cohort of type II diabetic patients: a global perspective.
Kidney Int 69, 2057–2063.
Paterson, A.D., Magistroni, R., He, N., et al. (2005). Progressive loss of
renal function is an age-dependent heritable trait in type 1 auto-
somal dominant polycystic kidney disease. J Am Soc Nephrol 16,
755–762.
Pedrini, M.T., Levey, A.S., Lau, J., Chalmers, T.C., and Wang, P.H.
(1996). The effect of dietary protein restriction on the progression
of diabetic and nondiabetic renal diseases:  a meta-analysis. Ann
Intern Med 124, 627–632.
Pei, Y., Lan, Z., Wang, K., et al. (2012). A missense mutation in PKD1
attenuates the severity of renal disease. Kidney Int 81, 412–417.
Pei, Y., Paterson, A.D., Wang, K.R., et  al. (2001). Bilineal disease and
trans-heterozygotes in autosomal dominant polycystic kidney dis-
ease. Am J Hum Genet 68, 355–363.
Peral, B., Ong, A.C., San Millan, J.L., Gamble, V., Rees, L., and Harris,
P.C. (1996). A stable, nonsense mutation associated with a case
of infantile onset polycystic kidney disease 1 (PKD1). Hum Mol
Genet 5, 539–542.
Persu, A., Devuyst, O., Lannoy, N., et al. (2000). CF gene and cystic fibrosis
transmembrane conductance regulator expression in autosomal dom-
inant polycystic kidney disease. J Am Soc Nephrol 11, 2285–2296.
Persu, A., Duyme, M., Pirson, Y., et  al. (2004). Comparison between
siblings and twins supports a role for modifier genes in ADPKD.
Kidney Int 66, 2132–2136.
Persu, A., Stoenoiu, M.S., Messiaen, T., et al. (2002). Modifier effect of
ENOS in autosomal dominant polycystic kidney disease. Hum Mol
Genet 11, 229–241.
Peters, D.J., and Breuning, M.H. (2001). Autosomal dominant polycys-
tic kidney disease: modification of disease progression. Lancet 358,
1439–1444.
Pettitt, D.J., Saad, M.F., Bennett, P.H., Nelson, R.G., and Knowler, W.C.
(1990). Familial predisposition to renal disease in two generations
of Pima Indians with type 2 (non-insulin-dependent) diabetes mel-
litus. Diabetologia 33, 438–443.
Pezzolesi, M.G., Poznik, G.D., Mychaleckyj, J.C., et  al. (2009).
Genome-wide association scan for diabetic nephropathy suscepti-
bility genes in type 1 diabetes. Diabetes 58, 1403–1410.
Phillips, D.I., Barker, D.J., Hales, C.N., Hirst, S., and Osmond, C. (1994).
Thinness at birth and insulin resistance in adult life. Diabetologia
37, 150–154.
Phimister, E.G. (2005). Genomic cartography—presenting the HapMap.
N Engl J Med 353, 1766–1768.
Placha, G., Canani, L.H., Warram, J.H., and Krolewski, A.S. (2005).
Evidence for different susceptibility genes for proteinuria and ESRD
in type 2 diabetes. ACDK 12, 155–169.
Pritchard, L., Sloane-Stanley, J.A., Sharpe, J.A., et al. (2000). A human
PKD1 transgene generates functional polycystin-1 in mice and is
associated with a cystic phenotype. Hum Mol Genet 9, 2617–2627.
Purcell, S., Neale, B., Todd-Brown, K., et al. (2007). PLINK: a tool set
for whole-genome association and population-based linkage analy-
ses. Am J Hum Genet 81, 559–575.
Qian, F., Boletta, A., Bhunia, A.K., et al. (2002). Cleavage of polycystin-1
requires the receptor for egg jelly domain and is disrupted by human
autosomal-dominant polycystic kidney disease 1-associated muta-
tions. Proc Natl Acad Sci U S A 99, 16981–16986.
Qian, F., Watnick, T.J., Onuchic, L.F., and Germino, G.G. (1996). The
molecular basis of focal cyst formation in human autosomal domi-
nant polycystic kidney disease type I. Cell 87, 979–987.
Qian, Q., Harris, P.C., and Torres, V.E. (2001). Treatment prospects
for autosomal-dominant polycystic kidney disease. Kidney Int 59,
2005–2022.
Qian, Q., Hunter, L.W., Li, M., et  al. (2003). Pkd2 haploinsufficiency
alters intracellular calcium regulation in vascular smooth muscle
cells. Hum Mol Genet 12, 1875–1880.
Quinn, M., Angelico, M.C., Warram, J.H., and Krolewski, A.S.
(1996). Familial factors determine the development of dia-
betic nephropathy in patients with IDDM. Diabetologia 39,
940–945.
Reeders, S.T., Breuning, M.H., Davies, K.E., et al. (1985). A highly poly-
morphic DNA marker linked to adult polycystic kidney disease on
chromosome 16. Nature 317, 542–544.
Remuzzi, G., Schieppati, A., and Ruggenenti, P. (2002). Clinical prac-
tice. Nephropathy in patients with type 2 diabetes. N Engl J Med
346, 1145–1151.
Rich, S.S. (2006). Genetics of diabetes and its complications. J Am Soc
Nephrol: JASN 17, 353–360.
Ritz, E., Ogata, H., and Orth, S.R. (2000). Smoking: a factor promoting
onset and progression of diabetic nephropathy. Diabetes Metab 26
Suppl 4, 54–63.
Ritz, E., Rychlik, I., Schomig, M., and Wagner, J. (2001). Blood pressure
in diabetic nephropathy—current controversies. J Intern Med 249,
215–223.
Robertson, L., Waugh, N., and Robertson, A. (2007). Protein restric-
tion for diabetic renal disease. Cochrane Database of Systematic
Reviews, CD002181.
Rocco, M.V., Chen, Y., Goldfarb, S., and Ziyadeh, F.N. (1992). Elevated
glucose stimulates TGF-beta gene expression and bioactivity in
proximal tubule. Kidney Int 41, 107–114.
Roglic, G., Colhoun, H.M., Stevens, L.K., Lemkes, H.H., Manes, C.,
and Fuller, J.H. (1998). Parental history of hypertension and
parental history of diabetes and microvascular complications in
insulin-dependent diabetes mellitus:  the EURODIAB IDDM
Complications Study. Diabet Med 15, 418–426.
Rosenblit, P.D. (2012). Do persons with diabetes benefit from combina-
tion statin and fibrate therapy? Curr Cardiol Rep 14, 112–124.
Rossetti, S., Burton, S., Strmecki, L., et al. (2002a). The position of the
polycystic kidney disease 1 (PKD1) gene mutation correlates with
the severity of renal disease. J Am Soc Nephrol 13, 1230–1237.
Rossetti, S., Chauveau, D., Kubly, V., et al. (2003). Association of muta-
tion position in polycystic kidney disease 1 (PKD1) gene and devel-
opment of a vascular phenotype. Lancet 361, 2196–2201.
Rossetti, S., Chauveau, D., Walker, D., et al. A complete mutation screen
of the ADPKD genes by DHPLC. Kidney Int 61, 1588–1599.
Rossetti, S., Consugar, M.B., Chapman, A.B., et  al. (2007).
Comprehensive molecular diagnostics in autosomal dominant poly-
cystic kidney disease. J Am Soc Nephrol 18, 2143–2160.
Rossetti, S., Kubly, V.J., Consugar, M.B., et al. (2009). Incompletely pen-
etrant PKD1 alleles suggest a role for gene dosage in cyst initiation
in polycystic kidney disease. Kidney Int 75, 848–855.
Rossetti, S., Strmecki, L., Gamble, V., et al. (2001). Mutation analysis of
the entire PKD1 gene:  genetic and diagnostic implications. Am J
Hum Genet 68, 46–63.
Satko, S.G., and Freedman, B.I. (2005). The familial clustering of
renal disease and related phenotypes. Med Clin North Am 89,
447–456.
Satko, S.G., Langefeld, C.D., Daeihagh, P., Bowden, D.W., Rich, S.S.,
and Freedman, B.I. (2002). Nephropathy in siblings of African
Americans with overt type 2 diabetic nephropathy. Am J Kidney
Dis 40, 489–494.
Savage, D.A., Patterson, C.C., Deloukas, P., et al. (2008). Genetic associ-
ation analyses of non-synonymous single nucleotide polymorphisms
in diabetic nephropathy. Diabetologia 51, 1998–2002.
G enetic s and G eno m ic s of C hronic K idney D i s ea s e   •   3 9 1
Schork, N.J., Murray, S.S., Frazer, K.A., and Topol, E.J. (2009). Common
vs. rare allele hypotheses for complex diseases. Curr Opin Genet
Dev 19, 212–219.
Schreuder, M.F., Nyengaard, J.R., Fodor, M., van Wijk, J.A., and
Delemarre-van de Waal, H.A. (2005). Glomerular number and
function are influenced by spontaneous and induced low birth
weight in rats. J Am Soc Nephrol 16, 2913–2919.
Seaquist, E.R., Goetz, F.C., Rich, S., and Barbosa, J. (1989). Familial clus-
tering of diabetic kidney disease. Evidence for genetic susceptibility
to diabetic nephropathy. N Engl J Med 320, 1161–1165.
Shields, J., and Maxwell, A.P. (2010). Managing diabetic nephropathy.
Clin Med 10, 500–504.
Shimazaki, A., Kawamura, Y., Kanazawa, A., et al. (2005). Genetic varia-
tions in the gene encoding ELMO1 are associated with susceptibil-
ity to diabetic nephropathy. Diabetes 54, 1171–1178.
Somlo, S., Rutecki, G., Giuffra, L.A., Reeders, S.T., Cugino, A., and
Whittier, F.C. (1993). A kindred exhibiting cosegregation of an
overlap connective tissue disorder and the chromosome 16 linked
form of autosomal dominant polycystic kidney disease. J Am Soc
Nephrol 4, 1371–1378.
Spielman, R.S., and Ewens, W.J. (1998). A sibship test for linkage in the
presence of association:  the sib transmission/disequilibrium test.
Am J Hum Genet 62, 450–458.
Spielman, R.S., McGinnis, R.E., and Ewens, W.J. (1993). Transmission
test for linkage disequilibrium:  the insulin gene region and
insulin-dependent diabetes mellitus (IDDM). Am J Hum Genet
52, 506–516.
Stanescu, H.C., Arcos-Burgos, M., Medlar, A., et  al. (2011). Risk
HLA-DQA1 and PLA(2)R1 alleles in idiopathic membranous
nephropathy. N Engl J Med 364, 616–626.
Stratton, I.M., Adler, A.I., Neil, H.A., et al. (2000). Association of gly-
caemia with macrovascular and microvascular complications of type
2 diabetes (UKPDS 35): prospective observational study. BMJ 321,
405–412.
Streets, A.J., Ibraghimov-Beskrovnaya, O., and Ong, A.C. (2002).
Polycystin-1 mediates cell-cell adhesion in PKD1 transgenic and
non-transgenic lines. J Am Soc Nephrol 13, 46A.
Sutton, K.A., Jungnickel, M.K., and Florman, H.M. (2008). A
polycystin-1 controls postcopulatory reproductive selection in
mice. Proc Natl Acad Sci U S A 105, 8661–8666.
Tanaka, N., Babazono, T., Saito, S., et  al. (2003). Association of sol-
ute carrier family 12 (sodium/chloride) member 3 with diabetic
nephropathy, identified by genome-wide analyses of single nucleo-
tide polymorphisms. Diabetes 52, 2848–2853.
Tarnow, L., Groop, P.H., Hadjadj, S., et al. (2008). European Rational
Approach for the Genetics of Diabetic Complications—
EURAGEDIC:  patient populations and strategy. Nephrol Dial
Transpl 23, 161–168.
Telmer, S., Christiansen, J.S., Andersen, A.R., Nerup, J., and Deckert, T.
(1984). Smoking habits and prevalence of clinical diabetic micro-
angiopathy in insulin-dependent diabetics. Acta Med Scand 215,
63–68.
Thomas, M.C., Groop, P.H., and Tryggvason, K. (2012). Towards under-
standing the inherited susceptibility for nephropathy in diabetes.
Curr Opin Nephrol Hy 21, 195–202.
Tonstad, S. (2009). Cigarette smoking, smoking cessation, and diabetes.
Diabetes research and clinical practice 85, 4–13.
Torra, R., Badenas, C., Perez-Oller, L., et  al. (2000). Increased preva-
lence of polycystic kidney disease type 2 among elderly polycystic
patients. Am J Kidney Dis 36, 728–734.
Torres, V.E., Wang, X., Qian, Q., Somlo, S., Harris, P.C., and Gattone, V.H.,
2nd (2004). Effective treatment of an orthologous model of autoso-
mal dominant polycystic kidney disease. Nat Med 10, 363–364.
Tuomilehto, J., Borch-Johnsen, K., Molarius, A., et al. (1998). Incidence
of cardiovascular disease in Type 1 (insulin-dependent) dia-
betic subjects with and without diabetic nephropathy in Finland.
Diabetologia 41, 784–790.
UK Prospective Diabetes Study (UKPDS) Group (1998). Intensive
blood-glucose control with sulphonylureas or insulin compared
with conventional treatment and risk of complications in patients
with type 2 diabetes (UKPDS 33). Lancet 352, 837–853.
Vlassara, H., Striker, L.J., Teichberg, S., Fuh, H., Li, Y.M., and Steffes, M.
(1994). Advanced glycation end products induce glomerular scle-
rosis and albuminuria in normal rats. Proc Natl Acad Sci U S A 91,
11704–11708.
Walker, D., Consugar, M., Slezak, J., et al. (2003). The ENOS polymor-
phism is not associated with severity of renal disease in polycystic
kidney disease 1. Am J Kidney Dis 41, 90–94.
Wanner, C., Krane, V., Marz, W., et al. (2005). Atorvastatin in patients
with type 2 diabetes mellitus undergoing hemodialysis. N Engl J
Med 353, 238–248.
Watnick, T.J., Torres, V.E., Gandolph, M.A., et al. (1998). Somatic mutation
in individual liver cysts supports a two-hit model of cystogenesis in
autosomal dominant polycystic kidney disease. Mol Cell 2, 247–251.
Wheeler, E., and Barroso, I. (2011). Genome-wide association studies
and type 2 diabetes. Brief Funct Genomic 10, 52–60.
Williams, R., and Airey, M. (2002). Epidemiology and public health
consequences of diabetes. Curr Med Res Opin 18 Suppl 1, s1–12.
Wolf, G., and Ritz, E. (2003). Diabetic nephropathy in type 2 diabe-
tes prevention and patient management. J Am Soc Nephrol 14,
1396–1405.
Woo, D.D., Nguyen, D.K., Khatibi, N., and Olsen, P. (1997). Genetic
identification of two major modifier loci of polycystic kidney disease
progression in pcy mice. J Clin Invest 100, 1934–1940.
Wu, G., D’Agati, V., Cai, Y., et al. (1998a). Somatic inactivation of Pkd2
results in polycystic kidney disease. Cell 93, 177–188.
Wu, G., Hayashi, T., Park, J.H., et al. (1998b). Identification of PKD2L,
a human PKD2-related gene:  tissue-specific expression and map-
ping to chromosome 10q25. Genomics 54, 564–568.
Wu, G., Markowitz, G.S., Li, L., et al. (2000). Cardiac defects and renal
failure in mice with targeted mutations in Pkd2. Nat Genet 24,
75–78.
3 9 2   •   G eno m ic s in C l inica l P ractice
Wu, G., Tian, X., Nishimura, S., et al. (2002). Trans-heterozygous Pkd1
and Pkd2 mutations modify expression of polycystic kidney disease.
Hum Mol Genet 11, 1845–1854.
Yamaguchi, T., Wallace, D.P., Magenheimer, B.S., Hempson, S.J.,
Grantham, J.J., and Calvet, J.P. (2004). Calcium restriction allows
cAMP activation of the B-Raf/ERK pathway, switching cells to a
cAMP-dependent growth-stimulated phenotype. J Biol Chem 279,
40419–40430.
Yamamoto, T., Nakamura, T., Noble, N.A., Ruoslahti, E., and Border,
W.A. (1993). Expression of transforming growth factor beta is ele-
vated in human and experimental diabetic nephropathy. Proc Natl
Acad Sci U S A 90, 1814–1818.
Yamamoto, T., Noble, N.A., Cohen, A.H., et al. (1996). Expression of
transforming growth factor-beta isoforms in human glomerular dis-
eases. Kidney Int 49, 461–469.
Yu, W., Clyne, M., Khoury, M.J., and Gwinn, M. (2010). Phenopedia
and Genopedia:  disease-centered and gene-centered views of the
evolving knowledge of human genetic associations. Bioinformatics
26, 145–146.
Yuasa, T., Venugopal, B., Weremowicz, S., Morton, C.C., Guo, L., and
Zhou, J. (2002). The sequence, expression, and chromosomal local-
ization of a novel polycystic kidney disease 1-like gene, PKD1L1, in
human. Genomics 79, 376–386.
Zatz, R., Dunn, B.R., Meyer, T.W., Anderson, S., Rennke, H.G., and
Brenner, B.M. (1986). Prevention of diabetic glomerulopathy by
pharmacological amelioration of glomerular capillary hypertension.
J Clin Invest 77, 1925–1930.
Zerres, K., Rudnik-Schoneborn, S., and Deget, F. (1993). Childhood
onset autosomal dominant polycystic kidney disease in sibs: clinical
picture and recurrence risk. German Working Group on Paediatric
Nephrology [Arbeitsgemeinschaft fur Padiatrische Nephrologie.]
J Med Genet 30, 583–588.
Ziyadeh, F.N., and Wolf, G. (2008). Pathogenesis of the podocytopathy
and proteinuria in diabetic glomerulopathy. Curr Diabetes Rev 4,
39–45.
 25.
GENETICS AND GENOMICS IN CLINICAL
HEMATOLOGY, I: HEMOSTASIS AND THROMB OSIS
John H. McVey
INTRODUCTION
Vertebrates have evolved a complex system to prevent
blood loss that involvescoordinate vascular wall muscle
contraction, cell aggregation (platelets), and thedeposition
of a clottable protein (fibrin). To achieve this, a complex
network ofpositive and negative regulated reactions have
evolved that result in controlledfibrin deposition and plate-
let activation only at the site of vascular injury, without-
compromising blood flow through either the uninjured or
damaged blood vessels.
The importance of such a system to multicellular
organisms that possess a high-pressurevascular system is
evidenced by the recent demonstration that thecoagula-
tion network is present in its entirety in teleosts (bony fish
such as zebrafish [Danio rerio] and Japanese puffer fish
[Fugu rubripes]) and must thereforehave evolved before the
divergence of teleosts and tetrapods 430  million yearsago
(Hanumanthaiah et al., 2002; Davidson, Hirt, et al., 2003;
Davidson, Tuddenham,et al., 2003).
T H E C OAGU L AT I O N N ET WO R K
T H E P RO C OAGU L A N T R E S P O N S E
Blood coagulation in vivois initiated by the exposure of
factor (F) VII to cells thatexpress the integral membrane
protein tissue factor (TF). The primary control ofhemo-
stasis is the segregation of cells that express functional
TF from othercomponents of thecoagulation network
that are present in blood. TF is constitutivelyexpressed
at biologicalboundaries such as skin, organ surfaces,
vascular adventitia,and epithelial-mesenchymal sur-
faces. The TF expression pattern has beendescribed as
forming a “hemostatic envelope”(Drake et  al., 1989),
which ensuresthat, following disruption of vascular
393
integrity, FVII/FVIIa in blood is exposed tocells that
express TF, leading to the initiation of blood coagula-
tion (Figure 25.1).
Conversely, it also ensures that inappropriate initiation
of intravascular coagulationdoes not occur within intact
vasculature. The formation of the TF-FVII complexpro-
motes the activation of FVII. The TF-FVIIa complex cata-
lyzes the activation ofFIX and FX.
In the absence of its cofactor FVa, FXa generates only
trace amounts ofthrombin. Although insufficient to ini-
tiate significant fibrin polymerization, traceamounts of
thrombin formed in the “initiation” stage of coagulation
are able toactivate FV and FVIII by limited proteolysis in
a positive-feedback loop. In the“amplification” phase of
coagulation, FVIIIa forms a complex with FIXa (thete-
nase complex) and activates sufficient FXa, which in con-
cert with FVa (theprothrombinase complex) leads to the
explosive generation of thrombin thatimmediately leads
to the formation of a fibrin clot. Thrombin also activates
FXI toFXIa in a further positive-feedback loop, result-
ing in further generation of FIXaindependently of the
TF-FVIIa complex.
A key feature of these processes is the assembly of
multiprotein complexeson a negatively charged phos-
pholipid surface. Each of these complexes consists ofa
cofactor (TF, FVa, FVIIIa), an enzyme (FVIIa, FIXa,
FXa), and a substrate that isa zymogen (FIX, FX, and
prothrombin) of a serine protease. The product of onere-
action becomes theenzyme in the next complex.Platelets
activated at sites of vascular injury play key roles in nor-
malhemostasis. By adhering to the exposed subendothe-
lium and aggregating, theycreate a physical barrier that
limits blood loss. In addition, platelets acceleratethrom-
bin generation by providing a surface that promotes the
activation of FX andprothrombin. Furthermore, they
release procoagulant factors that contribute to thelocal
coagulation response.
 TF expressing cell
TF TF

TF
TF
FVII TF
TF
FIX FX

FXI TF-FVIIa TFPI TF

FXIa TF-FVIIa-FXa
TFPI
AT
FIXa
FVIIIa
FXIa-AT FIXa-AT
PT KR KR SP

FXa
FVa FVai
FVIIIai
Serpin AT
Thrombin
FXa-AT
FVIII FV

AT

cc Thrombin-AT
α PS
cc APC
β FBG FXIIIa FXIII
cc
γ FBG

Fibrinogen Fibrin

TG

TG

PC
TM
bin
Endothelial cell

rom
Th

CR
Thrombin

EP
Platelet
Lys
Lys

FDPs PD-Fibrin XL-Fibrin

TM Fibrin
TAFI
Lys

Lys

Lys
Lys

Lys

Plasmin Plasmin tPA

Lys

Lys

Serpin
TAFIa
Serpin

PAI-1
Plasmin- α2AP
α2AP
tPA-PAI-1
PLG

Figure 25.1 
The hemostasis network. The process of blood coagulation is initiated by the exposure of cells expressing TF to flowing blood.
Thrombin generation is propagated by a series of positive-feedback loops, leading to fibrin deposition. This process is controlled by a series of
negative-feedback steps: The initiation complex is inhibited by the formation of the quaternary complex TF-FVIIa-FXa-TFPI, and the active
proteases FIXa, FXa, and thrombin are inactivated by the serpin AT. In addition, thrombin initiates a negative-feedback pathway by activating PC,
leading to the inhibition of FVa and FVIIIa. The fibrin clot is degraded by plasmin.All abbreviations are given in Table 25.1: Protein names are
colored thus: functionally active proteins, red; inactivated or inhibited proteins, blue; precursors and fibrinogen-derived components, black. Lines
and arrows are colored thus: dashed red lines, positive procoagulant feedback loops; dashed blue lines, negative feedback and inhibitory loops; solid
black lines, interactions or processes.

The importance of providing a negatively charged in Man [OMIM]: 262890), which is characterized by a fail-
phospholipid surface fortheassembly of the procoagulant ure toexposephosphatidylserine on the outer leaflet of the
response is seen in the extremely rare bleedingdisorder plasma membrane and isassociated with a moderate bleed-
known as Scott syndrome (Online Mendelian Inheritance ing tendency.

3 9 4   •   G eno m ic s in C linical P ractice

 T H E A N T I C OAGU L A N T R E S P O NS E
Following the initiation of coagulation, various inhibitory
mechanisms preventextension of the coagulation process
beyond the site of vascular injury that mightotherwise
result in unnecessary occlusion of the blood vessel. Tissue
factorpathway inhibitor (TFPI),associated with glycosami-
noglycans or glycosylphosphatidylinositol(GPI) anchored-
proteins on the endothelial cell surface, rapidly inactivates
the initiation complex byforming a quaternary inhibited
complex (TF-FVIIa-FXa-TFPI). Thrombinstimulates both
endothelial cells and platelets to release further TFPI.
Thrombin generated at the endothelial surface binds
thrombomodulin (TM), an integral membrane protein
expressed specifically on all endothelial cells, and activates
protein C (PC). The activation of PC is promoted by endo-
thelial cellPC receptor (EPCR), a GPI-anchored protein
expressed on a subset of endothelial cells, which provides a
direct binding site for PC onendothelial cells and increases
the affinity of the thrombin-TM complex for PC.The ana-
phase protein complex (APC) with its cofactor protein S
(PS) rapidly inactivates theprocoagulant cofactors FVa and
FVIIIa by specific proteolysis, forming a negativefeedback
loop. The activated coagulation proteases FIXa, FXa, FXIa,
and thrombinare all inhibited by antithrombin (AT). The
rate of inhibition by AT is substantiallyincreased by binding
glycosaminoglycans on the surface of endothelial cells.
T H E FI B R I N O LY T I C R E S P O NS E
Formation of fibrin triggers activation of the fibrinolytic sys-
tem and generation ofthe active fibrinolytic enzyme plasmin
that degrades fibrin into soluble fragments,disintegrating
the clot. Plasmin is formed from plasminogen by limited
proteolysisby the action of tissue-type plasminogen activa-
tor (tPA) or urokinase-typeplasminogen activator (uPA);
tPA is the most important activator in the circulation.
Plasminogen activator inhibitor (PAI)-1 and PAI-2 inhibit
the activities of theseactivators, while plasmin is primar-
ily inhibited by plasmin inhibitor (alpha2-antiplasmin),
however, tPA is a serine protease synthesized, stored, and
secreted by endothelialcells. In the absence of fibrin, tPA
has low activity toward plasminogen, but thisactivity is
increased up to three-fold in the presence of fibrin. The
rate-enhancingeffect of fibrin ensures that the formation
of plasmin and the generation offibrinolytic activity are
restricted to the location of a fibrin clot.
Thrombin-activatable fibrinolysis inhibitor (TAFI) is a
carboxypeptidase,synthesized in the liverthat circulates in
blood as a zymogen and is activated by asingle proteolytic
G enetic s and G eno m ic s in C linical H e m atology, I :   H e m o s ta s i s and T hro m b o s i s   •   395
cleavage mediated by thrombin in complex with TM.
TAFIainhibits fibrinolysis by removing the C-terminal
lysine residues formed by limitedplasmin proteolysis of
fibrin, thus removing binding sites for plasminogen and
tPA.The rate-enhancement effect of fibrin is reduced, and
fibrinolysis is downregulated.
M O L E C U L A R G E N ET I C S
The blood coagulation network is a finely balanced system
that ensures that clotformation occurs rapidly and efficiently
at the site of vascular injury but localizes the procoagulant
response in order to minimize blood loss without compro-
misingblood flow generally. Anticoagulant and fibrinolytic
components preventinappropriate clot formation or exten-
sion of the procoagulant response beyond thesite of vas-
cular injury. Deficiency of components of the coagulation
network canlead to either a bleeding phenotype or a throm-
botic phenotype, depending on therole of the deficient fac-
tor within the network.
The bleeding disorders are typically single-gene dis-
orders with a defect inthe gene encoding the coagulation
factor; the exceptions being combineddeficiency of FV and
FVIII, where the defect lies in two genes involved inintracel-
lular trafficking (LMAN1 and MCFD2)(Neerman-Arbez
et al.,1999;Nichols et al., 1999;Zhang et al., 2003; Zhang
et  al., 2006); multiple vitamin K–dependent factor defi-
ciencies type I and II, where the defect is in gammacarbox-
ylase (Wu et  al., 1997; Brenner et  al., 1998;Spronk et  al.,
2000; Mousallem et  al., 2001), or VKORC1 (Li et  al.,
2004; Rost et al., 2004), respectively; and Scottsyndrome,
an extremely rare, moderate bleeding disorder characterized
by a failure to expose phosphatidylserine on the outer leaflet
of the plasma membrane wherethe defect lies in the ABCA1
gene (Albrecht et al., 2005).
The genes encoding thehuman coagulation/fibrino-
lytic proteins were identified and characterized (often-
completely sequenced) in the 1980s. The completion of
the Human Genome Project in 2003 therefore had less
impact in this area than in many other disciplines. Many
mutationshave been identified and characterized,and
locus-specific mutation databases havebeen established
for many of the disorders (Table 25.1). The development
of high-throughput PCR–based strategies and improve-
ments in DNA sequencing technologies have led to an
explosion in mutation detection and characterization. All
types of mutations havebeen identified: missense, nonsense,
splice site, and promoter mutations, as well asinsertions,
deletions, and rearrangements. The spectrum of mutations
Table 25.1  PROTEINS IN THE HEMOSTATIC NETWORK ASSOCIATED WITH BLEEDING OR THROMBOSIS

COMMON NAME ABBREV- SUB-UNIT GENE GENE GENE NO. OF MRNA AMINO MR OF OMIM MAIN ACTION LMD
IATION SYMBOL LOCATION SIZE EXONS SIZE ACIDS MONOMER
(KBP) (BP) (MATURE) (KDA)
Tissue Factor TF F3 1p13 12.6 6 1852 263 44 134390 Cofactor for FVII/FVIIa
Prothrombin FII F2 11p11.1 20.3 14 2028 579 72 176930 Clots FBG, activates PC,
FXI, TAFI
Factor V FV F5 1q23 72.4 25 7009 2196 330 227400 Cofactor for FXa
Factor VII FVII F7 13q34 15.1 8 2880 416 50 227500 Activates FIX & FX
Factor VIII FVIII F8 Xq28 187.1 26 8957 2332 330 306700 Cofactor for FIXa http://hadb.org.uk/
Factor IX FIX F9 Xq27 32.8 8 2831 415 56 306900 Activates FX http://factorix.org/
Factor X FX F10 13q34 26.8 8 1884 445 59 227600 Activates prothrombin
Factor XI FXI F11 4q35 25.9 15 2979 607 80* 264900 Activates FIX http://www.factorxi.com/
Factor XIII** FXIII A F13A1 6p25 177.8 15 3834 731 75** 134570 Crosslinks fibrin http://www.f13-database.de/
(A chain) (xhgmobrswxgori45zk5jre45)/
content.aspx?menu=1,6
Factor XIII** FXIII B F13B 1q31 28 12 2190 641 80** 134580 Stabilizes FXIII A chain http://www.f13-database.de/
(B chain) (xhgmobrswxgori45zk5jre45)/
content.aspx?menu=1,6
Fibrinogen FGN Α FGA 4q32 7.8 6 3828 866 68*** 134820 Mechanical stabilization http://www.geht.org/
(α chain)*** of clot databaseang/fibrinogen/
Fibrinogen FGN Β FGB 4q32 9.8 8 3693 491 52*** 134830 Mechanical stabilization http://www.geht.org/
(β chain)*** of clot databaseang/fibrinogen/
Fibrinogen FGN Γ FGG 4q32 23.6 10 1753 453 49*** 134850 Mechanical stabilization http://www.geht.org/
(γ chain)*** of clot databaseang/fibrinogen/
von Willebrand factor VWF VWF 12p13 176 52 9028 2050 255 193400 Cell adhesion & FVIII http://www.vwf.group.shef.
carrier ac.uk/
Protein C PC PROC 2q14.2 10.8 9 1759 419 62 176860 Inactivation of FVa and
FVIIIa
Protein S PS PROS1 3q11.2 101.9 15 3477 676 69 176880 Inactivation of FVa and http://www.isth.
FVIIIa org/?ProteinSDeficiency
Antithrombin AT SERPINC1 1q23 21 9 1684 464 58 107300 Inhibits thrombin, FIX,
FX, FXI
Plasminogen PLG PLG 6q27 51.1 14 2905 791 92 173350 Dissolution of clot in
wound repair
Tissue plasminogen TPA PLAT 8p11.1 32.7 14 2933 562 69 173370 Plasma activator of
activator plasminogen
Plasminogen activator PAI-1 SERPINE1 7q22 12.3 9 3320 379 52 173360 Inhibition of TPA and
inhibitor 1 UPA
Alpha2-antiplasmin α2-AP SERPINF2 17p13 13.3 9 2210 452 67 262850 Inhibition of plasmin
Thrombin-activatable TAFI CPB2 13q14 52.4 11 1984 401 60 603101 Inhibition of fibrinolysis
fib. inhibitor
 Deletions
Insertions Promoter
1% 6%
10% Splice site
9%
Nonsense
4%
Missense
70%
Figure 25.2 
Spectrum of mutation types causing FVII deficiency. Data
obtained from the FVII mutation database (McVey et al., 2001).
represented inthe FVII mutation database (McVey et  al.,
2001) is typical of thosefound in other coagulation factor
gene disorders (Figure 25.2). Recurrent mutations inunre-
lated individuals arise at functionally important residues
and in CpGdinucleotides, which are known hotspots for
mutation due to deamination ofmethylated cytosine resi-
dues. Arginine residues occupy critically important func-
tional sites in the coagulation factors, and four of the six
codons encoding arginine contain the dinucleotide CpG.
A mutation hotspot, specific to hemophilia A, is therecur-
rent rearrangement responsible for 50% of all severe cases
in whichhomologous recombination of a sequence within
intron 22 of the F8gene and twoextragenic and telomeric
copies of the sequence results in an inversion that disrupt-
sthe F8gene.
THE BLEEDING DISORDER S
The most common inherited bleeding disorder is von
Willebrand’s disease (VWD). Von Willebrand factor
(VWF), which is the protein deficient or defective inpa-
tients with VWD, has two distinct roles in hemostasis. First,
it is responsiblefor enabling platelets to adhere to collagen
exposed at wound sites under conditionsof shear. Second,
VWF is the plasma carrier for FVIII, thus deficiency of
VWFleads to deficiency of FVIII. Mutations in the FVIII
binding site within VWF may result in structurally and func-
tionally normal VWF, except that it does not bind FVIII and
presents as mild hemophilia A, but with autosomal recessive
G enetic s and G eno m ic s in C linical H e m atology, I :   H e m o s ta s i s and T hro m b o s i s   •   397
rather than X-linked inheritance. VWD is an autosomal
disease with variableexpressivity and reduced penetrance. It
has been classified into six distinct types:types 1 and 3 result
from quantitative deficiency, whereas the four type 2 vari-
ants (2A, 2B, 2M, and 2N) are characterized by qualitative
abnormalities of VWF. Types 1,2A, 2B, and 2M VWD are
inherited as autosomal dominant, whereas types 2N and 3
areinherited in autosomal recessive manner. Four hundred
and three unique variants have been described that cause
VWD (https://grenada.lumc.nl/LOVD2/VWF/home.
php?select_db=VWF; Oct 2012).
FVII deficiency is an autosomal recessive disorder
with a highly variablephenotype. In contrast to the other
bleeding disorders, the residual clotting-factoractivity does
not accurately predict the clinical phenotype. Complete
absence ofFVII activity is usually incompatible with life,
and individuals die shortly afterbirth, due to severe hemor-
rhage. The majority of individuals with mutations intheir
F7gene(s) are asymptomatic and only come to light during
preoperative coagulation-factor screening. A severe bleed-
ing phenotype is only observed inindividuals with residual
FVII activity below 2%; however, a considerableproportion
of individuals with a mild to moderate bleedingphenotype
have similarFVII activity levels. The severity of the bleed-
ing in individuals with no residualFVII activity is consistent
with the key role of the TF-FVIIa complex in initiating-
blood coagulation in vivo. Surprisingly, no congenital
abnormalities in the geneencoding TF have been described
to date. Mutations in the TF gene might havebeenpredicted
to lead to either a bleeding (loss of function) or prothrom-
botic (gainof function) phenotype. Targeted disruption
of the mouse TF gene (f3) results inembryonic lethality
of f3–/– embryos at embryonic days 9.5–10.5 (Bugge et al.,
1996;Carmeliet et  al.,1996;Toomey et  al., 1996). Hence,
loss in early pregnancy mostprobably accounts for the lack
of TF-null individuals in clinical practice.
Deficiencies of FVIII (hemophilia A) and FIX (hemo-
philia B) share an indistinguishable clinical phenotype char-
acterized by bleeding into muscles, joints,and other organs,
with consequent damage. They are both X-linked recessive-
disorders affecting one in 5000 and one in 30,000 males,
respectively. Clinically,hemophilia is noted to be severe,
moderate, or mild, and this correlates with theresidual level
of activity of the affected factor, which in turn relates to the
precisegene defect. Thus, severely affected patients bleed
spontaneously and have lessthan 2% of the normal level
of factor in their blood; moderately affected patientsbleed
after minor trauma and have 2–5% of the normal level; and
mildly affectedindividuals have more than 5% of the nor-
mal level and bleed only after severe trauma.
 The severity of the bleeding in hemophilia A and B sup-
ports a key role for theFIXa-FVIIIa (tenase) complex in the
amplification phase of blood coagulation.Half of all severe
hemophilia A is caused by spontaneous homologousrecom-
bination resulting in inversion of the F8gene (Figure 25.3).
Two types of inversioninvolving a sequence in intron 22
and either the proximal (int22h-2) or the distal(int22h-3)
of two extragenic copies of this sequence were thought to
be responsiblefor the two forms of the inversion observed
in hemophilia A  patients(Naylor et  al., 1995). However,
more recent genomic sequence data following thecomple-
tion of the DNA sequence of the human X chromosome
(Ross et  al., 2005)indicates that recombination leading
to inversion can only occur between the intron22 and
distal copies of the sequence (Ross and Bentley, 2005).
Recombinationbetween intron 22 and the proximal cop-
ies of the sequence would be predicted tolead to duplica-
tion and deletion, since they are in the same orientation to
each other.The new sequence information also revealed that
int22h-2 and int22h-3 form thearms of a large palindrome.
It has therefore been suggested that recombinationbetween
the arms of the palindrome may occur, creating alleles
where either theint22h-2 or the int22h-3 occupies the prox-
imal position, thus explaining theobserved rearrangements
(Bagnall et al., 2005).
A second inversion, found in 5% of severe hemophilia
A cases, involves a 1 kb sequence in intron 1 (int1h-1) and
a repeated copy (int1h-2) 140 kb telomeric of the F8 gene
between the C6.1A and VBP1 genes. The inversion results
in the production of two chimeric mRNAs that comprise
either exon 1 of the F8 gene and exons 2–6 of the VBP1
gene or exons of the C6.1A gene and exons 2–26 of the F8
gene. The inversion causes severe hemophilia A because the
coding sequence for FVIII (exons 2–26) are downstream of
the translation stop codon of the C6.1A gene (Brinke et al.,
1996; Bagnall et al., 2002).
FXI deficiency (hemophilia C) is an autosomal bleeding
disorder that isnot completely recessive since heterozygotes
have a mild but definite bleedingtendency. FXI circulates
in plasma as a homodimer. The majority of identifiedmuta-
tions in the F11 gene associated with deficiency either pre-
vents or greatlyreduces protein expression. The dominant
negative effect of mutations maytherefore be due to a fail-
ure to secrete heterodimers of mutant and wild-type FXI
or tothe secretion of functionally compromised heterodi-
mers. The severity of bleedingin this disorder is relatively
mild, suggesting that the back-activation of FXI bythrom-
bin may only be required in severe trauma. In the Ashkenazi
Jewishpopulation, where the gene frequency for FXI muta-
tions is as high as 8%, fourdifferent mutations have been
3 9 8   •   G eno m ic s in C linical P ractice
identified:  type I  mutation occurs in the 5′ splicedonor
site of intron N; type II mutation in exon 5 is a nonsense
mutation (FXIE117X); type III mutation is a missense
mutation (FXI F283L) in exon9; type IVmutation is a 14
base-pair deletion at the junction of exon 14 and intron
N.  Theprevalence of these mutations is due to a founder
effect within this population, andthese mutations account
for only a small percentage of disease alleles in non-Jewish
patients (reviewed in O’Connell, 2003).
Deficiency of either FV, FX, or PT is exceedingly rare,
presumablybecause total deficiency is incompatible with
life, since complete gene knockout ofthese genes in mice
is embryonically lethal (Cui et  al., 1996;Sun et  al., 1998;
Dewerchin et al., 2000). Fibrinogen deficiency is also rare;
however, total deficiency offibrinogen in mice is compat-
ible with normal in uterodevelopment, although theydo
have a bleeding phenotype (Suh et al., 1995). Of note, some
types ofdysfibrinogenemia are associated with a throm-
botic rather than a bleedingtendency.Deficiency of PAI-1
results in a failure to control plasmingeneration adequately.
Excess plasmin increases fibrin degradation and clot
disintegration,leading to a bleeding phenotype. In contrast,
mice deficient for PAI-1 (serpine1–/– )do not show sponta-
neous bleeding or delayed bleeding even after amputation
of thetail (Carmeliet et  al., 1993). Deficiency of PAI-1 is
extremely rare, and once again, itshould be noted that raised
levels of PAI-1 are associated with a thrombotictendency.
T H E T H RO M B OT I C D I S O R D E R S
The failure to control the generation of thrombin or
impaired neutralization ofthrombin causes most inherited
thrombophilia. AT, when bound toglycosaminoglycans
on the surface of endothelial cells, neutralizes theproco-
agulants thrombin, FIXa, FXa, and FXIa. PC controls the
generation ofthrombin by inactivating the procoagulant
cofactors FVa and FVIIIa when incomplex with its cofac-
tor PS. Deficiency of AT (SERPINC1), PC (PROC), or PS
(PROS1)are all associated with a thrombotic phenotype
but are relatively rare (de Lau et al., 2010). HomozygousAT
deficiency is probably incompatible with life. Individuals
with homozygous PCor PS deficiency present soon after
birth with purpura fulminans or severe venousthrombosis;
however, these are exceedingly rare. Individuals heterozy-
gous fordeficiency of AT, PC, or PS are all susceptible to
venous thrombosis.
The most frequently occurring inherited defects associ-
ated with athrombotic tendency are resistance to activated PC
caused by a pointmutation in the gene for FV (FV R506Q;
FV Leiden) and a mutation in theprothrombin gene (e.g., F2
 Xq28 Xq28 Xq28

3' 3' 3'

FVIII FVIII FVIII
Gene int22h-1 Gene int22h-1 Gene int22h-1

5' 5' 5'

int22h-2 int22h-3
int22h-3 int22h-2

Xq28 Xq28

3' 3'
FVIII FVIII
Gene int22h-3 Gene int22h-2
int22h-2 int22h-3

5' 5'

Xq28 Xq28
FVIII FVIII
Gene Gene
int22h-3 Ex23-26 int22h-2
Ex23-26
int22h-2 int22h-3

5' 5'

FVIII FVIII
Gene Gene
Ex1-22 Ex1-22
int22h-1 int22h-1

Figure 25.3 
Inversions in Xq28 causing hemophilia A. Proposed mechanism causing polymorphic inversion and recombination between int22h
repeats leading to inversion of the F8gene. The F8 gene is indicated by the large arrow. Inversions disrupting the F8gene result from recombination
between int22h-1 region (in intron 22 of the F8gene) and either int22h-2 or int22h-3, which lie 400 kb distal to F8. The three copies of the int22h
sequence are indicated by the red, blue, and green colored boxes, and their orientation relative to the intron 22 copy is indicated by the arrowhead.
The distal copies (int22h-2 and int22h-3) are in opposite orientation to each other and are flanked by a large, imperfect palindrome, indicated by
blue boxes. Proposed recombination between the arms of the palindrome would generate alleles where either the in22h-2 or the int22h-3 is most
telomeric and in the opposite orientation to int22h-1.
G20210A). The population frequency of FV Leiden is 5%in
ethnic Europeans. FVa is inactivated by activated PC through
proteolyticcleavage at R506–S507 and R1765–L1766. The
cleavage at R506 is rate limiting,and the mutation FV R506Q
is resistant to cleavage by activated PC. The mutationin the
prothrombin gene occurs in the 3′ untranslated region of the
gene and isassociated with raised prothrombin levels. Its fre-
quency in Europeans is approximately 2%.
Elevated levels of FVIII are a strong independent risk
factor for the development of venousthromboembolism;
however, the cause is unknown (reviewed in Seligsohn and
Lubetsky, 2001;Feinbloom and Bauer, 2005).In contrast, a
failure to appropriately regulate fibrinolysis and thus clot-
disintegration may also be associated with a thrombotic
tendency; however, theseconditions (deficiency of tPA and
plasminogen; increased PAI-1 activity) are veryrare.
C O M PA R AT I VE S E Q U E N C E A N A LYS I S
The impact of the completed sequence of the human
genome on our understandingof hemostasis has been much
less than for other inherited disorders, since most ofthe rel-
evant genes were cloned before the genomic era. However,
the identification andcharacterization of the genes respon-
sible for combined FV and FVIII deficiencyand of Type II
multiple coagulation factor deficiency were aided by this
genomicinformation. The completion of the sequence of
other vertebrate genomes has,however, provided a fascinat-
ing insight into the evolution of the coagulationnetwork.
In addition, thecomparative sequence information has been
invaluable ininterpreting naturally occurring mutations
in blood coagulation deficiencies(Gomez et  al., 2004).
Comparison of the amino acid sequence alignments has
alsoprovided important insights into structure–function
relationships of the coagulationfactors. This information
may aid the design of novel recombinant coagulationfac-
tors for replacement therapy, in particular FVIII molecules
with modified domains (Ward et  al., 2011).Comparative
sequence analysis of non-coding sequences may also fur-
therour understanding of regulatory gene sequences.
G E N O M E -W I D E A SS O C I AT I O N
STUDIES
Genome-wide association studies (GWAS) have made an
enormous impact on the identification of genetic loci asso-
ciated with human diseases and their quantitative risk fac-
tors (http://www.genome.gov/gwastudies/). The etiology
4 0 0   •   G eno m ic s in C linical P ractice
of thrombosis in either the arterial or the venous vascular
system is complex and multifactorial, involving both envi-
ronmental and genetic factors. The GWAS approach is
particularly suited to identifying the genetic component of
complex diseases—however,in order to provide adequate
statistical power to identify common variants,these studies
require the genotyping of unprecedented numbers of indi-
viduals. Atherothrombosis represents an advanced stage
of atherosclerosis, in which plaque rupture or erosion has
triggered coagulation leading to the formation of an arte-
rial thrombosis. It is perhaps not surprising,therefore, that
the susceptibility loci identified for coronary artery disease
and myocardial infarction are related to the initiation and
progression of the disease rather than to the thrombotic
end-point (Lotta, 2010).Despite the success of GWAS in
identifying over 34 different genetic loci at which common
variants have been associated with coronary artery disease,
it is likely that these loci only explain a modest fraction of
the total heritability, and it is also likely that some common
coagulation-factor variants may contribute to inherited risk
of arterial disease (Peden and Farrall, 2011).
In contrast, GWAS studies on venous thrombosis have
to date been extremely limited and involved the analy-
sis of much smaller cohorts(Morange and Trgouet, 2011;
Germain et al, 2011). Prior to GWAS studies, as discussed
above, a number of variants associated with venous throm-
bosis had been identified in the genes encoding SERPINC1,
PROC, PROS1, F2, FGG, F5, and ABO.The first three
genes harbor multiple private mutations rather than com-
mon variants. In contrast, the latter four genes harbor more
common variants,which result in variations in qualitative
or quantitative differences in coagulation factors. In the
GWAS study, only the common variants in the F5 gene and
ABO genes reached statistical significance, although a num-
ber of other loci were identified (HIVEP1, C4BPA, TC2N,
GP6 and F11). The novel gene loci, however, confer a mod-
est increase in the risk of venous thrombosis (odds ratios
of 1.10–1.35). GWAS studies have also been carried out
on quantitative traits believed to be associated with venous
thrombosis: activated partial thromboplastin time, protein
S and VWF plasma levels have contributed to the identifica-
tion of novel susceptibility loci for venous thrombosis.
P H A R M AC O G E N O M I C S A N D
PER SONALIZED MEDICINE?
Expectations are high that increasing knowledge of the
genetic basis of disease will eventually lead to person-
alized medicine; that is, preventative and therapeutic
interventions that are targeted to at-risk individuals on the
basis of their genetic profiles. However, many diseases, such
as cardiovascular disease are complex and multifactorial,
and as discussed above, the identification of genetic vari-
ants associated with an increased risk is not trivial and their
predictive value may be limited.Pharmacogenomics—the
study of genetic interactions with pharmacotherapy—is a
promising area for applying genetics to personalized medi-
cine. If it can be successfully implemented to deliver safer
and more effective individualized therapy, it has the poten-
tial to improve health outcomes. However, the journey from
identifying genetic variants associated with an alteration in
a drug target or metabolizing enzyme, to predicting a clini-
cally relevant therapeutic outcome, can be long and fraught.
Two examples from the field of hemostasis and thrombosis
illustrate the problems in validating and implementing per-
sonalized medicine based on genotype.
WA R FA R I N
Warfarin is prescribed to over 2  million patients in the
United States for the treatment of thromboembolic events
associated with atrial fibrillation, venous and arterial
thrombosis, orthopedic surgery, and prosthetic heart valves.
Warfarin is an oral anticoagulant that acts by interfering
with the synthesis of vitamin-K dependent clotting factors
(prothrombin, factors VII, IX, FX and protein C) in the
liver; however, clinical management is difficult because of
a narrow therapeutic index and marked inter-patient vari-
ability in drug pharmacokinetics and pharmacodynamics;
resulting in dose requirements that may vary more than
10-fold. Complications from inappropriate warfarin dos-
ing are one of the most common causes of emergency room
visits and hospitalization; the most common adverse effect
is major bleeding. The risk of complications is particularly
high in the first 90 days of therapy (the induction phase);
therefore, the anticoagulant effect needs to be closely moni-
tored.Warfarin dose requirements are influenced by dietary
intake of vitamin K, age, gender, medication, and genetic
variation. At least 30 genes may be involved in modulat-
ing the action of warfarin: including those involved in the
transport and metabolism of warfarin; those involved in the
transport of vitamin K, and those involved in the vitamin
K cycle. The most important gene in the pharmacokinetics
of warfarin is CYP2C9 (cytochrome P-450 [CYP] 2C9),
whereas the central gene in the pharmacodynamics of war-
farin is VKORC1 (vitamin K epoxide reductase complex
subunit 1 gene).
Warfarin is an equal mixture of the S and R enan-
tiomers, however the S-warfarin is three to five times
G enetic s and G eno m ic s in C linical H e m atology, I :   H e m o s ta s i s and T hro m b o s i s   •   401
more potent than R-warfarin. Metabolism of S-warfarin
is mediated by CYP2C9, whereas R-warfarin is medi-
ated by CYP2C19, CYP1A2, and CYP3A4. Many dif-
ferent polymorphisms in the CYP2C9 gene have been
described, with three alleles in particular having been
studied in detail in relation to warfarin metabolism. The
*1 allele is the wild-type and is associated with normal war-
farin metabolism; the CYP2C9*2(430C>T) allele and
the CYP2C9*3(1075A>C) variants are associated with
reduced enzymatic activity and therefore reduced warfa-
rin metabolism. These variants are relatively common in
Caucasians:  1% homozygous and 22% heterozygous for
the CYP2C9*2 allele and 0.4% homozygous and 15%
heterozygous for the CYP2C9*3 allele. Homozygosity for
CYP2C9*2 or CYP2C9*3 reduces enzyme activity to 12%
and 5% of wild-type levels respectively, increasing the risk
for bleeding with standard warfarin doses.
Vitamin K epoxide reductase (VKOR) is the tar-
get enzyme for warfarin and is involved in regenerat-
ing vitamin K from vitamin K epoxide generated in the
gamma-carboxylation of the coagulation factors. A  num-
ber of common polymorphisms have been identified in
VKORC1, some of which are associated with a need for
lower doses of warfarin.
In 2007 the US Food and Drug Administration
added pharmacogenomic information to the warfarin
label alerting clinicians to the impact of polymorphisms
in CYP2C9 and VKOCR1 on warfarin dose and the risk
of bleeding and in 2009 recommended genotyping prior
to initial dosing (http://www.accessdata.fda.gov/scripts/
cder/drugsatfda/index.cfm?fuseaction=Search.Label_
ApprovalHistory#labelinfo). However, the prospective
clinical studies to date,incorporating pharmacogenomic
prediction models for warfarin dosing, have been small and
therefore have not provided conclusive evidence for recom-
mending CYP2C9 and VKORC1 genotyping when initi-
ating a patient on warfarin.There are several large,ongoing
studies that will provide clinical validity for personalized
warfarin therapy.The clinical utility of pharmacogenomic
testing will ultimately depend on the availability, speed, and
cost of genotyping (see Titus, 2012, for discussion).
CLOPIDOGREL
Clopidogrel has become the mainstay oral anti-platelet
therapy to prevent recurrent ischemic eventsin patients
with acute coronary syndrome and/or undergoing coronary
stent implantation. The placement of a bare or drug-eluting
metal stent in the coronary artery to improve myocar-
dialblood flow increases the risk of platelet-mediated
thrombosis. Despite receiving a combination anti-platelet
therapy of aspirin and clopidogrel, in-stent thrombosis
occurs in 1–4% of patients undergoing stent implantation,
resultingin myocardial infarction in up to 80% and mortal-
ity in up to 40% of cases. With over 1 million individuals
undergoing coronary stenting in the United States every
year, in-stent thrombosis represents a catastrophic compli-
cation for thousands of individuals each year.
Thirty percent of individuals show a marked
inter-individual variability in the anti-platelet effects of
clopidogrel.The inter-patient variability in the response
to clopidogrel is multifactorial and involves environmen-
tal, clinical, and genetic factors. Clopidogrel is an inac-
tive prodrug that is metabolized to its active form in vivo
that then inhibits the ADP P2Y12 receptor, a key media-
tor of platelet activation. The bio-availability of the active
form of clopidogrel is affected by a number of different
enzymes. Its absorption in the intestine can be modulated
by the intestinal efflux transporter ABCB1 (also known as
P-glycoprotein or multi-drug resistance 1 protein MDR1).
Once absorbed, 85% of the drug is converted to an inactive
form by the action of esterases; the remaining 15% under-
goes a two-step activation process in the liver catalyzedby
cytochrome P450 enzymes and an esterase paraoxonase-1
(PON1). The pharmacologically active product binds
irreversibly to the platelet P2Y12 receptor inhibiting its
activation.
Insights into the basis of inter-individual heterogene-
ity in response to clopidogrel began with a candidate gene
approach that identified a common polymorphism(s)
in the cytochrome 2C19 gene (CYP2C19) associated
with impaired platelet responsiveness to clopidogrel.
Furthermore, GWAS identified a single genomic region on
chromosome 10q24 that included the cytochrome cluster
and specifically CYP2C19 as the only common variant
associated with the heterogeneous response to clopidogrel,
although this represented only a small part of the herita-
bility of the platelet response to clopidogrel.In 2010 the
U.S. Food and Drug Administration added a boxed warn-
ing to the clopidogrel label to alert clinicians to the impact
of CYP2C19genotype on the drug’s efficacy. Despite the
strong evidence of genotypical variation having a significant
impact on the effectiveness of clopidogrel, pharmacoge-
netic testing has not been introduced into routine manage-
ment of patients undergoing coronary artery stenting. Two
of the road-blocks to introduction are cost and the immedi-
ate availability of genetic test results, as clopidogrel treat-
ment is often initiated immediately on patient presentation
in the emergency room. Although more rapid “point-of-
care” tests are now available, there are those who argue that
4 0 2   •   G eno m ic s in C linical P ractice
a phenotypical platelet function assay, which is cheaper and
faster, should take into account both genotypical and other
influences on clopidogrel platelet action (Cuisset et  al.,
2012; Godschalk et al., 2012).
REFERENCES
Albrecht, C, McVey, JH, Elliott, JI, et al. (2005). Anovel missense muta-
tion in ABCA1 results in altered protein traffickingand reduced
phosphatidylserine translocation in a patient with Scottsyndrome.
Blood 106:542–549.
Bagnall, RD, Giannelli, F, and Green, PM (2005). Polymorphism and
hemophilia Acausing inversions in distal Xq28: a complex picture.
JThrombHaemost,3:2598–2599.
Brenner, B, Sanchez-Vega, B, Wu, SM, Lanir, N, Stafford, DW, and
Solera, J (1998). A missense mutation in gamma-glutamyl carbox-
ylase gene causescombined deficiency of all vitamin K-dependent
blood coagulation factors.Blood, 92:4554–4559.
Bugge, TH, Xiao, Q, Kombrinck, KW, et  al. (1996). Fatal embry-
onic bleeding events in mice lacking tissue factor, the cell-
associatedinitiator of blood coagulation. ProcNatlAcadSciUSA, 93:
6258–6263.
Carmeliet, P, Mackman, N, Moons, L, et al. (1996). Role of tissue factor
in embryonic bloodvessel development. Nature, 383:73–75.
Carmeliet, P, Stassen, JM, Schoonjans, L, et  al. (1993). Plasminogen
activator inhibitor-1gene-deficient mice. II. Effects on hemostasis,
thrombosis, andthrombolysis. JClinInvest, 92:2756–2760.
Cui, J, O’Shea, KS, Purkayastha, A, Saunders, TL, and Ginsburg, D
(1996). Fatalhaemorrhage and incomplete block to embryogenesis
in mice lackingcoagulation factor V. Nature, 384:66–68.
Cuisset, T, Morange, P-E, and Alessi, M-C (2012). Recent advances in
the pharmacogenetics of clopidogrel. Hum Genet, 131:653–664.
Davidson, CJ, Hirt, RP, Lal, K, et  al. (2003). Molecular evolution of
the vertebrate bloodcoagulation network. ThrombHaemost,89:
420–428.
Davidson, CJ, Tuddenham, EG, and McVey, JH (2003). 450  million
years ofhemostasis. ThrombHaemost, 1:1487–1494.
Dewerchin, M, Liang, Z, Moons, L, et al. (2000). Blood coagulation fac-
tor X deficiency causes partialembryonic lethality and fatal neonatal
bleeding in mice. ThrombHaemost, 83:185–190.
Drake, TA, Morrissey, JH, and Edgington, TS (1989). Selective cellular
expressionof tissue factor in human tissues. Implications for disor-
ders of hemostasisand thrombosis. AmJPathol, 134:1087–1097.
Feinbloom, D and Bauer, KA (2005). Assessment of hemostatic risk
factors inpredicting arterial thrombotic events. Arterioscler Thromb
Vasc Biol, 25: 2043–2053.
Godschalk, TC, Hackeng, CM, and ten Berg, JM (2012). Towards
personalized medicine based on platelet function testing for stent
thrombosis patients. Thrombosis, 617098.
Gomez, K, Laffan, MA, Kemball-Cook, G, et al. (2004). Two novel muta-
tions in severefactor VII deficiency. BrJHaematol, 126:105–110.
Germain, M, Saut, N, Greliche, N, et  al. (2011). Genetics of venous
thrombosis:  insights from a new genome wide association study.
PLoS One, 6:e25581.
Hanumanthaiah, R, Day, K, and Jagadeeswaran, P (2002). Comprehensive
analysis ofblood coagulation pathways in teleostei:  evolution of
coagulation factorgenes and identification of zebrafish factor VIII.
Blood Cells MolDis, 29:57–68.
De Lau, LML, Leebeek, FWG, de Maat, MPM, Koudstaal, PJ, and
Dippel, DWJ (2010). A review of hereditary and acquired coagu-
lation disorders in the aetiology of ischaemic stroke. Int J Stroke,
5:385–394.
Li, T, Chang, CY, Jin, DY, Lin, PJ, Khvorova, A, and Stafford, DW
(2004).Identification of the gene for vitamin K epoxide reductase.
Nature, 427:541–544.
Lotta, LA (2010). Genome-wide association studies in atherothrombo-
sis. Eur J Intern Med, 21:74–78.
Morange, PE and Tregouet, DA (2011). Lessons from genome-wide asso-
ciation studies in venous thrombosis. J Thromb Haemost, 9:258–264.
Mousallem, M, Spronk, HM, Sacy, R, Hakime, N, and Soute, BA (2001).
Congenital combined deficiencies of all vitamin K-dependent
­coagulationfactors. Thromb Haemost, 86:1334–1336.
Naylor, JA, Buck, D, Green, P, Williamson, H, Bentley, D, and Giannelli,
F (1995).Investigation of the factor VIII intron 22 repeated region
(int22h) and theassociated inversion junctions. Hum Mol Genet,
4:1217–1224.
Neerman-Arbez, M, Johnson, KM, Morris, MA, et al.(1999). Molecular
analysis of the ERGIC-53 gene in 35 families with combined factor
V-factor VIII deficiency. Blood, 93:2253–2260.
Nichols, WC, Terry, VH, Wheatley, MA, et al. (1999). ERGIC-53 gene
structure and mutation analysis in 19 combinedfactors V and VIII
deficiency families. Blood, 93:2261–2266.
O’Connell, NM (2003). Factor XI deficiency—from molecular genet-
ics to clinicalmanagement. Blood Coagul Fibrinolysis, 14 (Suppl. 1):
S59–S64.
Peden, JF and Farrall, M (2011). Thirty-five common variants for coro-
nary artery disease:  the fruits of much collaborative labour. Hum
Mol Genet, 20:R198–R205.
Ross, MT and Bentley, DR (2005). More on:  polymorphism and
hemophilia Acausing inversions in distal Xq28: a complex picture.
J Thromb Haemost,3:2600–2601.
Ross, MT, Grafham, DV, Coffey, AJ, et al. (2005). The DNA sequence of
the human X chromosome. Nature, 434:325–337.
Rost, S, Fregin, A, Ivaskevicius, V, et al. (2004). Mutations in VKORC1
cause warfarinresistance and multiple coagulation factor deficiency
type 2. Nature, 427:537–541.
G enetic s and G eno m ic s in C linical H e m atology, I :   H e m o s ta s i s and T hro m b o s i s   •   403
Seligsohn, U and Lubetsky, A (2001). Genetic susceptibility to venous
thrombosis. N Engl J Med, 344:1222–1231.
Spronk, HM, Farah, RA, Buchanan, GR, Vermeer, C, and Soute,
BA (2000).Novel mutation in the gamma-glutamyl carboxylase
gene resulting incongenital combined deficiency of all vitamin
K-dependent bloodcoagulation factors. Blood, 96:3650–3652.
Suh, TT, Holmback, K, Jensen, NJ, et al. (1995). Resolution of spontane-
ous bleeding eventsbut failure of pregnancy in fibrinogen-deficient
mice. Genes Dev, 9:2020–2033.
Sun, WY, Witte, DP, Degen, JL, et al. (1998). Prothrombin deficiency
resultsin embryonic and neonatal lethality in mice. Proc Natl Acad
Sci USA, 95:7597–7602.
Titus, K, (2012). Warfarin pharmacogenetics: a waiting game. http://
www.cap.org/apps/cap.portal?_nfpb=true&cntvwrPtlt_action
Override=%2Fportlets%2FcontentViewer%2Fshow&_window
Label=cntvwrPtlt&cntvwrPtlt%7BactionForm.contentReference
%7D=cap_today%2F0512%2F0512a_­warfarin.html&_state=
maximized&_pageLabel=cntvwr
Toomey, JR, Kratzer, KE, Lasky, NM, Stanton, JJ, and Broze, GJ, Jr.
(1996). Targeted disruption of the murine tissue factor gene results
in embryoniclethality. Blood, 88:1583–1587.
Ward, NJ, Buckley, SMK, Waddington, SN, et  al. (2011). Correction
of hemophilia A using codon optimized B-domain variants. Blood,
117:798–807.
Wu, SM, Stafford, DW, Frazier, LD, et al. (1997). Genomic sequence and
transcription start site for the human gamma-glutamyl carboxylase.
Blood, 89:4058–4062.
Zhang, B, Cunningham, MA, Nichols, WC, et al.(2003). Bleeding due
to disruption of a cargo-specific ER-to-Golgi transportcomplex.
Nat Genet, 34:220–225.
Zhang, B, McGee, B, Yamaoka, JS, et al. (2006). Combined deficiency
of factor V and factor VIII is due to mutations in either LMAN1 or
MCFD2. Blood, 107:1903–1907.
 26.
GENETICS AND GENOMICS IN CLINICAL
HEMATOLOGY, II: INHERITED DISORDER S
OF HEMOGLOBIN
Sir David J. Weatherall
INTRODUCTION
The inherited disorders of hemoglobin are by far the com-
monest single-gene disorders. It is currently estimated that
between 300,000 and 400,000 babies are born each year
with one of these disorders, the bulk of the births being in
the low- or middle-income countries.1 Because the high fre-
quency of these conditions reflects heterozygote resistance
to malaria, a high frequency of consanguinity in many of the
countries in which the disease is common, and the epidemi-
ological transition whereby improved nutrition and public
health facilities are allowing more affected babies to survive
to come for diagnosis and treatment,2 it seems likely that
the birth rate of these conditions will continue to expand
for the foreseeable future. Here, I shall review briefly what
is known about the molecular pathology, pathophysiology,
clinical features, and approaches to prevention and man-
agement of the inherited hemoglobin disorders, and then
define the questions that remain in this field and how they
might be approached by more recent technology derived
from genomics.
T H E S T RU C T U R E A N D G E N ET I C
CONTROL OF HEMOGLOBIN
T H E N O R M A L HUM A N
H E MO G L O B I NS
The different oxygen requirements during embryonic, fetal,
and adult life are met by the production of different struc-
tural hemoglobins at each stage of development. They all
have the same tetrameric structure, consisting of two differ-
ent pairs of globin chains, each carrying one heme moiety.
Adult and fetal hemoglobins have α chains combined with
404
β chains (HbA, α2β2), δ chains (HbA2, α2δ2) and γ chains
(HbF, α2γ2). In embryonic life, α-like chains called “ζ chains”
combine with γ chains to produce Hb Portland (ζ2,γ2), or
with ε chains to make Hb Gower 1 (ζ2ε2), while α and ε
chains form Hb Gower 2 (α2ε2). Human fetal hemoglobin is
heterogeneous, consisting of two varieties of gamma chain
that differ only in their amino-acid composition at position
136, which may be occupied by either glycine or alanine.
The functional properties of hemoglobin and how they
adapt to anemia have been reviewed recently.3
Genetic Control of Hemoglobin
The α and β-like globin chains are the products of two
gene families that are found on different chromosomes
(Figure  26.1). The β-like globin genes form a linked clus-
ter on chromosome 11, spread over approximately 60 kb.
The genes that form this cluster are arranged in the order
5´-ε-Gγ-Aγ-ψβ-δ-β-3´. The α-like genes also form a linked
cluster, in this case on chromosome 16, in the order
5´-ζ-ψζ-ψα1-α2-α1-θ-3´. The ψβ,ψζ and ψα1 genes are
“pseudogenes”; that is, they have strong sequence homology
with the β, ζ, and α genes, but contain a number of differ-
ences that prevent them from directing the synthesis of any
products. Like the θ gene, they may be remnants of genes
that were functional at an earlier state of human evolution.
The β-like globin genes contain two introns, one
of 122–130 base pairs between codons 30 and 31 and
one of 850–900 base pairs between codons 104 and
105. Similar, though smaller, introns are found in the
α and ζ globin genes. These introns and exons, together
with short non-coding sequences at the 5´ and 3´ ends
of the genes, represent the major functional regions of
the particular genes. However, there are also extremely
important regulatory sequences that subserve these

31 32 99 100
ζ ψζ ψα2 ψα1 α2 α1
CHROMOSOME 16
ζ2ε2 ζ2γ2
Hb Gower 1 Hb Portland
Embryo
Figure 26.1  Genetic control of hemoglobin production (from Weatherall and Clegg, 2001, ref. 6).
functions that lie outside the genes themselves. At the 5´
non-coding (flanking) regions of the globin genes there
are blocks of nucleotide homology. The first, the ATA
box, is about 30 bases upstream of the initiation codon.
The second, the CCAAT box, is about 70 base pairs
upstream from the 5´ end of the genes. About 80–100
bases further upstream, there is the sequence GGGGTG
or CACCC, which may be inverted or duplicated. These
three highly conserved DNA sequences, or promoter ele-
ments, are involved in the initiation of the transcription
of the individual genes. Finally, in the 3´ non-coding
region of all the globin genes, there is the sequence
AATAAA, which is the signal for cleavage and polyA
addition to RNA transcripts.
The globin gene clusters also contain several sequences
that constitute regulatory elements, which interact to pro-
mote erythroid-specific gene expression and coordination
of the changes in globin gene activity during develop-
ment. These include the globin genes themselves and their
promoter elements, enhancers, and “master” regulatory
sequences, called, in the case of the β globin gene cluster, the
“locus control region” (LCR), and, in the case of the α genes,
“HS40” (a nuclease-hypersensitive site 40 kb upstream from
the α globin genes). Each of these sequences has a modular
structure made up of an array of short motifs that represent
the binding sites for transcriptional activators or repressors.
The regulatory regions contain sequence motifs for vari-
ous ubiquitous and erythroid-restricted transcription fac-
tors. There are binding sites for these factors in each of the
globin gene promoters and the hypersensitive site regions
in the various regulatory elements. While the ubiquitous
factors are found in a variety of cell types, there are many
transcription factors that appear to be restricted to red-cell
precursors, including GATA-1, EKLF, and NF-E2, and oth-
ers. Binding of the specific factors activates the LCR, and it
seems likely that this complex becomes approximated to the
G enetics and G enomics in C linical H ematology, I I   •   4 0 5
1 KILOBASE
30 31 104 105
θ1 ε Gγ Aγ ψβ δ β
CHROMOSOME 11
α2ε2 α2γ2 α2β2 α2δ2
Hb Gower 2 HbF HbA HbA2
Fetus Adult
different genes in the β globin gene cluster during different
stages of maturation.
The mechanisms that control the switches from embry-
onic to fetal, and fetal to adult, hemoglobin production are
still not understood. A number of loci have been identified
that are involved in the regulation of persistent fetal hemo-
globin production in adult life. These are the subject of sev-
eral recent reviews4,5 and are considered further in a later
section. A more extensive review of the genetic control of
hemoglobin production in general has also been published
recently.3
T H E C L A S S I F I C AT I O N O F T H E
HEMOGLOBIN DISORDER S
The hemoglobin disorders are broadly divided into the thal-
assemias, diseases characterized by a reduced rate of synthe-
sis of one or another of the globin chains, and the structural
hemoglobin variants, which alter the morphology, stability,
or function of the hemoglobin molecule.6 These conditions
are summarized in Table 26.1.
The thalassemias are divided into the α and β thalas-
semias, in which there is defective α or β chain synthesis,
respectively. The β thalassemias are divided into the β+ and β0
thalassemias in which there is either a reduced or a complete
absence of β chain synthesis, respectively. Other important
forms of β thalassemia include the compound heterozygous
states for β thalassemia and structural hemoglobin variants
such as hemoglobins S, C, or E. By far the commonest of
these is hemoglobin E (HbE) β thalassemia. Hemoglobin
E, which occurs at a high frequency throughout many parts
of Asia, results from a mutation that activates a new splice
site; hence, it is produced at a reduced rate and behaves like
a mild form of β thalassemia. HbE β thalassemia is by far
the commonest form of the disease in many Asian countries
Table 26.1  GENETIC DISORDERS OF HEMOGLOBIN
Structural Variants
Common disease-associated: Hbs S, E, C, D
Rare variants causing polycythemia, methemoglo-
bin, hemolytic anemia
Rare variants of no clinical significance
Thalassemia
α thalassemia
α+ thalassemia. Deletion.
Non-deletion
α0 thalassemia
β thalassemia
β+ thalassemia
β0 thalassemia
δβ thalassemia
εγδβ thalassemia
Hereditary Persistence of Fetal Hemoglobin
Pancellular
Heterocellular
and constitutes about 50% of the severe forms of β thalas-
semia worldwide.
The α thalassemias are divided into the α+ thalas-
semias, which result from the deletion or inactivation
(non-deletion, or ND) of one of the pair of linked α glo-
bin genes (-α/αα or αND α/αα), and the α0 thalassemias in
which both of the linked α globin genes are lost by dele-
tion (--/αα). The compound heterozygous state for α+ and
α0 thalassemia results in a condition called “Hemoglobin H
(HbH) disease,” while the homozygous state for α0 thalas-
semia results in stillbirth, a condition called the “Hb Bart’s
hydrops fetalis syndrome.”
As well as these common forms of thalassemia, there are
rarer conditions that result from deletions involving one or
more of the globin genes.6 These include the δβ thalassemias
and the εγδβ thalassemias. There is also a heterogeneous
family of conditions called “hereditary persistence of fetal
hemoglobin” that result from various-sized deletions of the
β globin gene complex or from point mutations within this
complex.
There are several hundred structural hemoglobin vari-
ants, although the only common ones are hemoglobins S,
C, and E. Many of the others are harmless, although there
are rare groups of variants that result in chronic hemolytic
anemia, polycythemia, or congenital cyanosis.
Finally, there are two disorders that, although they
are associated with an α thalassemia-like phenotype, are
unrelated to the other common disorders of hemoglobin
4 0 6   •   G enomics in C linical P ractice
synthesis. They are associated with different degrees of men-
tal retardation: one is encoded on chromosome 16 and is
called “α thalassemia/mental retardation 16” (ATR16), and
the other is encoded on the X chromosome (ATRX).
WO R L D D I S T R I B U T I O N
The world distribution of the inherited hemoglobin dis-
orders, together with current estimates of the frequency of
births of infants with these conditions, are the subject of
several extensive reviews.1,2,6,7 In short, the β thalassemias
occur at varying frequencies in the tropical belt extending
from sub-Saharan Africa, through the Middle East and
Indian subcontinent, to Southeast and East Asia. The α+
thalassemias, which are by far the commonest forms of thal-
assemia, have a similar distribution. Although they occur
sporadically in most tropical populations, the α0 thalas-
semias only reach high frequencies in Southeast Asia and in
a few of the Mediterranean island populations. There is, of
course, a varying frequency of these conditions in countries
in which there is a high frequency of immigrants from these
tropical regions.
The sickle-cell gene has arisen at least twice during our
evolutionary past and perhaps even more than once in
Africa. As distinguished by haplotype constitutions, the
African form occurs widely in sub-Saharan Africa, parts
of North Africa, the Mediterranean, and the Middle East,
while the Asian-Indian form occurs widely in localized
parts of the Indian subcontinent and extends into several
Middle Eastern countries.6
Hemoglobin E, probably the commonest structural
hemoglobin variant, occurs at high frequencies in many
populations of the eastern half of the Indian subcontinent
and throughout Southeast Asia. Data from certain local-
ized regions, northeast Thailand for example, show that
the carrier frequency is in excess of 70% of the population.
Haplotype analysis suggests that this variant may have arisen
on more than one occasion. Hemoglobin C is restricted to
parts of West and North Africa.
Recent micro-mapping studies have indicated that
there is remarkable heterogeneity in the distribution of
the genes for thalassemia and abnormal hemoglobins, even
within short geographical distances, in high-frequency
populations.2,8,9 This may reflect heterogeneity in the pres-
ent or past distribution of malaria in these regions. Recent
studies also suggest that some of the variation in the dis-
tribution of the abnormal hemoglobin genes may reflect
complex epistatic interactions between their phenotypes
with respect to malaria resistance.10,11 For example, Africans
with either the sickle-cell trait or the α+ thalassemia trait
are both relatively resistant to the complications of P. falci-
parum malaria, yet if they inherit both traits, this resistant
phenotype is cancelled out.
M O L E C U L A R PAT H O L O GY,
PAT H O P H YS I O L O GY, A N D
C L I N I C A L M A N I F E S TAT I O N S
There is an extensive literature on the pathophysiology and
clinical features of the abnormal hemoglobin disorders,3,6
and these aspects will only be briefly summarized here.
THE THALASSEMIAS
The β thalassemias result from over 200 different mutations
involving the β-globin genes, either point mutations or,
less commonly, deletions.3,6,12 They cause either an absence
of β globin-chain synthesis (β0 thalassemia) or a variably
reduced rate of β-chain production (β+ thalassemia). This
results in imbalanced globin-chain synthesis with a vari-
able excess of α-chains that are unstable and go through a
series of phases with the production of methemoglobin,
hemichromes, and the formation of visible precipitates in
the red-cell precursors. As well as causing mechanical dam-
age to the latter, they cause a wide range of oxidant dam-
age to the membranes of the red-cell precursors, leading
to intramedullary destruction. Since a variable proportion
of red-cell precursors appears to retain the ability to pro-
duce HbF after birth, these red cells come under selection
because there is less globin-chain imbalance due to the pro-
duction of γ chains; hence, elevated levels of HbF are one of
the hallmarks of β thalassemia. The profound anemia that
follows causes increased erythropoietin production with
expansion of the ineffective bone marrow, leading to wide-
spread skeletal changes involving the skull and long bones.
Thus this disease is characterized by profound anemia from
early life and in many cases can only be controlled by life-
long transfusions. The expanded bone marrow, associated
with increased iron absorption, and regular blood transfu-
sions lead to the accumulation of iron, which causes dam-
age to the liver, endocrine glands, and ultimately the heart.
Unless this is controlled, death usually occurs in the second
decade.
The pathophysiology of α thalassemia is different. It
results from many different-sized deletions of one or both
of the linked pairs of α-globin genes, or point mutations
that inactivate them. The overall effect of imbalanced
globin-chain synthesis is different from that which occurs
G enetics and G enomics in C linical H ematology, I I   •   4 0 7
in β thalassemia. Excess β-chains produced in adult life
form β4 tetramers, or HbH, which do not precipitate to a
great degree in the bone marrow but enter the peripheral
blood, which, because of their instability, precipitate as
inclusion bodies and result in a shortened red-cell survival.
In fetal life, α thalassemia is characterized by the produc-
tion of excess gamma chains which form γ4 tetramers,
or Hb Bart’s, which is less unstable than HbH. There are
two major clinical disorders resulting from these different
mutations. The compound heterozygous states for α0 and
α+ thalassemia or α0 thalassemia and a non-deletion form
of α+ thalassemia result in HbH disease. This is character-
ized by a variable degree of hemolytic anemia and enlarge-
ment of the spleen; the interactions between the deletion
forms of α+ thalassemia and α0 thalassemia tend to produce
a milder phenotype than those between the deletion and
non-deletion forms. The homozygous state for α0 thalas-
semia is characterized by the complete absence of α-chain
synthesis and therefore, of the production of both hemoglo-
bins F and A. It results in a condition called the “Hb Bart’s
hydrops syndrome,” in which babies are stillborn or die just
after delivery; their hemoglobin pattern is characterized by
very high levels of Hb Bart’s and persistence of embryonic
hemoglobin production.
S T RU C T U R A L H E MO G L O B I N
VA R I A N T S
The homozygous state for the sickle-cell gene, sickle-cell
anemia, is a chronic hemolytic anemia reflecting the abnor-
mal properties of HbS, which give rise to a distortion of the
red cells, a shortened red-cell survival, and a widespread
spectrum of vascular abnormalities resulting from interac-
tions between the red cells and the vascular endothelium.
The disease has a very variable course, marked by intermit-
tent acute episodes, or crises, characterized by either wide-
spread bone pain, rapid enlargement of the spleen with
sequestration of the red cells, aplasia of the bone marrow as
the result of infection, or stroke. There is also a wide variety
of chronic complications, including aseptic necrosis of the
hip or shoulder joints, chronic renal damage, silent cerebral
infarcts, and many others. The pathophysiology, clinical
manifestations, and complications have been the subject of
extensive reviews.3,13
The homozygous state for HbC is characterized by mild
hemolytic anemia resulting from the relative insolubility of
the hemoglobin variant; the compound heterozygous state
for this variant and HbS, “HbSC disease,” is characterized
by a condition similar to sickle-cell anemia, although usu-
ally of a milder course. Hemoglobin E, in the heterozygous
or homozygous state, does not produce any significant clini- the concept of there being different levels of modification of
cal abnormalities, but when it is inherited in the compound disorders of this type.15,16 Primary modifiers reflect the inter-
heterozygous state with β thalassemia, HbE β thalassemia, action of milder β thalassemia genes with HbE. Secondary
it results in a variably severe form of β thalassemia with a modifiers are those that reduce the degree of globin-chain
remarkably diverse phenotype. As discussed earlier, this imbalance and include variations at loci involved in HbF
condition represents approximately 50% of the severe forms production and the co-inheritance of α thalassemia, which
of thalassemia globally. has a quite remarkable modifying effect on the phenotype
As well as these common structural hemoglobin vari- of Hb E β thalassemia patients. There are also tertiary modi-
ants, there are several groups of extremely rare variants fiers; that is, genetic variants that modify the complications
that, because of their abnormal structure and function, of Hb E β thalassemia, such as the high level of bilirubin
cause a broad spectrum of clinical disorders. They have and profound jaundice that affect some of these patients17
been reviewed in detail recently.3 These include variants and those that reflect the greater absorption of iron. As
that produce hereditary polycythemia, hereditary hemo- well as these genetic modifiers, there is increasing evidence
lytic anemia due to hemoglobin instability, and hereditary that variation in adaptation to anemia may be of critical
methemoglobinemia. importance,18 and that the environment, at least as reflected
through susceptibility to malaria, may be an added compo-
nent in the pattern of heterogeneity.19 However, it should
P H E N OT Y P I C A L D I VE R S I T Y O F be emphasized that, despite characterizing the modifiers of
THE HEMOGLOBIN DISORDER S HbE β thalassemia and other forms of β thalassemia at these
different levels (Figure 26.2), when added together, the vari-
One of the most remarkable features of the hemoglobin dis- ous modifiers still leave well over half of the phenotypical
orders is their marked phenotypical variability, even when heterogeneity unexplained.
they result from the same mutation(s) in the α or β globin The case of sickle-cell anemia is equally complex.
genes. Phenotypical heterogeneity of this type is particu- Interestingly, there is a significant phenotypical difference
larly well demonstrated in the different forms of β thalas- between the form that arose in Africa and that which arose
semia. Although the majority of cases of homozygosity for in a different genetic setting in India. In some, but by no
β thalassemia are profoundly anemic from the first year of
life, and require lifelong transfusions, there is a subset who
are less anemic, do not require regular transfusions, and αααα/αα
may grow and develop quite normally. This disorder, named ααα/αα
αα/αα
“β thalassemia intermedia,” has been widely studied with ––/αα
respect to the causes for its milder phenotype.6,14 In short, –α/αα β0, β+, β++
there are three different mechanisms that may underlie the
mild courses that have been identified. First, there may be Proteolysis EXCESS α-CHAINS γ-chains
mutations of the β globin gene that result in only a relatively (Hb F, α2γ2)
mild defect in β globin chain production. Second, at least Inclusion bodies
three different loci have been identified by family studies, at Ineffective erythropoiesis, hemolysis
which variability appears to result in unusually high levels
of HbF production; hence, less globin chain imbalance and Anemia
a milder phenotype. They are quantitative trait loci (QTL)
Bone Iron
involving HBS1L-MYB on chromosome 6, Xmn-I Gγ disease loading
Jaundice Infection
on chromosome 11, and BCL11A on 2p15.14 Finally, the
co-inheritance of α+ or `α0 thalassemia also results in less
VDR HFE UGT1A1 CO-SELECTION
globin-chain imbalance and has a moderate effect on the ESR1 HLA-DR
severity of the phenotype in this condition. Collagen TNF-α
ICAM-1
The phenotypical diversity of Hb E β thalassemia is even ADAPTATION
DARC
more striking, ranging from a condition requiring lifelong
transfusions to a disorder with completely normal growth ENVIRONMENT
and development, albeit at a relatively low hemoglobin Figure 26.2 
Layers of modifiers of the phenotype of hemoglobin E β
level. Extensive studies of this condition have resulted in thalassemia (modified from Weatherall, 2001, ref. 16).

4 0 8   •   G enomics in C linical P ractice

means all, features of the disease, the latter tends to run a
milder clinical course.13,20 The Indian form of sickle-cell
disease is characterized by a significantly higher level of
HbF production and also by a particularly high frequency
of α thalassemia in affected populations. The reason for
the unusually high level of HbF in this form of the disease
still remains to be explained. Studies over many years have
shown that unusually elevated levels of HbF in all forms of
sickle-cell anemia are associated, in general, with a milder
course.13,21 Even in the African form of the disease, the
co-inheritance of α thalassemia is associated with a reduc-
tion in some of the complications.22
It should be emphasized that most of the information
regarding the underlying basis for the clinical heterogene-
ity of these disorders has been obtained over many years
by relating genotype to phenotype in individual family or
population studies, and, in particular, by the careful analysis
of unusual phenotypes.
The Clinical Importance of the
Phenotypical Diversity
There are several reasons why it is important to obtain fur-
ther information about why these disorders are so pheno-
typically diverse. Currently, they are subject to prevention
programs based on public education and the application of
prenatal diagnosis. This approach has been remarkably suc-
cessful in the case of the β thalassemias, one of the major
reasons being that, in the majority of cases of homozygous
or compound heterozygous β thalassemia, lifelong treat-
ment by transfusions and agents to remove excess iron is
the likely prognosis. As mentioned earlier, the situation in
HbE β thalassemia is different, and it is difficult for genetic
counsellors to anticipate the future course of the illness.
The same issue applies to sickle-cell anemia, which is prob-
ably why prenatal diagnosis for this condition has not been
taken up as widely as that for thalassemia. These problems
may become ever more pressing if noninvasive approaches
to prenatal diagnosis become available.23
Another important issue for the future is that, at the
moment, the management of these diseases is entirely symp-
tomatic, and, except for bone marrow transplantation, there
are no radical cures. Granted, the symptomatic treatment
of both β thalassemia and sickle-cell anemia has improved,
but the patients can still look forward to a life of continu-
ous medical attention. Further work directed towards
the better treatment of these conditions, notably by gene
therapy or therapeutic agents directed at the production of
HbF, all offer potential side effects in the clinical trials that
will be necessary to assess their efficacy; again, it would be
G enetics and G enomics in C linical H ematology, I I   •   4 0 9
extremely helpful to better understand the reasons for the
clinical heterogeneity of these conditions so that patients
and their parents can be given more reliable advice about
the value of the experimental approaches and their poten-
tial risks.
H A S T H E A P P L I C AT I O N O F
G E N O M I C T E C H N O L O GY I M P R O VE D
O U R U N D E R S TA N D I N G O F
T H E H ET E R O G E N E I T Y O F T H E
HEMOGLOBIN DISORDER S?
There are numerous difficulties in defining the ideal condi-
tions to establish genome-wide association studies (GWAS)
for analysis of the reasons for the phenotypical heterogene-
ity of the important hemoglobin disorders. Surprisingly,
there are very few background data regarding heritability
and the different characteristics of these diseases. Indeed,
there is only one published twin study, which focused on
nine identical twins with sickle-cell disease in Jamaica;
there was strong concordance for many of the hematologi-
cal findings, but most of the clinical findings and compli-
cations were discordant.24 In the large Co-operative Study
of Sickle Cell Disease (CSSCD), an analysis of 104 pairs
of siblings resulted in high correlation coefficients for HbF
levels, painful crises, and stroke, with only moderate corre-
lations for the acute chest syndrome, leg ulcers, osteonecro-
sis, and priapism, with a very low correlation for survival.25
Another major difficulty is in the clarity of the definition
of the phenotypes that are required for successful GWAS
studies. As mentioned earlier, the phenotypes in many of
the hemoglobin disorders are quite unstable, and, without
following large numbers of patients carefully over many
years, it is difficult to establish clear phenotypical groups.
This has been a major difficulty with most of the published
GWAS data to date.
So far there have been only limited successes in the appli-
cation of GWAS to the hemoglobin disorders. One study
identified BCL11A and its association with elevated HbF
levels and hence its role in the modification of the pheno-
type of β thalassemia.26 A similar study identified BCL11A
and HBS1L-MYB as loci containing a single-nucleotide
polymorphism that is associated with elevated HbF levels
and a reduction in painful crises in patients with sickle-cell
anemia.27 Another identified the same loci that are involved
with HbF production and the α-globin genes as modifiers
of Hb E β thalassemia.28
With the exception of BCLA11A, the GWAS stud-
ies have so far confirmed earlier work based on more
conventional genetic analyses that showed quite clearly
that genetic factors that increase the level of HbF and the
co-inheritance of α thalassemia are the major modifiers
that are currently known to affect the phenotypes of both
β thalassemia and sickle-cell anemia. The limited number
of GWAS studies that have been carried out in the hemo-
globin field have been reviewed recently,29 and, although
there have been some hints about the importance of the
TGF-β super family and oxidative stress in the variable
pathogenesis of the sickling disorders, very few completely
new concepts about phenotypical variability have been
disclosed.
Another valuable approach to the phenotypical hetero-
geneity of the hemoglobin disorders is the search for patho-
logical polymorphisms among those in the same population
who do not have a hemoglobin disorder. For example,
it was found that 25% of the population of Sri Lanka are
homozygous for the polymorphism in the promoter of the
UGT1A1 gene, which is associated with mildly elevated
levels of serum bilirubin; that is, Gilbert’s syndrome. In
those with Hb E β thalassemia with this genotype, there
were extremely high levels of bilirubin and a very high risk
of gallstones.17 This polymorphism also occurs commonly
in those of African descent, and it was found that when it
was inherited together with the sickle-cell gene, there was
also a greatly increased risk of gallstones.30 Similarly, the
MYH9 and APOL1 genes are associated with renal disease
in African populations and have been strongly related to
sickle-cell disease nephropathy.31
THE FUTURE
The new genomics’ approaches to the search for an under-
standing of the heterogeneity of monogenic disease are
still in their infancy. But it is already clear that one of the
major issues will be the better definition of phenotypes to
match the remarkable advances in genetic analysis. This will
require the study of patient groups who have been followed
for many years, given the phenotypical instability of many
of these conditions over time, and it will require a much
closer level of cooperation between clinicians and molecu-
lar geneticists.
Finally, it should be remembered that, with rare excep-
tions, the known genetic modifiers of the hemoglobin dis-
orders had been determined before the era of genomics.
Hence the new technology should go hand in hand with
classical family studies, attention to the unusual phenotype,
and careful searches for pathological polymorphisms such
as those that are involved in bilirubin metabolism or renal
4 1 0   •   G enomics in C linical P ractice
function in the normal population that might interact with
the inherited hemoglobin disorders.
AC K N OW L E D G E M E N T S
I thank Jeanne Packer and Liz Rose for their help in prepar-
ing this manuscript.
REFERENCES
1. Christianson A, Howson CP, Modell B. March of Dimes Global
Report on Birth Defects. New York: March of Dimes Birth Defects
Foundation; 2006.
2. Weatherall DJ. The inherited diseases of hemoglobin are an emerg-
ing global health burden. Blood. 2010;115(22):4331–4336.
3. Steinberg MH, Forget BG, Higgs DR, Weatherall DJ, eds. Disorders
of Hemoglobin (2nd ed.). New York: Cambridge University Press;
2009.
4. Bauer DE, Orkin SH. Update on fetal hemoglobin gene regu-
lation in hemoglobinopathies. Current Opinion in Pediatrics.
2011;23(1):1–8.
5. Forget BG. Progress in understanding the hemoglobin switch. The
New England Journal of Medicine. 2011;365(9):852–854.
6. Weatherall DJ, Clegg JB. The Thalassaemia Syndromes (4th ed.).
Oxford, UK: Blackwell Science; 2001.
7. Weatherall DJ. Thalassemia as a global health problem: recent prog-
ress towards its control in the developing countries. Annals of the
New York Academy of Science. 2010;1202: 17–23.
8. Colah R, Gorakshakar A, Phanasgaonkar S, et  al. Epidemiology
of beta-thalassaemia in Western India:  Mapping the frequncies
and mutations in sub-regions of Maharashtra and Gujarat. British
Journal of Haematology. 2010;149(5):739–747.
9. O’Riordan S, Hien TT, Miles K, et  al. Large scale screening for
haemoglobin disorders in southern Vietnam:  implications for
avoidance and management. British Journal of Haematology.
2010;150(3):359–364.
10. Penman BS, Pybus OG, Weatherall DJ, Gupta S. Epistatic interac-
tions between genetic disorders of hemoglobin can explain why the
sickle-cell gene is uncommon in the Mediterranean. Proceedings
of the National Academy of Sciences of the United States of America.
2009;106(50):21242–21246.
11. Williams TN, Mwangi TW, Wambua S, et  al. Negative epistasis
between the malaria-protective effects of alpha+-thalassemia and
the sickle cell trait. Nature Genetics. 2005;37(11):1253–1257.
12. Huisman THJ, Carver MFH, Efremov GD. A Syllabus of
Human Hemoglobin Variants (2nd ed.). Augusta, GA: Sickle Cell
Foundation; 1998.
13. Serjeant GR, Serjeant BE. Sickle Cell Disease (3rd ed.). Oxford,
UK: Oxford University Press; 2001.
14. Thein SL. Genetic modifiers of the beta-haemoglobinopathies.
British Journal of Haematology. 2008;141(3):357–366.
15. Premawardhena A, Fisher CA, Olivieri NF, et al. Haemoglobin E β
thalassaemia in Sri Lanka. Lancet. 2005;366 1467–1470.
16. Weatherall DJ. Phenotype–genotype relationships in monogenic
disease:  lessons from the thalassaemias. Nature Reviews Genetics.
2001;2 245–255.
17. Premawardhena A, Fisher CA, Fathiu F, et al. Genetic determinants
of jaundice and gallstones in haemoglobin E beta thalassaemia.
Lancet. 2001;357(9272):1945–1946.
18. Allen A, Fisher C, Premawardena A, et al. Adaptation to anemia in
hemoglobin E-beta thalassemia. Blood. 2010;116(24):5368–5370.
 19. O’Donnell A, Premawardhena A, Arambepola M, et al. Interaction
of malaria with a common form of severe thalassemia in an Asian
population. Proceedings of the National Academy of Sciences of the
United States of America. 2009;106:18716–18721.
20. Serjeant GR. Geographic heterogeneity of sickle cell disease.
In: Steinberg MH, Forget BG, Higgs DR, Nagel RL, eds. Disorders
of Hemoglobin. Cambridge, UK:  Cambridge University Press;
2001:895–905.
21. Higgs DR, Wood WG. Genetic complexity in sickle cell disease.
Proceedings of the National Academy of Sciences of the United States of
America. 2008;105(33):11595–11596.
22. Higgs DR, Aldridge BE, Lamb J, et  al. The interaction of
alpha-thalassaemia and homozygous sickle-cell disease. New
England Journal of Medicine. 1982;306(24):1441–1446.
23. Lo YM, Chiu RW. Noninvasive approaches to prenatal diagno-
sis of hemoglobinopathies using fetal DNA in maternal plasma.
Hematology/Oncology Clinics of North America. 2010;24(6):
1179–1186.
24. Weatherall MW, Higgs DR, Weiss H, Weatherall DJ, Serjeant GR.
Phenotype/genotype relationships in sickle cell disease: a pilot twin
study. Clinical and Laboratory Haematology. 2005;27(6):384–390.
25. Lettre G. The search for genetic modifiers of disease severity in
the beta-hemoglobinopathies. In:  Weatherall DJ, Nathan DG,
Schechter AN, eds. Inherited Disorders of Hemoglobin. New York:
Cold Spring Harbor Publications; 2013;313–324. doi:10.1101/
chsperspect.a015032
G enetics and G enomics in C linical H ematology, I I   •   4 1 1
26. Uda M, Galanello R, Sanna S, et al. Genome-wide association study
shows BCL11A associated with persistent fetal hemoglobin and
amelioration of the phenotype of beta-thalassemia. Proceedings of
the National Academy of Sciences of the United States of America.
2008;105(5):1620–1625.
27. Lettre G, Sankaran VG, Bezerra MA, et al. DNA polymorphisms
at the BCL11A, HBS1L-MYB, and beta-globin loci associate
with fetal hemoglobin levels and pain crises in sickle cell disease.
Proceedings of the National Academy of Sciences of the United States of
America. 2008;105(33):11869–11874.
28. Nuinoon M, Makarasara W, Mushiroda T, et  al. A genome-wide
association identified the common genetic variants influence dis-
ease severity in beta0-thalassemia/hemoglobin E. Human Genetics.
2010;127(3):303–314.
29. Fertrin KY, Costa FF. Genomic polymorphisms in sickle cell dis-
ease:  implications for clinical diversity and treatment. Expert
Reviews in Hematology. 2010;3(4):443–458.
30. Haverfield EV, McKenzie CA, Forrester T, et  al. UGT1A1
variation and gallstone formation in sickle cell disease. Blood.
2005;105(3):968–972.
31. Ashley-Koch AE, Okocha EC, Garrett ME, et  al. MYH9 and
APOL1 are both associated with sickle cell disease nephropathy.
British Journal of Haematology. 2011;155(3):386–394.
 27.
GENETIC AND GENOMICS IN CLINICAL
HEMATOLOGY, III: ACUTE LEUKEMIAS
Kenneth Mills
INTRODUCTION
Acute myeloid leukemia (AML) is a genetically heteroge-
neous disorder characterized by arrest of differentiation
in the myeloid lineage coupled with an accumulation of
immature progenitors in the bone marrow, resulting in
hematopoietic failure. (AML is also characterized by the
somatic acquisition of genetic and epigenetic alterations
that perturb the normal mechanisms of self-renewal, pro-
liferation, and differentiation.) AML is the most common
form of acute leukemia in adults; AML accounts for only
around 10% of pediatric leukemia cases.1
The overall incidence of AML is approximately four in
100,000 in the United States and United Kingdom, with
an increasing incidence from one in 100,000 in people
younger than 35 years old at diagnosis to 15 in 100,000 in
those older than 75 years.2 Despite significant advances in
the understanding of the biology of adult acute myeloid
leukemia (AML), overall survival remains poor for the
approximately 12,000 people diagnosed each year in the
United States due chiefly to the high rate of relapse after
achieving complete remission as well as primary failure of
induction chemotherapy.
AML is characterized by the infiltration of bone mar-
row by abnormal hematopoietic progenitors that disrupts
normal production of erythroid, myeloid, and/or mega-
karyocytic cells. The deregulated proliferation and impaired
differentiation result in an increased survival advantage for
the leukemia cell.3–5 AML arises from a clone of poorly dif-
ferentiated hematopoietic stem cells that contain mutated
genes which confer an advantage for growth or survival. The
clonal leukemia cells develop, expand in number, and even-
tually dominate normal hematopoiesis.
Most patients have a presentation that arises from
the disruption of normal blood component production.
Laboratory studies usually reveal pancytopenia, although a
combination of anemia, thrombocytopenia, and leukopenia
412
or leukocytosis may exist. Blasts and other immature white
blood cells may be present in the peripheral blood, and
because of rapid cell turnover, lactic dehydrogenase and uric
acid levels may be elevated. In acute promyelocytic leukemia
(APL), coagulation profiles should be evaluated carefully as
disseminated intravascular coagulation (DIC) may occur
when these cells lyse and release factors to promote coagula-
tion. Leukocytosis may also be present and can cause head-
aches, altered mental status, dyspnea, and ultimately fatal
pulmonary or intracranial hemorrhages.
The cause of AML is unknown in most cases, but the
development has been strongly linked to radiation, chemi-
cal exposure (benzene), or prior use of alkylating agents.
Other cases can arise following exposure to alkylating
agents or topoisomerase II inhibitors used in the treatment
of other neoplasias.6,7,8 In addition; there was a 20-fold
increase of AML in Japanese atomic bomb survivors, while
the incidence of AML is 10 times higher in manufacturing
workers exposed to benzene.9,10
Some AML patients have evolved from a prior hemato-
logical disorder, such as chronic myeloid leukemia (CML),
myelodysplastic syndrome (MDS) or myeloproliferative
neoplasm. Patients with MDS may develop AML; approxi-
mately 25% of AML patients have a previous history of
MDS, this leukemia are generally designated as secondary
leukemia and are of interest due to the probable multiple
genetic aberrations required for the evolution into AML.
In recent years, an increasing number of gene mutations,
deregulated expression of genes, noncoding RNA and miR-
NAs and epigenetic changes have been identified which has
added to the molecular heterogeneity of the disease
D I S E A S E C L A S S I F I C AT I O N
AML had been classified in the 1970’s using the
French-American-British (FAB) system, that is based on
morphological grounds such as degree of lineage, commit-
ment, and differentiation.11 In this classification, the AML
FAB subtypes (M0–M7) were determined by cell morphol-
ogy with particular subtypes such as M3 defining the acute
promyelocytic leukemia (APL). In 2002, the World Health
Organization (WHO) introduced a new classification of
hematopoietic and lymphoid neoplasm by combining mor-
phology, genetic, immunophenotype, with biologic and
clinical information.12
The 2008 WHO released an updated Classification of
Tumors of Hematopoietic and Lymphoid Tissues,13 which
included new criteria as well as clarification and refinement
of the defining criteria for the all the hematopoietic malig-
nancies, including AML. The 2008 revised WHO classifi-
cation placed the most common and well-defined recurring
cytogenetic abnormalities into a separate group:  t(8;21)
(q22;q22); inv(16)(p13q22) or t(16;16)(p13;q22);
t(15;17)(q22;q12); and 11q23 (MLL) abnormalities. Some
of these chromosomal abnormalities correlate with specific
FAB subtypes, for instance, the translocation t(8;21) is
present in approximately 40% of the FAB-M2 leukemia.14
Translocations involving the retinoic receptor α (RARα) on
chromosome 17 (t(15;17)) is present in 98% of cases15 of
APL (APML) FAB-M3 and inv(16) or t(16;16) are involved
in 80% of the FAB-M4-Eo cases.16,17
In addition, in the 2008 WHO classification, new cytoge-
netic entities were added: t(6;9)(p23;q34); inv(3)(q21q26.2)
or t(3;3)(q21;q26.2); t(1;22)(p13;q13). Furthermore, for the
first time, two provisional molecular mutation entities are
added: AML with mutated NPM1; and AML with mutated
CEBPA. It was recommended that FLT3 mutations also be
examined in cases of cytogenetically normal AML.
C Y TO G E N ET I C S :  C L A S S I F I C AT I O N S
A N D R E C U R R E N T T R A N S L O C AT I O N S
Cytogenetics including fluorescence in situ hybridization
(FISH) provided an insight into the heterogeneity of the
disease by identifying two major groups of AML. In one
group, chromosomal aberrations occurred in approxi-
mately 50% of de novo AML. Within this group, at least
two further subtypes could be further distinguished: one
that had balanced aberrations mainly consisting of t(8;21),
t(15;17), and inv(16); and a second subgroup with unbal-
anced aberrations such as 5q-, 7q-, -5, -7, or complex cyto-
genetic abnormalities. The second major group consisted
of approximately 40–50% of AML cases in whom the
karyotype appears to be normal (cytogenetically normal,
or CN).
G enetic and G eno m ics in C linical H e m atology, I I I :   Acute L eu k e m ias   •   4 1 3
Balanced translocations result in the expression of
fusion proteins, which frequently contain a truncated tran-
scription factor joined with another, unrelated protein.
The balanced translocations result in a deregulation of
transcription factors and can be seen as a main pathogenic
event.18 The simultaneous occurrence of more than one of
these recurrent, balanced translocations is rare. The detec-
tion of subtype-specific abnormalities such as the t(15;17)
or t(8;21) translocations has resulted in the characteriza-
tion of the involved genes, the identification of abnormal
fusion proteins, and provided new diagnostic targets, but
has also identified distinct AML subtypes.
Based only on cytogenetic abnormalities, patients were
classified into three distinct prognostic subgroups or risk
groups: those with favorable or good risk, intermediate or
standard risk, and unfavorable, adverse, or poor risk disease.
The favorable risk group includes approximately 15% of
patients, with an age below 60 years, and is defined by the
presence of either t(15;17), t(8;21) and inv(16), or t(16;16)
cytogenetic translocations. Patients presenting with these
mutations usually have an increased rate of complete remis-
sion and are at relatively low risk of relapse.19 The poor prog-
nosis group of AML patients have cytogenetic abnormalities
such as deletion of chromosomes 5 or 7 (or the short arms of
the chromosomes), abnormalities in the long-arm of chro-
mosome 3, and AML with complex karyotype (three or
more cytogenetic aberrations).20,21 Cytogenetic mutations
involving the 11q23 region are also considered a marker
for poor prognosis. Overall, patients with adverse risk-type
mutations generally have a probability of five-year survival
of approximately 20%. Patients presenting with AML with
other types of cytogenetic abnormalities or without any
apparent cytogenetic abnormalities (i.e., an apparently
normal karyotype) are considered to have an intermediate
risk.22,21,23
In 2010, a refinement of the cytogenetic classification
for AML was published24 (Figure 27.1). In this refinement,
almost 6000 patients, 18–59  years old at diagnosis, were
classified into 54 cytogenetic subgroups. The favorable cyto-
genetic group maintained the t(15;17), inv(16) and t(8;21)
balanced reciprocal translocations, whilst in the adverse group
the list of abnormalities included t(3;5)(q25;q34), inv(3)
(q21q26)/t(3;3)(q21;q26), add(5q)/del(5q), -5, 7, add(7q)/
del(7q), t(6;11)(q27;q23), t(10;11)(p11_13;q23), other
t(11q23) [excluding t(9;11)(p21_22;q23) and t(11;19)
(q23;p13)], t(9;22)(q34;q11), _17, and abn(17p). Note that
this list now moved several of the 11q23 (MLL) translocations
from the previous intermediate group into the adverse group.
In addition, any patients without the favorable or adverse
cytogenetic abnormalities, but with four or more unrelated
 Grimwade et al., 2010 Dohner, et al., 2011$ Grossmann et al., 2012

Very • PML-RARA or CEPBA double

mutant
Favourable • 3-yr. OS: 82.9%
• t(8;21)*; RUNX1-RUNX1T1;
mut
• t(15;17); t(8;21); inv(16)/ inv(16)/t(16;16)*; CBFB-MYH11; NPM1
Favourable /FLT3-ITD-; Mutated CEBPA
Favourable t(16;16)* • RUNX1-RUNX1T1; CBFP-
• 3-yr. OS: 71.2% • 3-yr. OS: 64.1% MYH11; or NPMmut/FLT3-
Favourable ITD
negative

• 3-yr. OS: 62.6%

• NPM1mut/FLT3-ITD; FLT3-ITD+;
Intermediate - I NPM1-/FLT3-ITD–
• Entities not classified as • 3-yr. OS: 45.8% • CEPBA single mutant; FLT3
Intermediate favourable or adverse ITD+; NPM1mut/FLT3-ITD+;
Intermediate or wild type cases
• 3-yr. OS: 46.3% • t(9;11)(p22;q23); MLLT3-MLL; Cytogenetic • 3-yr. OS: 44.2%
abnormalities not classified as favorable or
Intermediate - II • adverse
• 3-yr. OS: 37.4% • MLL-PTD a/o RUNX1 a/o
• Defined abnormalities of Unfavourable ASXL1 mutated
mainly chromosomes 3, 5, • inv(3)/t(3;3); RPN1-EVI1; t(6;9); DEK- • 3-yr. OS: 21.9%
Adverse 11 and 17** NUP214; t(v;11) (v;q23); MLL rearranged; –5
Unfavourable or del(5q); –7; abnl(17P); complex karyotype
• 3-yr. OS: 19.3%
• 3-yr. OS: 21.5%
Very • TP53 mutated
Unfavourable • 3-yr. OS: 0%

$: t(15;17) excluded

* t(15;17)(q22;q21); t(8;21)(q22;q22); inv(16)(p13q22)/t(16;16)(p13;q22)

** abn(3q) excluding t(3;5)(q2125;q3135); inv(3)(q21q26)/t(3;3)(q21;p26); add(5q);

del(5q); –5; –7; add(7q)/del(7q); t(6:11)(q27;q23); t(10;11)(p1113;q23); t(11q23)
excluding t(9;11)(p2122;q23) and t(11;19)(q23;p13); t(9;22)(q34;q11);17/abn(17p);
Complex (4 unrelated abnormalities)

Figure 27.1 
Comparison of different risk stratification classifications, showing the changes from solely cytogenetic based classification to the molecular
risk stratification model.

abnormalities, were designated as a “complex” group. The
refinement of the cytogenetic risk categories resulted in the
reassignment of 299 patients (15% of the patients with abnor-
mal, non-favorable cytogenetic abnormalities) from interme-
diate to adverse (275 patients) or adverse to intermediate (24
patients).24
The European Leukemia Network (ELN) also pro-
posed a standardized reporting system for genetic abnor-
malities that would allow for a better comparison of studies
by including data from cytogenetic analysis and NPM1,
CEBPA, and FLT3 gene mutation analyses.25
A fully molecular-derived risk stratification model
has also been developed26 (Figure 27.1) in which conven-
tional cytogenetic analysis was substituted with molec-
ular analysis of the fusion transcripts (PML-RARA,
RUNX1-RUNX1T1, CBFB-MYH11) and integrated with
mutation analysis for CEBPA, NPM1, RUNX1, ASXL1,
FLT3-ITD and MLL-PTD. This molecularly based stratifi-
cation led to five categories’ being formulated.
T(15;17)—PM L-RA Rα
Acute promyelocytic leukemia (APL: FAB subtype M3) is
a relatively rare malignancy and accounts for approximately
10% of all AML cases.27 A striking feature of APL is that is
very uncommon in children younger than 10 years of age;
then its incidence increases steadily during the teen years to
reach a plateau during early adulthood, and remains con-
stant and then decreases after age 60 years.28 There is a sug-
gestion that APL may arise as a complication of previous
exposure to chemotherapy (particularly drugs targeting
topoisomerase II) or radiotherapy, particularly in patients
with a history of breast cancer.29,30
APL is characterized by translocations involving the
RARα gene located at 17q12. In the majority of the cases,
the partner chromosome in the translocation is chromo-
some 15, resulting in the fusion of large parts of RARα to
most of PML coding sequence, generating the PML-RARα
fusion protein.31,32 Rarer cases of APL show translocations
involving t(11;17)(q23;q12), t(5;17)(q35q12), t(17;17)
(q11;q12), or t(11;17)(q13;q12). These result in fusion
proteins that also contain RARα but with different partner
genes: PLZF (promyelocytic leukemia zinc finger), NPM
(nucleophosmin), Stat5b or NuMA.33–39
RARα (retinoic acid receptor alpha) is a nuclear hor-
mone receptor that acts as a hormone-dependent tran-
scriptional switch and in the absence of the ligand retinoic
acid transcription is repressed through the recruitment of

4 1 4   •   G eno m ics in C linical P ractice

co-repressors and histone deacetylases (HDACs). When
retinoic acid binds to RARα, the heterodimeric complex of
RARα and RXRα releases the co-repressors and HDACs,
converting the complex into a transcriptional activator,
illustrated in Figure 27.2. The partner in the fusion is the
PML gene; this is an organizer of nuclear domains that are
known as PML nuclear bodies (NB).
PML-RARα has a dual dominant-negative activity on
signaling by each of its partners that disrupts the retention
of the DNA-binding specificity; physiological concentra-
tion of retinoic acid (RA) cannot induce the conformational
change necessary to release co-repressors. The PML-RARα
complexes dull the RA-induced activation of many canoni-
cal RARα target genes in a dominant-negative manner.40
Therefore any RARα target genes remain repressed.41–43
Treatment with pharmacological levels of all-trans-retinol
acid (ATRA; tretinoin) reverses this process and restores
the transcriptional activity of the retinoic acid promoter,
resulting in the differentiation of the leukemia cells along
the granulocytic lineage.44 APL is a classic example of how
basic research can influence clinical therapy as a direct result
of identifying the disease-specific molecular defect and its
functional repair by a targeted therapeutic approach.
The introduction of ATRA into the therapy of APL rev-
olutionized the management and outcome of this disease.
After ATRA, arsenic trioxide (ATO) is probably the most
biologically active single drug in APL. Several treatment
strategies using these agents, usually in combination with
chemotherapy, have provided excellent therapeutic results
with survival rates exceeding 70% in multicenter clinical
trials. Cure of patients with APL depends not only on the
effective use of combination therapy involving differentiat-
ing and classical cytotoxic agents, but also, critically, upon
supportive care measures that take into particular account
the biology of the disease and the complications associ-
ated with molecularly targeted therapies. The clinical use
of arsenic trioxide is a more potent anti-APL agent than
retinoic acid (RA), although unlike RA it is unable to reac-
tivate PML-RARA-dependent transcription. Both RA and
arsenic degrade PML-RARA:  RA induces PML-RARA
degradation via the RARA moiety; whereas arsenic acts
on the PML part of the fusion protein, which leads to
enhanced apoptosis with a minor effect on leukemic cell
differentiation.45
The molecular monitoring of the disease has suggested
that a benefit can be gained from preemptive therapy in
patients who show a molecular relapse compared with those
whose treatment was initiated at the time of hematological
relapse.46–48 These studies have indicated that monitoring
the minimal residual disease (MRD) burden in the bone
G enetic and G eno m ics in C linical H e m atology, I I I :   Acute L eu k e m ias   •   4 1 5
marrow every three months using sensitive quantitative
RT-PCR allows the early detection of emergent disease.
The over-expression of PML-RARα alone is insufficient
to induce an undifferentiated phenotype in mouse bone
marrow: only 15–20% of the transplanted mice develop a
disease with undifferentiated characteristics, and only after
a long latency. This would suggest that a second event is
required for the development of APL.49 Co-expression of
PML-RARα with Flt3-ITD mutations enhances the fre-
quency of disease in recipient mice; mice reconstituted with
Flt3-ITD do not develop an acute leukemia-like disease;
again indicating mutational cooperation.50,51 In the trans-
location t(15;17), the reciprocal RARα-PML transcript52
is also expressed in some patients and may contribute to
leukemia development, as transgenic mice expressing both
PML-RARα and RARα-PML have an increased frequency
of APL.53
Rare genetic variants of APL have been reported, the
RARA fusion partner can determine the response to ATRA
and ATO.54 In this regard, NuMA-RARA, NPM1-RARA,
and FIP1L1-RARA are known to be ATRA-sensitive,
whilst others are ATRA-resistant (STAT5b-RARA),39 rela-
tively resistant (PLZF-RARA), or have an unknown sensi-
tivity to ATRA (PRKAR1A-RARA).55
In contrast to PML-RARα, PLZF-RARα, the prod-
uct of t(11;17), does not respond to high doses of retinoic
acid. Like RARα, PLZF can bind co-repressors by itself; a
mechanism that cannot be overcome by ATRA.41 PLZF
contains an oligomerization domain available for PLZF
or PLZF-RARα homodimers as well as for heterodimers
with regulatory proteins. Therefore, PLZF-RARα forms
high-molecular-weight nuclear complexes in vivo, displac-
ing the normal PLZF from its correct location in the cell.
This would suggest that the fusion protein disturbs both
RARα and PLZF function.56
An important target gene of both PML-RARα and
PLZF-RARα is cyclin A1, a homolog of cyclin A2 that is spe-
cifically expressed in hematopoietic cells.57 ATRA exposure
decreases cyclin A1 expression in the PML-RARα-positive
NB4 cell line, Cyclin A1 is induced by PML-RARα and
PLZF-RARα,58,59 while over-expression in a transgenic
mouse model induces a myeloproliferative disease with long
latency.60 These findings provide direct evidence for a role of
PML-RARα in accelerating cell-cycle progression.
C O R E B I N D I N G FAC TO R
T R A NS L O C AT I O NS
Other balanced, and favorable, cytogenetic abnormalities
are the t(8;21) and inv 16 translocations, which disrupt
 t(15;17)(q22;q12)
PML-RARa

Sin
NCoR
3

HAT
H
RXR

D
AC
RARA
DNA DNA
Physiological levels of RA
(10–9 M)

Nucleosomes with
deacetylated histone tails

Nucleosomes with
acetylated histone tails
PML-RARA

DNA DNA
Pharmacological levels of RA
(10–6 M)

Figure 27.2 
Schematic of the different molecular responses between RARA and PML-RARA in response to physiological and pharmacological levels
of ATRA.

the “CBF (core binding factor) complex,” and these trans-
locations account for 12% of AML cases, particularly in
patients less than 60 years of age at diagnosis.24 The CBF
consists of two subunits:  RUNX1 (also referred to as
AML1, CBPα2, and PEBP2αB) and is located on chro-
mosome 21q22 and CBFβ, which is encoded by a gene on
16q22. The two subunits form a heterodimer transcrip-
tion activator, although only the AML1 subunit contains
DNA-binding ability.
The most frequent translocation is the t(8;21) transloca-
tion, which has been reported to be present in 7% of younger
(<65 yrs. old) adult AML patients with this disease24 and
results in the formation of the fusion protein AML1-ETO
or RUNX1-RUNXT1. The Runt-related transcription fac-
tor 1 (RUNX1) (also known as AML1) or core-binding fac-
tor subunit alpha-2 (CBFA2) regulates the differentiation
of hematopoietic stem cells into mature blood cells. The
RUNX1T1, also known as AML1T1, CBFA2T1, ETO, or
MTG8 is a member of the myeloid translocation gene fam-
ily, which interacts with DNA-bound transcription factors
and recruits a range of co-repressors to facilitate transcrip-
tional repression.
The C-terminus of the transcriptional activator AML1
is replaced by the transcriptional repressor ETO, resulting in
the fusion protein AML1-ETO (CBFA2-CBFA2T1).61 The
ETO protein binds to co-repressors (e.g., NcoR, SMRT, and
mSin3) and histone deacetylases (HDACs).62,63 The recruit-
ment of HDACs to the complex changes the histone acety-
lation status and as a result alters the DNA conformation,
making it less accessible for the basal transcription machin-
ery and results in a repression of AML1 target genes. Target
genes include granulocyte–macrophage colony-stimulating
factor (GM-CSF),64 for interleukin-3 (IL-3) and NP3.65
AML1-ETO influences various other transcription fac-
tors, including MEF, c/EBPα, AP-1, and Pu.1.66–68 c/EBPα
is a major regulator of early granulocytic differentiation, and
AML-ETO physically interacts with c/EBPα and downregu-
lates it at the transcriptional level,67 which has an important
influence on granulocytic differentiation. An additional
function of AML1-ETO in leukemia transformation is

4 1 6   •   G eno m ics in C linical P ractice

through its influence on cellular proliferation and survival. It Glutamine,
Lysine rich Proline
Glutamine,
Aspartic acid
rich region
has recently been shown that AML1-ETO directly inhibits region rich region

the expression of p14ARF,69 and this has a role in inducing

AF6 (6q27)
p53-dependent proliferation arrest and apoptosis.70 Low lev-
els of p14ARF are predictors of poorer outcome for AML MLL-AF6
patients.71 AML1-ETO also has a further role in apoptosis as (t(6;11)(q27;q23)

it upregulates Bcl2,72 although contrasting results have been

observed in an established monoblastic leukemia cell line MLL (11q23)

apoptosis.73 The downregulation of c/EBPα by AML1-ETO, AT

Hooks
DNA Zinc
Methyl fingers
SET
domain
resulting in the release of c/EBPα-mediated cell cycle control, transferase

leads to enhanced cellular proliferation. MLL-AF9

(t(9;11)(p22;q23)
Point of fusion

Other CBF leukemia-associated translocations in AML

are inv(16) and t(16;16) in which CBFβ is fused to MYH11,
also referred to as the “smooth muscle myosin heavy chain” AF9 (9q22)

(SMMHC) gene,74 generating a CBFβ/SMMHC fusion Serine rich

protein. The fusion protein binds to AML1 and through Nuclear regions
Translocation
DNA contact acts as a repressor of AML1 function.75 In Sequence

addition, the SMMHC part of the fusion protein contains

Figure 27.3 
Molecular consequences of the t(6;11) and t(9;11)
a repressor domain that directly represses AML1 activity translocations involving the MLL gene and AF6 or AF9 genes,
when bound to transcriptional active sites.75 respectively.

to immortalize myeloid progenitor cells. It has been shown

M LL G E N E T R A NS L O C AT I O NS
A subgroup of AML patients have translocations involving
the MLL gene located on chromosome 11q23. The MLL
protein, the human homolog of Drosophila trithorax, con-
sists of 3969 amino acids, and the gene spans 90 kb. MLL
has histone methyltransferase activity and assembles in
very large multi-protein complexes composed of more than
29 proteins76 (Figure 27.3). The DNA-binding domain
of MLL contains three AT-hooks that change DNA con-
formation and thereby allow regulatory transcription fac-
tors to bind to DNA.77 In addition to the chromosomal
translocations involving numerous partner chromosomes
and genes, partial tandem duplications (PTDs) also occur.
Translocations involving the MLL gene are found in up
to 80% of leukemia diagnosed in children younger than
24  months.78,79 Other common translocations involving
chromosome 11q23 result from a fusion chromosomes 4,
9, or 19, and only occur in 2–5% of all adult AML.79 More
than 60 cytogenetic translocations involving the region
11q23 have been described, with at least 30 of the fusion
partners of MLL defined.80
All fusion proteins contain the N-terminal part, includ-
ing the DNA-binding domain and the repressor domain of
MLL, fused to the C-terminal part of the translocation part-
ner. Most of the known MLL fusion partners—for example,
ENL, AF9, and AF4—contain transcriptional activation
domains. MLL fusion proteins do regulate the expression of
HOX genes, and this is partially responsible for their ability
that MLL-ENL fails to immortalize myeloid progenitors
from HOXA9- or HOXA7-deficient bone marrow.81 MLL
regulates HOX genes’ expression through direct binding
to promoter sequences, and HOXA9 is highly expressed in
MLL leukemia.82
H OX G E N E T R A NS L O C AT I O NS
HOX gene-family members are also involved in
AML-associated chromosomal translocations. Examples
include the translocations t(7;11) and t(2;11) generating the
fusion proteins NUP98/HoxA9 and NUP98/HOXD13.
NUP98 encodes for a component of the nuclear core com-
plex. The fusion protein consists of the homeodomain
of HOXA9 and the phenylalanine-glycine repeats from
NUP98,83–85 which contains an interaction domain with
the transcriptional activator CBP/p300. Over-expression
of the fusion protein induces a myeloproliferative disease,
followed by the occurrence of AML in transplanted mice.86
Frequently Mutated Genes
Prior to 2008, mutations in FLT3, NPM1, and CEPBA had
been frequently associated with AML. Mutations in FLT3
were identified in 1996, but a prognostic relationship was
only identified several years later. Mutations in CEPBA
and NPM1, as well as FLT3, were used in the 2008 WHO
Classification of Tumors of Hematopoietic and Lymphoid

G enetic and G eno m ics in C linical H e m atology, I I I :   Acute L eu k e m ias   •   4 1 7

 NPM1
DNMT3A
FLT3_ITD
CEPBA
MLL_PTD
RUNX1
1 368

Mutation landscape map for six genes (NPM1, FLT3 ITD, DNMT3A, CEPBA, MLL-PTD, and RUNX1) across ~350 AML samples.
Figure 27.4 
(Data combined from89,97)

Tissues.13 However, the number of mutated genes in AML
has dramatically expanded in recent years as consequence of
the accessibility of next-generation sequencing.
The first two cancer patients to have their genome
fully sequenced were AML patients, and from the analysis,
IDH1/2 and DNMT3A were identified as novel frequently
mutated genes.87,88 Mutated genes have subsequently been
identified in tumor-suppressor genes, genes associated
with DNA methylation or chromatin modifications, genes
involved in signaling and transcription factors.89–96 Some of
the genes are mutually exclusive; for example, patients with
an NPM1 mutation very rarely, if ever, have a mutation in
TP53. Other combinations frequently occur; for example,
NPM1 mutations with FLT3-ITD or NPM1 mutation with
a DNMT3A gene mutation. A study of 200 AML patients
analyzed by next-generation sequencing showed that eight
patients had mutations in all three of the genes frequently
mutations: NPM1, FLT3, and CEPBA. Furthermore, one
patient did not have any mutations in 33 genes or groups of
genes with similar functions (e.g., spliceosomes or Ser-Thr
kinases).97 Two ways of illustrating the interactions are
shown in Figures 27.4 and 27.5.

Figure 27.5  Circos plot visualizing the data from Figure 27.4 showing how the different mutations co-occur with each other. (Combined data taken
from89,97)

4 1 8   •   G eno m ics in C linical P ractice

 FLT3
Activating mutations within the FLT3 gene were first
described in 1996, and the protein has been identified as
both a possible progression factor and a drug target in AML.
FLT3 (Fms-like tyrosine kinase 3, also designated Flk-2
or STK-1) is a member of the class III Receptor Tyrosine
Kinases (RTK) and shares high homology with the other
members of this family, the PDGF (platelet-derived growth
factor) receptors, c-FMS/M-CSF (macrophage colony–
stimulating factor) receptor, and Kit (receptor for SCF, the
stem cell factor). Common features of class III RTK include
an extra-cellular domain consisting of five to seven immuno-
globulin-like subdomains and a split kinase domain.98 The
gene encoding Flt3 is located on chromosome 13q12.99 The
gene contained 24 exons, with exons 14 and 15 encoding the
juxta-membrane region of the receptor and exon 20 encod-
ing the tyrosine kinase domain (TKD)100 (Figure 27.6).
FLT3 is predominantly, but not exclusively, expressed
in hematopoietic tissues but can be detected in placenta,
brain, testis, lymph nodes, thymus, and in the liver of 6- to
8-week-old mice.101 In hematopoiesis, FLT3 is expressed
in early multi-potent progenitor cells, but not in primitive
stem cells. FLT3 expression seems to be accompanied by a
loss of self-renewal capacity in the hematopoietic stem cell
compartment and remains detectable during monocytic
differentiation but is lost during B-lymphoid and granulo-
cytic differentiation.102–104
After ligand binding, FLT3 activates the signal path-
way via receptor auto-phosphorylation and through sub-
sequent binding of proteins, which specifically recognize
phosphorylated tyrosines on the receptor. These include the
Extracellular ligand binding
domain with immunoglobulin
like loops
Cell membrance
Juxta-membrane domain
Kinase domain
Kinase domain
Figure 27.6  Molecular structure of the FLT3 protein showing the location of the possible sites of mutation: ITD and activating point mutations.
G enetic and G eno m ics in C linical H e m atology, I I I :   Acute L eu k e m ias   •   4 1 9
recruitment and activation of the Src kinase family and the
activation of the Ras/MAPK signaling cascade via direct
interaction with Grb2.105 FLT3 induces the phosphoryla-
tion of other substrates such as Gab106 and Cbl proteins that
are involved in receptor internalization and degradation.107
A weak but consistent phosphorylation of STAT5a, but not
STAT5b, can be observed after ligand-induced activation of
FLT3, but it is not associated with STAT5 DNA-binding or
expression of STAT5 target genes.108
IT D MU TAT I O N S O F FLT3
In AML, FLT3 is highly expressed in 60–92% of the cases,109
and this is at a higher level than in normal hematopoietic
progenitors. Flt3 has been shown to have in-frame inser-
tions of several nucleotides in the mRNA sequence encod-
ing the juxtamembrane domain of the receptor in AML
patients.110 These insertions are internal tandem duplica-
tions (ITDs) of varying lengths and result in the repetition
of between 3 and 50 amino acids in the juxtamembrane
region of the FLT3 receptor. Most patients also express
the wild-type allele. The prevalence of FLT3-ITD in AML
patients has indicated an incidence of between 20% and
30%.111–114 The frequency of in FAB subgroups of AML dif-
fers, with a relatively high frequency in the APL.112 Highest
prevalence of FLT3-ITD mutations was seen in cases with
normal karyotype, but they occur rarely in CBF [inv(16)
or t(8;21)] leukemia,112 where conversely Kit mutations are
seen with high frequency.
In APL, FLT3-ITD mutations are associated with high
white blood cell counts, a high percentage of bone marrow
Internal Tandem Duplication (ITD):
average insertion size 45 b.p.
Point mutations: Mainly at codon D836
blasts, and high levels of LDH at diagnosis.113,115–117 The
presence of FLT3-ITD has been associated with an unfavor-
able prognosis in all FAB subgroups as a result of reduced
disease-free and overall survival.111,115,116 ITD mutations
have a higher incidence with increasing age.118
The ITD-mutated alleles and the wild-type allele are
usually transcribed so that both are co-expressed in AML
blasts. However, the wild-type allele is frequently lost in
some leukemia clones, resulting in loss of expression of
the wild-type mRNA, which has been shown to be due to
homologous recombination and subsequent loss of het-
erozygosity as a form of uniparental disomy.116,119 In some
patients, Flt3-ITD mutations were not detected at initial
diagnosis but were found to be present at relapse, while in
others they were present at first diagnosis and undetectable
at relapse.113,120 Therefore the use of FLT3-mutations for the
detection of minimal residual disease may be limited.
AC T I VAT I N G MU TAT I O NS O F F LT3
A different type of FLT3 mutation has also been identified
within the activation loop, or the tyrosine kinase domain
(TKD), in approximately 7% of AML patients.114,121 The
activation loop is a highly conserved protein structure in
the membrane-distal portion of the kinase domain of sev-
eral receptor tyrosine kinases. This type of mutation usually
affects the codon 835 (D835) where the point mutations
alter the amino acid at that codon. Few studies have cor-
related TKD mutations with prognosis.122 Numerous other
point mutations within the TKD have been observed near
codon 835; for example, at codons 836, 839, or 840.114,123
Analysis of the signal transduction and biological con-
sequences of activating FLT3 mutations124 showed that
activation of STAT5a is weak, whereas STAT5b is strongly
activated. Flt3-ITD activates different Bcl2 family mem-
bers, and consequently different anti-apoptotic pathways,
from those induced by wild-type Flt3.125
N PM1
The NPM1 gene encodes for a nuclear-cytoplasmic protein
that is involved in translocations and mutations in vari-
ous hematological malignancies. The discovery of NPM1
mutations arose from the observation of ectopic expres-
sion of nucleophosmin in the cytoplasm of leukemia cells.
This immune-histochemical finding led to the discovery
of several gene mutations mainly located in exon-12 that
usually result in changes at the C-terminus of the native
NPM1 protein; i.e., the disruption of the nucleolar localiza-
tion signal and the generation of a new additional nuclear
4 2 0   •   G eno m ics in C linical P ractice
export signal motif126 (Figure 27.7). This interferes with the
nucleo-cytoplasmic trafficking of the protein, leading to the
aberrant accumulation of nucleophosmin in the cytoplasm
of leukemic cells carrying NPM1 mutations.126
NPM1 mutations are the most common single-gene
abnormality in adult AML, accounting for about 30% of
all AML and 50–60% of CN-AML. Unlike other muta-
tions, those affecting the NPM1 gene appear to be specific
for AML and usually occur in patients with de novo disease.
NPM1 mutations in AML are also stable during the course
of the disease, being detected at relapse even many years
after the initial diagnosis, and as such may make a good tar-
get for minimal residual disease (MRD) monitoring during
the disease.
However, how NPM1 mutations are involved in AML
development is still unclear. The normal NPM1 protein
is involved in ribosome biogenesis, control of centro-
some duplication, and stabilization of ARF protein in the
nucleoli, with a possible role in regulating transcription and
apoptosis.127 A mouse model has shown that mutant NPM1
knock-in alleles are AML-initiating lesions causing HOX
gene overexpression, increased self-renewal, and expanded
myelopoiesis.128 Cooperation of NPM1 mutants with other
genetic lesions led to delayed-onset AML in about one-third
of animals. NPM1 mutations frequently associate with
mutations affecting the FLT3, DNMT3A, and IDH1 genes
that are likely to play a cooperative role in leukemogenesis.
Many studies have shown that patients with NPM1
mutations but without FLT3-ITD mutation have a lower
rate of relapse and improved overall survival129; although
these are mainly seen in younger adults (<60  years old).
However, the presence of NPM1 mutations is also an
important favorable prognostic factor in older (>60 years)
AML patients, irrespective of the FLT3-ITD status.130
Oligomerisation Histone DNA/RNA
molecular chaperone binding binding
t(5;17) t(3;5) Insertions
APL MDS/AML AML
Figure 27.7 
Molecular structure of NPM1 highlighting the location of the
t(3;5) and t(5;17) translocation and the different common types of 4
base pair insertions seen in AML.
 CE PBA
The CEBPA (CCAAT/enhancer binding protein alpha)
gene encodes a member of the basic region leucine zipper
(bZIP) transcription factor family that is critical for neu-
throphil development. Mutations of the CEBPA gene are
seen in around 10–15% of AML, pre-dominantly in cases
carrying normal cytogenetics or 9q deletion. CEBPA muta-
tions have been identified in AML that affect both the
N-terminus and C-terminus of the protein, as illustrated
in Figure 27.8. Nonsense N-terminal mutations result in
the formation of a CEBPA truncated isoform with domi-
nant negative properties; whilst the C-terminal mutations
result in CEBPA proteins with decreased DNA-binding or
dimerization activity.91
About one-third of CEBPA-mutated AML patients
carry a single (N or C) CEBPA mutation. The rest of the
cases are double-mutated with an N-terminal mutation on
one allele and a C-terminal mutation on the second allele;
therefore no wild-type CEPBA protein is detected. Due
the type of CEBPA mutations, they have been detected
using conventional techniques, but the development of
next-generation amplicon deep-sequencing may help in the
diagnosis and monitoring.131
Murine models to examine the CEBPA mutants in leu-
kemogenesis showed that the C-terminal and N-terminal
mutations contributed in a different way to stem cell
expansion, homeostasis, and myeloid programming.132
Mutations of GATA2 were found to be frequently associ-
ated with CEBPA double mutations, suggesting that these
genetic alterations may cooperate in AML development.
AML patients with the double CEBPA mutations are a dis-
tinct entity from those with single mutations: a view that
is supported by in clinical trials where the double mutants
Transcription activation
domains
TAD1
Wild type
ATG 1
42 kD
Mutant
N-Terminal
nonsense mutation
C Terminal missense
mutation
Figure 27.8  Molecular structure of the CEPBA gene.
G enetic and G eno m ics in C linical H e m atology, I I I :   Acute L eu k e m ias   •   4 2 1
emerged as an independent prognostic factor for favorable
outcome; and patients with double CEPBA mutants were
mutually exclusive from those with NPM1 mutations.133
MU TAT I O N S I N E P I G E N ET I C
MO D I F Y I N G G E N E S
IDH1, IDH2, DNMT3A, and TET2 genes are in a group
of commonly mutated genes associated with epigenetic
modifications.
IDH1 and IDH2
The NADP+-dependent IDH1 and IDH2 (isocitrate dehy-
drogenases 1 and 2) genes encode for cytosolic enzymes
that catalyze a reaction in the tricarboxylic acid cycle.
IDH1 mutations were discovered by massively parallel
sequencing of the entire genome of the leukemic cells and
matched normal skin from a patient with CN-AML; and
in AML, IDH1 mutations are clustered at codon 132,134
which is a highly conserved arginine residue that directly
binds to isocitrate. IDH1 mutations occur in about 6–8%
of unselected AML134 and 10–12% of CN-AML, and at
variable frequency in other tumors, including gliomas,
myelodysplastic syndromes, myeloproliferative disorders,
and adult acute lymphoblastic leukemia (ALL). Because
IDH1 mutations at arginine 132 (R132) are closely asso-
ciated with NPM1 mutations in CN-AML,134 they may
have a cooperative role in AML development. IDH1 and
IDH2 mutations are usually mutually exclusive in AML.
Both IDH1 and IDH2 confer to the enzyme a neomor-
phic activity that leads to the reduction of α-ketoglutarate
to d-2-hydroxyglutarate (2HG). Accumulation of the 2HG
DNA
binding Dimerisation
domain region
TAD2 DBD bZIP
ATG 1
30 kD
OUT OF FRAME
IN FRAME
metabolite in the cell may exert an oncogenic activity by
acting as competitive inhibitor for α-KG-dependent dioxy-
genases,135 such as histone demethylases and the ten-eleven
translocation (TET) gene family of 5-methylcytosine
hydroxylases, leading to aberrant DNA methylation. This
is consistent with the clinical observation that IDH1 muta-
tions are mutually exclusive with TET2 mutations.135 The
prognostic impact of these mutations in the molecular sub-
sets of CN-AML remains unclear, as some studies claim
that IDH1 mutations have a negative prognostic effect in
the favorable-risk group of NPM1-mutated/FLT3-ITD–
negative AML,134 whilst others report inferior outcomes for
IDH1R132 mutations only in CN-AML that were NPM1
wild-type/FLT3-ITD–negative.
Following the discovery of IDH1 mutations, sequenc-
ing of IDH2 identified two mutations, IDH2R140 and
IDH2R172, in 9–11% of unselected AML.134 The muta-
tions are usually in codon 140 and associated clustered with
NPM1 mutations, whilst the IDH2R172 mutations are
mutually exclusive with other known mutations.129
The two types of mutations in IDH2 have dif-
ferent prognostic value. IDH2R172 may represent a
biologically and clinically distinct entity that is charac-
terized by unique gene- and miRNA-expression profile,
lower complete remission (CR) rate, and inferior sur-
vival.136,137 However, IDH2R140 mutations were pre-
dictive for inferior outcome in the favorable-risk group
of NPM1-mutated/ FLT3-ITD–negative AML. But
other studies have shown no impact of IDH2 mutations
on survival, whilst another study reported that younger
adult AML patients with an IDH2R140 mutations had
a significantly better prognosis than those with either
IDH2R172 or IDH1 mutations.136
DNMT3A Mutations
Human DNA methylation is regulated by a group of
DNA methyltransferase genes: DNMT1, DNMT3A, and
DNMT3B, enzymes that catalyze the transfer of a methyl
group onto the 5′ position of cytosine cytosine-guanine
(CpG) dinucleotides. DNMT3A and DNMT3B are pri-
marily involved in de novo methylation, whilst DNMT1
acts predominantly as maintenance methyltransferase.138
Mutations of the DNMT3A gene in AML were discov-
ered by using next-generation sequencing approaches in
about 20% of AML, and they are more frequently associ-
ated with CN-AML and appear stable during AML evo-
lution.139,140 About 95% of DNMT3A mutations occur in
the PDH and methyl-transferase domains, with a muta-
tion at R882H (arginine at 882 changed to a histidine)
4 2 2   •   G eno m ics in C linical P ractice
accounting for about 80% of cases (Figure 27.9). R882H,
at the dimer interface of the enzyme, and the other muta-
tions located at the tetramer interface disrupt tetrameriza-
tion, which is in turn critical for methylation of multiple
CpG sites. DNMT3A gene mutations occur in hemato-
logical malignancies other than AML, including 3–8%
of myelodysplastic syndromes and about 20% of T-cell
acute lymphoblastic leukemia. Several studies have failed
to identify a significant correlation between the DNMT3A
mutation status, gene expression signature, and DNA
methylation patterns. Although DNMT3A mutations
cause loss of methylase activity, and this has been suggested
to result into hypo-methylation and uncontrolled expres-
sion of multiple HOX genes. The upregulation of HOX
genes may not result from DNMT3A mutations but be
a reflection of the strong association of DNMT3A muta-
tions with NPM1 mutations,141 which are characteristically
associated with a HOX gene signature. As further evidence,
HOX genes are not overexpressed in DNMT3A-mutated
AML without NPM1 mutations. DNMT3A R882H
mutations are associated with mutations affecting the
genes NPM1, FLT3-ITD, BCOR, and IDH1,141 point-
ing to an active co-operation of these molecular events in
AML development. Interestingly, AML patients harboring
DNMT3A mutations are usually older than AML patients
with wild-type DNMT3A, and this may contribute to the
association of DNMT3A mutations with an inferior sur-
vival in AML. However, the clinical utility of DNMT3A
as prognostic marker for treatment decisions in CN-AML
remains to be defined. DNMT3A may exert a negative
impact on the high-risk category of CN-AML, for which
more intensive treatments are already recommended, but it
not certain whether DNMT3A mutations may also predict
response to hypo-methylating agents.
Methyl- Zinc Pro-Trp-Trp-Pro motif
transferase finger (PWWP) domain
Nonsense Missense
mutations mutations
Frame shift
insertions
mutations
Molecular structure of the DNMT3A gene showing the site
Figure 27.9 
and type of some of the mutations observed in AML.
 TET2 Mutations
TET1 and TET2 (TET gene family member 1 or 2) encodes
for proteins that are involved in epigenetic regulation by
converting 5-methylcytosine to 5-hydroxymethylcytosine,
a critical step in regulation of DNA methylation. TET2
mutations have been detected in 7.6% of AML, are associ-
ated with CN-AML, and are mutually exclusive with IDH
mutations.142 The prognostic impact of TET2 mutations in
AML is unclear, since they have been associated with poor
outcome in one study,143 but not in another.142
ASXL1 and Other Polycomb-Group Genes
In addition to mutations in genes that affect DNA cytosine
modifications, discussed above, genes that affect histone
post-translational modifications have also been found to be
repeatedly mutated in AML. The best-studied of such genes
is MLL, which is affected by translocations (see above) as
well as in-frame duplications in AML (see below).
Mutations in the Polycomb group of proteins, in par-
ticular the ASX1 gene, have been found in patients with all
myeloid malignancies. ASXL1 is affected by somatic dele-
tions as well as point mutations in patients with all myeloid
malignancies. Mutations in ASXL1 were initially identi-
fied96 in patients with myelodysplastic syndrome (MDS) and
chronic myelomonocytic leukemia. However, mutations in
ASXL1 are present in 6% to 30% of all AML patients, with
a positive correlation of ASXL1 mutations with age: 16.2%
to 25% of AML patients >60 years of age had ASXL1 muta-
tions, compared with 3.6% to 8% of patients <60 years of
age. AML patients with a preceding history of MDS have
an enrichment of ASXL1 mutations144; and these mutations
are associated with adverse outcome in all studies to date
in AML, MDS, and myeloproliferative neoplasm. ASXL1
has been shown to interact with the nuclear de-ubiquitinase
BAP1 in a biochemical complex that serves to remove a
ubiquitin from histone H2A at lysine 11950 (H2AK119);
however, it has been reported that mutations in ASXL1 also
result in loss of ASXL1 protein expression.51 In vivo stud-
ies have reported that loss of ASXL1 expression resulted in
accelerated disease latency and increased disease burden.
M LL MU TAT I O NS
In addition to the MLL gene being involved in 11q24 chro-
mosomal translocations (see above and Figure 27.3), partial
tandem duplications (PTD) have been described in AML.
PTDs result in the duplication of a portion of the gene
and have been associated with trisomy 11145,146 and around
G enetic and G eno m ics in C linical H e m atology, I I I :   Acute L eu k e m ias   •   4 2 3
5–11% of patients with cytogenetically normal AML. The
incidence of MLL PTD in FLT3-ITD positive patients is
significantly higher than in FLT3-ITD negative patients.146
An association of IDH1R132 and MLL-partial tandem
duplication (PTD) in CN-AML has been also reported.147
The clinical significance of MLL-PTD in CN-AML
patients remains controversial, having been associated with
inferior outcome, or even having no prognostic impact.
E P I G E N ET I C T H E R A P I E S
The high frequency of somatic alterations in epigenetic
modifiers in AML patients, combined with the established
clinical importance of DNA methyltransferase inhibitors
in the clinical management of patients with myeloid malig-
nancies, has led to a great interest in the development of
novel epigenetic therapies. These can include demethylation
agents and histone deacetylase inhibitors, but the armory is
being extended to include: BET inhibitors that may be use-
ful for both MLL-translocated and MLL-wild type AML
by targeting MYC gene transcription; EZH2, TET2, and
IDH1/2 inhibitors; and LSD1 inhibitors in combination
with ATRA in patients with MLL translocations.
OT H E R MU TAT E D G E N E S
WT1 Mutations
WT1 mutations are detectable in 10–13% of CN-AML,
although their clinical significance is unclear, but they have
been associated with either an inferior outcome or no prog-
nostic impact.148
RUNX1 Mutations
RUNX1 is involved in the chromosomal translocation
[t(8;21)] gene mutations, however mutations have also
been reported that cluster in the Runt domain of the gene.
The mutations have been found in undifferentiated AML
(M0 FAB) and in association with trisomies 13 and 21, but
in general predict an inferior outcome in two studies.149
BCOR Mutations
Mutations in the BCOR (BCL6 co-repressor) gene were
discovered by the whole-exome sequencing of a single
CN-AML patient.89 As a result of follow-up studies, BCOR
mutations were identified in ~4% of all CN-AML. BCOR
mutations may act by interfering with epigenetic mechanism
and are frequently associated with DNMT3A mutations.
 KIT Mutations
Activating mutations of KIT, also a class  III RTK, occur
in subgroups of AML affecting two regions of the recep-
tor: the juxta-membrane region and also in residues within
in the activation loop.150 These mutations are very frequent
in the AML patients with a translocation affecting the
CBF:  t(8;21) and inv(16) chromosomal translocations.
In one study, 24% of patients with inv(16) had an exon 8
mutation in KIT, but in contrast only 2% in patients with
t(8;21) had a KIT gene mutation. However, mutations of
residue 816 in KIT exon 17 occurred with similar frequen-
cies (7.9% and 10.6%, respectively).151 This indicates that
different mutations in tyrosine kinase receptors may occur
in the different types of AML, and they are probably depen-
dent on, and highly specific for, accompanying molecular
alterations in transcription factors.
RAS
Activating RAS mutations have been described in many
reports. These are usually at codon 12 and 13 of N-RAS,152
also codon 61 of N-RAS, but rarely in K-RAS, and very
rarely in H-RAS. Overall, the incidence of N-RAS muta-
tions is between 10% and 27%.153 The presence of RAS
mutations has been reported to correlate with low blast
counts and an improved survival of patients, but the con-
trast has also been reported in patients with RAS mutations
having a worse outcome compared to outcomes in patients
without tyrosine kinase or RAS mutations.153
RAS mutations MDSs patients correlated with an
increased risk of progression to AML,153,154 but the RAS
mutation status in patients at first diagnosis and at relapse
revealed a marked instability of these mutations, with loss
of mutations in some cases, sometimes replaced by muta-
tions in different codons.155 This would suggest that RAS
mutations are not an initiating event but may be important
for disease progression.
Mutations within the FLT3, KIT, and RAS genes have
been shown to have a strong involvement in the develop-
ment of AML. But it is likely that mutations and the sub-
sequent activation of other, as yet unidentified, signaling
pathways also occur.
The over-expression of non-mutated receptor tyrosine
kinases may result in ligand-independent activation, such
as might occur in cases of FLT3 over-expression.156 High
expression of wild-type FLT3 in AML blasts can induce
spontaneous FLT3 auto-phosphorylation,156 which may
contribute to leukemia transformation.
4 2 4   •   G eno m ics in C linical P ractice
The transcription factors Pu.1 and GATA-1 are impor-
tant in myeloid differentiation and the observed func-
tional inactivation by AML-associated fusion proteins, and
although they have been examined, no involvement in chro-
mosomal translocations has been observed, but mutations
have been identified in subgroups of AML patients.
Pu.1
Mutations of Pu.1 were observed in approximately 7%
of AML patients157 disrupting the function of one allele
without generating a dominant negative protein. The
multi-lineage importance of Pu.1 was confirmed by the
finding of mutations in patients with an immature pheno-
type (FAB M0), myelomonocytic or monocytic subtypes
(FAB M4 or M5), or with erythroleukemia (FAB M6).
MO L EC U L A R EVO LU T I O N I N A M L
The advent of next-generation sequencing (NGS) has been
able to give valuable insights into the clonal mutational
spectrum and evolution in AML. Several of the mutations
discussed above were initially identified by next-generation
sequencing and then subsequently validated and the inci-
dence confirmed in larger cohorts of AML patients.
However, the NGS analysis has also identified that each
patient has a unique spectrum of random, preexisting back-
ground mutations in the hematopoietic cell that acquired
the potentially initiating mutation.158–161 These studies have
also suggested that relapse clone in AML originates not
only from the larger dominant clone present at diagnosis,
but from the rarer clones, at diagnosis, that evolve into more
resistant and aggressive clones. Alternative therapies that
can completely eradicate all the leukemic clones need to be
developed to improve outcomes by reducing relapse.
G E N E -E X P R E S S I O N C L A S S I FI C AT I O N
IN AML
The first approach to disease classification, based on micro-
arrays, was described in 1999162 and used an unsupervised,
class discovery approach, identifying previously unrecog-
nized subtypes, to uncover the distinction between AML
and acute lymphoblastic leukemia (ALL). The authors
speculated that162 a larger sample would enable finer
sub-classifications to be identified that would correspond to
existing sub-classifications for AML or define new group-
ings. This concept sparked numerous publications163,164
trying to identify gene-expression profiles or signatures
that could distinguish between the major sub-classes
within AML.
The use of microarrays for a global approach to leukemia
classification was suggested by Haferlach and colleagues165
in 2005, in which they stated that a large multi-center
comparison assessing microarray diagnosis with standard
diagnostics was required. A major step toward this type of
diagnostic classification was the launch of an international
multi-center clinical research program (the MILE Study) to
assess the application of a microarray test and its potential
for use in the diagnosis and sub- classification of hemato-
logical malignancies. The MILE Study research program
was launched in 11 centers: seven European centers in asso-
ciation with the European Leukemia Network (ELN),166
three centers from the United States, and one in Singapore.
This study included ~3000 patients (from all types of leu-
kemia) and assessed the clinical accuracy of gene-expression
profiles of 16 acute and chronic leukemia subclasses,
including MDS, with current gold-standard routine diag-
nostic work-up.167 This study showed a high sensitivity and
specificity of the gene-profiling approach for disease classi-
fication, and although it was overtaken by next-generation
sequencing approaches, it has been an invaluable source of
data for further studies.
C O N C LU S I O N
The genomics of AML is one of the fastest moving areas
of cancer research. It is now one of the best molecularly
understood malignant diseases, and its dissection and
understanding of the molecular abnormalities continues
to be a paradigm for researching numerous other diseases.
The molecular phenotype now has the potential to be used
for diagnosis, risk prediction, therapeutic decisions, and
achieving an understanding of the development of resis-
tance and relapse.
REFERENCES
1. Lightfoot, T. Aetiology of childhood leukemia. Bioelectromagnetics.
Suppl 7, S5–S11. 2005.
2. Dores, G. M., Devesa, S. S., Curtis, R. E., Linet, M. S., and Morton,
L. M. Acute leukemia incidence and patient survival among children
and adults in the United States, 2001–2007. Blood. 119, 34–43.
2011. doi:10.1182/blood-2011-04-347872.
3. Stone, R. M., O’Donnell, M. R., and Sekeres, M. A. Acute myeloid
leukemia. Hematology Am. Soc. Hematol. Educ. Program. 98–117.
2004.
4. Lowenberg, B., Griffin, J.  D., and Tallman, M.  S. Acute myeloid
leukemia and acute promyelocytic leukemia. Hematology Am. Soc.
Hematol. Educ. Program. 82–101. 2003.
G enetic and G eno m ics in C linical H e m atology, I I I :   Acute L eu k e m ias   •   4 2 5
5. Bishop, J. F. The treatment of adult acute myeloid leukemia. Semin.
Oncol. 24(1), 57–69. 1997.
6. Stanulla, M., Wang, J., Chervinsky, D.  S., Thandla, S., and Aplan,
P.  D. DNA cleavage within the MLL breakpoint cluster region is
a specific event which occurs as part of higher-order chromatin
fragmentation during the initial stages of apoptosis. Mol. Cell Biol.
17(7), 4070–4079. 1997.
7. Stanulla, M., Wang, J., Chervinsky, D.  S., and Aplan, P.  D.
Topoisomerase II inhibitors induce DNA double-strand breaks at
a specific site within the AML1 locus. Leukemia. 11(4), 490–496.
1997.
8. Blanco, J. G., Dervieux, T., Edick, M. J., et al. Molecular emergence
of acute myeloid leukemia during treatment for acute lymphoblas-
tic leukemia. Proc. Natl. Acad. Sci. U. S. A. 98(18), 10338–10343.
8-28-2001.
9. Deininger, M. W. N., Bose, S., Gora-Tybor, J., Yan, X-H., Goldman,
J.  M., and Melo, J.  V. Selective induction of leukemia-associated
fusion genes by high-dose ionizing radiation. Cancer Res. 58(3),
421–425. 1998.
10. Andersen, M. K., Christiansen, D. H., Jensen, B. A., Ernst, P., Hauge,
G., and Pedersen-Bjergaard, J. Therapy-related acute lymphoblastic
leukemia with MLL rearrangements following DNA topoisomerase
II inhibitors, an increasing problem: report on two new cases and
review of the literature since 1992. Br. J. Haematol. 114(3), 539–
543. 2001.
11. Bennett, J.  M., Catovsky, D., Daniel, M.  T., et  al. Proposals for
the classification of the acute leukemias. French-American-British
(FAB) co-operative group. Br. J.  Haematol. 33(4), 451–458.
1976.
12. Vardiman, J.  W., Harris, N.  L., and Brunning, R.  D. The World
Health Organization (WHO) classification of the myeloid neo-
plasms. Blood. 100(7), 2292–2302. 10-1-2002.
13. Swerdlow, S. H., Campa, E., Harris, N. L., et al. WHO Classification
of Tumours of Haematopoietic and Lymphoid Tissues. Lyon,
France: IARC; 2008.
14. Miyoshi, H., Shimizu, K., Kozu, T., Maseki, N., Kaneko, Y., and
Ohki, M. t(8;21) breakpoints on chromosome 21 in acute myeloid
leukemia are clustered within a limited region of a single gene,
AML1. Proc. Natl. Acad. Sci. U.  S. A. 88(23), 10431–10434.
12-1-1991.
15. Warrell, R. P., Jr., De The, H., Wang, Z. Y., and Degos, L. Acute pro-
myelocytic leukemia. N. Engl. J. Med. 329, 177–189. 1993.
16. Le Beau, M.  M., Larson, R.  A., Bitter, M.  A., Vardiman, J.  W.,
Golomb, H.  M., and Rowley, J.  D. Association of an inversion of
chromosome 16 with abnormal marrow eosinophils in acute myelo-
monocytic leukemia. A  unique cytogenetic-clinicopathological
association. N. Engl. J. Med. 309(11), 630–636. 9-15-1983.
17. Larson, R.  A., Williams, S.  F., Le Beau, M.  M., Bitter, M.  A.,
Vardiman, J. W., and Rowley, J. D. Acute myelomonocytic leukemia
with abnormal eosinophils and inv(16) or t(16;16) has a favorable
prognosis. Blood. 68(6), 1242–1249. 1986.
18. Rowley, J. D. Chromosome translocations: dangerous liaisons revis-
ited. Nat. Rev. Cancer. 1(3), 245–250. 2001.
19. Zwaan, ChM and Kaspers, G.  J. Possibilities for tailored and tar-
geted therapy in paediatric acute myeloid leukemia. Br. J. Haematol.
127(3), 264–279. 2004.
20. Kalwinsky, D.  K., Raimondi, S.  C., Schell, M.  J., et  al. Prognostic
importance of cytogenetic subgroups in de novo pediatric acute
nonlymphocytic leukemia. J. Clin. Oncol. 8(1), 75–83. 1990.
21. Grimwade, D., Walker, H., Oliver, F., et al. The importance of diag-
nostic cytogenetics on outcome in AML: analysis of 1,612 patients
entered into the MRC AML 10 trial. The Medical Research Council
Adult and Children’s Leukemia Working Parties. Blood. 92(7),
2322–2333. 10-1-1998.
22. Mrozek, K., Heinonen, K., de la, Chapelle A., and Bloomfield, C. D.
Clinical significance of cytogenetics in acute myeloid leukemia.
Semin. Oncol. 24(1), 17–31. 1997.
 23. Lowenberg, B., Downing, J. R., and Burnett, A. Acute myeloid leu-
kemia. N. Engl. J. Med. 341(14), 1051–1062. 9-30-1999.
24. Grimwade, D., Hills, R. K., Moorman, A. V., et al., and on behalf of
the National Cancer Research Institute Adult Leukemia Working
Group. Refinement of cytogenetic classification in acute myeloid
leukemia: determination of prognostic significance of rare recurring
chromosomal abnormalities among 5876 younger adult patients
treated in the United Kingdom Medical Research Council trials.
Blood. 116(3), 354–365. 7-22-2010.
25. Dohner, H., Estey, E.  H., Amadori, S., et  al. Diagnosis and man-
agement of acute myeloid leukemia in adults:  recommendations
from an international expert panel, on behalf of the European
LeukemiaNet. Blood. 115(3), 453–474. 1-21-2010.
26. Grossmann, V., Schnittger, S., Kohlmann, A., et al. A novel hierar-
chical prognostic model of AML solely based on molecular muta-
tions. Blood. 120(2963), 2972. 8-20-2012.
27. Grimwade, D., Walker, H., Harrison, G., et al. The predictive value
of hierarchical cytogenetic classification in older adults with acute
myeloid leukemia (AML): analysis of 1065 patients entered into the
United Kingdom Medical Research Council AML11 trial. Blood.
98(5), 1312–1320. 9-1-2001.
28. Vickers, M., Jackson, G., and Taylor, P. The incidence of acute pro-
myelocytic leukemia appears constant over most of a human lifes-
pan, implying only one rate limiting mutation. Leukemia. 14(4),
722–726. 2000.
29. Pulsoni, A., Stazi, A., Cotichini, R., et al. Acute promyelocytic leu-
kemia:  epidemiology and risk factors. A  report of the GIMEMA
Italian archive of adult acute leukemia. GIMEMA Cooperative
Group. Eur. J. Haematol. 61(5), 327–332. 1998.
30. Pulsoni, A., Pagano, L., Lo Coco, F., et al. Clinicobiological features
and outcome of acute promyelocytic leukemia occurring as a sec-
ond tumor: the GIMEMA experience. Blood. 100(6), 1972–1976.
9-15-2002.
31. Pandolfi, P. P. In vivo analysis of the molecular genetics of acute pro-
myelocytic leukemia. Oncogene. 20(40), 5726–5735. 9-10-2001.
32. Pandolfi, P. P. Oncogenes and tumor suppressors in the molecular
pathogenesis of acute promyelocytic leukemia. Hum. Mol. Genet.
10(7), 769–775. 2001.
33. Chen, S.  J., Zelent, A., Tong, J.  H., et  al. Rearrangements of the
retinoic acid receptor alpha and promyelocytic leukemia zinc finger
genes resulting from t(11; 17)(q23;q21) in a patient with acute pro-
myelocytic leukemia. J. Clin. Invest. 91, 2260–2267. 1993.
34. Chen, Z., Brand, N.  J., Chen, A., et  al. Fusion between a novel
Kruppel-like zinc finger gene and the retinoic acid receptor-alpha
locus due to a variant t(11;17) translocation associated with acute
promyelocytic leukemia. EMBO J. 12, 1161–1167. 1993.
35. Redner, R. L., Chen, J. D., Rush, E. A., Li, H., and Pollock, S. L. The
t(5;17) acute promyelocytic leukemia fusion protein NPM-RAR
interacts with co-repressor and co-activator proteins and exhibits
both positive and negative transcriptional properties. Blood. 95(8),
2683–2690. 4-15-2000.
36. Redner, R. L., Contis, L. C., Craig, F., Evans, C., Sherer, M. E., and
Shekhter-Levin, S. A novel t(3;17)(p25;q21) variant translocation
of acute promyelocytic leukemia with rearrangement of the RARA
locus. Leukemia. 20(2), 376–379. 2006.
37. Wells, R. A., Hummel, J. L., De Koven, A., et al. A new variant trans-
location in acute promyelocytic leukemia:  molecular characteriza-
tion and clinical correlation. Leukemia. 10(4), 735–740. 1996.
38. Wells, R. A., Catzavelos, C., and Kamel-Reid, S. Fusion of retinoic
acid receptor alpha to NuMA, the nuclear mitotic apparatus pro-
tein, by a variant translocation in acute promyelocytic leukemia.
Nat. Genet. 17(1), 109–113. 1997.
39. Arnould, C., Philippe, C., Bourdon, V., Grégoire, M. J., Berger, R.,
and Jonveaux, P. The signal transducer and activator of transcrip-
tion STAT5b gene is a new partner of retinoic acid receptor alpha
in acute promyelocytic-like leukemia. Hum. Mol. Genet. 8(9),
1741–1749. 1999.
4 2 6   •   G eno m ics in C linical P ractice
40. Melnick, A. and Licht, J. D. Deconstructing a disease: RARalpha,
its fusion partners, and their roles in the pathogenesis of acute pro-
myelocytic leukemia. Blood. 93(10), 3167–3215. 5-15-1999.
41. Grignani, F., De Matteis, S., Nervi, C., et al. Fusion proteins of the
retinoic acid receptor-alpha recruit histone deacetylase in promyelo-
cytic leukemia. Nature 391, 815–818. 1998.
42. He, L. Z., Guidez, F., Tribioli, C., Peruzzi, D., Ruthardt, M., Zelent,
A., and Pandolfi, P.  P. Distinct interactions of PML-RARalpha
and PLZF-RARalpha with co-repressors determine differential
responses to RA in APL. Nat. Genet. 18, 126–135. 1998.
43. Lin, R. J., Nagy, L., Inoue, S., Shao, W., Miller, W. H., Jr., and Evans,
R. M. Role of the histone deacetylase complex in acute promyelo-
cytic leukemia. Nature. 391, 811–814. 1998.
44. Fenaux, P., Chomienne, C., and Degos, L. Treatment of acute pro-
myelocytic leukemia. Best. Pract. Res. Clin. Haematol. 14(1), 153–
174. 2001.
45. Ablain, J. and De The, H. Revisiting the differentiation paradigm in
acute promyelocytic leukemia. Blood. 117, 5795–5802. 3-28-2011.
46. Grimwade, D., Jovanovic, J. V., Hills, R. K., et al. Prospective mini-
mal residual disease monitoring to predict relapse of acute promy-
elocytic leukemia and to direct pre-emptive arsenic trioxide therapy.
J. Clin. Oncol. 27(22), 3650–3658. 8-1-2009.
47. Esteve, J., Escoda, L., Martin, G., et  al. Outcome of patients with
acute promyelocytic leukemia failing to front-line treatment with
all-trans retinoic acid and anthracycline-based chemotherapy
(PETHEMA protocols LPA96 and LPA99):  benefit of an early
intervention. Leukemia. 21(3), 446–452. 2007.
48. Lo, Coco F., Diverio, D., Avvisati, G., et  al. Therapy of molecular
relapse in acute promyelocytic leukemia. Blood. 94(7), 2225–2229.
10-1-1999.
49. Grignani, F., Valtieri, M., Gabbianelli, M., et al. PML/RAR alpha
fusion protein expression in normal human hematopoietic progeni-
tors dictates myeloid commitment and the promyelocytic pheno-
type. Blood. 96(4), 1531–1537. 8-15-2000.
50. Le Beau, M. M., Bitts, S., Davis, E. M., and Kogan, S. C. Recurring
chromosomal abnormalities in leukemia in PML-RARA transgenic
mice parallel human acute promyelocytic leukemia. Blood. 99(8),
2985–2991. 4-15-2002.
51. Le Beau, M. M., Davis, E. M., Patel, B., Phan, V. T., Sohal, J., and
Kogan, S.  C. Recurring chromosomal abnormalities in leukemia
in PML-RARA transgenic mice identify cooperating events and
genetic pathways to acute promyelocytic leukemia. Blood. 102(3),
1072–1074. 8-1-2003.
52. Alcalay, M., Orleth, A., Sebastiani, C., et  al. Common themes in
the pathogenesis of acute myeloid leukemia. Oncogene. 20(40),
5680–5694. 9-10-2001.
53. Pollock, J.  L., Westervelt, P., Walter, M.  J., Lane, A.  A., and Ley,
T. J. Mouse models of acute promyelocytic leukemia. Curr. Opin.
Hematol. 8(4), 206–211. 2001.
54. Mistry, A. R., Pedersen, E. W., Solomon, E., and Grimwade, D. The
molecular pathogenesis of acute promyelocytic leukemia: implica-
tions for the clinical management of the disease. Blood. Rev. 17(2),
71–97. 2003.
55. Catalano, A., Dawson, M.  A., Somana, K., et  al. The PRKAR1A
gene is fused to RARA in a new variant acute promyelocytic leuke-
mia. Blood. 110(12), 4073–4076. 12-1-2007.
56. Dong, S., Zhu, J., Reid, A., et al. Amino-terminal protein-protein
interaction motif (POZ-domain) is responsible for activities of the
promyelocytic leukemia zinc finger–retinoic acid receptor-alpha
fusion protein. Proc. Natl. Acad. Sci. U. S. A. 93, 3624–3629. 1996.
57. Yasmeen, A., Berdel, W.  E., Serve, H., and Muller-Tidow, C.  E-
and A-type cyclins as markers for cancer diagnosis and prognosis.
Expert. Rev. Mol. Diagn. 3(5), 617–633. 2003.
58. Muller, C., Readhead, C., Diederichs, S., et al. Methylation of the
cyclin A1 promoter correlates with gene silencing in somatic cell
lines, while tissue-specific expression of cyclin A1 is methylation
independent. Mol. Cell Biol. 20(9), 3316–3329. 2000.
 59. Muller, C., Yang, R., Park, D.  J., Serve, H., Berdel, W.  E., and
Koeffler, H. P. The aberrant fusion proteins PML-RAR alpha and
PLZF-RAR alpha contribute to the overexpression of cyclin A1
in acute promyelocytic leukemia. Blood. 96(12), 3894–3899.
12-1-2000.
60. Liao, C., Wang, X. Y., Wei, H. Q., et al. Altered myelopoiesis and
the development of acute myeloid leukemia in transgenic mice
overexpressing cyclin A1. Proc. Natl. Acad. Sci. U.  S. A.98(12),
6853–6858. 6-5-2001.
61. Meyers, S., Lenny, N., and Hiebert, S. W. The t(8;21) fusion protein
interferes with AML-1B-dependent transcriptional activation. Mol.
Cell Biol. 15(4), 1974–1982. 1995.
62. Amann, J. M., Nip, J., Strom, D. K., et al. ETO, a target of t(8;21)
in acute leukemia, makes distinct contacts with multiple histone
deacetylases and binds mSin3A through its oligomerization domain.
Mol. Cell Biol. 21(19), 6470–6483. 2001.
63. Gelmetti, V., Zhang, J., Fanelli, M., Minucci, S., Pelicci, P.  G.,
and Lazar, M.  A. Aberrant recruitment of the nuclear receptor
corepressor-histone deacetylase complex by the acute myeloid leuke-
mia fusion partner ETO. Mol. Cell Biol. 18(12), 7185–7191. 1998.
64. Frank, R., Zhang, J., Uchida, H., Meyers, S., Hiebert, S.  W., and
Nimer, S. D. The AML1/ETO fusion protein blocks transactiva-
tion of the GM-CSF promoter by AML1B. Oncogene. 11(12),
2667–2674. 12-21-1995.
65. Westendorf, J.  J., Yamamoto, C.  M., Lenny, N., Downing, J.  R.,
Selsted, M.  E., and Hiebert, S.  W. The t(8;21) fusion prod-
uct, AML-1-ETO, associates with C/EBP-alpha, inhibits C/
EBP-alpha-dependent transcription, and blocks granulocytic differ-
entiation. Mol. Cell Biol. 18(1), 322–333. 1998.
66. Vangala, R.  K., Heiss-Neumann, M.  S., Rangatia, J.  S., et  al. The
myeloid master regulator transcription factor PU.1 is inactivated by
AML1-ETO in t(8;21) myeloid leukemia. Blood. 101(1), 270–277.
1-1-2003.
67. Mueller, B.  U. and Pabst, T. C/EBPalpha and the pathophysiol-
ogy of acute myeloid leukemia. Curr. Opin. Hematol. 13(1), 7–14.
2006.
68. Muller-Tidow, C., Wang, W., Idos, G. E., et al. Cyclin A1 directly
interacts with B-myb and cyclin A1/cdk2 phosphorylate B-myb at
functionally important serine and threonine residues: tissue-specific
regulation of B-myb function. Blood. 97(7), 2091–2097. 4-1-2001.
69. Linggi, B., Muller-Tidow, C., van de, Locht L., et  al. The t(8;21)
fusion protein, AML1 ETO, specifically represses the transcription
of the p14(ARF) tumor suppressor in acute myeloid leukemia. Nat.
Med. 8(7), 743–750. 2002.
70. Hiebert, S. W., Reed-Inderbitzin, E. F., Amann, J., Irvin, B., Durst, K.,
and Linggi, B. The t(8;21) fusion protein contacts co-repressors and
histone deacetylases to repress the transcription of the p14(ARF)
tumor suppressor. Blood Cells Mol. Dis. 30(2), 177–183. 2003.
71. Muller-Tidow, C., Metzelder, S.  K., Buerger, H., et  al. Expression
of the p14ARF tumor suppressor predicts survival in acute myeloid
leukemia. Leukemia. 18(4), 720–726. 2004.
72. Klampfer, L., Zhang, J., Zelenetz, A.  O., Uchida, H., and Nimer,
S.  D. The AML1/ETO fusion protein activates transcription of
BCL-2. Proc. Natl. Acad. Sci. U.  S. A. 93(24), 14059–14064.
11-26-1996.
73. Burel, S. A., Harakawa, N., Zhou, L., Pabst, T., Tenen, D. G., and
Zhang, D. E. Dichotomy of AML1-ETO functions: growth arrest
versus block of differentiation. Mol. Cell Biol. 21(16), 5577–5590.
2001.
74. Liu, P., Tarle, S. A., Hajra, A., et al. Fusion between transcription
factor CBF beta/PEBP2 beta and a myosin heavy chain in acute
myeloid leukemia. Science. 261, 1041–1044. 1993.
75. Lutterbach, B., Hou, Y., Durst, K. L., and Hiebert, S. W. The inv(16)
encodes an acute myeloid leukemia 1 transcriptional corepressor.
Proc. Natl. Acad. Sci. U. S. A. 96(22), 12822–12827. 10-26-1999.
76. Nakamura, T., Mori, T., Tada, S., et al, and Canaani, E. ALL-1 is a
histone methyltransferase that assembles a supercomplex of proteins
G enetic and G eno m ics in C linical H e m atology, I I I :   Acute L eu k e m ias   •   4 2 7
involved in transcriptional regulation. Mol. Cell 10(5), 1119–1128.
2002.
77. Aravind, L. and Landsman, D. AT-hook motifs identified in a
wide variety of DNA-binding proteins. Nucleic Acids Res. 26(19),
4413–4421. 10-1-1998.
78. Satake, N., Maseki, N., Nishiyama, M., et al. Chromosome abnor-
malities and MLL rearrangements in acute myeloid leukemia of
infants. Leukemia. 13(7), 1013–1017. 1999.
79. Schoch, C., Schnittger, S., Klaus, M., Kern, W., Hiddemann, W.,
and Haferlach, T. AML with 11q23/MLL abnormalities as defined
by the WHO classification: incidence, partner chromosomes, FAB
subtype, age distribution, and prognostic impact in an unselected
series of 1897 cytogenetically analyzed AML cases. Blood. 102(7),
2395–2402. 10-1-2003.
80. Huret, J.  L., Senon, S., Bernheim, A., and Dessen, P. An Atlas on
genes and chromosomes in oncology and haematology. Cell Mol.
Biol. (Noisy.-le-grand) 50(7), 805–807. 2004.
81. Ayton, P.  M. and Cleary, M.  L. Transformation of myeloid pro-
genitors by MLL oncoproteins is dependent on Hoxa7 and Hoxa9.
Genes Dev. 17(18), 2298–2307. 9-15-2003.
82. Armstrong, S. A., Staunton, J. E., Silverman, L. B., et al. MLL trans-
locations specify a distinct gene expression profile that distinguishes
a unique leukemia. Nat. Genet. 30(1), 41–47. 2002.
83. Ghannam, G., Takeda, A., Camarata, T., Moore, M. A., Viale, A.,
and Yaseen, N.  R. The oncogene Nup98-HOXA9 induces gene
transcription in myeloid cells. J. Biol. Chem. 8, 1741–1749. 2003.
84. Kawakami, K., Miyanishi, S., Nishii, K., et al. A case of acute myeloid
leukemia with t(7;11)(p15;p15) mimicking myeloid crisis of chronic
myelogenous leukemia. Int. J. Hematol. 76(1), 80–83. 2002.
85. Taketani, T., Taki, T., Ono, R., Kobayashi, Y., Ida, K., and Hayashi, Y.
The chromosome translocation t(7;11)(p15;p15) in acute myeloid
leukemia results in fusion of the NUP98 gene with a HOXA clus-
ter gene, HOXA13, but not HOXA9. Gene. Chromosome. Canc.
34(4), 437–443. 2002.
86. Kasper, L.  H., Brindle, P.  K., Schnabel, C.  A., Pritchard, C.  E.,
Cleary, M. L., and van Deursen, J. M. CREB binding protein inter-
acts with nucleoporin-specific FG repeats that activate transcription
and mediate NUP98-HOXA9 oncogenicity. Mol. Cell Biol. 19(1),
764–776. 1999.
87. Mardis, E.  R., Ding, L., Dooling, D., et  al. Recurring mutations
found by sequencing an acute myeloid leukemia genome. N. Engl.
J. Med. 361(11), 1056–1066. 2009.
88. Ley, T. J., Ding, L., Walter, M. J., et al. DNMT3A mutations in acute
myeloid leukemia. N. Engl. J. Med. 363(25), 2424–2433,2010.
89. Grossmann, V., Tiacci, E., Holmes, A.  B., et  al. Whole-exome
sequencing identifies mutations of BCOR in acute myeloid leuke-
mia with normal karyotype. Blood. 118, 6153–6163. 10-19-2011.
90. Tyner, J.  W., Erickson, H., Deininger, M.  W. N., et  al.
High-throughput sequencing screen reveals novel, transforming
RAS mutations in myeloid leukemia patients. Blood. 113(8), 1749–
1755. 2-19-2009.
91. Wouters, B. J., Lowenberg, B., Erpelinck-Verschueren, C. A. J., Van
Putten, W. L. J., Valk, P. J. M., and Delwel, R. Double CEBPA muta-
tions, but not single CEBPA mutations, define a subgroup of acute
myeloid leukemia with a distinctive gene expression profile that is
uniquely associated with a favorable outcome. Blood. 113(13),
3088–3091. 2009.
92. Dicker, F., Haferlach, C., Sundermann, J., et al. Mutation analysis for
RUNX1, MLL-PTD, FLT3-ITD, NPM1 and NRAS in 269 patients
with MDS or secondary AML. Leukemia. 24(8), 1528–1532. 2010.
93. Schnittger, S., Dicker, F., Kern, W., et al. RUNX1 mutations are fre-
quent in de novo AML with non complex karyotype and confer an
unfavourable prognosis. Blood. 117(8), 2348–2357. 2010.
94. Chou, W. C., Chou, S. C., Liu, C. Y., et al. TET2 mutation is an
unfavorable prognostic factor in acute myeloid leukemia patients
with intermediate-risk cytogenetics. Blood. 118(14), 3803–3810.
2011.
 95. Gelsi-Boyer, V., Brecqueville, M., Devillier, R., Murati, A.,
Mozziconacci, M. J., and Birnbaum, D. Mutations in ASXL1 are
associated with poor prognosis across the spectrum of malignant
myeloid diseases. J. Hematol. Oncol. 5, 12. 2012.
96. Schnittger, S., Eder, C., Jeromin, S., et al. ASXL1 exon 12 muta-
tions are frequent in AML with intermediate risk karyotype and
are independently associated with an adverse outcome. Leukemia.
27(1), 82–91. 2013.
97. The Cancer Genome Atlas Research Network. Genomic and
epigenomic landscapes of adult de novo acute myeloid leuke-
mia. N. Engl. J.  Med. 368(22), 2059–2074. 2013. doi:10.1056/
NEJMoa1301689.
98. Rosnet, O., Schiff, C., Pebusque, M. J., et al. Human FLT3/FLK2
gene: cDNA cloning and expression in hematopoietic cells. Blood.
82(4), 1110–1119. 8-15-1993.
99. Carow, C. E., Kim, E., Hawkins, A. L., et al. Localization of the
human stem cell tyrosine kinase-1 gene (FLT3) to 13q12-->q13.
Cytogenet. Cell Genet. 70(3–4), 255–257. 1995.
100. Abu-Duhier, F.  M., Goodeve, A.  C., Wilson, G.  A., Care,
R.  S., Peake, I.  R., and Reilly, J.  T. Genomic structure of human
FLT3:  implications for mutational analysis. Br. J.  Haematol.
113(4), 1076–1077. 2001.
101. Maroc, N., Rottapel, R., Rosnet, O., et al. Biochemical character-
ization and analysis of the transforming potential of the FLT3/
FLK2 receptor tyrosine kinase. Oncogene. 8(4), 909–918. 1993.
102. Adolfsson, J., Borge, O. J., Bryder, D., et al. Upregulation of Flt3
expression within the bone marrow Lin(-)Sca1(+)c-kit(+) stem
cell compartment is accompanied by loss of self-renewal capacity.
Immunity. 15(4), 659–669. 2001.
103. Christensen, J. L. and Weissman, I. L. Flk-2 is a marker in hema-
topoietic stem cell differentiation:  a simple method to isolate
long-term stem cells. Proc. Natl. Acad. Sci. U. S. A. 98(25), 14541–
14546. 12-4-2001.
104. Rappold, I., Ziegler, B.  L., Kohler, I., et  al. Functional and phe-
notypic characterization of cord blood and bone marrow subsets
expressing FLT3 (CD135) receptor tyrosine kinase. Blood. 90(1),
111–125. 7-1-1997.
105. Marchetto, S., Fournier, E., Beslu, N., et al. SHC and SHIP phos-
phorylation and interaction in response to activation of the FLT3
receptor. Leukemia. 13(9), 1374–1382. 1999.
106. Zhang, S. and Broxmeyer, H. E. Flt3 ligand induces tyrosine phos-
phorylation of gab1 and gab2 and their association with shp-2,
grb2, and PI3 kinase. Biochem. Biophys. Res. Commun. 277(1),
195–199. 10-14-2000.
107. Lavagna-Sevenier, C., Marchetto, S., Birnbaum, D., and Rosnet,
O. The CBL-related protein CBLB participates in FLT3 and
interleukin-7 receptor signal transduction in pro-B cells. J. Biol.
Chem. 273(24), 14962–14967. 6-12-1998.
108. Mizuki, M., Fenski, R., Halfter, H., et  al. Flt3 mutations from
patients with acute myeloid leukemia induce transformation of
32D cells mediated by the Ras and STAT5 pathways. Blood.
96(12), 3907–3914. 12-1-2000.
109. Stacchini, A., Fubini, L., Severino, A., Sanavio, F., Aglietta, M.,
and Piacibello, W. Expression of type III receptor tyrosine kinases
FLT3 and KIT and responses to their ligands by acute myeloid leu-
kemia blasts. Leukemia. 10(10), 1584–1591. 1996.
110. Nakao, M., Yokota, S., Iwai, T., et al. Internal tandem duplication
of the FLT3 gene found in acute myeloid leukemia. Leukemia.
10(12), 1911–1918. 1996.
111. Abu-Duhier, F.  M., Goodeve, A.  C., Wilson, G.  A., et  al. FLT3
internal tandem duplication mutations in adult acute myeloid leu-
kemia define a high-risk group. Br. J. Haematol. 111(1), 190–195.
2000.
112. Schnittger, S., Schoch, C., Dugas, M., et al. Analysis of FLT3 length
mutations in 1003 patients with acute myeloid leukemia: correla-
tion to cytogenetics, FAB subtype, and prognosis in the AMLCG
4 2 8   •   G eno m ics in C linical P ractice
study and usefulness as a marker for the detection of minimal resid-
ual disease. Blood. 100(1), 59–66. 7-1-2002.
113. Kottaridis, P.  D., Gale, R.  E., Frew, M.  E., et  al. The presence
of a FLT3 internal tandem duplication in patients with acute
myeloid leukemia (AML) adds important prognostic informa-
tion to cytogenetic risk group and response to the first cycle of
chemotherapy: analysis of 854 patients from the United Kingdom
Medical Research Council AML 10 and 12 trials. Blood. 98(6),
1752–1759. 9-15-2001.
114. Mills, K.  I., Gilkes, A.  F., Walsh, V., Sweeney, M., and Gale, R.
Rapid and sensitive detection of internal tandem duplication
and activating loop mutations of FLT3. Br. J. Haematol. 130(2),
203–208. 2005.
115. Kiyoi, H., Naoe, T., Nakano, Y., et  al. Prognostic implication of
FLT3 and N-RAS gene mutations in acute myeloid leukemia.
Blood. 93(9), 3074–3080. 5-1-1999.
116. Kottaridis, P. D., Gale, R. E., and Linch, D. C. Prognostic impli-
cations of the presence of FLT3 mutations in patients with acute
myeloid leukemia. Leuk. Lymphoma 44(6), 905–913. 2003.
117. Gale, R. E., Hills, R., Pizzey, A. R., et al. The relationship between
FLT3 mutation status, biological characteristics and response to
targeted therapy in acute promyelocytic leukemia. Blood. 106(12),
3768–3776. 8-16-2005.
118. Stirewalt, D. L., Kopecky, K. J., Meshinchi, S., et al. FLT3, RAS,
and TP53 mutations in elderly patients with acute myeloid leuke-
mia. Blood. 97(11), 3589–3595. 6-1-2001.
119. Fitzgibbon, J., Smith, L.  L., Raghavan, M., et  al. Association
between acquired uniparental disomy and homozygous gene muta-
tion in acute myeloid leukemias. Cancer Res. 65(20), 9152–9154.
10-15-2005.
120. Shih, L.  Y., Huang, C.  F., Wu, J.  H., et  al. Heterogeneous pat-
terns of FLT3 Asp(835) mutations in relapsed de novo acute
myeloid leukemia: a comparative analysis of 120 paired diagnos-
tic and relapse bone marrow samples. Clin. Cancer Res. 10(4),
1326–1332. 2-15-2004.
121. Abu-Duhier, F.  M., Goodeve, A.  C., Wilson, G.  A., Care, R.  S.,
Peake, I. R., and Reilly, J. T. Identification of novel FLT-3 Asp835
mutations in adult acute myeloid leukemia. Br. J.  Haematol.
113(4), 983–988. 2001.
122. Frohling, S., Schlenk, R.  F., Breitruck, J., et  al. Prognostic sig-
nificance of activating FLT3 mutations in younger adults (16 to
60  years) with acute myeloid leukemia and normal cytogenet-
ics:  a study of the AML Study Group, Ulm. Blood. 100(13),
4372–4380. 12-15-2002.
123. Spiekermann, K., Bagrintseva, K., Schoch, C., Haferlach, T.,
Hiddemann, W., and Schnittger, S. A new and recurrent activat-
ing length mutation in exon 20 of the FLT3 gene in acute myeloid
leukemia. Blood. 100(9), 3423–3425. 11-1-2002.
124. Hayakawa, F., Towatari, M., Kiyoi, H., et  al. Tandem-duplicated
Flt3 constitutively activates STAT5 and MAP kinase and intro-
duces autonomous cell growth in IL-3-dependent cell lines.
Oncogene. 19(5), 624–631. 2-3-2000.
125. Minami, Y., Yamamoto, K., Kiyoi, H., Ueda, R., Saito, H., and
Naoe, T. Different antiapoptotic pathways between wild-type
and mutated FLT3: insights into therapeutic targets in leukemia.
Blood. 102(8), 2969–2975. 10-15-2003.
126. Falini, B., Bolli, N., Shan, J., et  al. Both carboxy-terminus NES
motif and mutated tryptophan(s) are crucial for aberrant nuclear
export of nucleophosmin leukemic mutants in NPMc+ AML.
Blood. 107(11), 4514–4523. 2006.
127. Grisendi, S., Mecucci, C., Falini, B., and Pandolfi, P.  P.
Nucleophosmin and cancer. Nat. Rev. Cancer. 6(7), 493–505.
2006.
128. Vassiliou, G. S., Cooper, J. L., Rad, R., et al. Mutant nucleophos-
min and cooperating pathways drive leukemia initiation and pro-
gression in mice. Nat. Genet. 43(5), 470–475.2011.
 129. Dohner, H. and Gaidzik, V. I. Impact of genetic features on treat-
ment decisions in AML. Hematology Am. Soc. Hematol. Educ.
Program. 2011, 36–42. 2011.
130. Becker, H., Marcucci, G., Maharry, K., et al. Favorable prognostic
impact of NPM1 mutations in older patients with cytogenetically
normal de novo acute myeloid leukemia and associated gene- and
microRNA-expression signatures: a Cancer and Leukemia Group
B study. J. Clin. Oncol. 28(4), 596–604. 2-1-2010.
131. Fasan, A., Alpermann, T., Haferlach, C., et al. Frequency and prog-
nostic impact of CEBPA proximal, distal and core promoter meth-
ylation in normal karyotype AML: a study on 623 cases. PLoS One
8(2), e54365. 2013.
132. Kirstetter, P., Schuster, M.  B., Bereshchenko, O., et  al. Modeling
of C/EBPalpha mutant acute myeloid leukemia reveals a common
expression signature of committed myeloid leukemia-initiating
cells. Cancer Cell 13(4), 299–310. 2008.
133. Green, C. L., Koo, K. K., Hills, R. K., Burnett, A. K., Linch, D. C.,
and Gale, R.  E. Prognostic significance of CEBPA mutations in a
large cohort of younger adult patients with acute myeloid leukemia:
impact of double CEBPA mutations and the interaction with FLT3
and NPM1 mutations. J. Clin. Oncol. 28(16), 2739–2747. 6-1-2010.
134. Paschka, P., Schlenk, R. F., Gaidzik, V. I., et al. IDH1 and IDH2
mutations are frequent genetic alterations in acute myeloid leuke-
mia and confer adverse prognosis in cytogenetically normal acute
myeloid leukemia with NPM1 mutation without FLT3 internal
tandem duplication. J. Clin. Oncol. 28(22), 3636–3643. 8-1-2010.
135. Lu, C., Ward, P. S., Kapoor, G. S., et al. IDH mutation impairs his-
tone demethylation and results in a block to cell differentiation.
Nature. 483(7390), 474–478. 3-22-2012.
136. Green, C.  L., Evans, C.  M., Hills, R.  K., Burnett, A.  K., Linch,
D. C., and Gale, R. E. The prognostic significance of IDH1 muta-
tions in younger adult patients with acute myeloid leukemia is
dependent on FLT3/ITD status. Blood. 118(2), 409–412.2010.
137. Marcucci, G., Maharry, K., Wu, Y. Z., et al. IDH1 and IDH2 gene
mutations identify novel molecular subsets within de novo cytoge-
netically normal acute myeloid leukemia: a Cancer and Leukemia
Group B study. J. Clin. Oncol. 28(14), 2348–2355.2010.
138. Ooi, S. K., O’Donnell, A. H., and Bestor, T. H. Mammalian cyto-
sine methylation at a glance. J. Cell Sci. 122(Pt 16), 2787–2791.
8-15-2009.
139. Gaidzik, V.  I., Schlenk, R.  F., Paschka, P., et  al. Clinical impact
of DNMT3A mutations in younger adult patients with acute
myeloid leukemia:  results of the AML Study Group (AMLSG).
Blood. 121(23), 4769–4777. 6-6-2013.
140. Ley, T.  J., Ding, L., Walter, M.  J., et  al. DNMT3A mutations in
acute myeloid leukemia. N. Engl. J.  Med. 363(25), 2424–2433.
12-16-2010.
141. Ribeiro, A.  F., Pratcorona, M., Erpelinck-Verschueren, C., et  al.
Mutant DNMT3A: a marker of poor prognosis in acute myeloid
leukemia. Blood. 119(24), 5824–5831. 6-14-2012.
142. Gaidzik, V.  I., Paschka, P., Spath, D., et  al. TET2 mutations in
acute myeloid leukemia (AML):  results from a comprehensive
genetic and clinical analysis of the AML study group. J. Clin.
Oncol. 30(12), 1350–1357. 4-20-2012.
143. Metzeler, K.  H., Maharry, K., Radmacher, M.  D., et  al. TET2
mutations improve the new European LeukemiaNet risk classifica-
tion of acute myeloid leukemia: a Cancer and Leukemia Group B
study. J. Clin. Oncol. 29(10), 1373–1381. 4-1-2011.
144. Boultwood, J., Perry, J., Pellagatti, A., et  al. Frequent mutation
of the polycomb-associated gene ASXL1 in the myelodysplas-
tic syndromes and in acute myeloid leukemia. Leukemia. 24(5),
1062–1065. 2010.
145. Schnittger, S., Schoch, C., Kern, W., et  al. Nucleophosmin
gene mutations are predictors of favourable prognosis in acute
myelogenous leukemia with a normal kayotype. Blood 106(12),
3733–3739. 2005.
G enetic and G eno m ics in C linical H e m atology, I I I :   Acute L eu k e m ias   •   4 2 9
146. Steudel, C., Wermke, M., Schaich, M., et al. Comparative analy-
sis of MLL partial tandem duplication and FLT3 internal tan-
dem duplication mutations in 956 adult patients with acute
myeloid leukemia. Gene. Chromosome. Canc. 37(3), 237–251.
2003.
147. Schnittger, S., Haferlach, C., Ulke, M., Alpermann, T., Kern, W.,
and Haferlach, T. IDH1 mutations are detected in 6.6% of 1414
AML patients and are associated with intermediate risk karyo-
type and unfavorable prognosis in adults younger than 60  years
and unmutated NPM1 status. Blood. 116(25), 5486–5496.
12-16-2010.
148. Paschka, P., Marcucci, G., Ruppert, A.  S., et  al. Wilms tumor 1
gene mutations independently predict poor outcome in adults
with cytogenetically normal acute myeloid leukemia: a Cancer and
Leukemia Group B study. J. Clin. Oncol. 26(28), 4595–46022008.
doi:10.1200/JCO.2007.15.2058.
149. Schnittger, S., Dicker, F., Kern, W., et  al. RUNX1 mutations
are frequent in de novo AML with noncomplex karyotype and
confer an unfavorable prognosis. Blood. 117(8), 2348–2357.
2-24-2011.
150. Nagata, H., Worobec, A.  S., Oh, C.  K., et  al. Identification of a
point mutation in the catalytic domain of the protooncogene c-kit
in peripheral blood mononuclear cells of patients who have masto-
cytosis with an associated hematologic disorder. Proc. Natl. Acad.
Sci. U. S. A. 92(23), 10560–10564. 11-7-1995.
151. Care, R. S., Valk, P. J., Goodeve, A. C., et al. Incidence and prog-
nosis of c-KIT and FLT3 mutations in core binding factor (CBF)
acute myeloid leukemias. Br. J. Haematol. 121(5), 775–777. 2003.
152. Bos, J. L., Toksoz, D., Marshall, C. J., et al. Amino-acid substitu-
tions at codon 13 of the N-ras oncogene in human acute myeloid
leukemia. Nature. 315, 726–730. 1985.
153. Bowen, D. T., Frew, M. E., Hills, R., et al. RAS mutation in acute
myeloid leukemia is associated with distinct cytogenetic subgroups
but does not influence outcome in patients <60 yrs. Blood. 106,
2113–2119. 6-16-2005.
154. Paquette, R. L., Landaw, E. M., Pierre, R. V., et al. N-ras mutations
are associated with poor prognosis and increased risk of leukemia
in myelodysplastic syndrome. Blood. 82, 590–599. 1993.
155. Farr, C. J., Saiki, R. K., Erlich, H. A., McCormick, F., and Marshall,
C. J. Analysis of RAS gene mutations in acute myeloid leukemia by
polymerase chain reaction and oligonucleotide probes. Proc. Natl.
Acad. Sci. U. S. A. 85, 1629–1633. 1988.
156. Armstrong, S.  A., Golub, T.  R., and Korsmeyer, S.  J.
MLL-rearranged leukemias: Insights from gene expression profil-
ing. Semin. Hematol. 40(4), 268–273. 2003.
157. Mueller, B. U., Pabst, T., Osato, M., et al. Heterozygous PU.1 muta-
tions are associated with acute myeloid leukemia. Blood. 100(3),
998–1007. 8-1-2002.
158. Parkin, B., Ouillette, P., Li, Y., et al. Clonal evolution and devo-
lution after chemotherapy in adult acute myelogenous leukemia.
Blood. 121(2), 369–377. 1-10-2013.
159. Welch, J. S., Ley, T. J., Link, D. C., et al. The origin and evolution
of mutations in acute myeloid leukemia. Cell. 150(2), 264–278.
7-20-2012.
160. Walter, M.  J., Shen, D., Ding, L., et  al. Clonal architecture of
secondary acute myeloid leukemia. N. Engl. J.  Med. 366(12),
1090–1098. 3-22-2012.
161. Ding, L., Ley, T. J., Larson, D. E., et al. Clonal evolution in relapsed
acute myeloid leukemia revealed by whole-genome sequencing.
Nature. 481(7382), 506–510. 1-26-2012.
162. Golub, T. R., Slonim, D. K., Tamayo, P., et al. Molecular classifica-
tion of cancer: class discovery and class prediction by gene expres-
sion monitoring. Science. 286(5439), 531–537. 10-15-1999.
163. Bullinger, L., Dohner, K., Bair, E., et  al. Use of gene-expression
profiling to identify prognostic subclasses in adult acute myeloid
leukemia. N. Engl. J. Med. 350(16), 1605–1616. 4-15-2004.
 164. Valk, P. J., Verhaak, R. G., Beijen, M. A., et al. Prognostically use-
ful gene-expression profiles in acute myeloid leukemia. N. Engl.
J. Med. 350(16), 1617–1628. 4-15-2004.
165. Haferlach, T., Kohlmann, A., Schnittger, S., et al. A global approach
to the diagnosis of leukemia using gene expression profiling. Blood.
106, 1189–1198. 5-5-2005.
166. European Leukemia Net. http://www.leukemia-net.org
167. Haferlach, T., Kohlmann, A., Wieczorek, L., et  al. Clinical util-
ity of microarray-based gene expression profiling in the diagnosis
and subclassification of leukemia:  report from the International
Microarray Innovations in Leukemia Study Group. J. Clin. Oncol.
28(15), 2529–2537. 2010.

4 3 0   •   G eno m ics in C linical P ractice

 28.
GENETICS AND GENOMICS OF CHRONIC
INFLAMMATORY DISORDER S, I: INFLAMMATORY
B OWEL DISEASE
Saad Pathan and Derek P. Jewell
INTRODUCTION
Ulcerative colitis (UC) and Crohn’s disease (CD) are
chronic inflammatory diseases of the intestine that can
begin at any age, even in children. Both diseases are com-
monly known as inflammatory bowel diseases (IBD).
They represent a major challenge, because their causation
is unclear, medical treatment is far from satisfactory, and
many patients require major surgery. The turning point in
the genomic portrait of IBD was facilitated with the high-
resolution detection of loci that individually have a very
modest influence. There is evidence for over 163 distinct
genetic loci, and this has, to date, been unprecedented for a
complex disease. More specifically, there are at least 30 dis-
tinct loci that are associated with the phenotype of CD, and
23 loci with UC. Moreover, the boundaries between CD
and UC are blurred by the 110 loci that are shared. There
are also remarkable parallels and paradoxes vis-à-vis other
diseases, such as inflammatory, autoimmune, infectious,
and primary immunodeficiencies. The evidence implies that
an unbalanced triggering of the mucosal immune system
towards commensal intestinal flora results in the develop-
ment of chronic inflammatory diseases. The selection pres-
sures for the common IBD-associated variants are likely to
have been protection from infectious diseases. The disease-
associated variants within the genes, such as NOD2, IL23R,
ATG16L1, and CARD9, have the potential to adversely
affect innate and adaptive immunity, as well as muco-
sal barrier integrity and autophagy. A variety of causative
variants have also been discovered within the associated
loci that could render individuals susceptible to IBD and
might influence the behavior of the phenotype. The associa-
tion between the exceedingly rare homozygous mutations
and unusual phenotypes adds an additional dimension to
probe the biology, as exemplified by the interleukin IL-10
431
receptor and IL-10 cytokine deficiencies. The new insights
from unravelling the multitude of disease mechanisms are a
prerequisite for novel, personalized therapeutic strategies.
A COMPLEX CLINICAL
PREDISPOSITION WITH COMPLEX
C O M P L I C AT I O N S
In 1859, Samuel Wilks, at Guy’s Hospital in London, differ-
entiated UC from bacterial dysentery. By 1931, Sir Arthur
Hurst described clinical and sigmoidoscopic features of the
disease and showed that the rectum was always affected,
with extension proximally to involve a variable length of the
colon. The disease is most commonly seen in individuals
between 20 and 40 years of age, but can be present in the
first few months of life or in adults in their eighties. The
disease is characterized by a relapsing and remitting course
in the majority of patients.
The formal recognition of CD as distinct from UC
had to await the finding in 1932 by Crohn, Ginzburg,
and Oppenheimer. The granulomatous (aggregation of
macrophages together with T lymphocytes) nature of the
disease was recognized, but only in approximately 65% of
patients, and it soon became clear that the disease could
affect any part of the gastrointestinal tract. Nevertheless,
it is by no means clear that UC and CD are two homoge-
neous disease entities, and the terms may well encompass
a variety of disorders with shared clinical characteristics
(Table 28.1). When CD is confined to the colon, it may
well be difficult to distinguish it from UC in 5–10%—
these patients are currently labelled as “colitis not yet clas-
sified.” Colonic CD and UC patients are both associated
with a greater overall relative risk of colorectal cancer
(Ekbom et al. 1990b, 1990a; Grivennikov et al. 2010). In
Table 28.1  COMPARISON OF CROHN’S DISEASE WITH ULCERATIVE COLITIS

CROHN’S DISEASE ULCERATIVE COLITIS

Genetic Epidemiology
Prevalence 27–106 per 100, 000 80–150 per 100,000
Incidence 4–10 per 100, 000 6–15 per 100, 000
Sibling affected risk ratio 30 to 42 4
Concordance between identical twins 37% 10%
Concordance between non-identical twins 10% 3%
Mode of inheritance Complex Complex
Loci predominantly linked with either NOD2, ATG16L1, IRGM HNF4A, LAMB1, CDH1, and GNA12
Crohn’s or ulcerative colitis
MHC HLA DR7, DRB3*0301 and DQ7 (asso- HLA class II phenotypes DR2, DR9 and
ciated with colonic Crohn’s) DRB1*0103
Predisposing environmental factor Smoking Smoking
Protective environmental factor
Clinical Features
Bloody diarrhea Less common Common
Abdominal mass Common Rare
Perianal disease Common Less common
Malabsorption Frequent (ileal disease) Never
Radiological/Endoscopic Features
Rectal involvement Frequently spared Invariable
Distribution Segmental discontinuous Continuous
Mucosa Cobblestones, fissure ulcers Fine ulceration, “double contour”
Strictures (narrowing) Common Rare
Fistulas (opening) Frequent Rare
Histological Features
Distribution Transmural Mucosal
Cellular infiltrate Lymphocytes, plasma cells, macrophages Lymphocytes, polymorphs, plasma cells,
eosinophils
Glands Gland preservation Mucus depletion, gland destruction, crypt abscess
Special Features Aphthoid ulcers, granulomas, None
histiocyte-lined fissures
Adapted from the Oxford Textbook of Medicine.

mice affected with colitis, the increased risk of cancer can
be due to the toxicity of the altered microbial composition
(Arthur et al. 2012) as well as modulation by interleukin
22 (IL-22), a cytokine of the IL-10 superfamily (Huber
et al. 2012).
E P I D E M I O L O GY
Over the last century, CD and UC have manifested an increas-
ing incidence in various global populations, with a current esti-
mated prevalence of 0.15% in northwest Europe and North
America (Binder 2004). The environmental influences associ-
ated with modern urban developments have coincided with
increasing incidence and prevalence of IBD in industrialized
countries (see Figure 28.1). The clinical manifestation of IBDs
is likely to be due to a provocation of the mucosal immune sys-
tem by commensal intestinal flora in susceptible individuals—
in most individuals, these would be non-pathogenic (Xavier
and Podolsky 2007). Accumulating evidence from genetic
epidemiology implies that the host genome predisposes to
autoimmune and/or inflammatory diseases (Jostins et  al.
2012). However, genetics alone is not adequate in explaining
the disease, since animal models that are genetically altered
for predisposition to colitis do not develop the phenotype
when kept in a germ-free environment (Blumberg et al. 1999;
Strober 1985).

4 3 2   •   G enomics in C linical P ractice

 SNP8 SNP12 SNP13
4.5 0.9 2.7 SNP8 SNP12 SNP13
2.7 1.2 1.2 3.3 0.6 4.8
1.8 0 1.7
SNP8 SNP12 SNP13
7.1 1.8 4.7
SNP8 SNP12 SNP13 5.4 0.3 2.3
12.9 5.2 10.3 SNP8 SNP12 SNP13 SNP8 SNP12 SNP13
4.2 0.7 0.7 12.5 3.3 9.4 8.5 4.1 11
SNP8 SNP12 SNP1.73 5.2 1.4 1.6 6.1 1.3 4.2
8.2 5.7 5.7
TNFSF15, HNF4A SNP8 SNP12 SNP13
2.9 0.9 1.7
5.3 7.9 2.4
SNP8 SNP12 SNP13
2.9 2.9 0.7
10.6 8.5 9.6
6 2.3 1.4
TNFSF15, HNF4A
High incidence and
prevalence

SNP8 SNP12 SNP13 SNP8 SNP12 SNP13

6 0 1 11.2 2.4 6.7
3 0.5 4 4.5 0.7 1

Figure 28.1 
The regions known for high incidence and prevalence of inflammatory bowel diseases are highlighted on the global atlas. The allele
frequencies (shown in the lighter font showing the cases and the darker font showing the controls) indicate association of the heterozygote
NOD2 mutations Arg702Trp (SNP8), Gly908Arg (SNP12), and the frameshift 3020 insC (SNP13) among populations of European descent
and European admixture populations affected with Crohn’s disease (see section “Epidemiology of NOD2 in CD”; Hampe et al., 2002; Cavanaugh
et al., 2003; Fidder et al., 2003; Palmieri et al., 2003). In contrast, a core TNFSF15 haplotype is associated with CD in both Caucasians and
Japanese populations (see section “TNFSF15”). Furthermore, the HNF4A locus is also associated with UC in both Caucasian and Japanese
populations.

NAT U R E VE R S US NU RT U R E An alternative to a comparison of the concordance rates

between MZ and DZ twins affected with IBD is the mea-
The manifestation of familial clustering as observed by
surement of risk-to-relative ratio for a relative of type R,
Kirsner in 1963 set the foundation for subsequent genetic
known as λR; this method also yields an assessment of the
investigations. Furthermore, within families, there was a
genetic component (Risch 1990). The λR quantifies the
striking concordance of clinical characteristics (Bayless
familial component of a discrete trait that describes the rela-
et al. 1996; Satsangi et al. 1996a). A more rigorous valida-
tionship:  first-degree siblings, closely related family mem-
tion was provided from twin studies that compared disease
bers, and/or other relatives. Therefore, if the risk for a relative
concordance in monozygotic (MZ) and dizygotic (DZ)
or affected sib is 3–3.5% for CD, and the population preva-
twins, because familial clustering does not entirely exclude
lence is 0.1–0.2%, then the λS value is 15–35 (Satsangi et al.
environmental influences (Halfvarson et al. 2005; Orholm
1994). The value between 6 and 9 for UC (Orholm et al.
et al. 2000; Thompson et al. 1996; Tysk et al. 1988). The
1991) implies that the genetic influence is not as powerful as
twin studies provide an indication of the maximum risk of
it is for CD. This measurement could include discrepancies
occurrence. The concordance rates among MZ twins with
due to the introduction of a bias if the family members are
CD and UC range from 37% to 10% respectively, while
exposed to the same environmental factors (Guo 1998). The
the corresponding concordance rates among DZ twins are
λS value may fluctuate among different populations.
only 10% to 3%. Higher concordance rates are seen in CD
than in UC twin pairs. Since twins in general share similar
MARKER S FOR MAPPING
environments during childhood, the higher disease con-
cordance rates in MZ twins are a very strong argument for Segregation analysis followed by homozygosity map-
genetic susceptibility. ping with genetic markers, such as microsatellites, and

G enetics and G enomics of C hronic I nflammatory D isorders , I   •   4 3 3

subsequently sequencing the candidate genes, such as
the IL10R and IL10, have implicated mutations with an
unusual and severe form of enterocolitis (Glocker et  al.
2009b; Glocker et  al. 2010). These patients from consan-
guineous families manifested an autosomal-recessive severe
enterocolitis within the first year of life. In addition to these
exceedingly rare mutations, relatively rare polymorphisms
within the IL10R and IL10 genes have also been associated
with very early onset of UC (Moran et al. 2013). Originally,
the IL10 association was detected within a European UC
case-control study (Franke et al. 2008), and subsequently it
has been confirmed within a broader spectrum of IBD from
a number of different geographical populations ( Jostins
et al. 2012).
The IBD susceptibility genes tend to aggregate in fami-
lies rather than segregate within families. Linkage analysis
with evenly spaced microsatellite markers has also been
applied in a collection of families, followed by denser link-
age mapping and further fine mapping by association with
single-nucleotide polymorphisms (SNPs) and other micro-
satellite markers. Such approaches led to the identification
of NOD2 at 16q12 (see NOD2 section below; Hugot et al.
2001). Genome-wide linkage studies were undertaken to
map genes that would otherwise not be considered, as well as
for enabling the coverage of the exhaustive list of candidate
genes and loci that require systematic analysis. Advances
in high-throughput genotyping to search for susceptibil-
ity loci were originally pioneered with the implementation
of fluorescently labelled genotyping microsatellite mark-
ers in a genome-wide linkage study in families with type 1
diabetes (Davies et al. 1994). Initially, for IBD, there was a
general lack of replication for the numerous loci of linkage
that were reported, and this added to the burden of drawing
conclusions from genome search results, as was the case for
other complex diseases (Altmuller et al. 2001). Nevertheless,
there was rapid replication of the novel, highly significant
16q12 locus (Hugot et al. 1996). To date, there have been
up to a dozen independent genome search results that
have been reported from multiply affected IBD families of
European ancestry, and an emerging pattern of replication
now reaffirms that the following loci are significantly linked
with IBD (Figure 28.2): 16q, 12q, 6p, 14q, 5q, 19p, 1p, 16p,

Meta-Analyses of IBDs

1 2 3 4 5 6 7 8 9 10 11

M
H
C
IL23R

TNFSF15
CARD9

IL10
ATG16L1

Significant and Confirmed Confirmed Confirmed Confirmed

replicated locus locus from region from region from region from
from independent meta-analysis GWAS meta- GWAS meta- GWAS meta-
genome-wide linkage of linkage in analysis in CD analysis in UC analysis in IBDs
studies in IBDs IBDs

12 13 14 15 16 17 18 19 20 21 22

NOD2 HNF4A

Figure 28.2 
The meta-analysis of GWAS has confirmed 110 (dark shaded line) IBD loci, and 30 CD-specific (medium shaded line) and 23
UC-specific (light shaded line). Linkage to IBD at the MHC locus (6p) was also confirmed by a meta-analysis (adapted from van Heel et al.,
2004).

4 3 4   •   G enomics in C linical P ractice

and 3p (Ahmad et al. 2004; Gaya et al. 2006). Furthermore,
two meta-analyses by pooling the results from the different
genome searches have shown that 6p is significantly linked
with disease, and that 16q and 19p show suggestive linkage
(van Heel et  al. 2004; Williams et  al. 2002). The reasons
for these loci not being consistently replicated in all of the
independent investigations could be due to one or more of
the following: a false positive; a true locus that is popula-
tion specific; a study with low power; a variable number of
CD and UC patients; different criteria for diagnosis; differ-
ent markers; and variation in genotyping quality. Similarly,
pooling the data for meta-analysis can be hindered by these
differences between the different investigations.
Unlike the SNPs within the NOD2 gene that were asso-
ciated with CD ( Jostins et al. 2012), the exceedingly rare
disease-causing mutations, such as that within the IL10
and IL10R genes that segregated with a very extreme phe-
notype, were not identical by descent between different
families (Glocker et al. 2009b; Glocker et al. 2010)—even
if the defective gene was the same. Therefore, family studies
using linkage analysis, rather than population case-control
studies, are more appropriate in assessing contribution to
disease, because the very rare disease allele that is associ-
ated with the disease would be more likely to occur in cer-
tain families rather than that at the population level, and
it would not be sufficient to lead to association at the level
of the population. Thus, although linkage analysis is appli-
cable for an exceedingly rare disease-causing mutation, a
population case-control study for association would not be
appropriate for some of the very rare mutations; for exam-
ple, those described within the IL10 and IL10R genes in
colitis (Glocker et al. 2009b; Glocker et al. 2010).
Genome-wide association studies (GWAS) marked a
new era for case-control investigations with SNP markers in
CD, as a successor to genome-wide search with dinucleotide
repeat sizing for linkage analysis in families. The prelude to
the prolific nature of SNP genotyping was highlighted by
the identification of the association between the tumor
necrosis factor ligand superfamily member 15 (TNFSF15)
gene at 9q32 and 94 CD patients compared to 752 healthy
controls in a Japanese study (Yamazaki et al. 2005). For fine
mapping and GWAS, the availability of a higher density of
SNPs has offered a higher resolution and has been better at
detecting association in a region of short-range linkage dis-
equilibrium (LD) (Kruglyak 1999). Furthermore, in popu-
lations of European descent, this locus has been confirmed
in IBD ( Jostins et  al. 2012). Thus, in contrast to NOD2
mutations and the IBD5 haplotype, a core TNFSF15 hap-
lotype associated with IBD is found in both Caucasian and
Japanese populations (see Figure 28.1).
G enetics and G enomics of C hronic I nflammatory D isorders , I   •   4 3 5
Directly genotyping variants with potential for func-
tion, such as non-synonymous genome-wide SNPs, has
successfully implicated an autophagy-related gene (see
Autophagy section further on), the autophagy 16–like gene
(ATG16L1) at chromosome 2q37 in a German population
case-control investigation (Hampe et al. 2007). Microarray
platforms that utilize genome-wide haplotype tagged SNPs
derived from the HapMap project have also efficiently led
to the identification of novel genetic variants that are associ-
ated with IBD. This was exemplified by the implication of
a subunit of the receptor for the proinflammatory cytokine
interleukin-23 (IL23R) gene at chromosome 1p31 with
non-coding variants that increased the risk of disease, and
a rare coding variant that protected from disease in North
American populations of European ancestry (Duerr et  al.
2006). The other novel genome-wide significant associated
regions from an extension to this investigation included
paired-like homeobox 2B (PHOX2B) 4p13, neutrophil
cytosolic factor 4 (NCF4) at 22q12, a predicted gene
(FAM92B) at 16q24, an intergenic locus at 10q21, as well
as additional evidence for ATG16L1 (Rioux et  al. 2007).
A similar approach in a Belgian investigation discovered a
novel association of CD with a gene desert region with no
recognized genes other than CpG islands on chromosome
5p13 (Libioulle et  al. 2007), and expression quantitative
trait loci (eQTL) analysis revealed the disease-associated
polymorphism that regulated the expression of the PTGER
gene. A  larger scale GWAS with high-density randomly
genotyped SNPs in the United Kingdom had indepen-
dently unravelled another novel autophagy-inducing gene,
the p47 immunity-related GTPase (IRGM) gene on chro-
mosome 5q33 (Parkes et  al. 2007; WTCCC 2007)  (also
see CNV and autophagy in the next section). This investi-
gation also revealed a novel association with gene deserts on
chromosome 1q. The other novel loci included 3p21, 5q33,
10q24, and 21q22. Also, the analysis of the GWAS results
among several diseases showed a novel association with the
T cell protein tyrosine phosphatase (PTPN2) gene locus at
chromosome 18p in both CD and type 1 diabetes, which
subsequently was replicated in both diseases (Parkes et al.
2007; Todd et al. 2007). There was a consistent pattern of
replication between the different GWAS, but the intersec-
tion between genome-wide linkage and association was only
visible at the 16q and 5q loci (see Figure 28.2). This cor-
relation also seemed to occur at 6p and 3p—these loci are
gene-dense and show higher levels of LD that make it dif-
ficult to identify the causative variants. However, the other
loci show no correlation between the previously under-
taken genome-wide linkage scans and the recently under-
taken GWAS. The meticulous genotyping of microsatellite
markers can yield equivalent information in comparison to
that of a higher density of SNP markers for linkage analysis,
thus their use might not have been a major reason for the
lack of correlation between genome-wide linkage and asso-
ciation studies. In retrospect, an over-optimistic estimation
of weak replication between genome-wide linkage scans to
reaffirm loci might be one of the contributing factors that
led to a lack of correlation between the more robust GWAS.
In addition, the regions of aggregation in linkage are
broader, and their boundaries are not precisely demarcated.
The first generation of commercially available SNP
microarray platforms had a similar overall coverage, but
the coverage was incomplete in different regions of the
genome (Barrett and Cardon 2006). Several different sys-
tematic meta-analyses and follow-up studies have combined
the results from the different GWAS platforms and new
cohorts, and refined the locations of overlap between the
IBD loci (Anderson et al. 2011; Barrett et al. 2008; Beckly
et  al. 2008; Cummings et  al. 2007b; Cummings et  al.
2007a; Franke et al. 2010; Jostins et al. 2012). The largest
meta-analysis of inflammatory bowel diseases has included a
total of over 75,000 cases and controls derived from 15 pre-
viously conducted GWAS, in addition to a follow-up with
a customized high-density fine-mapping array known as
the “ImmunoChip” ( Jostins et al. 2012). The high-density
markers for the ImmunoChip were selected from the associ-
ated regions of a number of different immune diseases. These
data presented substantial evidence for the sharing of bio-
logical pathways with other diseases, such as inflammatory
diseases, autoimmune diseases, mycobacterial infections,
and primary immunodeficiencies. The shared associations
with other immune-mediated diseases suggested that infec-
tious diseases are likely to be the strongest selection pres-
sures. The efficient genotyping of surrogate SNPs from
the “HapMap3” reference set enabled the prediction of
1.23 million genotypes, and the assigned genotypes within
IBD GWAS regions were subsequently validated by further
genotyping with the ImmunoChip. The two major pheno-
types of IBD—CD and UC—have 110 loci that are shared
between them. More specifically, 30 are associated with
Crohn’s disease; 23, with ulcerative colitis.
Other markers, such as trinucleotide repeats, might be
of direct relevance in IBD susceptibility, if the expansion
of their repeat size is associated with disease (Polito et al.
1996), similar to the “anticipation” that has previously been
described for Huntington’s disease (Vegter-van der Vlis et al.
1976). There is, however, no conclusive evidence regarding
the role of anticipation in IBD, since this observation could
merely be an earlier diagnosis in the later-generation mem-
bers of a family who are under clinical investigation, rather
4 3 6   •   G enomics in C linical P ractice
than genuinely earlier onset of disease than in the preceding
generation (Faybush et al. 2002).
Further insights of the contribution of copy number vari-
ants (CNV) in IBD have been gained from the identifica-
tion of a deletion within the IRGM gene after fine-mapping
a tagging SNP marker that was in perfect LD with a dele-
tion CNV (McCarroll et al. 2008). The deletion polymor-
phism was found to influence the expression of the IRGM
gene. Subsequently, a reevaluation of the genome for CNVs
suggested that, except for the CNV at the IRGM and HLA
genes, the other common CNVs are not likely to be associ-
ated with IBD (Wellcome Trust Case Control et al. 2010).
The IBD-associated SNP markers explain less than 30%
of the estimated inheritance, and further mapping inves-
tigations in the genetics of IBD are not likely to discover
many more than the 163 distinct loci that have already
been discovered ( Jostins et al. 2012). Therefore the “miss-
ing inheritance” might be due to a number of different
reasons. These include the possibility that an overestima-
tion of the genetic component from the twin studies was
due to previous methodological limitations (Halfvarson
2011). Undetected monogenic mutations may also have a
reduced penetrance for the complete syndrome (Casanova
and Abel 2009). The rare variants (minor allele frequency
[MAF] <0.5%) that have a modest influence (odds
ratios: 1.1 to 1.5) are not likely to capture the association
by case-control studies (Manolio et  al. 2009; Pritchard
2001). There is also the possibility that a single-marker
model that assumed complete LD could have missed some
of the distinct genetic contributions, such as another dis-
tinct gene association, CYLD, in addition to the NOD2 at
16q (Elding et al. 2011). Methylated regions of the genome
may also reveal the missing link between the influences of
nature and nurture (Cooke et al. 2012; Nimmo et al. 2012).
Moreover, epigenetics may mediate between the influences
of an altered composition of the microbiome and the host
genome ( Jenke and Zilbauer 2012). Another reason for the
“missing inheritance” from the GWAS studies could be that
gene–gene interactions are computationally challenging to
assess and require very large sample sizes (Zuk et al. 2012).
Pathway-based analysis of GWAS markers with a prior
biological understanding of gene function is likely to have
greater power to detect associations (Wang et al. 2010).
Although the high-resolution SNP markers can be rela-
tively abstract for mapping, they have been pivotal to paint-
ing the genomic portrait of polygenic IBD. However, as an
extension to the deep resequencing of the GWAS-associated
regions (Rivas et al. 2011), the whole genome resequencing
at base pair resolution could be an ultimate entity for map-
ping; and, although it may not be a complete blueprint for
IBD, it is likely to be a starting point for unravelling IBD’s
complexity.
T H E I D E N T I FI C AT I O N O F T H E NO D2
GENE
The initial genome-wide linkage analysis search for IBD
susceptibility made a great leap forward with the semi-auto-
mated genotyping of 270 fluorescently labelled dinucleotide-
repeat microsatellite markers in 78 multiple sibs affected
with CD (Hugot et al. 1996). This model-free approach
identified a novel significant linkage at chromosome 16q.
Subsequent to the completion of other genome searches, it
was found that this locus was much more consistently repli-
cated than were other regions that were reported to confer
susceptibility to IBD (Cho et al. 1998; Hampe et al. 1999b;
Williams et al. 2002). Furthermore, the locus-specific sib-
ling locus was estimated at 1.3 as predicted by the propor-
tion of expected allele-sharing by affected sib pairs identical
by descent (i.e., 25%) to the observed allele-sharing (19.2%)
(Hugot et al. 1996). From this locus value, it was calculated
(Risch 1987) that the 16q might contribute less than 20% of
the genetic influence in CD (Hugot et al. 2003). Linkage of
the disease to 16q has been confirmed by two meta-analyses
of published genome-wide screens (van Heel et al. 2004;
Williams et al. 2002). This seminal discovery is regarded
as the first novel, unequivocal identification of a suscep-
tibility gene by using nonparametric approaches (Hugot
et al. 2001). Equally remarkable were the results from the
positional candidate-gene approach that was undertaken
simultaneously by an independent research group (Ogura et
al. 2001), and subsequently replicated (Hampe et al. 2001).
The rationale for selecting NOD2 as a positional candidate
gene related to functional investigations that were emerging
for the role of this gene in innate immunity (Ogura et al.,
2001). Deep sequencing has also enabled the identification
of other relatively rare variants within the NOD2 genes that
are associated with CD (Rivas et al. 2011). The successful
genetic incrimination of NOD2 gene in CD has opened
new questions of how the mutations cause disease and their
implications for the epidemiology of disease. Other NOD-
like receptors could also be considered as candidates for
influencing CD (Cummings et al. 2010).
NO D2 A N D I N NAT E I M MU N IT Y
The NOD2 gene belongs to a family of cytosolic
pattern-recognition receptors (Inohara and Nunez 2003).
Its role in innate immunity is based on its ability to recog-
nize conserved structures within the gut flora (Inohara et al.
G enetics and G enomics of C hronic I nflammatory D isorders , I   •   4 3 7
2003). There are now more than 20 different NOD-like
human proteins, and they show varying degrees of homol-
ogy to plant cytosolic R proteins. An acronym put forward
to describe NOD-like human proteins is CATERPILLAR,
because leucine-rich repeats are found in the carboxy-termini
(CATERPILLAR:  CArd, Transcription Enhancer, R
(Purine)-binding, pyrin, and many Leucine repeats [Harton
et al. 2002]).
The NOD2 gene consists of a central nucleotide-binding
domain (NBD), a leucine-rich repeat, and two N-terminal
CARD domains. The three common variants of NOD2
(Figure 28.3) are located within the leucine-rich repeat; this
includes the Arg702Trp (tryptophan substituted for argi-
nine at codon 702), the Gly908Arg (arginine substituted
for glycine at codon 908), and Leu1007finsC (frameshift
mutation truncates the ending 3% of protein). Subsequently,
a number of relatively rare SNPs and many other private
mutations in the NOD2 gene have been found in patients
with CD (Rivas et al. 2011). Other novel, rare mutations in
the NBD were found to segregate with Blau syndrome: an
autosomal granulomatous disease (Kanazawa et  al. 2005).
These mutations in the NOD2 gene have been shown to
provide a gain of function with increased nuclear factor-kB
(NF-kB) activation in Blau syndrome (Miceli-Richard et al.
2001). However, not all granulomatous diseases are associ-
ated with NOD2 mutations, as they were not found in sar-
coidosis (Ho et al. 2005; Schurmann et al. 2003), although
association has been found in patients with early onset of
disease (Kanazawa et al. 2005).
NOD2 is an intracellular pattern-recognition protein
for bacterial detection (Chamaillard et  al. 2003b). The
recognition of the peptidoglycan component of bacterial
cell walls depends on the detection of muramyl dipep-
tide (MDP), a minimal motif that is an almost-universal
constituent of peptoglycans found in gram-negative and
gram-positive bacteria (Girardin et al. 2003; Inohara et al.
2003). Specifically, it is the leucine-rich repeat region of
NOD2 that has a role in binding to MDP, and consequently
activating NF-κB, through a number of other intracellular
molecules (Chamaillard et al. 2003b; Girardin et al. 2003;
Inohara and Nunez 2003). Paradoxically, in vitro experi-
ments have shown that common variants of NOD2 asso-
ciated with CD decrease NF-κB activation (Bonen et  al.
2003; Chamaillard et al. 2003a). It was hypothesized that
this contradiction between in vitro and in vivo findings
might be explained by NOD2-independent bacterial activa-
tion of NF-κB mediated by Toll-like receptors on macro-
phages (Inohara et al. 2002). Experimental results from in
vivo experiments have been consistent with the hypothesis
(Watanabe et al. 2004). Further insights have come from the
transfection of epithelial cell lines with the CD-associated
variants of NOD2. The transfected cells failed to kill intra-
cellular Salmonella typhimurium compared to cells trans-
fected with wild-type NOD2 (Hisamatsu et  al. 2003).
A  very novel mechanism whereby NOD2 mutations may
put individuals at risk of disease has recently been pro-
posed. The expression of NOD2 has been shown in human
Paneth cells (Lala et al. 2003). Subsequently, CD patients
with NOD2 mutations have been shown to have a reduced
release of the antimicrobial peptides, α-defensins 4 and 5,
from Paneth cells (Wehkamp et al. 2004; Wehkamp et al.
2005b; Wehkamp et  al. 2005a; Wehkamp et  al. 2005c;
Wehkamp and Stange 2005; Wehkamp et al. 2005d). Since
Paneth cells are predominantly found in the terminal ileum,
deficiency of antimicrobial peptides in the intestinal lumen
may allow bacterial-induced inflammation to occur.
Nod2 knockout murine models have also been used
to address the inconsistency of MDP activation between
in vitro and in vivo (Kobayashi et  al. 2005). It was found
that the detection of bacterial MDP was abolished in the
absence of Nod2 in mice. Furthermore, the Nod2-deficient
mice were more susceptible than wild-type mice to infection
with bacteria via the oral route, but not with intravenous
(IV) or intraperitoneal (IP) routes of infection. However,
the Nod2-deficient mice showed no evidence of intestinal
inflammation. These mice also showed a reduced release of
cryptdins from intestinal Paneth cells—cryptdins being the
murine equivalent of human defensins.
In contrast to the loss of function in cell lines trans-
fected with NOD2 variants or the NOD2 knockout mod-
els, knock-in of the NOD2 frameshift mutation has shown
a much more efficient induction of cytokine interleukin-1β
and NF-κB in response to MDP (Maeda et  al. 2005).
Perhaps the presence of an alternative pathway for positive
regulation of NF-κB activation in response to MDP could
offer a possible reconciliation between elevated NF-κB acti-
vation in the experimental models using NOD2 variants.
E P I D E M I O L O GY O F NO D2 I N C D
The functional insights have been further supported by the
compelling weight of replication of epidemiological disease
association with populations of European descent (Abreu
et al. 2002; Ahmad et al. 2002b; Cuthbert et al. 2002; Hampe
et  al. 2001; Jostins et  al. 2012; Ogura et  al. 2001). These
investigations have indicated that 10–30% of CD patients
are heterozygotes for one of the three common mutations
(see Figures 28.1 and 28.3). Another 3–15% of patients are
either homozygotes or compound heterozygotes. Although
this relative risk of 20–40 in homozygotes (with a 1/25
risk of developing disease) is likely to be detected within a
population case-control investigation, it would have been
unlikely to have influenced the high-linkage score that was
obtained from the genome search, because homozygotes
tend to reduce the LOD (log of the odds) score (Brant
and Shugart 2004). Furthermore, families unaffected by
disease, yet possessing the high-risk alleles, have also been
reported (Ahmad et al. 2002b; Linde et al. 2003; Radlmayr
et  al. 2002). Moreover, within different populations of
European descent, there is a variable association of these
alleles with disease (see Figure 30(II)–1), even though the
control allele frequencies within these populations are simi-
lar (Economou et  al. 2004). These differences in attribut-
able risk of the NOD2 disease-susceptibility variants could

Apoptosis and NF-kβ

activation Oligomerization Bacterial recognition

Arg703Cys Met863Val
Arg311Trp Ser431Leu Val793Met Asn852Ser
28 124 127 220 273 577 744 1020 1040

Nuclear
Leucine rich
CARD1 CARD2 binding
repeat region
domain

N terminal C terminal

Leu469Phe Gly908Arg
Arg702Trp
Arg334Glu
Arg334Trp
Leu1007finsC
Mutations in vitro show an in Mutations in vitro show a
increase in NF-kβ activation decrease in NF-kβ

Figure 28.3 
The three NOD2 domains include the following: caspase recruitment, nuclear binding, and leucine-rich domain. The three relatively
common CD-associated variants display a loss of function, but the rare mutations that are associated with Blau syndrome reveal a gain in function
from in vitro studies. Other relatively rare CD-associated variants are also shown.

4 3 8   •   G enomics in C linical P ractice

in some cases correlate with the apparent North–South
gradient (Vind et  al. 2005). There is a reduced associa-
tion of these susceptibility alleles with CD in Scotland
(Arnott et  al. 2004; Russell et  al. 2005), Ireland (Bairead
et  al. 2003), and Finland (Helio et  al. 2003). Moreover,
in northern Europe, the higher incidence and prevalence
of disease is likely to be due to other genetic and environ-
mental influences (Halfvarson et al. 2005). There is a nota-
bly reduced allele frequency of the disease-susceptibility
alleles within Afro-American communities (Kugathasan
et  al. 2005). There is also an absence, or an exceedingly
rare occurrence, of NOD2 disease-susceptibility alleles
within a number of Asian populations, such as the Koreans
(Croucher et  al. 2003; Lee et  al. 2005), Japanese (Inoue
et al. 2002; Yamazaki et al. 2002), and Chinese (Guo et al.
2004; Leong et al. 2003). In other communities, such as the
Ashkenazi Jews, the much higher incidence and prevalence
of CD could be irrespective of geographic and environmen-
tal influences in comparison to other groups (Zhou et  al.
2002). In general, however, these differences of disease inci-
dence and prevalence are usually thought to be due in part
to migration and changes in the environment (Lanzarotto
et al. 2005). Although the much higher Gly908Arg allele
frequencies within the Ashkenazi appear to correlate with
disease (Bonen et al. 2003), this particular genetic influence
does not alone explain the much higher prevalence of CD
within this community.
NO D2 MU TAT I O NS A N D P H E N OT Y P E
The association between the three common variants of
NOD2 has been consistently replicated with CD ( Jostins
et  al. 2012), but only a weak protective association with
UC was noted ( Jostins et al. 2012). Within CD (Ahmad
et al. 2002b; Cleynen et al. 2012), it was reported that the
association was particularly strong for ileal CD. This has
now been widely replicated and confirmed by meta-analysis
within populations of European descent (Economou et al.
2004). How NOD2 variants influence the anatomical site
of disease is unclear, but the possibility that Paneth cells are
involved, as discussed in the preceding section, is intrigu-
ing. Paneth cells are predominantly located in the distal
ileum, and it is possible that the lack of defensin production
in individuals with NOD2 mutations may impair ileal anti-
microbial defense mechanisms. NOD2 variants have also
been reported to predispose to early onset of disease and
to stricturing disease. However, these associations are not
completely established, although they received some sup-
port from the meta-analysis (Economou et al. 2004).
G enetics and G enomics of C hronic I nflammatory D isorders , I   •   4 3 9
T H E A N C E S TO R’S TA L E O F MU TAT I O NS
T H AT P R E D I S P O S E TO I B D
The NOD2 mutations that were associated with CD were
likely to postdate the “out-of-Africa migration,” since this
has so far not been detected in populations outside of
those of European ancestry (Inoue et al. 2002), and there
is reduced allele frequency within the African-American
communities (Kugathasan et al. 2005). The proline-to-ser-
ine amino acid substitution at position 268 of the NOD2
gene has also not been detected in various global popula-
tions other than in those of European descent (Inoue et al.
2002). The substitution at position 268 has, however, been
found in LD with all of the three common variants that
predispose to CD in European populations (Lesage et al.
2002). Although the evolutionary significance of this asso-
ciation is not understood (Cho 2001), the absence of the
variant allele at position 268 in various global communities
other than in those of European descent suggests that it is
not likely to precede the out-of-Africa migration. In addi-
tion, the LD between the variant allele at position 268 and
the three other known disease susceptibility alleles could
have occurred by chance alone, since there is limited hap-
lotype diversity within such a narrow region. Other, rarer
mutations have also been identified within the NOD2 gene
(Rivas et al. 2011). Each of the different mutations may
have a tale of its own about the protection from infectious
diseases. The three most commonly CD-associated NOD2
mutations were estimated to have arisen about 40,000 years
ago (Croucher et al. 2003; Schreiber et al. 2005), and the
very rare NOD2 mutations would have arisen more recently
(Rivas et al. 2011). The age of the mutation and geographi-
cal distribution of allele frequencies may yield clues to
the survival advantage of this mutation. Mycobacteria are
likely to have been a strong selective pressure for the IBD-
associated SNPs ( Jostins et al. 2012).
Since the CARD9 gene has been associated with fun-
gal infection in exceedingly rare immunodeficiency states
(Glocker et al. 2009a; Pathan et al. 2013), the relatively
common fungal infectious diseases might have also been
a selective pressure for the polymorphisms within the
CARD9 gene that were associated with IBD. Dectin-1
is also in the signaling pathway upstream from the adap-
tor CARD9 (Hsu et al. 2007), and the Dectin-1 gene was
associated with colitis patients’ requiring surgery (Iliev
et al. 2012). Furthermore, although previous research in
mice models has indicated that an imbalance of commensal
flora within gastrointestinal tract influences colitis, fungi
may also contribute to colitis (Iliev et al. 2012). The com-
plex haplotype structures of the major histocompatibility
complex (MHC) are also likely to have been under strong
selection pressures from infectious diseases (Cagliani et al.
2011; Jostins et al. 2012).
M A J O R H I S TO C O M PAT I B I L IT Y
C O M P L E X (6P21)
At human chromosome 6, the genetic region of the MHC
is known to confer crucial immunological function and
is linked with susceptibility to IBD (Ahmad et al. 2006a;
Ahmad et al. 2006b; Yap et al. 2004) and many other auto-
immune diseases. This locus has been the most investigated
over a period of 30 years, and was more recently researched
at a higher resolution with the tools of molecular genet-
ics. However, a good number of earlier investigations had
yielded inconsistent results (Stokkers et  al. 1999). Unlike
most other suspected autoimmune diseases, IBD has not
shown the same level of consistent linkage to the MHC
from independent genome-wide searches. In fact, it was
only when a denser set of markers was applied in the MHC
locus that it became possible to detect linkage (Hampe
et al. 1999a). The most consistent data within the human
leukocyte antigen (HLA) class  II region have concerned
HLA-DR2 and HLA-DRB1 0103. The DRB1 1502 allele
of DR2 has been found predominantly in Japanese popula-
tions of UC patients (OR 3.74, CI 2.20–6.38), but it has
also been reported from North America and the United
Kingdom (Ahmad et al. 2003a; Futami et al. 1995; Stokkers
et al. 1999; Trachtenberg et al. 2000; Yoshitake et al. 1999).
The higher frequency of this allele in the Japanese general
population, compared with its low frequency in Caucasians,
probably explains the more consistent data from Japan. The
association between UC and HLA-DRB1 0103 has been
convincingly replicated in several large studies, which have
also shown that patients possessing this allele are at risk of
having severe colitis with a risk of colectomy (Ahmad et al.
2003a; Bouma et al. 1997; Roussomoustakaki et al. 1997).
It has been shown that this allele also influences the natu-
ral history of CD of the colon; that is, it is associated with
severe disease and risk of colectomy early in the course
of disease (Hancock et  al. 2006; Silverberg et  al. 2003).
Whether this association with DRB1 0103 is causative or
whether this allele is in tight LD with a more relevant allele
is uncertain. However, DRB1 0103 appears to be present
on only two small haplotypes, and thus a genuine associa-
tion seems likely. The meta-analysis in 1999 also showed
that there was a negative association between HLA-DR4
and UC (OR 0.54, CI 0.43–0.68). Subsequently, this
protective effect was shown to be due to the DRB1 0401
molecular subtype, but only when it is present on the
4 4 0   •   G enomics in C linical P ractice
DRB1 0401-DQB1 0301 haplotype (Ahmad et al. 2003a;
Ahmad et al. 2003b). For CD, meta-analysis has confirmed
significant positive associations with DR7 (OR1.42, CI
1.16–1.74), DRB3 0301 (OR 2.18, CI 1.25–3.80), and
DQ4 (OR 1.88, CI 1.16–3.05), but, in contrast to UC,
a negative association with DR2. The associations with
DR7 and DR2 have been confirmed in a subsequent study
(Akolkar et al. 2001). Adjacent to the boundary between
the HLA class I and class III regions, there are members of
the non-classical MHC class-I-related chain (MIC) gene
family. MICA and MICB are polymorphic and are the only
two members of the family to encode functional transcripts.
They are predominantly expressed on the basal-lateral sur-
face of epithelial cells and interact with NKG2D and a vari-
ety of natural killer (NK) cells, T cells, and macrophages,
but particularly the CD8α T cells and the γT cells found
in the intraepithelial compartment. So far, no consistent
associations between UC or CD and polymorphisms in the
MICA or MICB genes have been reported, and the most
comprehensive study was negative (Ahmad et  al. 2002a).
Between the MHC class I and class II region is the highly
gene-dense class III region, spanning approximately 900 kb.
The genes for TNFa, lymphotoxin a, and heat-shock proteins
are found within class III, and polymorphisms within these
genes have been associated with IBD. For TNFa, a number
of promoter polymorphisms have been studied (at posi-
tions 1031, 308, and 857). In the UK, TNF-857C was asso-
ciated with CD patients not possessing NOD2 mutations
(van Heel et  al. 2002), but, in Australia, TNF-857C was
present only in patients with NOD2 variants (O’Callaghan
et  al. 2003). Published data on the other promoter poly-
morphisms have been equally inconsistent (Ahmad et  al.
2006a; Ahmad et al. 2006b). The common promoter haplo-
type (TNF 1031T, 863C, 857C, 308G, 380G, and 238G)
has been shown to be associated with distal UC, which is
stable over time and, at least for a minimum of a 10-year
follow-up, does not extend (Ahmad et  al. 2003a). The
mechanisms are unknown, but patients homozygous for
this haplotype might well have impaired TNF production,
which could influence disease activity. Polymorphisms in
the lymphotoxin a gene and within a number of heat-shock
protein genes are not consistently associated with either
UC or CD (reviewed in Yap et  al. 2004). Genes within
the HLA region on Chr 6p may also determine whether
patients with either UC or CD are at risk of developing
extraintestinal manifestations. Thus, the arthropathies,
uveitis, erythema nodosum, recurrent mouth ulcers, and
primary sclerosing cholangitis have been reported in asso-
ciation with class I and class II alleles (Ahmad et al. 2006b).
For example, the reactive large-joint arthropathy seen in
some patients with active colitis is strongly associated with
HLA-B 27, HLA-B 35, and HLA DRB1 0103, whereas
the symmetrical small-joint arthropathy is associated with
HLA-B 44 (Orchard et al. 2000). Uveitis has been associ-
ated with HLA-B27 and DRB1 0103, and erythema nodo-
sum with TNF- 1031C. The numbers are small, but it seems
likely that the phenotypical heterogeneity seen in the clinic
may partly be explained by genetic polymorphisms within
the HLA region.
Association with the HLA locus has now been confirmed
by the customized ImmunoChip and the meta-analysis
( Jostins et al. 2012). In addition to the HLA genes at the
Chr 6p locus, a protective association with a synonymous
SNP Q350Q within the BTNL2 gene has been consistently
replicated with UC (Pathan et  al. 2009). It is plausible
that a synonymous SNP can influence disease (Sauna and
Kimchi-Sarfaty 2011). A missense K196E BTNL2 variant
showed novel protective association in UC cases requiring
colectomy, but the same allele was weakly predisposing for
the ileal sub-phenotype in CD (Pathan et  al. 2009). The
sub-phenotypical studies require further replication.
T H E C AUS AT I V E VA R I A N T S A N D
F U N C T I O NA L I M P L I C AT I O NS
Of the total 163 IBD-associated variants there were only
29 missense variants; most of the other disease-associated
SNPs were within the non-coding regions, and these
include 64 SNPs that were correlated with gene expression
( Jostins et  al. 2012). In fact, even when deep sequencing
was undertaken within the regions of the GWAS-associated
genes, only a limited number of coding variants were iden-
tified, such as a CARD9 splice variant (Rivas et al. 2011).
However, even this relatively rare protective splice variant
within the CARD9 gene was distinct from the common
CARD9 predisposing variant that was previously reported
within a CD GWAS meta-analysis (Franke et  al. 2010).
The other notable coding variants that were identified from
resequencing the IBD-associated GWAS regions include
IL18RAP, CUL2, C1orf106, PTPN22, and MUC19 (Rivas
et al. 2011).
The systematic interrogation of non-coding
IBD-associated regions reveals the tendency to cluster
within the regulatory DNA marked by deoxyribonuclease
I  (DNase I) hypersensitive sites (DHSs) (Maurano et  al.
2012). About 88% of such DHSs have a role in fetal devel-
opment. DHSs are likely to correlate with IBD associations
that obstruct transcription factor recognition sequences,
frequently alter allelic chromatin states, and lead to the
deregulation of networks in IBD.
G enetics and G enomics of C hronic I nflammatory D isorders , I   •   4 4 1
The genetic studies alone are not likely to define the
causal variant when there is complete LD between a num-
ber of the plausible causative coding alleles, such as that
between the associated alleles spanning one of the hap-
lotypes between the ORMDL3 and GDSBM genes at
17q21 (Barrett et  al. 2008). The expression studies from
Epstein-Barr virus–transformed lymphoblastoid cells from
patients with CD and asthma have, however, presented
complementary evidence, by showing that the coding SNPs
with the strongest association were consistently associated
with levels of ORMDL3 transcripts rather than adjacent
genes (Barrett et al. 2008; Moffatt et al. 2007). Further func-
tional insights may also be possible from protein studies for
the ORMDL3 gene, which is a member of a conserved fam-
ily of endoplasmic reticulum membrane proteins.
The macrophage-stimulating protein (MST1) modu-
lates innate immunity and adaptive responses, and the cod-
ing variant R689C of the MST1 gene was one of the likely
causative variants within a region of high LD at 3p that was
associated with IBD (Goyette et al. 2008). In addition, at
3p there has been some evidence for another distinct associ-
ation within the receptor of biological MST1 gene (Beckly
et al. 2008; Pathan et al. 2010). Since this region of associa-
tion intersects with a region of linkage at 3p, it is likely that
there are also several other rare SNPs that could influence
the IBD (Satsangi et al. 1996b).
AU TO P H AGY
Functional links have been established between innate
immunity and autophagy (Cooney et  al. 2010; Travassos
et  al. 2010)—autophagy is more literally known as
“self-eating.” The microbial loads within infected cells are
degraded within the lysosomes by the process of autophagy.
This process recycles proteins and organelles for optimal
cellular balance between synthesis and reducing the pro-
teins into constituent components. Autophagy requires
the formation of double-membrane cytosolic vesicles called
autophagosomes that sequester cytoplasmic contents and
transport them to the lysosome for subsequent degradation.
The dynamic process of membrane expansion enables the
disposal of microbes of any size.
The autophagy genes that are associated with CD, such
as ATG16L1, ATG5, IRGM, and leucine-rich repeat kinase
2 (LRRK2) are expressed in the small intestinal Paneth
cells. NOD2 induces autophagy by directing ATG16L1
to the cell membrane within the vicinity of bacterial entry
(Cooney et al. 2010). The other autophagy genes, such as
RIPK2, ATG5, and ATG7, are also involved. In addition to
autophagy degrading the damaged organelles and proteins,
it also connects with the process of antigen presentation.
The bacterial handling and MHC class II antigen presenta-
tion occurs in human dendritic cells (DCs). The DCs of CD
patients expressing CD risk variant NOD2 or ATG16L1
have impaired autophagy induction and reduced bacterial
killing and defective antigen presentation (Cooney et  al.
2010). This correlates with the major allele in cases having
an impaired ability to capture bacteria.
NOD2 also has a role in the autophagic response to
bacteria engulfed by phagosomes (Travassos et  al. 2010).
The Crohn’s disease (CD) associated with the frame-shift
NOD2 mutation as well as the ATG16L1 variant results in
defective bacterial breakdown in lysosomes (Cooney et al.
2010). Furthermore, NOD2 and DCs can modulate Th17
responses (Brain et al. 2012). The IL-23 signaling pathway
stimulates antimicrobial peptide cryptidins in the intestinal
cells, which would otherwise result in the impairment of
bacterial destruction. Knockout mice for ATG16L1 that
are infected with the intestinal pathogen norovirus also
develop defects in Paneth cells (Cadwell et al. 2008).
Colonic epithelial IRGM gene over-expression was
observed in CD patients with a homozygous protective
IRGM synonymous SNP allele, but not in patients with the
risk allele (Brest et al. 2011). In addition, the microRNA-196
(miR-196) was found to have an impaired binding to
the allele that predisposes to CD. Therefore, either an
over-expression (Brest et al. 2011) or an under-expression
of the IRGM gene (McCarroll et al. 2008) could reflect an
impaired autophagy.
A DA P T I VE I M MU N E SYS T E M
The stimulation of adaptive immunity by the innate
immune system results in a specific response with a greater
effectiveness against infectious diseases than that of the
innate immune system. Adaptive and innate immunity are,
however, connected to each other in the most intimate way.
In fact, adaptive immunity depends on innate immunity.
NOD2, IRGM, ATG16L1, and IL-23R are examples of
genes that regulate innate and adaptive immune responses.
More specifically, for example, the activation of NOD2
signaling serves to block the expression of IL-12, and indi-
rectly IL-23, by the upregulation of a microRNA, miR-29
(Brain et  al. 2013). However, when there was a NOD2
frameshift mutation, the miR-29 was not upregulated and
this may explain how a mutated NOD2 gene can lead to
inflammation with elevated IL-23.
The IL-23R gene has one of the strongest associations
with CD ( Jostins et  al. 2012). Furthermore, other genes
within the IL23/Th17 signaling pathway, such as IL12B,
4 4 2   •   G enomics in C linical P ractice
JAK2, TYK2, and STAT3, were also associated with the
disease. The results from the mouse model, which had
also been studied extensively prior to the identification of
genetic association, explained the biology underlying the
genetic associations.
There are several other interconnected genes associated
with CD, such as the IL10, IL12B, and STAT3 (Glocker
et al. 2011). This pathway contributes to the clonal selec-
tion of lymphocytes and sets the foundation for adaptive
immunity with the expression of highly specific receptors
on B and T cells.
MU C O S A L BA R R I E R F U N C T I O N
For the UC-associated 20q locus, within populations of
both Europeans (Consortium et  al. 2009; Jostins et  al.
2012) and Japanese (Asano et al. 2009), an impaired barrier
function due to a defective HNF4A gene product is plausi-
ble. Mice models have also suggested that Hnf4a deficiency
could enhance epithelial permeability, and an unbalanced
Hnf4a gene expression may correlate with UC (Ahn et al.
2008). The other mucosal barrier genes that were specifi-
cally associated with UC include ECM1 (Anderson et al.
2011; Fisher et  al. 2008); LAMB1, CDH1, and GNA12
( Jostins et al. 2012).
T H E PA R A L L E L S A N D PA R A D OX E S
WIT H OT H E R D I S E A S E S
There is a substantial sharing of about 67% of the total 163
distinct loci associated with the phenotypes of both CD
and UC, and about 69% of the distinct IBD associations
transcend the boundaries between many other complex
diseases that have a very different clinical disease manifesta-
tion ( Jostins et al. 2012; Lees et al. 2011; Zhernakova et al.
2009). In certain cases, there are emerging patterns of sym-
metry between different diseases and the shared loci. For
example, the R620W PTPN22 variant predisposes to type
1 diabetes, systemic lupus erythematosus (SLE), and rheu-
matoid arthritis (RA); but it is protective for CD ( Jostins
et al. 2012; WTCCC 2007) and also weakly protective in
UC ( Jostins et al. 2012).
Intriguingly, the HNF4A gene locus is associated with
both maturity diabetes of the young (MODY) (Yamagata
et  al. 1996), the metabolic syndrome of type 2 diabetes
(T2D; Kathiresan et  al. 2009), and UC. Both MODY
and type 2 have some clinical similarities. However, the
HNF4A locus is not associated with type 1 diabetes (T1D).
The HNF4A gene is a plausible pleotropic gene. The other
T2D and metabolic syndrome genes that may have parallels
with IBD include CDKAL1, CAPN10, GCKR, THADA,
CPEB4, BTLN2, FADS1, and SMAD3 (Lees et al. 2011).
IBD is not likely to be a quintessential autoimmune
disease like systemic lupus erythematosus (SLE), anky-
losing spondylitis (AS), or rheumatoid arthritis (RA),
yet there is a very high degree of shared features with the
classical autoimmune diseases, such as the IL23R sig-
naling pathway in and AS, SLE, and RA. The different
autoimmune diseases may also cluster in affected patients
(Snook et al. 1989). Intriguingly, in Manitoba, Canada,
there is a very high incidence of IBD in the population
of European descent, but low in First Nations. The dif-
ference in the frequencies for the alleles that are involved
in bacteria-processing may be one of the contributing
factors for the protection from IBD in the First Nations
(Murdoch et al. 2012).
The sharing of loci between IBD and infectious diseases
are revelatory from the perspective of predicting the selec-
tion advantage of IBD variants. For example, a Chinese
GWAS of leprosy (Mycobacterium leprae) indicated that
the MHC genes were most significantly associated with
leprosy. The NOD2 variants in infectious diseases in Asian
populations were distinct from the variants associated with
CD in patients of European ancestry.
There are also shared loci between primary immuno-
deficiencies and IBD, such as the CARD9 gene (Pathan
et al. 2013), which is paradoxical, given that an inflam-
matory disease is an over-response rather than an under-
response, as is the case for an immunodeficiency. There
has also been a report of a monogenic form of immunode-
ficiency and IBD being associated with a splice-site muta-
tion in the NEMO gene on chromosome X (Orstavik
et al. 2006; Puel et al. 2006). Mutations in the NEMO
gene result in the variable phenotype of incontinentia
pigmenti (IP), an X-linked dominant disorder, predom-
inantly seen in females, as it is lethal in males. Turner’s
syndrome, a recognizable multiple-anomaly syndrome
due to a number of constitutional X chromosome abnor-
malities, was also associated with IBD. More than 20
reported cases have shown this association (Hayward et
al. 1996). The reported frequency of IBD has been found
to be much higher in patients affected with Turner’s syn-
drome than in the general population. X-linked chronic
granulomatous disease is also another monogenic immu-
nodeficiency in which patients frequently manifest a
granulomatous phenotype and Crohn’s-like disease
(Glocker and Grimbacher 2012).
For polygenic granulomatous diseases, a predispos-
ing association may be shared between complex BTNL2/
HLA haplotypes, sarcoidosis (Valentonyte et al. 2005), and
G enetics and G enomics of C hronic I nflammatory D isorders , I   •   4 4 3
primary biliary cirrhosis (PBC) (Hirschfield et  al. 2009),
and there may be parallels with ileal CD (Pathan et  al.
2009). In UC, which is a non-granulomatous disease, these
alleles were protective (Pathan et al. 2009).
C L I N I C A L I M P L I C AT I O N S
A N D T R A N S L AT I O N
The genomics revolution is inspiring new directions in
clinical translation. This includes the possibility of predict-
ing disease severity in newly diagnosed patients with the
transcriptional signature as a biomarker (Lee et al. 2011).
Furthermore, innovative approaches in proteomics may
enhance the phenotyping of patients according to their pat-
tern of protein expression (Roda et al. 2010).
Although these modestly influencing loci are not ade-
quate to predict disease risk in unaffected individuals, the
identification of genes involved in autophagy and in innate
as well as adaptive immunity may provide greater insight
into disease pathogenesis. The therapeutic intervention to
activate autophagy in either Paneth cells or macrophages
may decrease inflammation (Klionsky 2009). The detection
of differences in circulating microRNAs (miRNAs) in the
serum of IBD patients may also facilitate an earlier diagno-
sis (Coskun et al. 2012). The miRNAs may be able to exert
an influence in about one-third of the genes in the human
genome.
The discovery of the exceedingly rare IL-10 and IL-10
receptor polymorphisms that cause a severe phenotype have
been followed up with novel treatments that include hema-
topoietic stem cell transplantation (HSCT). Allogeneic
HSCT that restores IL-10 signaling in hematopoietic cells
seems to be a very promising new approach (Glocker et al.
2011).
The early detection of tumor cells and the monitoring of
circulating tumor cells (CTCs) in IBD patients with a pre-
disposition to colorectal cancer are likely to develop with
the emerging single-cell genomics technology (Navin and
Hicks 2011).
C O N C LU S I O N
Considerable progress has been made in understanding
the role of genetics in both CD and UC. This has been
achieved by assembling large cohorts of patients who have
been meticulously documented in terms of clinical pheno-
type and by the application of new technologies developed
for studying polygenic disorders. To date, the genome-wide
association studies have unearthed many unsuspected
associated genes and their regulatory regions. The genetic
knowledge so far has highlighted the important role of the
host’s innate immunity and its interaction with commensal
bacteria, including autophagy, and the intimate integration
between adaptive immunity, epithelial barrier function, and
IL-10 signaling. Further functional insights have also been
integrated from the expression microarrays and proteomic
technology. Genotyping and linkage analysis can still be
applied as a complementary approach to next-generation
exome-sequencing for the identification of monogenic
mutations. However, for the very modestly influencing
loci, linkage analysis is mostly of historical interest, and it
has been succeeded by high-resolution association studies.
There is optimism that the post-genomics revolution will be
translated into personalized therapy.
REFERENCES
Abreu, M.  T., et  al. (2002), “Mutations in NOD2 are associated
with fibrostenosing disease in patients with Crohn’s disease,”
Gastroenterology, 123 (3), 679–688.
Ahmad, T., Marshall, S.  E., and Jewell, D. (2006a), “Genetics of
IBD: Topic highlights for the role of the HLA complex,” World J
Gastroenterol, 12 (23), 3628–3635.
Ahmad, T., Marshall, S. E., and Jewell, D. (2006b), “Genetics of inflam-
matory bowel disease:  the role of the HLA complex,” World J
Gastroenterol, 12 (23), 3628–3635.
Ahmad, T., et al. (2004), “Clinical relevance of advances in genetics and
pharmacogenetics of IBD,” Gastroenterology, 126 (6), 1533–1549.
Ahmad, T., et al. (2003a), “The contribution of human leucocyte anti-
gen complex genes to disease phenotype in ulcerative colitis,” Tissue
Antigens, 62 (6), 527–535.
Ahmad, T., et al. (2002a), “High resolution MIC genotyping: design and
application to the investigation of inflammatory bowel disease sus-
ceptibility,” Tissue Antigens, 60 (2), 164–179.
Ahmad, T., et al. (2002b), “The molecular classification of the clinical man-
ifestations of Crohn’s disease,” Gastroenterology, 122 (4), 854–866.
Ahmad, T., et  al. (2003b), “Haplotype-specific linkage disequilibrium
patterns define the genetic topography of the human MHC,” Hum
Mol Genet, 12 (6), 647–656.
Ahn, S. H., et al. (2008), “Hepatocyte nuclear factor 4alpha in the intes-
tinal epithelial cells protects against inflammatory bowel disease,”
Inflamm Bowel Dis, 14 (7), 908–920.
Akolkar, P.  N., et  al. (2001), “The IBD1 locus for susceptibility to
Crohn’s disease has a greater impact in Ashkenazi Jews with early
onset disease,” Am J Gastroenterol, 96 (4), 1127–1132.
Altmuller, J., et  al. (2001), “Genomewide scans of complex human dis-
eases: true linkage is hard to find,” Am J Hum Genet, 69 (5), 936–950.
Anderson, C.  A., et  al. (2011), “Meta-analysis identifies 29 additional
ulcerative colitis risk loci, increasing the number of confirmed asso-
ciations to 47,” Nat Genet, 43 (3), 246–252.
Arnott, I.  D., et  al. (2004), “NOD2/CARD15, TLR4 and CD14
mutations in Scottish and Irish Crohn’s disease patients:  evidence
for genetic heterogeneity within Europe?” Genes Immun, 5 (5),
417–425.
Arthur, J.  C., et  al. (2012), “Intestinal inflammation targets
cancer-inducing activity of the microbiota,” Science, 338 (6103),
120–123.
4 4 4   •   G enomics in C linical P ractice
Asano, K., et  al. (2009), “A genome-wide association study identifies
three new susceptibility loci for ulcerative colitis in the Japanese
population,” Nat Genet, 41 (12), 1325–1329.
Bairead, E., et al. (2003), “Association of NOD2 with Crohn’s disease in a
homogenous Irish population,” Eur J Hum Genet, 11 (3), 237–244.
Barrett, J.  C. and Cardon, L.  R. (2006), “Evaluating coverage of
genome-wide association studies,” Nat Genet, 38 (6), 659–62.
Barrett, J. C., et al. (2008), “Genome-wide association defines more than
30 distinct susceptibility loci for Crohn’s disease,” Nat Genet, 40 (8),
955–962.
Bayless, T. M., et al. (1996), “Crohn’s disease: concordance for site and
clinical type in affected family members—potential hereditary
influences,” Gastroenterology, 111 (3), 573–579.
Beckly, J. B., et al. (2008), “Two-stage candidate gene study of chromo-
some 3p demonstrates an association between nonsynonymous vari-
ants in the MST1R gene and Crohn’s disease,” Inflamm Bowel Dis,
14 (4), 500–507.
Binder, V. (2004), “Epidemiology of IBD during the twentieth cen-
tury: an integrated view,” Best Pract Res Clin Gastroenterol, 18 (3),
463–479.
Blumberg, R.  S., Saubermann, L.  J., and Strober, W. (1999), “Animal
models of mucosal inflammation and their relation to human
inflammatory bowel disease,” Curr Opin Immunol, 11 (6), 648–656.
Bonen, D. K., et al. (2003), “Crohn’s disease-associated NOD2 variants
share a signaling defect in response to lipopolysaccharide and pepti-
doglycan,” Gastroenterology, 124 (1), 140–146.
Bouma, G., et al. (1997), “Evidence for genetic heterogeneity in inflam-
matory bowel disease (IBD):  HLA genes in the predisposition to
suffer from ulcerative colitis (UC) and Crohn’s disease (CD),” Clin
Exp Immunol, 109 (1), 175–179.
Brain, O., et  al. (2012), “Functional consequences of mutations in the
autophagy genes in the pathogenesis of Crohn’s disease,” Inflamm
Bowel Dis, 18 (4), 778–881.
Brain, O., et  al. (2013), “The intracellular sensor NOD2 induces
microRNA-29 expression in human dendritic cells to limit IL-23
release,” Immunity, 39(3), 521–536.
Brant, S. R. and Shugart, Y. Y. (2004), “Inflammatory bowel disease gene
hunting by linkage analysis: rationale, methodology, and present sta-
tus of the field,” Inflamm Bowel Dis, 10 (3), 300–311.
Brest, P., et al. (2011), “A synonymous variant in IRGM alters a bind-
ing site for miR-196 and causes deregulation of IRGM-dependent
xenophagy in Crohn’s disease,” Nat Genet, 43 (3), 242–245.
Cadwell, K., et al. (2008), “A key role for autophagy and the autophagy
gene Atg16l1 in mouse and human intestinal Paneth cells,” Nature,
456 (7219), 259–263.
Cagliani, R., et  al. (2011), “Balancing selection is common in the
extended MHC region but most alleles with opposite risk profile
for autoimmune diseases are neutrally evolving,” BMC Evol Biol, 11,
171.
Casanova, J.  L. and Abel, L. (2009), “Revisiting Crohn’s disease as a
primary immunodeficiency of macrophages,” J Exp Med, 206 (9),
1839–1843.
Chamaillard, M., et  al. (2003a), “Nods, Nalps and Naip:  intracellular
regulators of bacterial-induced inflammation,” Cell Microbiol, 5 (9),
581–592.
Chamaillard, M., et  al. (2003b), “An essential role for NOD1 in host
recognition of bacterial peptidoglycan containing diaminopimelic
acid,” Nat Immunol, 4 (7), 702–707.
Cho, J. H. (2001), “The Nod2 gene in Crohn’s disease: implications for
future research into the genetics and immunology of Crohn’s dis-
ease,” Inflamm Bowel Dis, 7 (3), 271–275.
Cho, J. H., et al. (1998), “Identification of novel susceptibility loci for
inflammatory bowel disease on chromosomes 1p, 3q, and 4q: evi-
dence for epistasis between 1p and IBD1,” Proc Natl Acad Sci U S A,
95 (13), 7502–7507.
Cleynen, I., et  al. (2012), “Genetic factors conferring an increased
susceptibility to develop Crohn’s disease also influence
 disease phenotype:  results from the IBDchip European Project,”
Gut doi:10.1136/gutjnl-2011-300777.
Consortium, Uk Ibd Genetics, et al. (2009), “Genome-wide association
study of ulcerative colitis identifies three new susceptibility loci,
including the HNF4A region,” Nat Genet, 41 (12), 1330–1334.
Cooke, J., et  al. (2012), “Mucosal genome-wide methylation changes
in inflammatory bowel disease,” Inflamm Bowel Dis, 18 (11),
2128–2137.
Cooney, R., et  al. (2010), “NOD2 stimulation induces autophagy in
dendritic cells influencing bacterial handling and antigen presenta-
tion,” Nat Med, 16 (1), 90–97.
Coskun, M., et al. (2012), “MicroRNAs in inflammatory bowel disease—
pathogenesis, diagnostics and therapeutics,” World J Gastroenterol,
18 (34), 4629–4634.
Croucher, P.  J., et  al. (2003), “Haplotype structure and association to
Crohn’s disease of CARD15 mutations in two ethnically divergent
populations,” Eur J Hum Genet, 11 (1), 6–16.
Cummings, J. R., et al. (2007a), “Contribution of the novel inflamma-
tory bowel disease gene IL23R to disease susceptibility and pheno-
type,” Inflamm Bowel Dis, 13 (9), 1063–1068.
Cummings, J. R., et al. (2010), “The genetics of NOD-like receptors in
Crohn’s disease,” Tissue Antigens, 76 (1), 48–56.
Cummings, J. R., et al. (2007b), “Confirmation of the role of ATG16L1
as a Crohn’s disease susceptibility gene,” Inflamm Bowel Dis, 13 (8),
941–946.
Cuthbert, A. P., et al. (2002), “The contribution of NOD2 gene muta-
tions to the risk and site of disease in inflammatory bowel disease,”
Gastroenterology, 122 (4), 867–874.
Davies, J. L., et al. (1994), “A genome-wide search for human type 1 dia-
betes susceptibility genes,” Nature, 371 (6493), 130–136.
Duerr, R. H., et al. (2006), “A genome-wide association study identifies
IL23R as an inflammatory bowel disease gene,” Science, 314 (5804),
1461–1463.
Economou, M., et  al. (2004), “Differential effects of NOD2 variants
on Crohn’s disease risk and phenotype in diverse populations:  a
meta-analysis,” Am J Gastroenterol, 99 (12), 2393–2404.
Ekbom, A., et  al. (1990a), “Ulcerative colitis and colorectal cancer.
A population-based study,” N Engl J Med, 323 (18), 1228–1233.
--- (1990b), “Increased risk of large-bowel cancer in Crohn’s disease with
colonic involvement,” Lancet, 336 (8711), 357–359.
Elding, H., et  al. (2011), “Dissecting the genetics of complex inheri-
tance: linkage disequilibrium mapping provides insight into Crohn
disease,” Am J Hum Genet, 89 (6), 798–805.
Faybush, E. M., et al. (2002), “Generational differences in the age at diag-
nosis with IBD: genetic anticipation, bias, or temporal effects,” Am J
Gastroenterol, 97 (3), 636–640.
Fisher, S.  A., et  al. (2008), “Genetic determinants of ulcerative colitis
include the ECM1 locus and five loci implicated in Crohn’s disease,”
Nat Genet, 40 (6), 710–712.
Franke, A., et al. (2008), “Sequence variants in IL10, ARPC2 and mul-
tiple other loci contribute to ulcerative colitis susceptibility,” Nat
Genet, 40 (11), 1319–1323.
Franke, A., et al. (2010), “Genome-wide meta-analysis increases to 71 the
number of confirmed Crohn’s disease susceptibility loci,” Nat Genet,
42 (12), 1118–1125.
Futami, S., et  al. (1995), “HLA-DRB1*1502 allele, subtype of DR15,
is associated with susceptibility to ulcerative colitis and its progres-
sion,” Dig Dis Sci, 40 (4), 814–818.
Gaya, D. R., et al. (2006), “New genes in inflammatory bowel disease: les-
sons for complex diseases?” Lancet, 367 (9518), 1271–1284.
Girardin, S. E., et al. (2003), “Nod2 is a general sensor of peptidoglycan
through muramyl dipeptide (MDP) detection,” J Biol Chem, 278
(11), 8869–8872.
Glocker, E. and Grimbacher, B. (2012), “Inflammatory bowel disease: is
it a primary immunodeficiency?” Cell Mol Life Sci, 69 (1), 41–48.
Glocker, E.  O., et  al. (2011), “IL-10 and IL-10 receptor defects in
humans,” Ann N Y Acad Sci, 1246, 102–107.
G enetics and G enomics of C hronic I nflammatory D isorders , I   •   4 4 5
Glocker, E.  O., et  al. (2010), “Infant colitis—it’s in the genes,” Lancet,
376 (9748), 1272.
Glocker, E.  O., et  al. (2009a), “A homozygous CARD9 mutation in a
family with susceptibility to fungal infections,” N Engl J Med, 361
(18), 1727–1735.
Glocker, E. O., et al. (2009b), “Inflammatory bowel disease and muta-
tions affecting the interleukin-10 receptor,” N Engl J Med, 361 (21),
2033–2045.
Goyette, P., et al. (2008), “Gene-centric association mapping of chromo-
some 3p implicates MST1 in IBD pathogenesis,” Mucosal Immunol,
1 (2), 131–138.
Grivennikov, S.  I., Greten, F.  R., and Karin, M. (2010), “Immunity,
inflammation, and cancer,” Cell, 140 (6), 883–899.
Guo, Q.  S., et  al. (2004), “NOD2 3020insC frameshift mutation
is not associated with inflammatory bowel disease in Chinese
patients of Han nationality,” World J Gastroenterol, 10 (7),
1069–1071.
Guo, S.  W. (1998), “Inflation of sibling recurrence-risk ratio, due to
ascertainment bias and/or overreporting,” Am J Hum Genet, 63 (1),
252–258.
Halfvarson, J. (2011), “Genetics in twins with Crohn’s disease: less pro-
nounced than previously believed?” Inflamm Bowel Dis, 17 (1),
6–12.
Halfvarson, J., et al. (2005), “CARD15/NOD2 polymorphisms do not
explain concordance of Crohn’s disease in Swedish monozygotic
twins,” Dig Liver Dis, 37 (10), 768–772.
Hampe, J., et  al. (1999a), “Linkage of inflammatory bowel disease to
human chromosome 6p,” Am J Hum Genet, 65 (6), 1647–1655.
Hampe, J., et al. (1999b), “A genomewide analysis provides evidence for
novel linkages in inflammatory bowel disease in a large European
cohort,” Am J Hum Genet, 64 (3), 808–816.
Hampe, J., et  al. (2001), “Association between insertion mutation in
NOD2 gene and Crohn’s disease in German and British popula-
tions,” Lancet, 357 (9272), 1925–1928.
Hampe, J., et al. (2007), “A genome-wide association scan of nonsynony-
mous SNPs identifies a susceptibility variant for Crohn disease in
ATG16L1,” Nat Genet, 39 (2), 207–211.
Hancock, L., et al. (2006), “Clinical and molecular characteristics of iso-
lated colonic Crohn’s disease,” Gut, 55 (Suppl II) A1).
Harton, J.  A., et  al. (2002), “Cutting edge:  CATERPILLER:  a
large family of mammalian genes containing CARD, pyrin,
nucleotide-binding, and leucine-rich repeat domains,” J Immunol,
169 (8), 4088–4093.
Hayward, P.  A., Satsangi, J., and Jewell, D.  P. (1996), “Inflammatory
bowel disease and the X chromosome,” Q J Med, 89 (9), 713–718.
Helio, T., et al. (2003), “CARD15/NOD2 gene variants are associated
with familially occurring and complicated forms of Crohn’s disease,”
Gut, 52 (4), 558–562.
Hirschfield, G.  M., et  al. (2009), “Primary biliary cirrhosis associated
with HLA, IL12A, and IL12RB2 variants,” N Engl J Med, 360 (24),
2544–2555.
Hisamatsu, T., et al. (2003), “CARD15/NOD2 functions as an antibac-
terial factor in human intestinal epithelial cells,” Gastroenterology,
124 (4), 993–1000.
Ho, L.  P., et  al. (2005), “CARD 15 gene mutations in sarcoidosis,”
Thorax, 60 (4), 354–355.
Hsu, Y. M., et al. (2007), “The adaptor protein CARD9 is required for
innate immune responses to intracellular pathogens,” Nat Immunol,
8 (2), 198–205.
Huber, S., et  al. (2012), “IL-22BP is regulated by the inflammasome
and modulates tumorigenesis in the intestine,” Nature, 491 (7423),
259–263.
Hugot, J.  P., Zouali, H., and Lesage, S. (2003), “Lessons to be learned
from the NOD2 gene in Crohn’s disease,” Eur J Gastroenterol
Hepatol, 15 (6), 593–597.
Hugot, J. P., et al. (1996), “Mapping of a susceptibility locus for Crohn’s
disease on chromosome 16,” Nature, 379 (6568), 821–823.
Hugot, J.  P., et  al. (2001), “Association of NOD2 leucine-rich repeat
variants with susceptibility to Crohn’s disease,” Nature, 411 (6837),
599–603.
Iliev, I.  D., et  al. (2012), “Interactions between commensal fungi and
the C-type lectin receptor Dectin-1 influence colitis,” Science, 336
(6086), 1314–1317.
Inohara, N. and Nunez, G. (2003), “NODs:  intracellular proteins
involved in inflammation and apoptosis,” Nat Rev Immunol, 3 (5),
371–382.
Inohara, N., Ogura, Y., and Nunez, G. (2002), “Nods: a family of cyto-
solic proteins that regulate the host response to pathogens,” Curr
Opin Microbiol, 5 (1), 76–80.
Inohara, N., et al. (2003), “Host recognition of bacterial muramyl dipep-
tide mediated through NOD2. Implications for Crohn’s disease,” J
Biol Chem, 278 (8), 5509–5512.
Inoue, N., et al. (2002), “Lack of common NOD2 variants in Japanese
patients with Crohn’s disease,” Gastroenterology, 123 (1), 86–91.
Jenke, A.  C. and Zilbauer, M. (2012), “Epigenetics in inflammatory
bowel disease,” Curr Opin Gastroenterol, 28 (6), 577–584.
Jostins, L., et  al. (2012), “Host-microbe interactions have shaped the
genetic architecture of inflammatory bowel disease,” Nature, 491
(7422), 119–124.
Kanazawa, N., et al. (2005), “Early-onset sarcoidosis and CARD15 muta-
tions with constitutive nuclear factor-kappaB activation: common
genetic etiology with Blau syndrome,” Blood, 105 (3), 1195–1197.
Kathiresan, S., et al. (2009), “Common variants at 30 loci contribute to
polygenic dyslipidemia,” Nat Genet, 41 (1), 56–65.
Klionsky, D. J. (2009), “Crohn’s disease, autophagy, and the Paneth cell,”
N Engl J Med, 360 (17), 1785–1786.
Kobayashi, K.  S., et  al. (2005), “Nod2-dependent regulation of innate
and adaptive immunity in the intestinal tract,” Science, 307 (5710),
731–734.
Kruglyak, L. (1999), “Prospects for whole-genome linkage disequi-
librium mapping of common disease genes,” Nat Genet, 22 (2),
139–144.
Kugathasan, S., et al. (2005), “Comparative phenotypic and CARD15
mutational analysis among African American, Hispanic, and
White children with Crohn’s disease,” Inflamm Bowel Dis, 11 (7),
631–638.
Lala, S., et al. (2003), “Crohn’s disease and the NOD2 gene: a role for
Paneth cells,” Gastroenterology, 125 (1), 47–57.
Lanzarotto, F., Akbar, A., and Ghosh, S. (2005), “Does innate immune
response defect underlie inflammatory bowel disease in the Asian
population?” Postgrad Med J, 81 (958), 483–485.
Lee, G. H., et al. (2005), “[Frequency analysis of NOD2 gene mutations
in Korean patients with Crohn’s disease],” Korean J Gastroenterol,
45 (3), 162–168.
Lee, J. C., et al. (2011), “Gene expression profiling of CD8+ T cells pre-
dicts prognosis in patients with Crohn disease and ulcerative colitis,”
J Clin Invest, 121 (10), 4170–4179.
Lees, C. W., et al. (2011), “New IBD genetics: common pathways with
other diseases,” Gut, 60 (12), 1739–1753.
Leong, R.  W., et  al. (2003), “NOD2/CARD15 gene polymorphisms
and Crohn’s disease in the Chinese population,” Aliment Pharmacol
Ther, 17 (12), 1465–1470.
Lesage, S., et  al. (2002), “CARD15/NOD2 mutational analysis and
genotype-phenotype correlation in 612 patients with inflammatory
bowel disease,” Am J Hum Genet, 70 (4), 845–857.
Libioulle, C., et  al. (2007), “Novel Crohn disease locus identified by
genome-wide association maps to a gene desert on 5p13.1 and mod-
ulates expression of PTGER4,” PLoS Genet, 3 (4), e58.
Linde, K., et  al. (2003), “Card15 and Crohn’s disease:  healthy homo-
zygous carriers of the 3020insC frameshift mutation,” Am
J Gastroenterol, 98 (3), 613–617.
Maeda, S., et al. (2005), “Nod2 mutation in Crohn’s disease potentiates
NF-kappaB activity and IL-1beta processing,” Science, 307 (5710),
734–738.
4 4 6   •   G enomics in C linical P ractice
Manolio, T. A., et al. (2009), “Finding the missing heritability of com-
plex diseases,” Nature, 461 (7265), 747–753.
Maurano, M.  T., et  al. (2012), “Systematic localization of common
disease-associated variation in regulatory DNA,” Science, 337
(6099), 1190–1195.
McCarroll, S.  A., et  al. (2008), “Deletion polymorphism upstream of
IRGM associated with altered IRGM expression and Crohn’s dis-
ease,” Nat Genet. 40(9), 1107–1112.
Miceli-Richard, C., et  al. (2001), “CARD15 mutations in Blau syn-
drome,” Nat Genet, 29 (1), 19–20.
Moffatt, M.  F., et  al. (2007), “Genetic variants regulating ORMDL3
expression contribute to the risk of childhood asthma,” Nature, 448
(7152), 470–473.
Moran, C. J., et al. (2013), “IL-10R polymorphisms are associated with
very-early-onset ulcerative colitis,” Inflamm Bowel Dis, 19 (1),
115–123.
Murdoch, T. B., et al. (2012), “Prevalence of genetic variants associated
with inflammatory bowel disease in a healthy First Nations cohort,”
CMAJ, 184 (8), E435–E441.
Navin, N. and Hicks, J. (2011), “Future medical applications of single-cell
sequencing in cancer,” Genome Med, 3 (5), 31.
Nimmo, E.  R., et  al. (2012), “Genome-wide methylation profiling in
Crohn’s disease identifies altered epigenetic regulation of key host
defense mechanisms including the Th17 pathway,” Inflamm Bowel
Dis, 18 (5), 889–899.
O’Callaghan, N. J., et al. (2003), “Association of TNF-alpha-857C with
inflammatory bowel disease in the Australian population,” Scand J
Gastroenterol, 38 (5), 533–534.
Ogura, Y., et al. (2001), “A frameshift mutation in NOD2 associated with
susceptibility to Crohn’s disease,” Nature, 411 (6837), 603–606.
Orchard, T.  R., et  al. (2000), “Clinical phenotype is related to HLA
genotype in the peripheral arthropathies of inflammatory bowel dis-
ease,” Gastroenterology, 118 (2), 274–278.
Orholm, M., et  al. (2000), “Concordance of inflammatory bowel dis-
ease among Danish twins. Results of a nationwide study,” Scand J
Gastroenterol, 35 (10), 1075–1081.
Orholm, M., et al. (1991), “Familial occurrence of inflammatory bowel
disease,” N Engl J Med, 324 (2), 84–88.
Orstavik, K. H., et al. (2006), “Novel splicing mutation in the NEMO
(IKK-gamma) gene with severe immunodeficiency and heteroge-
neity of X-chromosome inactivation,” Am J Med Genet A, 140 (1),
31–39.
Parkes, M., et  al. (2007), “Sequence variants in the autophagy gene
IRGM and multiple other replicating loci contribute to Crohn’s dis-
ease susceptibility,” Nat Genet, 39 (7), 830–832.
Pathan, S., et al. (2010), “Putative genetic loci and confirmation of asso-
ciation in IBD,” Gastroenterology, 138 (5), S–6.
Pathan, S., et al. (2009), “Confirmation of the novel association at the
BTNL2 locus with ulcerative colitis,” Tissue Antigens, 74 (4),
322–329.
Pathan, S., et al. (2013), “Deep dermatophytosis and inherited CARD9
deficiency,” N Engl J Med, 369 (18), 1704–1714.
Polito, J. M., 2nd, et al. (1996), “Preliminary evidence for genetic antici-
pation in Crohn’s disease,” Lancet, 347 (9004), 798–800.
Pritchard, J. K. (2001), “Are rare variants responsible for susceptibility to
complex diseases?” Am J Hum Genet, 69 (1), 124–137.
Puel, A., et al. (2006), “The NEMO mutation creating the most-upstream
premature stop codon is hypomorphic because of a reinitiation of
translation,” Am J Hum Genet, 78 (4), 691–701.
Radlmayr, M., et  al. (2002), “The c-insertion mutation of the NOD2
gene is associated with fistulizing and fibrostenotic phenotypes in
Crohn’s disease,” Gastroenterology, 122 (7), 2091–2092.
Rioux, J.  D., et  al. (2007), “Genome-wide association study identifies
new susceptibility loci for Crohn disease and implicates autophagy
in disease pathogenesis,” Nat Genet, 39 (5), 596–604.
Risch, N. (1987), “Assessing the role of HLA-linked and unlinked deter-
minants of disease,” Am J Hum Genet, 40 (1), 1–14.
Risch, N. (1990), “Linkage strategies for genetically complex traits.
II. The power of affected relative pairs,” Am J Hum Genet, 46 (2),
229–241.
Rivas, M. A., et al. (2011), “Deep resequencing of GWAS loci identifies
independent rare variants associated with inflammatory bowel dis-
ease,” Nat Genet, 43 (11), 1066–1073.
Roda, G., et al. (2010), “New proteomic approaches for biomarker dis-
covery in inflammatory bowel disease,” Inflamm Bowel Dis, 16 (7),
1239–1246.
Roussomoustakaki, M., et al. (1997), “Genetic markers may predict dis-
ease behavior in patients with ulcerative colitis,” Gastroenterology,
112 (6), 1845–1853.
Russell, R.  K., et  al. (2005), “Genotype-phenotype analysis in
childhood-onset Crohn’s disease:  NOD2/CARD15 variants
consistently predict phenotypic characteristics of severe disease,”
Inflamm Bowel Dis, 11 (11), 955–964.
Satsangi, J., et al. (1994), “Genetics of inflammatory bowel disease,” Gut,
35 (5), 696–700.
Satsangi, J., et  al. (1996a), “Clinical patterns of familial inflammatory
bowel disease,” Gut, 38 (5), 738–741.
Satsangi, J., et al. (1996b), “Two stage genome-wide search in inflamma-
tory bowel disease provides evidence for susceptibility loci on chro-
mosomes 3, 7, and 12,” Nat Genet, 14 (2), 199–202.
Sauna, Z. E. and Kimchi-Sarfaty, C. (2011), “Understanding the contri-
bution of synonymous mutations to human disease,” Nat Rev Genet,
12 (10), 683–691.
Schreiber, S., et  al. (2005), “Genetics of Crohn disease, an archetypal
inflammatory barrier disease,” Nat Rev Genet, 6 (5), 376–388.
Schurmann, M., et al. (2003), “CARD15 gene mutations in sarcoidosis,”
Eur Respir J, 22 (5), 748–754.
Silverberg, M.  S., et  al. (2003), “A population- and family-based study
of Canadian families reveals association of HLA DRB1*0103 with
colonic involvement in inflammatory bowel disease,” Inflamm Bowel
Dis, 9 (1), 1–9.
Snook, J. A., de Silva, H. J., and Jewell, D. P. (1989), “The association of
autoimmune disorders with inflammatory bowel disease,” Q J Med,
72 (269), 835–840.
Stokkers, P. C., et al. (1999), “HLA-DR and -DQ phenotypes in inflam-
matory bowel disease: a meta-analysis,” Gut, 45 (3), 395–401.
Strober, W. (1985), “Animal models of inflammatory bowel disease—an
overview,” Dig Dis Sci, 30 (12 Suppl), 3S-10S.
Thompson, N. P., et al. (1996), “Genetics versus environment in inflam-
matory bowel disease:  results of a British twin study,” BMJ, 312
(7023), 95–96.
Todd, J. A., et al. (2007), “Robust associations of four new chromosome
regions from genome-wide analyses of type 1 diabetes,” Nat Genet,
39 (7), 857–864.
Trachtenberg, E.  A., et  al. (2000), “HLA class  II haplotype associa-
tions with inflammatory bowel disease in Jewish (Ashkenazi) and
non-Jewish Caucasian populations,” Hum Immunol, 61 (3), 326–333.
Travassos, L.  H., et  al. (2010), “Nod1 and Nod2 direct autophagy by
recruiting ATG16L1 to the plasma membrane at the site of bacterial
entry,” Nat Immunol, 11 (1), 55–62.
Tysk, C., et  al. (1988), “Ulcerative colitis and Crohn’s disease in an
unselected population of monozygotic and dizygotic twins. A study
of heritability and the influence of smoking,” Gut, 29 (7), 990–996.
Valentonyte, R., et al. (2005), “Sarcoidosis is associated with a truncating
splice site mutation in BTNL2,” Nat Genet, 37 (4), 357–364.
van Heel, D. A., et al. (2004), “Inflammatory bowel disease susceptibility
loci defined by genome scan meta-analysis of 1952 affected relative
pairs,” Hum Mol Genet, 13 (7), 763–770.
van Heel, D. A., et al. (2002), “Inflammatory bowel disease is associated
with a TNF polymorphism that affects an interaction between the
OCT1 and NF(-kappa)B transcription factors,” Hum Mol Genet, 11
(11), 1281–1289.
G enetics and G enomics of C hronic I nflammatory D isorders , I   •   4 4 7
Vegter-van der Vlis, M., Volkers, W. S., and Went, L. N. (1976), “Ages
of death of children with Huntington’s chorea and of their affected
parents,” Ann Hum Genet, 39 (3), 329–334.
Vind, I., et  al. (2005), “NOD2/CARD15 gene polymorphisms in
Crohn’s disease:  a genotype-phenotype analysis in Danish and
Portuguese patients and controls,” Digestion, 72 (2–3), 156–163.
Wang, K., Li, M., and Hakonarson, H. (2010), “Analysing biological
pathways in genome-wide association studies,” Nat Rev Genet, 11
(12), 843–854.
Watanabe, T., et al. (2004), “NOD2 is a negative regulator of Toll-like
receptor 2-mediated T helper type 1 responses,” Nat Immunol, 5 (8),
800–808.
Wehkamp, J. and Stange, E. F. (2005), “NOD2 mutation and mice: no
Crohn’s disease but many lessons to learn,” Trends Mol Med, 11 (7),
307–309.
Wehkamp, J., Fellermann, K., and Stange, E. F. (2005a), “Human defen-
sins in Crohn’s disease,” Chem Immunol Allergy, 86, 42–54.
Wehkamp, J., et  al. (2005b), “Defensin deficiency, intestinal microbes,
and the clinical phenotypes of Crohn’s disease,” J Leukoc Biol, 77 (4),
460–465.
Wehkamp, J., et al. (2005c), “Mechanisms of disease: defensins in gas-
trointestinal diseases,” Nat Clin Pract Gastroenterol Hepatol, 2 (9),
406–415.
Wehkamp, J., et al. (2004), “NOD2 (CARD15) mutations in Crohn’s
disease are associated with diminished mucosal alpha-defensin
expression,” Gut, 53 (11), 1658–1664.
Wehkamp, J., et  al. (2005d), “Reduced Paneth cell alpha-defensins
in ileal Crohn’s disease,” Proc Natl Acad Sci U S A, 102 (50),
18129–18134.
Wellcome Trust Case Control, Consortium, et al. (2010), “Genome-wide
association study of CNVs in 16,000 cases of eight common diseases
and 3,000 shared controls,” Nature, 464 (7289), 713–720.
Williams, C.  N., et  al. (2002), “Using a genome-wide scan and
meta-analysis to identify a novel IBD locus and confirm previously
identified IBD loci,” Inflamm Bowel Dis, 8 (6), 375–381.
WTCCC (2007), “Genome-wide association study of 14,000 cases of
seven common diseases and 3,000 shared controls,” Nature, 447
(7145), 661–678.
Xavier, R. J. and Podolsky, D. K. (2007), “Unravelling the pathogenesis
of inflammatory bowel disease,” Nature, 448 (7152), 427–434.
Yamagata, K., et  al. (1996), “Mutations in the hepatocyte nuclear
factor-4alpha gene in maturity-onset diabetes of the young
(MODY1),” Nature, 384 (6608), 458–460.
Yamazaki, K., et  al. (2002), “Absence of mutation in the NOD2/
CARD15 gene among 483 Japanese patients with Crohn’s disease,”
J Hum Genet, 47 (9), 469–472.
Yamazaki, K., et  al. (2005), “Single nucleotide polymorphisms in
TNFSF15 confer susceptibility to Crohn’s disease,” Hum Mol
Genet, 14 (22), 3499–3506.
Yap, L. M., Ahmad, T., and Jewell, D. P. (2004), “The contribution of
HLA genes to IBD susceptibility and phenotype,” Best Pract Res
Clin Gastroenterol, 18 (3), 577–596.
Yoshitake, S., et al. (1999), “HLA class  II alleles in Japanese patients
with inflammatory bowel disease,” Tissue Antigens, 53 (4 Pt 1),
350–358.
Zhernakova, A., van Diemen, C.  C., and Wijmenga, C. (2009),
“Detecting shared pathogenesis from the shared genetics of
immune-related diseases,” Nat Rev Genet, 10 (1), 43–55.
Zhou, Z., et al. (2002), “Variation at NOD2/CARD15 in familial and
sporadic cases of Crohn’s disease in the Ashkenazi Jewish popula-
tion,” Am J Gastroenterol, 97 (12), 3095–3101.
Zuk, O., et  al. (2012), “The mystery of missing heritability:  Genetic
interactions create phantom heritability,” Proc Natl Acad Sci U S A,
109 (4), 1193–1198.
 29.
GENETICS AND GENOMICS OF CHRONIC
INFLAMMATORY DISORDER S, II: RHEUMATOID
ARTHRITIS AND RELATED ARTHROPATHIES
Kate McAllister and Stephen Eyre
INTRODUCTION
Chronic inflammatory joint diseases of non-infective ori-
gin are common and cause significant long-term morbidity.
Rheumatoid arthritis (RA) and most of the related forms
of inflammatory arthritis are complex diseases, caused by
a combination of both genetic and environmental factors.
Although they are often referred to as “autoimmune disor-
ders,” it is far from clear what triggers the abnormal inflam-
matory processes that characterize these conditions (see also
Chapter 38). There has, however, been substantial progress
in determining the nature of genetic susceptibility to these
potentially crippling disorders. After 30 years in which the
human leukocyte antigen (HLA) genes have stood out as the
only oasis in a genetic desert, genome-wide association stud-
ies (GWAS) have made tremendous progress in identifying a
wealth of genetic variants that underlie these complex traits,
so much so, that there are now 60 genetic loci associated with
onset of RA.1 Since the application of genome-wide associa-
tion studies, it has also become evident that there are emerg-
ing patterns of genetic overlap between many autoimmune
diseases, including RA and the related arthropathies. An
example of this is the shared association of three genetic loci
(TNIP1, TNFAIP3, and IL2-IL21) between RA and pso-
riatic arthritis (PsA), indicating the likelihood of common
molecular mechanisms’ underlying disease. Determining
both the existence of shared genetic pathways and the genes
that have key phenotype-determining roles may enable effec-
tive management of these complex diseases (see Box 29.1).
R H E U M ATO I D A RT H R I T I S ( R A )
RA is a chronic immune-mediated inflammatory disorder
predominantly, but not exclusively, affecting synovial joints
448
(Figure 29.1). Systemic features, including weight loss, mal-
aise, low-grade fever, and vasculitis, are also common but
variable accompaniments. Many people contend that RA is
not a discrete clinical entity, given the heterogeneity seen
in the pattern of joint inflammation, autoantibodies, dis-
ease progression, the presence of extra-articular features,
and differential response to therapy. Rheumatoid synovitis
typically affects the diarthrodial joints of the hands and feet
first, causing progressive cartilage and bone destruction in
the affected joints (Figure 29.2). Subsequently, the more
proximal joints, including the hip and knee, may become
involved, leading to a severe functional disability. In the
course of the disease, any synovial joint can be affected,
including the temporomandibular and even the cricoaryte-
noid joints.
RA is one of the most common autoimmune disorders,
affecting approximately 1% of most populations worldwide.
The peak age at onset is in the fifth decade, although it can
start at virtually any age. The incidence of RA is typically
two to three times more common in women, and although
the reasons for this are unknown, both genetic (X-linked)
and hormonal influences have been suggested. RA usually
has a subacute (20%) or insidious (60%) onset with inflam-
matory symptoms of pain, swelling, and stiffness, typically
in the extremities, and gradually involving more joints over
weeks or months. Approximately 5% of patients have an
acute, severe polyarticular onset, and a few start with pal-
indromic rheumatism characterized by short-lived episodes
(<48 hours) of synovitis, recurring at variable intervals. The
course of the disease is variable; mild, self-limiting disease is
common, but the development of bony erosions early in the
course of the disease indicates a poor prognosis.
The presence of certain types of autoantibody in
the blood is not only highly predictive for the develop-
ment of RA but also often predates the onset of clinical
 Box 29.1  GLOSSARY OF GENETIC TER MS (SEE ALSO CHAPTER 50)

3C Chromosome conformation capture. This is a technique that can study the spatial 3D conformation
of chromatin within a cell.
Epigenetics The study of heritable changes to gene function that do not involve a change in the DNA sequence.
eQTL Expression quantitative-trait loci are genomic loci that that explain variation in gene-expression levels.
Exome sequencing A sequencing strategy to obtain the DNA sequence corresponding to all exons, excluding introns and
non-coding genomic sequence.
GWA or GWAS Genome wide association study. A study in which specific genetic markers (usually SNPs) across the entire
genome are scanned in both cases and controls in order to find genetic variations associated with disease.
Genome-wide Statistical term indicating that a genetic association is probably real, rather than being a false positive. To
significance account for multiple testing, 5 x 10-8 is used to claim statistical significance for genetic association. This is
based on testing 1 million SNPs with Bonferroni correction.
HapMap A catalogue of common genetic variants to help identify similarities and differences between human beings.
HLA Human leukocyte antigen. Human version of the major histocompatibility complex, a region on chromo-
some 6 containing many genes of immunological function.
Genotype imputation The prediction of genotypes at SNPs that have not been assayed directly in an association study.
Linkage disequilibrium Non-random association of alleles at genetic loci such that combinations of alleles occur more or less fre-
quently in a population than would be expected from the random formation of haplotypes.
MAF Minor allele frequency. Frequency at which the least common allele occurs in a given population.
Non-synonymous SNP A nucleotide mutation that alters the amino acid sequence of a protein.
SNP Single nucleotide polymorphism. A single base change in the DNA sequence. The most common type of
genetic variation among people.

manifestations by months or years. These antibodies include
the classic immunoglobulin M (IgM) rheumatoid factor
(RF) found in approximately 85% of the patients, and the
more specific anti-cyclic citrullinated peptide antibodies
(anti-CCP), the latter being detectable early in the course
of disease and also shown to be strong predictors of disease
severity and radiological damage.2 Both tests are often used
simultaneously to better predict the diagnosis and progno-
sis of patients with RA.
Bone loss is often a complication of RA; it can be
both focal and more generalized.3 Juxta-articular osteope-
nia adjacent to inflamed joints and the presence of focal

Figure 29.1 
Rheumatoid arthritis of the hands showing volar subluxation
at the wrist and metacarpophalangeal joints with associated soft-tissue Figure 29.2 
Classic juxta-articular erosions in rheumatoid arthritis
swelling. occurring at the point of synovial insertion (arrows).

G enetics and G enomics o f C hronic I n f l ammatory D isorders , I I   •   4 4 9

erosions of subchondral bone at the joint margins occur in
the areas of direct invasion by synovial pannus. This is medi-
ated through the action of osteoclast-like cells believed to
be derived from macrophage/monocyte precursors under
the influence of growth factors such as RANKL (recep-
tor activator of [nuclear factor kB] NFkB ligand), and
proinflammatory cytokines such as tumor necrosis factor
(TNF) produced in large quantities locally in the inflamed
synovium by activated macrophages. The inflammatory
process can also lead to a more generalized bone loss, caus-
ing osteopenia and osteoporosis of the axial and appendicu-
lar skeleton.4 As well as the established role of inflammation
in the destruction of bone, anti-CCP have also been shown
to induce bone erosion and secondary osteoporosis in RA
patients.5
Extra-articular manifestations of RA are common,
affecting around 40% of patients. Although RA is more
common in females, many studies have reported that
extra-articular features are more common in males and
occur more frequently in patients with more severe dis-
ease.6 The most common extra-articular feature occurring
in around 20–30% of seropositive cases are rheumatoid
nodules (RN).7 There is no unifying classification for these
extra manifestations; they can affect almost any organ, and
include, but are not limited to, nodules, vasculitis, serositis,
pulmonary fibrosis, and hematological, neurological, ocu-
lar, and cardiovascular abnormalities. These extra-articular
manifestations have a defining impact on disease outcome,
including increased mortality not only in relation to the
general population, but also in comparison to RA in gen-
eral. Standardized mortality rates are increased two or three
times, largely due to increased cardiovascular disease and
infections.
The classic cardiovascular manifestations of RA include
pericarditis, nodular valve disease, and major-vessel vasculi-
tis. It is also increasingly apparent that the risk of myocar-
dial infarction or congestive heart failure in RA patients
is significantly higher than in age- and sex-matched indi-
viduals. RA has been shown to increase the overall risk of
cardiovascular mortality by up to 50%.7 The excess risk of
cardiovascular disease seems to be independent of the tra-
ditional risk factors such as smoking, hypertension, and
body mass index (BMI).8,9 Nontraditional risk factors that
are emerging as contributors to atherogenesis in rheumatic
diseases include abnormal rheological characteristics such
as erythrocyte deformability and erythrocyte nitric oxide
production.10 The reasons for increased cardiovascular
morbidity and mortality are likely to be multifactorial, but
it is likely that the chronic systemic inflammation found
in rheumatoid disease promotes endothelial dysfunction
4 5 0   •   G enomics in C l inica l P ractice
and induces the accelerated atherosclerosis observed in RA
patients.11 Clearly, aggressive control of classic risk factors
for cardiovascular disease is indicated in RA along similar
lines to those recommended for diabetes mellitus, but there
is also emerging evidence that the use of disease-modifying
therapies such as methotrexate and biologics agents, includ-
ing TNF inhibitors with potent anti-inflammatory proper-
ties, may reduce cardiovascular risk.12,13
Rheumatoid vasculitis affects patients with more active
disease and accounts for a prevalence of fewer than 5% of
all RA cases. Clinical vasculitis can take one of many forms,
ranging from benign nail-fold lesions through isolated digi-
tal infarcts, to systemic involvement with major organ dam-
age:  such as mononeuritis multiplex, massive skin ulcers,
myocardial infarction, and bowel necrosis. Peripheral neu-
ropathy in RA is usually mild but has been associated with
excess mortality in RA.14 It is suggested that major vascu-
litic features are more common in those who are homozy-
gous for HLA-DRB1*04 alleles.15
Felty syndrome is less commonly encountered and
affects around 1% of those with RA, typically those with
longstanding disease. Felty syndrome describes the triad
of rheumatoid arthritis, neutropenia, and splenomegaly,
which is often associated with high titers of RF and other
extra-articular manifestations.16 It is particularly strongly
associated with the HLA-DRB1*0401 allele.17
D I AG N O S I S
Over the last decade it has been consistently demonstrated
that early therapeutic intervention is crucial in the manage-
ment and treatment of RA,18 and as a result, new classifica-
tion criteria have been developed through a collaborative
effort between the American College of Rheumatology
(ACR) and the European League Against Rheumatism
(EULAR)19 (Table 29.1).
The new 2010ACR/EULAR criteria have replaced
the 1987 criteria, which, although well accepted, are often
criticized because of their lack of focus on patients with
early disease symptoms. The 2010 criteria are intended to
be applicable to patients who have clinical synovitis in one
joint that is not better explained by any other diagnosis.
Once this has been determined, patients who then achieve
a point total of 6/10 or higher across four different areas of
diagnosis are classified with “definite RA.” The four areas
of diagnosis include joint involvement, serological param-
eters (including anti-CCP and RF), markers of inflamma-
tion (including ESR and CRP), and symptom duration.
Differentiating early RA from self-limiting or other forms
Table 29.1  THE 2010 CLASSIFICATION CRITERIA FOR RHEUMATOID ARTHRITIS DEVELOPED BY AMERICAN
COLLEGE OF RHEUMATOLOGY/EUROPEAN LEAGUE AGAINST RHEUMATISM

SCORE
Target population (Who should be tested?): patients who—
Have at least one joint with definite clinical synovitis (swelling)*
With the synovitis not better explained by another disease†
Classification criteria for RA (score-based algorithm: add score of categories A–D; a score of 6/10 is needed for classification
of a patient as having definite RA)‡
A. Joint involvement§
1 large joint¶ 0
2–10 large joints 1
1–3 small joints (with or without involvement of large joints) 2
4–10 small joints (with or without involvement of large joints) 3
>10 joints (at least 1 small joint)** 5
B. Serology (at least 1 test result is needed for classification)††
Negative RF and negative anti-CCP 0
Low-positive RF or low-positive anti-CCP 2
High-positive RF or high-positive anti-CCP 3
C. Acute-phase reactants (at least one test result is needed for classification)‡‡
Normal CRP and normal ESR 0
Abnormal CRP or abnormal ESR 1
D. Duration of symptoms§§
<6 weeks 0
≥6 weeks 1

* The criteria are aimed at classification of newly presenting patients. In addition, patients with erosive disease typical of rheumatoid arthritis (RA) with a history com-
patible with prior fulfilment of the 2010 criteria should be classified as having RA. Patients with longstanding disease, including those whose disease is inactive (with
or without treatment) and who, based on retrospectively available data, have previously fulfilled the 2010 criteria, should be classified as having RA.
† Differential diagnoses vary among patients with different presentations, but may include conditions such as systemic lupus erythematosus, psoriatic arthritis, and
gout. If it is unclear which relevant differential diagnoses to consider, an expert rheumatologist should be consulted.
‡ Although patients with a score of 6/10 are not classifiable as having RA, their status can be reassessed and the criteria might be fulfilled cumulatively over time.
§ “Joint involvement” refers to any swollen or tender joint on examination, which may be confirmed by imaging evidence of synovitis. Distal interphalangeal joints, first
carpometacarpal joints, and first metatarsophalangeal joints are excluded from assessment. Categories of joint distribution are classified according to the location and
number of involved joints, with placement into the highest category possible based on the pattern of joint involvement.
¶ “Large joints” refers to shoulders, elbows, hips, knees, and ankles.
# “Small joints” refers to the metacarpophalangeal joints, proximal interphalangeal joints, second through fifth metatarsophalangeal joints, thumb interphalangeal
joints, and wrists.
** In this category, at least one of the involved joints must be a small joint; the other joints can include any combination of large and additional small joints, as well as
other joints not specifically listed elsewhere (e.g., temporomandibular, acromioclavicular, sternoclavicular, etc.).
†† “Negative” refers to international unit (IU) values that are less than or equal to the upper limit of normal (ULN) for the laboratory and assay; “low-positive” refers
to IU values that are higher than the ULN but 3 times the ULN for the laboratory and assay; “high-positive” refers to IU values that are 3 times the ULN for the labo-
ratory and assay. When rheumatoid factor (RF) information is only available as positive or negative, a positive result should be scored as low-positive for RF.
‡‡ “Normal/abnormal” is determined by local laboratory standards. CRP—C-reactive protein; ESR—erythrocyte sedimentation rate.
§§ “Duration of symptoms” refers to patient self-report of the duration of signs or symptoms of synovitis (e.g., pain, swelling, tenderness) of joints that are clinically
involved at the time of assessment, regardless of treatment status.
SOURCE: Aletaha et al. “2010 Rheumatoid Arthritis Classification Criteria.” Arthritis & Rheumatism. 2010; 62(9):2574. DOI 10.1002/art.27584. Copyright 2010,
American College of Rheumatology. Reprinted with permission by John Wiley & Sons.

of inflammatory arthritis is challenging. Radiographic
changes are often absent at presentation, and conventional
radiographic methods do not provide much information
about the soft tissues and synovium.20 Ultrasonography and
magnetic resonance imaging (MRI) have shown promising
results for detecting synovitis and early erosive change, but
their utility in the management of RA is currently limited
by cost and availability.21–23 The new classification criteria
home in on RA characteristics that emerge early in the dis-
ease course, with the aim of identifying patients early in
their disease in order to widen the window of opportunity
for early treatment intervention and higher chances of
remission. In the new criteria, the diagnostic and prog-
nostic importance of anti-CCP as a serological marker in
addition to RF has been realized. Anti-CCP antibodies
can present early in the disease, often before clinical onset,
and the titer of anti-CCP has been shown to increase over
time, predating the diagnosis of RA.24,25 The anti-CCP
test has high specificity (95% for RA), and the presence of
anti-CCP antibodies has also been shown to predict the
development of erosive disease with a higher risk of joint
damage.26 A strong association of anti-CCP with the major

G enetics and G enomics o f C hronic I n f l ammatory D isorders , I I   •   4 5 1

histocompatibility complex (MHC) is often reported. G E N ET I C S O F R H EUM ATO I D
The extent to which either RF or anti-CCP antibodies are A RT H R IT I S
involved in the pathogenesis of RA is unclear but demands
RA is a classic example of a complex disease with a genetic
further exploration.
component characterized by incomplete penetrance,
To further advance early diagnosis, interest has grown in
genetic heterogeneity, and a role for many disease genes.
“prediction modelling” to estimate someone’s lifetime risk
Evidence from familial studies has long indicated a role
of developing RA. Despite combining all known genetic
of genetics in RA. An initial genetic link was established
and environmental risk factors into a prediction model,
with the observation of clustering in cases, with siblings of
the accuracy of such efforts remains modest. Most predic-
affected patients having an increased relative risk compared
tion models rely heavily on the human leukocyte antigen
to the general population. A  standard parameter used to
(HLA) alleles, but the incorporation of non-HLA loci into
quantify the role of genetic factors is the sibling recurrence
the model provides minimal additional prediction benefit,
risk (λs), which is the ratio of the lifetime risk of disease in
probably owing to their modest effect sizes.27 Currently,
siblings of affected patients compared to the lifetime risk
risk prediction is not accurate enough to be used in the
in the general population. RA is commonly cited as having
general population, but it may provide greater benefit in
a modest λs (5–10), which is quite common for complex
targeted screening approaches such as in first-degree rela-
diseases.30 Studies in twins have also added evidence to sup-
tives of RA patients. Better characterization of both genetic
port a genetic component. It has been broadly accepted that
and environmental risk factors will be necessary, alongside
the monozygotic concordance rate (15%) is approximately
the incorporation of additional factors such as epigenetic
four times that of dizygotic twins (3.6%).31 The relatively
information, biomarkers, and clinical predictors in order
low concordance rate between monozygotic twins supports
to improve the utilization of genetic profiling in a wider
the idea of an additional environmental component to dis-
setting.
ease susceptibility. Environmental triggers are supported
by a number of theories, but only smoking remains as a
validated risk factor.32,33 From family and twin studies, the
P O P U L AT I O N P R E VA L E N C E
estimated heritability (proportion of phenotypical variance
attributed to genetics rather than the environment) of RA
The overall world prevalence of RA is relatively con-
is around 60%. Uncovering the causal genes and molecular
stant at approximately 0.5–1%. There are some impor-
biology driving this genetic risk of RA will enable a bet-
tant differences in the prevalence of RA in ethnic groups
ter understanding of the etiology of the disease, providing
that illustrate the potential importance of environmen-
evidence of changes that occur before disease onset, unlike
tal and genetic factors in its etiology. For example, RA
many disease biomarkers, which are often a consequence of
is more common in some native Amerindian tribes
disease status.
(Yakima, Pima, Tlingit, and Chippewa).28 Whether
Over the last 60 years, many different strategies have been
environmental factors contribute to this is not known,
employed to investigate the genetic basis of RA and assem-
but the high frequency of certain RA-associated HLA
ble the ever-expanding list of over 60 currently confirmed
class  II alleles could explain this in the Chippewa
RA associated risk variants. The approaches can be broadly
(HLA-DRB1*04) and Yakima (HLA-DRB1*1402).
divided into the earlier hypothesis-driven approaches,
The prevalence of RA has also been shown to be low
including candidate-gene and linkage studies, and the more
in some areas of rural Africa compared to their urban
recent (post-2007) hypothesis-free genome-wide associa-
counterparts, suggesting that environmental factors
tion studies (GWAS), which have proved to be the most
could play a key role to explain the difference between
powerful and successful.
genetically related communities.29 Other than these
examples, there is relatively little consistent evidence
that shows unusually high or low RA-prevalence rates L I N K AG E A N D
between different ethnic groups. It remains difficult to C A N D I DAT E - G E N E S T U D I E S
make accurate comparisons of RA prevalence because
of different case-ascertainment criteria and assessment The use of traditional linkage and candidate-gene stud-
methodologies. Adoption of more standardized meth- ies to uncover genetic susceptibility to RA (pre-GWAS)
ods would make data assessment of the prevalence and led to the discovery of only two unequivocally associated
incidence of RA more reliable. genetic loci:  human leukocyte antigen (HLA, odds ratio

4 5 2   •   G enomics in C l inica l P ractice

[OR] ~4)34 and protein tyrosine phosphatase non-receptor
type 22 (PTPN22, OR ~1.8).35 Despite these successes,
candidate-gene and linkage study designs were generally
underpowered for the detection of variants of small effect
sizes and often led to many inconsistent results that proved
difficult to replicate. The reliance of these approaches
on a  priori evidence of a biological hypothesis has also
restricted the findings in complex diseases whose under-
lying biology remains relatively unknown. These study
designs have proved most successful in Mendelian disor-
ders, with the detection of variants with large effect sizes
and high penetrance.
G E N O M E -WI D E A S S O C I AT I O N
STUDIES
The completion of the Human Genome Project (HGP)
led to the full release of the genome sequence in 2003 and,
alongside vast technological advances, paved the way for-
ward for the GWAS approach.36 The HGP identified single
nucleotide polymorphisms (SNPs) as the most common
type of sequence variation (~90%). As well as being highly
abundant, SNPs are evolutionarily stable, making them
excellent genotypic markers. GWAS is based on genetic asso-
ciation determined if SNPs are seen more or less frequently
in cases (affected individuals) compared to controls (unaf-
fected individuals) in the same population. GWAS enables
the unbiased, systemic search of the entire genome without
any need for prior biological knowledge. The development
of a detailed map of genetic variation across four popula-
tions was achieved with the International HapMap Project
(2005), leading to the determination of linkage disequilib-
rium (LD) across the whole genome, enabling the design
of “tag” SNPs that capture the full genetic architecture of
the genome. In GWAS, hundreds of thousands of variants
across the genome are tested. To account for multiple test-
ing, the significance threshold for these studies is stringent
(P < 5 × 10–8). The success of GWAS is evident, not only
from the discovery of a plethora of risk variants in RA, but
in an online GWAS catalogue that currently documents
the discovery of 7226 trait-associated SNPs at the accepted
threshold of genome-wide significance (P = 5 × 10–8) as of
August 2012.
A groundbreaking GWAS was undertaken by the
Wellcome Trust Case Control Consortium (WTCCC) in
2007. This study consisted of 14,000 cases representative
of seven common diseases, including RA, and 3000 shared
controls. A case-control comparison led to the replication of
the robustly associated RA-susceptibility loci (HLA-DRB1,
P = 7.5 × 10–27) and (PTPN22, P = 5.6 × 10–25).37 Together,
G enetics and G enomics o f C hronic I n f l ammatory D isorders , I I   •   4 5 3
it is estimated these loci explain around 50% of the heri-
table risk of RA. In addition to these key findings, numer-
ous loci with suggestive significance were reported (P = 1 ×
10–5 – 5 × 10–7) and taken forward for a validation in 5063
RA cases and 3849 healthy controls. One SNP, rs6920220
mapping between two genes (OLIG3 and TNFAIP3), was
replicated (trend P = 1.1 × 10–8).38 A further 49 SNPs of
nominal significance (P = 1 × 10–4 – 1 × 10–5) were assessed
in a validation of 4,106 RA cases and a reference group of
11,238 healthy controls, resulting in replication of three
SNPs (rs4750316, rs1678542, rs3218253) that all map to
genes with plausible biological relevance (PRKCQ, KIF5A,
and IL2RB, respectively).39 GWAS validation studies have
since become a customary approach to ensure that findings
can be independently replicated in other cohorts and addi-
tionally confirmed in other populations to rule out false-
positive associations.
Since the initial WTCCC, numerous independent
GWAS have been carried out in RA, resulting in the dis-
covery of many additional novel RA associations. With the
exception of HLA and PTPN22, the emerging novel loci
have been of modest effect sizes (OR~1.2). As individual
studies often suffer from small sample sizes, it has become
more common to use a GWAS meta-analysis approach
that involves combining data from several studies in order
to increase the statistical power to detect variants of these
more modest effect sizes. Meta-analysis can have its limita-
tions, however: namely, the different methodologies used
across studies and the possibility of population stratifica-
tion when combining studies from different populations.
If these limitations are accounted and controlled for, then
meta-analysis can be a very successful strategy to increase
the power of statistical analyses. Two of the most notable
GWAS meta-analyses in RA were undertaken in different
populations. The first was conducted in 2010 and led to
the identification of seven novel RA-susceptibility loci
in people of European descent.40 This study was made up
of fourteen distinct sample collections from a variety of
countries, giving a combined total (including the repli-
cation cohort) of 12,307 cases and 28,975 controls. The
genome-wide associated SNPs identified to date, local-
ise to immune-related genes including IL6ST, SPRED2,
RBPJ, CCR6, IRF5, and PXK. In 2012, Okada and col-
leagues used data from a Japanese population. This study
analyzed 4,074 individuals with RA and 16,891 controls
initially, followed by a replication in 5,277 cases and
21,684 controls. Nine novel RA-risk alleles were identi-
fied that mapped to genes including B3GNT2, ANXA3,
CSF2, CD83, NFKBIE, ARID5B, PDE2A-ARAP1,
PLD4, and PTPN2.41
 It has become apparent that there is genetic heterogene-
ity amongst different ethnic populations. PADI4, respon-
sible for the in situ conversion of arginine to citrulline by
deamination, was initially found and replicated in the
Japanese population but only recently shows evidence for
association in Caucasian populations, after years of incon-
sistent findings.42–45 It is yet to be determined whether
the association signal is shared between ethnic popula-
tions. Although some genes have emerged as being com-
mon across ethnic groups, some variants have been found
to be restricted to specific ethnic groups. One example is
PTPN22. The PTPN22 gene encodes the protein Lyp, an
important negative regulator of T-cell activation. The pres-
ence of the PTPN22 1858 T/T genotype, encoding an
R620W amino acid substitution, increases the risk for RA
more than two-fold. The PTPN22 risk allele (C1858T) is
present in up to 17% of Caucasian patients but is absent
in East Asian populations, highlighting the likelihood of
genetic heterogeneity between Europeans and Asians.46
This example shows that risk alleles can be present in one
population but not the other.47 It can also be possible that
gene–gene or gene–environment interactions within dif-
ferent populations could explain different associations
with RA.
Bioinformatic analysis has also been successfully
employed in the discovery of RA risk loci. A  statistical
text-mining approach—gene relationships across implicated
loci (GRAIL)—was used to score moderately associated
genetic loci (n = 179) from the meta-GWAS for functional
relationships to genes in 16 pre-established RA loci. Of the
179 loci tested, 22 scored significantly for functional con-
nectivity; consequently, representative SNPs for each of the
regions were chosen and genotyped in an independent set
of 7,957 cases and 11,958 controls. Three of these SNPs
replicated successfully and implicated CD2-CD58, CD28,
PRDM1, respectively, as likely candidate genes.48
Despite the success of GWAS, the significantly associ-
ated loci still only explain a portion (60%) of the full herita-
bility, giving rise to widespread discussion of many different
explanations for this “hidden heritability.” One theory that
is often favored is a “rare variant” model wherein some of
the variance is explained by rare but highly penetrant vari-
ants.49 One of the limitations of GWAS is that it is designed
to detect the effects of common genetic variation, with a
minor allele frequency (MAF >5%), with limited power
for rare variants (MAF <1%). A recent pooled sequencing
study in RA assessed the contribution of rare coding vari-
ants in 25 RA candidate genes in 500 cases and 650 controls
of European ancestry. In RA patients, there was an accumu-
lation of rare missense variants in the IL2RA and IL2RB
4 5 4   •   G enomics in C l inica l P ractice
RA-susceptibility genes, demonstrating that there is a
potential role for rare coding variants in RA susceptibility.50
Further studies are required to determine how much of the
“hidden heritability” can be attributed to rare variants, but
results from other diseases suggest that they will explain
only a modest fraction. Other research into the heritabil-
ity of RA includes the role of epigenetics and the explora-
tion of gene–gene and gene–environment interactions, but
it will require further research before we can establish how
much each of these factors explains the “missing” heritabil-
ity to disease.51
P O S T- GWA S
By 2011, 36 risk loci were associated with RA across multi-
ple populations (31 in European ancestry). With few excep-
tions (PTPN22, PADI4, CCR6, IL2RA, and TNFAIP3)
these RA-susceptibility loci have yet to be experimentally
linked to biological function. SNPs identified by GWAS
are not necessarily the causal variants, with high corre-
lation between markers in the identified regions due to
strong LD, meaning a number of variants can be implicated
equally. Post-GWAS determining the causal variant and
assigning functionality to susceptibility SNPs has become
the focus of research. Despite the “modest effect sizes” of
RA risk variants, there are “backward validation” examples
of how biological therapies have been successfully used to
treat target genes and pathways identified through GWAS
(Tocilizumab—IL6, Abatacept—CTLA4, Adalimumab—
Anti-TNF). This highlights how targeting these “modest”
findings can have a tremendous biological impact by identi-
fying novel pathways and targets that might be amenable to
therapeutic intervention.
Methods are constantly being developed to charac-
terize the role of trait-associated SNPs. SNPs in exons
(protein-coding regions) can directly alter protein function
or stability. However, in RA, most of the associated SNPs
reside in non-coding regulatory regions of the genome and
may influence phenotype thorough a variety of mecha-
nisms, such as splicing, transcriptional activity, microRNA
regulation, post-translational modification, or even epigen-
etic modifications. First, though, in order to validate the
hypothesis of the associated SNP’s impacting on a causal
gene then going on to affect a biological pathway, asso-
ciation signals from GWAS need to be refined, before we
embark on expensive functional studies.
Fine-mapping is an initial step in the functional charac-
terization of loci. Instead of using a “tag” SNP approach like
GWAS, fine-mapping involves dense genotyping of every
known variant to help refine the GWAS signal in order to
identify the causal variant. The ImmunoChip, a custom SNP
array with 185,806 SNPs, was designed with fine-mapping
of loci as one of its key aims. The ImmunoChip project
was made possible through collaboration with investiga-
tors from 12 auto-immune diseases (AID). The chip was
designed to leverage the extensive shared overlap of loci
seen between these diseases, including RA. This is evi-
denced by the clustering of AID in families and also from
the shared association of many RA risk variants with other
AID. A total of 186 confirmed susceptibility loci across all
AID were selected for dense genotyping and fine-mapped
in 11,475 RA patients and 15,870 controls of European
descent. In total, 14 novel RA loci were identified, bring-
ing the total of RA susceptibility variants in European
populations to 46. Fine-mapping also localized association
signals for 19/46 to single genes. In addition, the analysis
highlighted seven non-synonymous exonic SNPs strongly
associated with disease, with likely functional capacity.44
Without fine-mapping, a different genetic region would
have been implicated as being causal in these regions, lead-
ing to erroneous downstream experiments.
The challenge of post–fine genetic mapping in RA, as in
all complex diseases, is to determine which genes the asso-
ciated variants affect, their mode of action, and how these
results can be translated into clinical utility. Great strides
have already been made in determining which cell types are
critical in RA susceptibility. Utilizing gene-expression data
that defined cell subtypes and overlaying the RA suscepti-
bility loci, it was found the greatest correlation was with
CD4+ T-cells.52 This evidence was further strengthened
using epigenetic chromatin marks that define gene activity
and are unique to cell subsets.53 Again, the greatest correla-
tion was seen between RA loci and regions defining CD4+
T-cells. This, then, gives us confidence that RA suscepti-
bility is triggered by a set of genes acting in CD4+ T-cell
subsets, providing a framework for future functional work.
It is now essential to definitively link all RA-susceptibility
SNPs to genes, using evidence gathered from experiments
such as expression quantitative trait loci (eQTL) analy-
sis, to identify whether SNPs alter gene expression54 and
chromosome conformation capture (3C) which enables
the interrogation of the three dimensional organization of
chromosomes within a cell.55 Once genes are identified, the
effect of the causal SNP on the target gene, as well as the
molecular mechanism involved, can be determined, utiliz-
ing such techniques as chromatin immunoprecipitation
(ChIP) and reporter-gene assays. Evidence gathered in these
experiments will determine the genes and genetic pathways
that are perturbed before the onset of disease, giving clini-
cal utility in terms of targeting genes for novel therapies,
G enetics and G enomics o f C hronic I n f l ammatory D isorders , I I   •   4 5 5
repositioning existing therapies from other diseases, and
stratifying patients into more homogeneous subtypes, lead-
ing to a better understanding of patient outcomes at base-
line and targeted therapeutic intervention.
H L A A N D R A S US C E P T I B I L IT Y
HLA remains the strongest genetic link to RA, with
the genetic contribution from this locus estimated to be
between 30% and 50%.56 The association of RA with
HLA-DRB1 was first discovered by Stastny. In 1976,
Stastny demonstrated that, in allogeneic mixed lym-
phocyte reactions, lymphocytes from patients with RA
tended to be mutually non-stimulatory compared to con-
trols.57 This observation could be explained on the basis
that the Dw4 specificity (later named DR4) was mark-
edly over-represented in the RA population. Cloning and
sequencing experiments have further characterized alleles
at the DR4 locus, a HLA-DR serotype. HLA-DR is a part
of the highly polymorphic MHC class II gene family that
resides on chromosome 6. MHC class II molecules are het-
erodimeric (α and β chain) transmembrane glycoproteins,
typically but not exclusively displayed at the surface of spe-
cialized antigen-presenting cells of the immune system. The
α and β chain of each MHC II molecule is encoded by sepa-
rate genes in the MHC class II regions. Three MHC class II
isotypes exist in humans, designated HLA-DR, HLA-DQ,
and HLA-DP. All three isotypes serve similar functions and
are critically involved in regulating the response to antigenic
peptides presented to T cells. Of the HLA-DR complex,
DRB1 is the most prevalent beta subunit.
With advancement of molecular typing techniques in
the 1980s, it was discovered that associated HLADR alleles
share an amino acid motif at positions 70–74 of the third
hyper-variable region of the DRB1 chain, commonly termed
the “shared epitope” (SE).58 Although studies in many pop-
ulations worldwide have shown association between RA
and various HLADRB1*04 alleles, this association is sub-
ject to ethnic variability, and not all HLA-DRB1*04 alleles
contribute equally to susceptibility. HLADRB1*0401 and
*0404 are most common in European populations. In con-
trast, in East Asian populations, DRB1*0401 is relatively
rare, and DRB1*0405 is the most frequent RA-associated
DRB1*04 allele. HLA-DRB1 0101 is also associated with
RA in many populations, although the attributable risk is
relatively weak compared to DRB1 0401 or 0404, particu-
larly in northern Europe.
Many studies have demonstrated that the SE alone
does not fully explain the association seen at the HLA and
have suggested that other independent genes within the
HLA complex may explain some of the association. The
highly polymorphic nature of the HLA, strong linkage
disequilibrium (LD) across the region, and the complex-
ity of genotyping have previously precluded widespread
interrogation of this region. However, the field of statisti-
cal genetics utilizing genetic imputation has recently seen
an advance in the understanding of the HLA–RA associa-
tion by fine-mapping and redefining the signal seen within
the MHC in seropositive patients.59 In this study, 19,992
anti-CCP positive individuals were used to define the refer-
ence data. These individuals had genotype data available on
2,537 SNPs that tagged the MHC and classic HLA alleles
at 4-digit resolution. The imputation method SNP2HLA
enabled the imputation of both classical HLA alleles as
well as amino acid polymorphisms in the HLA protein.
The strongest association in the HLA was determined at
amino acid 11 in the HLADRβ1 and not at the SE. To test
for independent effects, a series of conditional and haplo-
type analyses was performed, and this revealed additional
associations at amino acids 71 and 74 of the HLADRB1,
within the traditional SE region, as well as HLA-B (posi-
tion 9)  and HLADPB1 (position 9). Each of these asso-
ciations is located in the antigen peptide-binding groove,
suggesting that they could alter the affinity of antigen bind-
ing. A further study in a Norwegian population has found
an association independent of HLADRB1 at the HLA-C
locus in anti-CCP-positive RA.60
The precise role of the association of the HLA with RA
remains uncertain; the HLA could be critical in influencing
the binding of potentially arthritogenic peptides, in shap-
ing the T-cell repertoire necessary for the development of
autoimmunity, or by acting itself as an autoantigen. Despite
the identification of a plethora of putative autoantigens,
there has been little evidence implicating any of them in
the pathogenesis of RA.61 It is hoped that the rejuvena-
tion of the SE hypothesis will help facilitate new modelling
approaches and the application of panels of antigen micro-
arrays to help decipher and understand the role of the HLA
in RA pathogenesis.62
H L A A N D S EV E R I T Y
It now appears that, in addition to contributing to suscepti-
bility to RA, HLA-DRB1 genes play a role in determining
severity of the disease. HLA DRB1*0401/*0404 individu-
als have been convincingly shown to confer the greatest
risk for both early disease-onset and erosive disease. The
SE remains the single confirmed genetic marker of disease
severity; however, it still only explains a small proportion
of the total variance in the Larsen score. Improvement of
4 5 6   •   G enomics in C l inica l P ractice
the prediction of disease course and severity will enable
more effective management of patients. Although early
and aggressive treatment is necessary in patients who are
likely to develop erosive disease, it is important to distin-
guish them from patients who could achieve low disease
activity and remain non-erosive with minimal treatment.
Preliminary work has shown that HLADRB1 haplotypes
based on amino acid positions 11, 71, and 74 with a range
of effects on disease susceptibility (risk, neutral, protective)
also define severity haplotypes. If this holds true, these find-
ings could be utilized in a clinical setting to stratify patients
into risk categories prior to commencement of treatment.
PAT H O L O GY
The normal synovial membrane, which is one to two cell
layers thick, is transformed in the rheumatoid joint into a
proliferating cell mass, pannus, which destroys surround-
ing tissue and bone. A cellular composition of rheumatoid
synovitis is relatively simple and consists of resident cells and
inflammatory infiltrate. Synovial membrane includes resi-
dent fibroblast-like (type A) and macrophage-like (type B)
synoviocytes and endothelial cells. A  complex cascade of
cytokine-mediated events alters the phenotype of synovi-
ocytes (especially type B cells), and they show increased
expression of oncogenes and resist apoptosis.63 As a result,
the synoviocytes become highly proliferative, resulting in
widespread joint inflammation and matrix degradation.64
A  broad display of macrophage and fibroblast cytokines
like interleukin-1 (IL-1), IL-6, IL-12, IL-15, IL17, IL-18,
TNF-, granulocyte-macrophage colony-stimulating factor
(GM-CSF) and many more are produced by RA synovium
and play a central role in maintaining synovial inflamma-
tion.65 The inflammatory infiltrate in RA synovitis consists
of T cells, B cells, dendritic cells, macrophages, plasma cells,
mast cells, and granulocytes. Histology has shown that
synovial infiltrates can form different architectural pat-
terns. The most common form of rheumatoid synovitis is
a diffuse infiltrate, wherein inflammatory cells are scattered
among synovial cells without forming any more-organized
structures. In approximately 40% of patients, inflammatory
cells are organized into follicular structures and can form
germinal centers. In a minority of patients, the formation
of granulomas, resembling rheumatoid nodules, has been
observed. These different patterns are important, as they
may influence therapeutic response.
T lymphocytes have a central role in the formation and
maintenance of chronic inflammatory processes in RA.
So it came as a surprise that RA synovial fluid and tissue
have relatively low concentrations of T cell cytokines such
as IL-2 and interferon-γ (IFN-γ), in contrast to the mac-
rophage and fibroblast products.66 Besides the accumu-
lation of activated T cells in the RA synovium, the entire
T-cell pool phenotype and homeostasis is also abnormal.
Significant telomere-shortening found in the CD4+ T
cells of RA patients indicates that they have undergone
intensive replication.67 CD4+ T cells express CCR5 and
CXCR3 chemokine receptors. These chemoattractants
mark the subsets of cells with the capacity for migration to
inflammatory sites.68 In addition to the aberrations in the
global and the synovial T-cell receptor repertoire, clonal
proliferation of T-cell population in RA patients has been
observed. Proliferating CD4 cells are functionally differ-
ent from classic T-helper cells and are uniformly present
in the circulation and infiltrate into synovial lesions.69,70 A
set of clonally expanded CD4+ T cells lacks surface expres-
sion of the major co-stimulatory molecule CD28. This is
caused by a transcriptional block of the CD28 gene, due to
the lack of two gene-specific nuclear proteins that are neces-
sary for the initiation of CD28 transcription.71 The CD4+
CD28-T cells are potentially cytotoxic, as they produce
inflammatory cytokines (IFN-γ) and contain intracellular
cytolytic substances (perforin and granzyme B) necessary
to lyse target cells. In the absence of the CD28 molecule,
these unusual cells use alternative co-stimulatory pathways
and de novo express killer-cell immunoglobulin-like recep-
tors (KIR2DS2, NKG2D). These cells may be related to
natural killer (NK) cells, as they express MHC class I–rec-
ognizing receptors, allowing them to participate in innate
and adaptive immune processes. Furthermore, the increased
frequency of CD4+ CD28 cells is a predictor of progressive
erosive disease in early RA.72
TNF and IL-1 are the two main cytokines contribut-
ing to the joint inflammation in RA. TNF is a proinflam-
matory cytokine, produced by monocytes, T lymphocytes,
macrophages, and fibroblasts. TNF stimulates prostaglan-
din E2 and collagenase-release from human synovial cells
and potentiates osteoclastogenesis. IL-2 is typically synthe-
sized and released simultaneously with TNF, and the two
cytokines share similar biological properties. Elevated con-
centrations of proinflammatory cytokines have been found
in peripheral blood and synovial fluid of RA patients, and
the concentrations correlate with disease activity and bone
resorption.73,74 Both IL-1 and TNF, as well as GM-CSF,
stimulate fibroblast growth and pannus formation. These
inflammatory cytokines also interact with each other and
promote the inflammatory process. Normal bone remodel-
ing depends on the coordinated activity of the osteoblasts
and osteoclasts. Under normal physiological conditions,
osteoclasts are primarily responsible for bone resorption.
G enetics and G enomics o f C hronic I n f l ammatory D isorders , I I   •   4 5 7
These multinucleated cells, derived from the monocyte/
macrophage line, also mediate focal bone resorption in
RA. Osteoclast precursors have been identified in the bone
resorption lacuna adjacent to the invading pannus.75 The
differentiation of osteoclast precursors into mature bone
resorbing cells depends on the cytokine network includ-
ing TNF, IL-1, and IL-6. Osteoblasts and bone stromal
cells produce a cytokine RANKL, which, in conjugation
with M-CSF, is the essential factor for osteoclast differen-
tiation. RANKL binds to its cell-surface receptor RANK,
located on dendritic, T-lymphocytes, osteoclast precursors,
and osteoclasts. RANK mediates its signal transduction via
TNF receptor-associated factor (TRAF) proteins, leading
to differentiation of precursors, fusion to form multicellular
osteoclasts, and the activation of the osteoclasts to resorb
bone.76 Osteoblasts also secrete osteoprotegerin (OPG),
a soluble decoy receptor for RANKL. By binding to
RANKL with high affinity, OPG prevents RANKL from
interacting with its cognate receptor (RANK). It is the rela-
tive amounts of RANKL and OPG that regulate osteoclast
activity.77 Confirming evidence for the role of RANKL in
bone resorption comes from studies of rat adjuvant arthri-
tis.78 Arthritic rats treated with OPG at the onset of dis-
ease had a minimal loss of cortical and trabecular bone in
contrast to the untreated control animals, whose bone loss
was severe. OPG treatment also prevented osteoclast accu-
mulation. However, the treatment did not decrease joint
inflammation, indicating that the effect was specific only for
osteoclast-mediated bone loss. RANKL-knockout mice, in
whom inflammatory arthritis was induced, demonstrated a
reduced degree of bone erosions compared to the wild-type
animals.79 In addition, RA synovial tissue produces other
cytokines, like IL-1α, IL-1β, and IL-11, that enhance osteo-
clast formation, activity, and survival.80,81
T R E AT M E N T
RA treatment aims to relieve symptoms, improve overall
function, and prevent disease progression. There is little
evidence that standard pre-biologic treatments directed
toward symptom relief significantly lessen joint destruc-
tion and disease progression.82 Previously, the well-accepted
“step-up” or pyramidal approach to treating RA under-
went a significant reconsideration when observational and
MRI studies revealed that joint erosions, and more impor-
tantly, functional decline, emerge early in the disease.83
As a result, standard disease-modifying antirheumatic
drugs (DMARD) were increasingly introduced early in
the disease course, and prescribing combinations of differ-
ent DMARDs became common practice.84,85 Treatment
with methotrexate (MTX), leflunomide, and sulfasala-
zine (SSZ) has been shown to reduce swollen/tender joint
counts as well as decrease radiographic progression86–88;
however, a third of patients do not respond to treatment
with traditional DMARDS.89,90 In the last decade, signifi-
cant advances have been made, and RA treatment has been
revolutionized by the development and introduction of
biologics agents. There are at least 10 biologics therapies
now approved for use that target specific molecules on the
cells of the immune system that are involved in the patho-
genesis of RA. The most widely used biologics are the TNF
antagonists such as etanercept, infliximab, adalimumab,
certolizumab, and golimumab.91
Etanercept is a human recombinant fusion protein com-
bining two soluble p75 TNF receptors linked to the Fc part
of human IgG1. This compound blocks the activity of TNF
(and also the related cytokine lymphotoxin a) by competi-
tively blocking the association of TNF with its cell-surface
receptors.92 Infliximab is a mouse/human chimeric mono-
clonal antibody that binds with high affinity and selectivity
to TNF. Like etanercept, infliximab prevents the associa-
tion of TNF with its cell-surface receptor and blocks the
further signaling activity. Anakinra is another class of bio-
logic therapy that has been approved for the treatment of
RA. Anakinra is a human recombinant IL-1 receptor antag-
onist and closely resembles naturally occurring IL-receptor
antagonist. Anakinra competitively inhibits binding of
IL-1 to its cell-surface receptor, thus blocking the biologi-
cal activity of IL-1.93 Promising results in treating refractory
RA have also been reported with rituximab.94,95 Rituximab
is a chimeric monoclonal antibody directed against CD20.
The CD20 antigen is present on the cell surface of differ-
entiating B cells, though its expression ceases when B cells
terminally differentiate into plasma cells. Rituximab causes
prolonged and sustained improvement in disease activity
and is useful in patients who have failed anti-TNF treat-
ment.96 Other biological agents modulating the inflamma-
tory response that have proved beneficial in rheumatoid
arthritis include antibodies targeting the soluble IL-6 recep-
tor97and CTLA4 1g fusion protein, blocking interactions
between the co-stimulatory molecules CD28 and CD80/
CD86.98,99 There has also been the recent development of
two tyrosine kinase (TK) inhibitors. Tyrosine kinases are
enzymes that play diverse roles in critical cellular and bio-
logical functions. Numerous kinases have been implicated
in the role of immune defense and inflammation, but it is
the spleen TYK and the Janus kinase ( JAK) family of mol-
ecules that are the targets of the novel TK inhibitors (fos-
tamatinib and tofacitinib, respectively). Both of these have
been approved as second-line therapies and have achieved a
4 5 8   •   G enomics in C l inica l P ractice
level of clinical efficacy comparable to that of the anti-TNF
agents.100
The introduction of biologics therapies has resulted in
fewer patients’ progressing to disability and has seen an
increase in the number of patients with low disease activity
and remission, although many patients (around 30–40%)
will still fail to respond to these new-line therapies and
remain with active disease.101 In addition, biologics therapies
are expensive and have associated side effects and adverse
events. As well as emphasizing the need for novel therapies
with different modes of action, this supports the great inter-
est in trying to identify predictors of drug response, leading
to stratified medicine and tailored therapies. The ultimate
aim is to get the right drug to the right patient the first time,
to give the best benefit and provide a better long-term out-
look. Despite there being no current reliable and definite
predictor of drug response, larger and well characterized
datasets are being collected to address this goal.
REFERENCES
1. Viatte S, Plant D, Raychaudhuri S. Genetics and epigenetics of rheu-
matoid arthritis. Nat Rev Rheumatol. March 2013;9(3): 141–153.
2. Taylor P, Gartemann J, Hsieh J, Creeden J. A systematic review
of serum biomarkers anti-cyclic citrullinated peptide and rheu-
matoid factor as tests for rheumatoid arthritis. Autoimmune Dis.
2011;2011:815038.
3. Deal C. Bone loss in rheumatoid arthritis:  systemic, periarticular,
and focal. Curr Rheumatol Rep. June 2012;14(3):231–237.
4. Goldring SR, Gravallese EM. Mechanisms of bone loss in inflamma-
tory arthritis: diagnosis and therapeutic implications. Arthritis Res.
2000;2(1):33–37.
5. Kocijan R, Harre U, Schett G. ACPA and bone loss in rheumatoid
arthritis. Curr Rheumatol Rep. October 2013;15(10):366.
6. Young A, Koduri G. Extra-articular manifestations and compli-
cations of rheumatoid arthritis. Best Pract Res Clin Rheumatol.
October 2007;21(5):907–927.
7. Vina-Zubieta JA, Choi HK, Sadatsafavi M, Etminan M, Esdaile JM,
Lacaille D. Risk of cardiovascular mortality in patients with rheu-
matoid arthritis: a meta-analysis of observational studies. Arthritis
Rheum. December 15, 2008;59(12):1690–1697.
8. Gabriel SE, Crowson CS, O’Fallon WM. Comorbidity in arthritis.
J Rheumatol. November 1999;26(11):2475–2479.
9. Nicola PJ, Maradit-Kremers H, Roger VL, et al. The risk of conges-
tive heart failure in rheumatoid arthritis: a population-based study
over 46 years. Arthritis Rheum. February 2005;52(2):412–420.
10. Gabriel SE, Crowson CS. Risk factors for cardiovascular disease in rheu­
matoid arthritis. Curr Opin Rheumatol. March 2012;24(2): 171–176.
11. Ridker PM, Rifai N, Rose L, Buring JE, Cook NR. Comparison
of C-reactive protein and low-density lipoprotein cholesterol lev-
els in the prediction of first cardiovascular events. N Engl J Med.
November 14, 2002;347(20):1557–1565.
12. Jacobsson LT, Turesson C, Gulfe A, et  al. Treatment with tumor
necrosis factor blockers is associated with a lower incidence of
first cardiovascular events in patients with rheumatoid arthritis.
J Rheumatol. July 2005;32(7):1213–1218.
13. Gonzalez-Juanatey C, Testa A, Garcia-Castelo A, Garcia-Porrua C,
Llorca J, Gonzalez-Gay MA. Active but transient improvement of
 endothelial function in rheumatoid arthritis patients undergoing
long-term treatment with anti-tumor necrosis factor alpha antibody.
Arthritis Rheum. June 15, 2004;51(3):447–450.
14. Turesson C, O’Fallon WM, Crowson CS, Gabriel SE, Matteson EL.
Occurrence of extra-articular disease manifestations is associated
with excess mortality in a community based cohort of patients with
rheumatoid arthritis. J Rheumatol. January 2002;29(1): 62–67.
15. Weyand CM, Hicok KC, Conn DL, Goronzy JJ. The influence of
HLA-DRB1 genes on disease severity in rheumatoid arthritis. Ann
Intern Med. November 15, 1992;117(10):801–806.
16. Campion G, Maddison PJ, Goulding N, et  al. The Felty syn-
drome:  a case-matched study of clinical manifestations and out-
come, serologic features, and immunogenetic associations. Medicine
(Baltimore). March 1990;69(2):69–80.
17. Lanchbury JS, Jaeger EE, Sansom DM, et al. Strong primary selec-
tion for the Dw4 subtype of DR4 accounts for the HLA-DQw7
association with Felty’s syndrome. Hum Immunol. September
1991;32(1):56–64.
18. Emery P. Evidence supporting the benefit of early intervention in
rheumatoid arthritis. J Rheumatol Suppl. November 2002;66:3–8.
19. Aletaha D, Neogi T, Silman AJ, et  al. 2010 rheumatoid arthritis
classification criteria:  an American College of Rheumatology/
European League Against Rheumatism collaborative initiative. Ann
Rheum Dis. September 2010;69(9):1580–1588.
20. van der Heijde DM, van Leeuwen MA, van Riel PL, van de Putte
LB. Radiographic progression on radiographs of hands and feet
during the first 3  years of rheumatoid arthritis measured accord-
ing to Sharp’s method (van der Heijde modification). J Rheumatol.
September 1995;22(9):1792–1796.
21. Hoving JL, Buchbinder R, Hall S, et  al. A comparison of mag-
netic resonance imaging, sonography, and radiography of the hand
in patients with early rheumatoid arthritis. J Rheumatol. April
2004;31(4):663–675.
22. Lopez-Ben R, Bernreuter WK, Moreland LW, Alarcon GS.
Ultrasound detection of bone erosions in rheumatoid arthritis:  a
comparison to routine radiographs of the hands and feet. Skeletal
Radiol. February 2004;33(2):80–84.
23. Szkudlarek M, Narvestad E, Klarlund M, Court-Payen, Thomsen
HS, Ostergaard M. Ultrasonography of the metatarsophalangeal
joints in rheumatoid arthritis:  comparison with magnetic reso-
nance imaging, conventional radiography, and clinical examination.
Arthritis Rheum. July 2004;50(7):2103–2112.
24. Burr ML, Viatte S, Bukhari M, et al. Long-term stability of anti-cyclic
citrullinated peptide antibody status in patients with early inflam-
matory polyarthritis. Arthritis Res Ther. 2012;14(3):R109.
25. Rantapaa-Dahlqvist S, de Jong BA, Berglin E, et  al. Antibodies
against cyclic citrullinated peptide and IgA rheumatoid factor
predict the development of rheumatoid arthritis. Arthritis Rheum.
October 2003;48(10):2741–2749.
26. Niewold TB, Harrison MJ, Paget SA. Anti-CCP antibody testing as
a diagnostic and prognostic tool in rheumatoid arthritis. QJM. April
2007;100(4):193–201.
27. Yarwood A, Han B, Raychaudhuri S, et al. A weighted genetic risk
score using all known susceptibility variants to estimate rheuma-
toid arthritis risk. Ann Rheum Dis. October 3, 2013. doi:10.1136/
annrheumdis-2013-204133. [Epub ahead of print]
28. Ferucci ED, Templin DW, Lanier AP. Rheumatoid arthritis in
American Indians and Alaska Natives:  a review of the literature.
Semin Arthritis Rheum. February 2005;34(4):662–667.
29. Silman AJ, Ollier W, Holligan S, et  al. Absence of rheuma-
toid arthritis in a rural Nigerian population. J Rheumatol. April
1993;20(4):618–622.
30. Lawrence JS. Heberden Oration, 1969. Rheumatoid arthritis—
nature or nurture? Ann Rheum Dis. July 1970;29(4):357–379.
31. Silman AJ, MacGregor AJ, Thomson W, et  al. Twin concordance
rates for rheumatoid arthritis: results from a nationwide study. Br J
Rheumatol. October 1993;32(10):903–907.
G enetics and G enomics o f C hronic I n f l ammatory D isorders , I I   •   4 5 9
32. Hutchinson D, Moots R. Cigarette smoking and severity of rheu-
matoid arthritis. Rheumatology (Oxford). December 2001;40(12):
1426–1427.
33. Silman AJ. Smoking and the risk of rheumatoid arthritis. J
Rheumatol. November 1993;20(11):1815–1816.
34. MacKay K, Eyre S, Myerscough A, et  al. Whole-genome linkage
analysis of rheumatoid arthritis susceptibility loci in 252 affected
sibling pairs in the United Kingdom. Arthritis Rheum. March
2002;46(3):632–639.
35. Michou L, Lasbleiz S, Rat AC, et al. Linkage proof for PTPN22,
a rheumatoid arthritis susceptibility gene and a human autoim-
munity gene. Proc Natl Acad Sci U S A. January 30, 2007;104(5):
1649–1654.
36. International Human Genome Sequencing Consortium. Nature.
2004;431:931–945. doi:10.1038/nature03001.
37. Genome-wide association study of 14,000 cases of seven common
diseases and 3,000 shared controls. Nature. June 7, 2007;447(7145):
661–678.
38. Thomson W, Barton A, Ke X, et al. Rheumatoid arthritis association
at 6q23. Nat Genet. December 2007;39(12):1431–1433.
39. Barton A, Thomson W, Ke X, et  al. Rheumatoid arthritis suscep-
tibility loci at chromosomes 10p15, 12q13 and 22q13. Nat Genet.
October 2008;40(10):1156–1159.
40. Stahl EA, Raychaudhuri S, Remmers EF, et al. Genome-wide associ-
ation study meta-analysis identifies seven new rheumatoid arthritis
risk loci. Nat Genet. June 2010;42(6):508–514.
41. Okada Y, Terao C, Ikari K, et al. Meta-analysis identifies nine new
loci associated with rheumatoid arthritis in the Japanese population.
Nat Genet. May 2012;44(5):511–516.
42. Barton A, Bowes J, Eyre S, et  al. A functional haplotype of the
PADI4 gene associated with rheumatoid arthritis in a Japanese pop-
ulation is not associated in a United Kingdom population. Arthritis
Rheum. April 2004;50(4):1117–1121.
43. Caponi L, Petit-Teixeira E, Sebbag M, et al. A family based study
shows no association between rheumatoid arthritis and the
PADI4 gene in a white French population. Ann Rheum Dis. April
2005;64(4):587–593.
44. Eyre S, Bowes J, Diogo D, et al. High-density genetic mapping iden-
tifies new susceptibility loci for rheumatoid arthritis. Nat Genet.
December 2012;44(12):1336–1340.
45. Martinez A, Valdivia A, Pascual-Salcedo D, et  al. PADI4 poly-
morphisms are not associated with rheumatoid arthritis in the
Spanish population. Rheumatology (Oxford). October 2005;44(10):
1263–1266.
46. Gregersen PK, Lee HS, Batliwalla F, Begovich AB. PTPN22: setting
thresholds for autoimmunity. Semin Immunol. August 2006;18(4):
214–223.
47. Yamada R, Yamamoto K. Mechanisms of disease: genetics of rheu-
matoid arthritis—ethnic differences in disease-associated genes. Nat
Clin Pract Rheumatol. November 2007;3(11):644–650.
48. Raychaudhuri S, Thomson BP, Remmers EF, et al. Genetic vari-
ants at CD28, PRDM1 and CD2/CD58 are associated with
rheumatoid arthritis risk. Nat Genet. December 2009;41(12):
1313–1318.
49. Gibson G. Rare and common variants: twenty arguments. Nat Rev
Genet. February 2011;13(2):135–145.
50. Diogo D, Kurreeman F, Stahl EA, et  al. Rare, low-frequency, and
common variants in the protein-coding sequence of biological can-
didate genes from GWASs contribute to risk of rheumatoid arthri-
tis. Am J Hum Genet. January 10, 2013;92(1):15–27.
51. Eichler EE, Flint J, Gibson G, et al. Missing heritability and strate-
gies for finding the underlying causes of complex disease. Nat Rev
Genet. June 2010;11(6):446–450.
52. Hu X, Kim H, Stahl E, Plenge R, Daly M, Raychaudhuri S.
Integrating autoimmune risk loci with gene-expression data iden-
tifies specific pathogenic immune cell subsets. Am J Hum Genet.
October 7, 2011;89(4):496–506.
 53. Trynka G, Sandor C, Han B, et al. Chromatin marks identify criti-
cal cell types for fine mapping complex trait variants. Nat Genet.
February 2013;45(2):124–130.
54. Nica AC, Dermitzakis ET. Expression quantitative trait
loci:  present and future. Philos Trans R Soc Lond B Biol Sci.
2013;368(1620):20120362.
55. Elzo de Wit, Wouter de Laat. A decade of 3C technologies: insights
into nuclear organization. Genes Dev. January 1, 2012;26(1):11–24.
doi:10.1101/gad.179804.111.
56. Jawaheer D, Li W, Graham RR, et  al. Dissecting the genetic
complexity of the association between human leukocyte anti-
gens and rheumatoid arthritis. Am J Hum Genet. September
2002;71(3):585–594.
57. Stastny P. Mixed lymphocyte cultures in rheumatoid arthritis. J Clin
Invest. May 1976;57(5):1148–1157.
58. Gregersen PK, Silver J, Winchester RJ. The shared epitope hypoth-
esis. An approach to understanding the molecular genetics of sus-
ceptibility to rheumatoid arthritis. Arthritis Rheum. November
1987;30(11):1205–1213.
59. Raychaudhuri S, Sandor C, Stahl EA, et  al. Five amino acids in
three HLA proteins explain most of the association between
MHC and seropositive rheumatoid arthritis. Nat Genet. March
2012;44(3):291–296.
60. Nordang GB, Flam ST, Maehlen MT, Kvien TK, Viken MK,
Lie BA. HLA-C alleles confer risk for anti-citrullinated peptide
antibody-positive rheumatoid arthritis independent of HLA-DRB1
alleles. Rheumatology (Oxford). November 2013;52(11): 1973–1982.
61. Corrigall VM, Panayi GS. Autoantigens and immune pathways in
rheumatoid arthritis. Crit Rev Immunol. 2002;22(4):281–293.
62. Yeste A, Quintana FJ. Antigen microarrays for the study of autoim-
mune diseases. Clin Chem. July 2013;59(7):1036–1044.
63. Yamanishi Y, Boyle DL, Green DR, et al. p53 tumor suppressor gene
mutations in fibroblast-like synoviocytes from erosion synovium
and non-erosion synovium in rheumatoid arthritis. Arthritis Res
Ther. 2005;7(1):R12–R18.
64. Ritchlin C. Fibroblast biology. Effector signals released by the syno-
vial fibroblast in arthritis. Arthritis Res. 2000;2(5):356–360.
65. Bucala R, Ritchlin C, Winchester R, Cerami A. Constitutive pro-
duction of inflammatory and mitogenic cytokines by rheumatoid
synovial fibroblasts. J Exp Med. March 1, 1991;173(3):569–574.
66. Firestein GS, Varo-Gracia JM, Maki R. Quantitative analysis of
cytokine gene expression in rheumatoid arthritis. J Immunol. May
1, 1990;144(9):3347–3353.
67. Wagner UG, Koetz K, Weyand CM, Goronzy JJ. Perturbation of
the T cell repertoire in rheumatoid arthritis. Proc Natl Acad Sci U S
A. November 24, 1998;95(24):14447–14452.
68. Qin S, Rottman JB, Myers P, et  al. The chemokine receptors
CXCR3 and CCR5 mark subsets of T cells associated with certain
inflammatory reactions. J Clin Invest. February 15, 1998;101(4):
746–754.
69. Goronzy JJ, Bartz-Bazzanella P, Hu W, Jendro MC, Walser-Kuntz
DR, Weyand CM. Dominant clonotypes in the repertoire of periph-
eral CD4+ T cells in rheumatoid arthritis. J Clin Invest. November
1994;94(5):2068–2076.
70. Waase I, Kayser C, Carlson PJ, Goronzy JJ, Weyand CM.
Oligoclonal T cell proliferation in patients with rheumatoid arthri-
tis and their unaffected siblings. Arthritis Rheum. June 1996;39(6):
904–913.
71. Vallejo AN, Nestel AR, Schirmer M, Weyand CM, Goronzy JJ.
Aging-related deficiency of CD28 expression in CD4+ T cells is
associated with the loss of gene-specific nuclear factor binding activ-
ity. J Biol Chem. April 3, 1998;273(14):8119–8129.
72. Goronzy JJ, Matteson EL, Fulbright JW, et al. Prognostic markers
of radiographic progression in early rheumatoid arthritis. Arthritis
Rheum. January 2004;50(1):43–54.
73. Eastgate JA, Symons JA, Wood NC, Grinlinton FM, di Giovine
FS, Duff GW. Correlation of plasma interleukin 1 levels with
4 6 0   •   G enomics in C l inica l P ractice
disease activity in rheumatoid arthritis. Lancet. September 24,
1988;2(8613):706–709.
74. Neidel J, Schulze M, Lindschau J. Association between degree of
bone-erosion and synovial fluid-levels of tumor necrosis factor alpha
in the knee-joints of patients with rheumatoid arthritis. Inflamm
Res. May 1995;44(5):217–221.
75. Bromley M, Woolley DE. Histopathology of the rheumatoid lesion.
Identification of cell types at sites of cartilage erosion. Arthritis
Rheum. August 1984;27(8):857–863.
76. Gravallese EM, Kantrowitz FG. Arthritic manifestations of
inflammatory bowel disease. Am J Gastroenterol. July 1988;83(7):
703–709.
77. Lacey DL, Timms E, Tan HL, et  al. Osteoprotegerin ligand is a
cytokine that regulates osteoclast differentiation and activation.
Cell. 1998 April 17;93(2):165–176.
78. Kong YY, Feige U, Sarosi I, et al. Activated T cells regulate bone loss
and joint destruction in adjuvant arthritis through osteoprotegerin
ligand. Nature. November 18, 1999;402(6759):304–309.
79. Pettit AR, Ji H, von SD, et al. TRANCE/RANKL knockout mice
are protected from bone erosion in a serum transfer model of arthri-
tis. Am J Pathol. November 2001;159(5):1689–1699.
80. Deleuran BW, Chu CQ, Field M, et al. Localization of interleukin-1
alpha, type 1 interleukin-1 receptor and interleukin-1 receptor
antagonist in the synovial membrane and cartilage/pannus junction
in rheumatoid arthritis. Br J Rheumatol. December 1992;31(12):
801–809.
81. Deleuran BW, Chu CQ, Field M, et al. Localization of tumor necro-
sis factor receptors in the synovial tissue and cartilage-pannus junc-
tion in patients with rheumatoid arthritis. Implications for local
actions of tumor necrosis factor alpha. Arthritis Rheum. October
1992;35(10):1170–1178.
82. Callahan LF, Pincus T, Huston JW, III, Brooks RH, Nance EP,
Jr., Kaye JJ. Measures of activity and damage in rheumatoid arthri-
tis: depiction of changes and prediction of mortality over five years.
Arthritis Care Res. December 1997;10(6):381–394.
83. McQueen FM, Stewart N, Crabbe J, et  al. Magnetic resonance
imaging of the wrist in early rheumatoid arthritis reveals a high prev-
alence of erosions at four months after symptom onset. Ann Rheum
Dis. June 1998;57(6):350–356.
84. Calguneri M, Pay S, Caliskaner Z, et  al. Combination therapy
versus monotherapy for the treatment of patients with rheu-
matoid arthritis. Clin Exp Rheumatol. November 1999;17(6):
699–704.
85. Hochberg MC. Early aggressive DMARD therapy: the key to slow-
ing disease progression in rheumatoid arthritis. Scand J Rheumatol
Suppl. 1999;112:3–7.
86. Rich E, Moreland LW, Alarcon GS. Paucity of radiographic pro-
gression in rheumatoid arthritis treated with methotrexate as the
first disease modifying antirheumatic drug. J Rheumatol. February
1999;26(2):259–261.
87. Smolen JS, Kalden JR, Scott DL, et  al. Efficacy and safety of
leflunomide compared with placebo and sulphasalazine in active
rheumatoid arthritis:  a double-blind, randomised, multicentre
trial. European Leflunomide Study Group. Lancet. January 23,
1999;353(9149):259–266.
88. Sharp JT, Strand V, Leung H, Hurley F, Loew-Friedrich I. Treatment
with leflunomide slows radiographic progression of rheumatoid
arthritis:  results from three randomized controlled trials of leflu-
nomide in patients with active rheumatoid arthritis. Leflunomide
Rheumatoid Arthritis Investigators Group. Arthritis Rheum. March
2000;43(3):495–505.
89. Nagashima M, Matsuoka T, Saitoh K, Koyama T, Kikuchi
O, Yoshino S. Treatment continuation rate in relation to effi-
cacy and toxicity in long-term therapy with low-dose metho-
trexate, sulfasalazine, and bucillamine in 1,358 Japanese
patients with rheumatoid arthritis. Clin Exp Rheumatol. May
2006;24(3):260–267.
 90. Scott DL. Biologics-based therapy for the treatment of rheumatoid
arthritis. Clin Pharmacol Ther. January 2012;91(1):30–43.
91. Wilkie WS, Schwieterman P. Strategies for the management of
rheumatoid arthritis. Orthopedics. February 2012;35(2): 125–130.
92. Moreland LW, Baumgartner SW, Schiff MH, et  al. Treatment of
rheumatoid arthritis with a recombinant human tumor necrosis
factor receptor (p75)-Fc fusion protein. N Engl J Med. July 17,
1997;337(3):141–147.
93. Hannum CH, Wilcox CJ, Arend WP, et  al. Interleukin-1 recep-
tor antagonist activity of a human interleukin-1 inhibitor. Nature.
January 25, 1990;343(6256):336–340.
94. Edwards JC, Cambridge G. Sustained improvement in rheumatoid
arthritis following a protocol designed to deplete B lymphocytes.
Rheumatology (Oxford). February 2001;40(2):205–211.
95. Leandro MJ, Edwards JC, Cambridge G. Clinical outcome in 22
patients with rheumatoid arthritis treated with B lymphocyte deple-
tion. Ann Rheum Dis. October 2002;61(10):883–888.
96. Edwards JC, Szczepanski L, Szechinski J, et  al. Efficacy of
B-cell-targeted therapy with rituximab in patients with rheuma-
toid arthritis. N Engl J Med. June 17, 2004;350(25): 2572–2581.
G enetics and G enomics o f C hronic I n f l ammatory D isorders , I I   •   4 6 1
97. Maini RN, Taylor PC, Szechinski J, et al. Double-blind random-
ized controlled clinical trial of the interleukin-6 receptor antago-
nist, tocilizumab, in European patients with rheumatoid arthritis
who had an incomplete response to methotrexate. Arthritis Rheum.
September 2006;54(9):2817–2829.
98. Westhovens R, Cole JC, Li T, et al. Improved health-related qual-
ity of life for rheumatoid arthritis patients treated with abata-
cept who have inadequate response to anti-TNF therapy in a
double-blind, placebo-controlled, multicentre randomized clinical
trial. Rheumatology (Oxford). October 2006;45(10):1238–1246.
99. Sacre SM, Andreakos E, Taylor P, Feldmann M, Foxwell BM.
Molecular therapeutic targets in rheumatoid arthritis. Expert Rev
Mol Med. August 24, 2005;7(16):1–20.
100. Gomez-Puerta JA, Mocsa A. Tyrosine kinase inhibitors for
the treat­ment of rheumatoid arthritis. Curr Top Med Chem.
2013;13(6):760–773.
101. Tan RJ, Gibbons LJ, Potter C, et al. Investigation of rheumatoid
arthritis susceptibility genes identifies association of AFF3 and
CD226 variants with response to anti-tumour necrosis factor treat-
ment. Ann Rheum Dis. June 2010;69(6):1029–1035.
 30.
GENETICS AND GENOMICS OF CHRONIC
INFLAMMATORY DISORDER S, III: BRONCHIAL ASTHMA
William Cookson
INTRODUCTION
Bronchial asthma is characterized by an abnormal mucosa
of the airways, inflammation, and symptoms of wheezing
and shortness of breath. Asthma affects more than one child
in 10 in the developed world, and there are 300  million
cases worldwide.1 Although some effective therapies for
mild asthma exist, severe asthma remains difficult to treat.
The societal cost of the disease is substantial in Westernized
societies,2,3 and it is likely to become overwhelmingly
important in the developing world as asthma rises in preva-
lence there.
Asthma should be considered a syndrome rather than
a disease. Childhood-onset disease has a strong association
with atopy (diagnosed by positive skin tests, or elevations of
allergen-specific and total serum immunoglobulin E [IgE]
concentrations) and is more common in males. Adult-onset
disease typically comes on after 40 years of age, is more com-
mon in women, is not associated with atopy beyond the
general population prevalence of the condition, is at times
associated with cigarette smoking, and is often resistant to
therapy.
Occupational asthma is attributable to, or is made
worse by, environmental exposures in the workplace and
may be present in 10–15% of cases of adult disease.4 It
serves as an exemplar for dose–response relationships in
allergen-induced disease (for example, in bakers) and dis-
ease induced by small molecules (such as isocyanate paint
additives). Other clinical variants of asthma may include
severe disease, therapy-resistant asthma, brittle asthma, and
asthma that results in fixed airway diseases.
Asthma runs strongly in families and has a heritability
(variance attributable to genetics) of up to 60%.5 Genetic
studies therefore offer a structured means of understand-
ing the causes of asthma as well as identifying targets for
treating the syndrome. As with other common complex
diseases, large-scale genetic studies have led to considerable
462
advances in the understanding of asthma. In general, robust
genetic effects have been identified that carry substantial
population-attributable risks. Several genes act in pathways
that communicate the presence of epithelial damage to the
adaptive immune system, providing a new focus for effec-
tive therapies. Genetic findings have also led to a reassess-
ment of the primacy of atopy in both asthma and eczema
(atopic dermatitis), suggesting that atopy is secondary to
epithelial or other events in both diseases.
Despite these advances, only a small component of the
overall genetic contribution to asthma has been identi-
fied. This missing heritability may be due to multiple small
effects (polygenes) or rare, highly penetrant mutations, or
epigenetic modifications of gene function.
Asthma was a rare disorder in the nineteenth century,
and a sharp increase in the prevalence of asthma has been
observed in many countries during the last decades of the
twentieth century. Phase I  of the International Study of
Asthma and Allergies in Childhood (ISAAC I) found that
the prevalence of symptoms of asthma in children differed as
much as 20-fold between study centers around the world.1
It has been established with clarity that a rich microbial
environment in childhood confers significant protection
from the development of asthma, hay fever, and atopy,6–8
particularly in the context of farming environments across
Europe.8 Recent studies in Poland indicate that the preva-
lence of asthma and atopy are profoundly altered by a rapid
movement of the population from traditional rural to
urban dwelling following Poland’s entry into the European
Community.9
These findings indicate the presence of strong environ-
mental factors underlying disease susceptibility, and suggest
that asthma may be prevented by discovery of these factors.
The search for the exact microbial modifiers is becoming
progressively refined.10,11
This review will describe in detail the progress made by
genetic studies of asthma and will outline what are likely to
be the next frontiers in genetic and epigenetic investigations.
I will conjecture about the implications of specific loci for
therapeutic intervention and interpret the overall findings
in the context of investigations of the airway microbiome.
G E N ET I C S
G E N O M E -WI D E A S S O C I AT I O N S T U D I E S
Large-Scale Studies
Adequately powered genome-wide association studies
(GWAS), in which hundreds of thousands of genetic mark-
ers are genotyped in thousands of cases and controls, have
been the method of choice for identifying complex dis-
ease genes.12 Several of these have now been completed for
asthma, with gratifying and consistent results.
A first-generation GWAS of approximately 1000
children with asthma and 1000 controls showed in 2007
that markers on chromosome 17q21 were strongly asso-
ciated with childhood-onset asthma.13,14 The GWAS
was extended by measuring global gene expression in
Epstein-Barr virus–transformed lymphoblastoid cell
lines (LCL) from asthmatic children and their siblings.15
This showed that the asthma-associated SNPs were also
strongly associated (P = 10–23) with transcript abundance
of ORMDL3.13 Subsequently, we and others have shown
that the locus also regulates expression of the neighboring
gene, GSDMB.16,17
A multidisciplinary study to identify the genetic and
environmental causes of asthma in the European commu-
nity (GABRIEL collaboration) was a large international
initiative that completed a second-generation GWAS of
10,000 asthmatics and 16,000 controls, including those
with adult-onset asthma as well as childhood disease.18
Associations were found on chromosome 2 within IL1RL1/
IL18R1; on chromosome 6 within HLA-DQ; on chromo-
some 9 flanking IL33; on chromosome 15 within SMAD3,
and on chromosome 22 within IL2RB. Association within
the chromosome 17q21 ORMDL3/GSDMB locus was
confined to childhood-onset disease, and the HLA-DQ
locus was the only significant hit in the adult-onset group.
Markers at two additional genes, SLC22A5 and RORA,
were identified just below genome-wide significance in the
whole group, and markers within TSLP were seen at a simi-
lar level of significance in the patients with the most severe
disease. Associations within IL1RL1 and IL33 had been
identified previously, through the intermediate phenotype
of elevated eosinophil counts,19 and TSLP variants had
G e n etic s a nd G e no m ic s o f C h ronic I n f l a m m atory D i s ord e r s , I I I   •   4 6 3
also been previously associated with include SNPs within
PDE4D20 and DENND1B.21
The study of populations with diverse ancestry and
environmental exposures adds depth to genetic studies.
A  GWAS of pediatric asthma in the Japanese population
showed pediatric asthma to be associated primarily to the
major histocompatibility complex (MHC), with asso-
ciation to HLA-DP alleles DPA1*0201 and DPB1*0901,
which are in strong linkage-disequilibrium with each other
were strongly associated with pediatric asthma.22 This study
found no effects from the ORMDL3 locus, although asso-
ciation to this locus had been confirmed in other Japanese
subjects (Hirota PMID 18155279). A study of asthma in
Japanese adults showed the strongest associations to the
MHC, between the complement genes and the HLA-DRA
locus, and confirmed association to the TSLP locus.23 These
results suggest a differing balance of etiologies to asthma in
Japan and Europe. The HLA-DP association may indicate
the presence of an antigen driving the process, and it may
be relevant that HLA-DP associations to aspirin-induced24
and isocyanate asthma25 have been observed previously,
both involving small molecules.
A well-powered meta-analysis of GWAS in ethnically
diverse North American populations from the EVE consor-
tium (meta-analysis of genome-wide association studies of
asthma in ethnically diverse North American populations)
confirmed the previously reported loci near ORMDL3,
IL1RL1, TSLP, and IL33, and showed that these loci were
associated with asthma risk in Hispanic and African-American
groups. The authors also identified association to PYHIN1,
with the association being specific to individuals of African
descent.26 This result has yet to be confirmed.
More recently, an Australian study that combined new
data with published results showed new associations to the
interleukin-6 receptor (IL6R) gene and markers on chro-
mosome 11q13.5 near the leucine-rich repeat containing 32
gene (LRRC32, also known as GARP). The LRRC32 locus
was significantly associated with atopic status among asth-
matics.27 It had first been found to be a susceptibility locus
for atopic dermatitis28 and was shown in the ALSPAC pop-
ulation to be a novel susceptibility factor for atopic asthma
and hay fever.29
The asthma genetics community is currently consoli-
dating all the data from these international GWAS, so the
number of asthmatics included in a single meta-analysis
will rise from 10,000 to 25,000. The experience with other
­diseases30 indicates that this exercise may double the num-
ber of verified asthma-susceptibility loci. It is likely, how-
ever, that the strongest and most consistent genetic effects
on asthma have already been identified.
 The summary from all of these studies is that ORMDL3
single biggest risk factor for childhood asthma; that
childhood-onset asthma consistently shows the strongest
genetic effects18; that IL33, IL33R/IL18R, and TSLP com-
municate epithelial damage to the adaptive immune system;
that SMAD3 and 1L2RB downregulate inflammation; and
that the MHC has diverse effects that may center on the
presentation of specific foreign antigens.
Heterogeneity and Atopy
These studies point to a heterogeneity underlying asthma
that has long been suspected by clinicians. The ORMDL3
locus is confined to childhood-onset disease and is associ-
ated with severity31 but not with atopy. The MHC is the
strongest risk factor for adult-onset disease and shows
different allelic relationships to disease in different popu-
lations and age groups. The LRRC32 locus may have its
strongest effects in atopic asthma in conjunction with
atopic dermatitis.
GWAS studies for total serum IgE levels have been suc-
cessfully carried out, identifying effects around FCER1A,
STAT6, and IL13.18,28,32,33 With the exception of the IL13
locus, it is notable that the genes that regulate IgE produc-
tion have a minimal effect on asthma susceptibility, and vice
versa.18 Whilst these results do not negate the importance of
atopic mechanisms in asthma, they do indicate that atopy is
secondary to asthma, rather than its driving force. Similarly,
disorders in the epidermal barrier proteins SPINK534 and
FLG35 lead to strong atopic manifestations that must be sec-
ondary to inflammatory events that follow barrier failure.
Overlap of GWAS Hits with Inflammatory
Bowel Disease
Several of the robust asthma associations are also found in
studies of inflammatory bowel disease (IBD).36 Clinically,
IBD is well recognized to be accompanied by low-grade
airway inflammation. Loci that contribute to both diseases
include SMAD3, ORMDL3, HLA-DR, and DENNDB1.36
The overlap suggests that both diseases may result from the
disordered interactions between bacteria and mucosal sur-
faces. The functions of these genes and the potential role of
the airway microbiome are described below. Notably, genes
that identify defective processing of intracellular bacteria
in Crohn’s disease (such as NOD2, IRGM, and ATG16L1)
are not implicated in asthma. It is also of interest that
genetic evidence has demonstrated the importance of the
barrier function to the development of ulcerative colitis,
although the genes involved (HNF4A, LAMB1, CDH1,
4 6 4   •   G e no m ic s in C l inic a l P r actic e
and GNA12) are different from those that mediate the bar-
rier function in asthma and atopic dermatitis.36
F U N C T I O NS O F A S T H M A-A S S O C I AT E D
GENES
ORMDL3 and GSDMB
The association signals on human chromosome 17
with asthma are maximal within an island of linkage-
disequilibrium that contains ORMDL3 and GSDMB.
Asthma-associated SNPs were strongly associated with
transcript abundance of ORMDL3 and more weakly with
the neighboring GSDMB, so it is feasible that the locus
contributes to asthma through the differential regulation of
both genes. A single promoter upstream of ORMDL3 is a
promising target region to search for functional SNPs that
affect this regulation.
The locus covers an area of approximately 200 kb.
ORMDL3 and GSDMB reside in one island of linkage dis-
equilibrium that contains all the maximally associated SNPs.
However, statistically independent associations are detectable
telomerically near the GSDMA and PSDM3 genes,13 which
may make additional contributions to asthma susceptibility.
ORMDL3 protein is found in the membranes of the
endoplasmic reticulum (ER). ER stress is linked to cellular
responses to inflammation.37 Enforced X-box-binding pro-
tein 1 (XBP1S) is a transcriptional activator that is a key
element in the development of the ER. ORMDL3 has been
found to be upregulated in XBP-1(S) transuded NIH-3T3
fibroblasts.38 It has been shown that changes in ORM gene
expression cause dysregulation of sphingolipid metabo-
lism.39 Sphingolipids are amphiphatic molecules formally
derived from sphingosine. Sphingosine phosphorylation
leads to sphingosine-1-phosphate (S1P) and acylation to
ceramide. Sphingolipids mediate cell survival, proliferation,
apoptosis, differentiation, and cell-cycle arrest.40 Clinical
observation shows that sphingosines and ceramide are
increased in asthmatic airways.41
It has also been suggested that ORMDL3 modifies
SERCA in the ER resulting in an unfolded-protein response
and inducing inflammation.42 A  recent study showed
mononuclear cells significantly increased IL-17 secretion in
17q21 risk allele–carrier children.43 All these results show
that ORMDL3 may influence multiple pathways in the ER
that mediate inflammation during asthma.
The GSDM family of genes have been best studied in
mice. They are expressed predominantly in the gastrointes-
tinal (GI) tract and in the skin44 in a highly tissue-specific
manner.45 In humans, GSDMA and GSDMB are expressed
in the gastrointestinal and bronchial epithelium (http://
www.proteinatlas.org/). Members of the gene family may
have a role in the regulation of apoptosis.46
IL33, IL18R1, and IL1RL1
IL33, IL18, and IL1 belong to the IL1 family of cytokines
that alter host responses to inflammatory and infectious
challenges. They exert their functions through a family of
receptors that belong to the Toll-like receptor-IL-1 recep-
tor (TLR-IL-1R) superfamily. An important feature of IL1
receptor signaling is the activation of transcription factor
NF-κB and mitogen-activated protein (MAP) kinases p38,
JNK, and ERK1/2.47
IL33 was originally identified as a nuclear factor in vas-
cular endothelial cells,48 and was subsequently detected in
airway epithelial cells.49,50 The activities of IL33 as a nuclear
factor remain unclear,51 but it may repress gene expression
of pro-inflammatory factors. In contrast to inducible cyto-
kines, IL33 is constitutively expressed and is thought to
function as an endogenous danger signal or alarmin to alert
the immune system after endothelial or epithelial cell dam-
age during trauma or infection.52 IL33 is formed as a prodo-
main containing polypeptide, which after activation with
caspase-1, is realized as mature IL33. A mouse gene knock-
out has shown Il33 works as a crucial amplifier of innate
immunity.53 IL-33 expression is induced by a range of envi-
ronmental and endogenous triggers, suggesting an essential
role during infection, inflammation, and tissue damage.54
IL33 activates a heterodimeric receptor complex con-
taining IL1RL1 (ST2) and IL-1 receptor accessory pro-
tein (IL1RAP), leading to activation of NF-κB and MAP
kinases and drives production of Th2-associated cytokines
such as IL4, IL5, and IL13.49
The IL18R1 gene, along with four other members of
the interleukin 1 receptor family (IL1R2, IL1R1, ILRL2
[IL-1Rrp2], and IL1RL1 [T1/ST2]), form a cluster on
chromosome 2q. IL18R1 and IL1RL1 flank each other
with the same orientation of translation. They are within
the same island of linkage disequilibrium, and it has not yet
been possible to assign the genetic effects at this locus to
one gene or the other. It is possible that both genes may be
co-regulated.
As stated above, IL1RLI encodes the receptor of IL33.
IL18 is closely related to IL3349 and synergizes with IL12 to
induce the production of interferon gamma and to promote
TH1 responses.55 These loci therefore identify a pathway for
the communication of epithelial damage to the adaptive
immune system and a potential switch point for choosing
between TH1 or TH2 responses.
G e n etic s a nd G e no m ic s o f C h ronic I n f l a m m atory D i s ord e r s , I I I   •   4 6 5
Thymic stromal lymphopoietin (TSLP)
Despite the initial identification of TSLP in the culture
supernatant of a thymic stromal cell line, this cytokine is
expressed mainly by epithelial cells at barrier surfaces (skin,
gut and lung).56,57 The cell populations with the highest
known co-expression of the TSLP receptor (TSLPR) and
its associated subunit IL-7Rα are myeloid dendritic cells
(DCs).56 Treatment of human DCs with TSLP induces
improved survival, upregulation of major histocompatibil-
ity complex class II and co-stimulatory molecules, and the
production of a variety of chemokines.56 It promotes TH2
cytokine–associated inflammation by directly promot-
ing the effector functions of CD4+ TH2 cells, basophils,
and other granulocyte populations while simultaneously
limiting the expression of DC-derived proinflammatory
cytokines and promoting regulatory T-cell responses in
peripheral tissues.57
SMAD3
SMAD3 encodes SMAD (“mothers against decapentaple-
gic homolog”) family member 3 and came to attention
for its role in modifying tumour growth57,58,59 through
the transforming growth factor-beta (TGFB) pathway.60
SMAD3 is concentrated in the nuclei of bronchial epithe-
lial cells and macrophages (http://www.proteinatlas.org/)
and functions as a transcriptional modulator activated by
TGFB. TGFB family members exert fundamental effects
on the maintenance of immune function in the lung,61
and TGFB signaling pathways are activated after allergen
challenge in mild asthma.62 A  mouse knockout of Smad3
showed accelerated wound healing and an impaired local
inflammatory response,63 even though mice lacking Smad3
may exhibit increased baseline levels of proinflammatory
cytokines in their lungs.64 Smad3 signaling is required for
myogenic differentiation of myoblasts,65 perhaps pointing
to a role in airway smooth muscle hypertrophy.
IL2RB
IL12RB encodes the beta receptor of IL2. IL2 is secreted by
antigen-activated T cells. It controls the survival and prolif-
eration of regulatory T cells66 and plays a crucial role in the
maintenance of natural immunological self-tolerance.67 The
IL2 receptor is composed of α (CD25), β (CD122), and γ
chains.66 The β chain (IL2RB) is a signal transduction ele-
ment that is also present in the IL15 receptor. It belongs to
the type I cytokine receptor family and is devoid of intrinsic
kinase activity.68 The receptor modulates T cell–mediated
immune responses through endocytosis, whereby ectodo-
main shedding of IL2Rβ generates an intracellular fragment
with biological functions.69 In a murine model of asthma,
local blockade of Il2rb restored an immunosuppressive
cytokine milieu that ameliorated lung inflammation.70
IL4 and IL13
IL4 and IL13 are adjacent to each other on chromo-
some 5 and downstream of RAD50. The locus is excep-
tional in showing strong association to IgE in addition to
doctor-diagnosed asthma. The 3′ end of RAD50 contains
several conserved non-coding sequences and enhancer ele-
ments that act as a locus-control region for IL4 and IL13.71
RAD50 encodes a gene that mediates DNA double-stranded
break repair, and it is most likely that the effects at this locus
are mediated through IL4 and IL4 expression.
The Major Histocompatibility Complex
The MHC has long been a focus of interest in understand-
ing the genetics of asthma, and it is one of few candidate
regions that have survived the stringent criteria imposed
by GWAS. As described above, most MHC associations to
asthma and related phenotypes have emerged around the
HLA class  II genes, with apparently independent effects
for adult-onset asthma in Europeans,18 asthma in Japanese
adults,23 asthma in Japanese children,22 and total serum IgE
levels in Europeans.18 Previously, HLA-DQ and HLA-DR
alleles have been associated with IgE responses to individ-
ual allergens,72–74 and TNF alleles in the same region have
been related to the presence of asthma and its response to
therapy.75 Intriguingly, polymorphisms within the MHC
are strongly related to the levels of gene expression of many
genes, including the classical HLA antigens.15
The HLA class  II proteins classically restrict immune
responses to particular external antigens, and it is possible
that the GWAS findings are pointing to different types
of antigen, perhaps bacterial (see below) as well as typical
allergens. A  concerted collaborative international effort is
needed to disentangle these observations.
P H A R M AC O G E N ET I C S A N D A S T H M A
Pharmacogenetic studies may define variants that modify
an individual patient’s response to therapy. Small or moder-
ate beneficial effects will require studies that are as large as
current GWAS for disease susceptibility. GWAS of thera-
peutic response have to date been most successful in show-
ing side effects of drugs to result from variation in their
4 6 6   •   G e no m ic s in C l inic a l P r actic e
metabolism.76 A recent study of the response of asthmatics
to glucocorticoid therapy has shown that patients who were
homozygous for an allele mapped to glucocorticoid-induced
transcript 1 gene (GLCCI1) had an improvement in respi-
ratory function that was only about one-third of that seen
in similarly treated subjects who were homozygous for the
wild-type allele (3.2 ± 1.6% vs. 9.4 ± 1.1%). The same sub-
jects’ risk of a poor response was significantly higher, with
genotype accounting for about 6.6% of overall inhaled glu-
cocorticoid response variability.77 These associations were
supported by functional studies that showed the risk allele
to be associated with decrements in GLCCI1 expression.
These intriguing results are worthy of replication.
E P I G E N ET I C S
Epigenetic effects, mediated through mechanisms other
than sequence variation, are another possible cause of famil-
ial clustering. The patterns of gene expression that deter-
mine cellular type and function become stably restricted
during development, partly through methylation of CpG
sequences and gene silencing.78 Islands of CpG sequences
are positioned at the 5′ UTRs of many human genes.79
About one-fifth of islands are variably methylated, and in
one-third of these, methylation status correlates with tran-
script abundance.80 Abnormalities of DNA methylation
are well recognized in single-gene disorders and in can-
cer,81 and it is postulated that epigenetic changes in meth-
ylation may be of importance to common human diseases81
such as asthma and chronic obstructive pulmonary disease
(COPD).
Whilst it is very likely that genome-wide studies of
methylation status at various loci may identify new genes
and pathways that mediate airway diseases, it is impor-
tant to recognize that age,80,82,83 sex,80,82 genetic polymor-
phism,84,85 and environmental factors83,85 have all been
strongly associated with altered methylation at selected loci.
The relative contribution of these factors to methylation
at loci genome-wide is not known, and it is not certain to
what extent true epigenetic inheritance with transmission
between generations takes place. These factors will have to
be taken into account if methylation changes at individual
loci are to help us understand complex diseases.
T HE M I C R O B I O ME
Asthma is due to complex interactions between genes and
the environment. Whilst the genetic studies described
above have shown the central importance of the respiratory
epithelium, a worthwhile understanding of the causes of
asthma needs to reconcile consistent epidemiological indi-
cations of the importance of the microbiome to the disease.
These include the protection afforded by a rich microbial
environment in early life,10,86 observations that the bron-
chial tree contains characteristic flora that are disturbed by
the presence of pathogens such as Haemophilus infuenzae in
asthma,11,87 birth cohort studies showing that the presence
of the same pathogens in throat swabs predicts the later
development of asthma,88 and recognition that these bac-
teria have consistently been associated with exacerbations
of asthma.89
Ninety percent of the cells in the human body are
microorganisms including bacteria, parasites, and archaea.90
These microorganisms are commensal on body surfaces
exposed to the external environment, including the gut,
respiratory tract, and skin. Although most bacteria are not
cultivable with standard methods,91 the membership of
complex microbial communities can be quantified and clas-
sified by DNA sequencing of the conserved bacterial 16S
rRNA gene.92,93 Bacteria are classified by these sequences
into operational taxonomic units (OTUs). OTUs approxi-
mate closely, but not completely, to taxonomy derived from
classical techniques, but sequences of other regions may be
necessary for precise discrimination at the species level.
A small study from our group compared the airway
microbiota at three levels in adult patients with asthma,
the related condition of COPD, and controls.11 Bronchial
lavage from asthmatic children and controls was also inves-
tigated.11 The bronchial tree was not sterile, and contained
a mean of 2,000 bacterial genomes per square centimeter
surface sampled. Pathogenic proteobacteria, particularly
Haemophilus spp., were much more frequent in bronchi of
adult asthmatics or patients with COPD than in controls.
Highly significant increases in proteobacteria were also seen
in asthmatic children.
A recent study used a microarray technique to test for
relationships between the composition of the airway bacte-
rial microbiota and clinical features of asthma, establishing
that bacterial burden and diversity were higher among asth-
matic patients compared to controls, although not identify-
ing the organisms responsible for this effect.87
Of potential importance is the finding that organisms
commonly found in healthy airways (such as Bacteroidetes,
particularly Prevotella spp.) were significantly reduced in
asthmatic airways.11 As described above, genetic studies
indicate an overlap in mechanisms underlying asthma and
IBD. It is increasingly recognized that a normal bacterial
flora is essential in maintaining a healthy bowel mucosa,94
and similar mechanisms are likely to be important in the
G e n etic s a nd G e no m ic s o f C h ronic I n f l a m m atory D i s ord e r s , I I I   •   4 6 7
airways.95–97 Murine studies have shown that sterility of the
airways and intestinal tract results in markedly enhanced
inflammatory responses to a variety of stimuli.96,98
The study of the role of the microbiome of asthma is in
its infancy. Even at this early stage, however, there are many
questions that suggest a structured set of experiments that
will elucidate how the microbiome operates in health and
disease.
C L I N I C AL I M P L I C AT I O N S
G E N ET I C R I S K
The primary motivation for genetic investigations of
asthma has been to develop a systematic understanding
of otherwise impenetrable disease processes. The success
of GWAS for complex diseases and the rapidly declining
costs of sequencing individual human genomes have led to
a high level of public interest in using DNA to provide an
estimated individual risk of disease, with the expectation
that high-risk individuals may benefit from preventive and
therapeutic interventions.99 It is appropriate that the value
of individual genetic screening has been questioned on the
grounds of interpretation of statistical risk estimates,99 the
quality control of genotyping, and the need to establish
the validity of predictive tests from appropriately designed
trials.100
The outstanding finding from asthma GWAS has been
the GSDMB-ORMDL3 locus on chromosome 17q21.13,18
In the initial GWAS,13 the odds ratio (OR) for the most
strongly associated SNP in cases of childhood asthma
recruited through hospital clinics and compared to con-
trols was 1.84 per additional risk allele. The per-allele OR
at this locus for childhood-onset asthma was 1.32 in the
GABRIEL consortium GWAS, which contained a high
proportion of asthmatics from epidemiological surveys,18
and was 2.02 in severe therapy-resistant asthmatics attend-
ing tertiary referral clinics.101 These substantial differences
in ORs reflect the different ascertainment criteria and types
of asthma for each study, and show that genetic risk is high-
est in children with early-onset disease that is resistant to
therapy, and that environmental factors are important mod-
ifiers of this risk.
The size of these effects can be put into perspective
by considering well-known epidemiological effects on
asthma. The OR for the protective effect of farming on the
prevalence of childhood asthma in the PARSIFAL study
(Prevention of allergy risk factors for sensitization in chil-
dren related to farming and anthroposophic lifestyle) was
0.62 (inverse 1.61) and was 0.86 (1.16) in the GABRIEL
advanced survey.10 At school age, the typical ORs associ-
ated with either parent’s smoking were 1.24 for wheezing
in the last year and 1.21 for asthma.102 Although these fig-
ures are derived from other populations, genetic risks seem
comparable to those conferred by known environmental
factors.
Genetic Risks in Populations
The population-attributable risk (PAR; also known as the
population-attributable risk fraction, PARF) is widely used
to estimate the burden of a risk factor in population surveys.
Applied to genetic studies, the PARF estimates the fraction
of cases that would not occur if no one in the population
carried the risk allele.99 Usage of the PARF to model how
genetic variants at the six GABRIEL GWAS significant
loci impacted jointly on the burden of childhood disease
in the GABRIEL populations18 indicated that 38% of
childhood-onset asthma was attributable to SNP combina-
tions at these loci, and in an independent population-based
replication sample, the same SNPs accounted for 49% of
asthma risk from birth to middle age.18 These findings indi-
cated that the genetic variants at these loci and their pre-
sumed functional outcomes had a significant impact on the
burden of asthma in the community. A  large PARF does
not imply that the measured genotypes can provide clini-
cally useful risk predictions,99 but it does suggest that thera-
pies directed at the functional consequences of the genes
identified by the GWAS are likely to be of benefit to many
asthmatics.
Genetic Risk in Individuals
The GABRIEL GWAS is the largest set of data analyzed
to date for the ability of genetic variants to determine the
risk of an individual developing asthma.18 We assessed
individual risk for the seven SNPs associated with child-
hood asthma in a classification analysis based on a logisti-
cal regression model. Taking as a cut-point the risk score
exceeded by 25% of the non-asthmatic population (speci-
ficity, 75%), the sensitivity of classifying asthma was just
35% (false-negative rate, 65%). Assuming the prevalence of
asthma in the general population is between 5% and 10%,
the positive predictive value of a genetic risk score above
this arbitrary cut-point would be in the range from 6.9% to
13.5%. The clear conclusion from this is that the common
genetic polymorphisms measured or imputed by GWAS are
not useful in predicting which children or infants are at risk
of developing asthma.
4 6 8   •   G e no m ic s in C l inic a l P r actic e
Familial Risk and Heritability
A third component of genetic risk relates to families. The
increase in risk in the siblings of affected individuals is
known as the familial relative risk (FRR). A related mea-
surement is that of heritability, which is defined as the
variance in a complex trait due to inherited factors. The
heritability of childhood asthma has been estimated to
be as high as 60%, but the variance in asthma prevalence
accounted for by the loci in the GABRIEL large-scale
GWAS was only 4%. This is consistent with other com-
plex human traits such as height,103 Crohn’s disease,104 and
obesity,105 and the search for the “missing heritability” is
a major current focus for geneticists who study complex
diseases.106
Applying heritability estimates from rigorously ascer-
tained family studies to general population samples with
their attendant environmental variation is inherently
unreliable. Additionally, GWAS are directed at SNPs with
minor allele frequencies greater than 5% or 10%, because
the power to detect genetic associations is only high when
a disease and its underlying susceptibility alleles are both
common,107 and because parametric statistical tests of
association are unreliable when allele counts are sparse.
However, the clustering of asthma within families may be
due to private mutations that are rare in the general popu-
lation. The search for missing heritability, and for the abil-
ity to predict individual risks of disease, should perhaps
therefore be concentrated on rare mutations with high
penetrance.
T H E R A P EU T I C TA RG ET S
The increasing scale and level of international coopera-
tion have provided several clear directions for therapeu-
tic interventions in asthma. It is not safe to assume that
GWAS will identify targets that can be accessed either by
small molecules (drugs) or biologics (antibodies and pro-
teins). Of the top GWAS hits for asthma, it is perhaps
ORMDL3 that offers the most promise. ORM proteins
act as a brake on sphingolipid metabolism with increased
dihydrospingosine and ceramide levels observed in knock-
down cells.39 Sphingolipids have a wide range of actions in
modifying inflammatory processes, providing a potential
mechanism and a therapeutic target for modulating airway
inflammation.108
The IL33/IL18 ST2/IL18R axis may be crucial in deter-
mining the type of adaptive immune response to airway
damage and should be the focus of therapeutic endeavors.
Despite much effort, however, the exact roles of IL33 and its
receptor in inflammation have been difficult to define with
precision. IL33 knockouts have been difficult to develop,
the nuclear binding sites for IL33 are not known, there has
not yet been a clear demonstration of free IL33 in inflamed
airways, and it is not yet a definite target for therapy. Based
on the results of murine knockouts, IL18 and its receptor
IL18R may be more tractable therapeutic targets, but the
field requires the development of an appropriate model sys-
tem for IL18/IL18R interactions.
TSLP has a well-defined biology, and it represents a
good target for biologic therapies. It may be anticipated
that an anti-TSLP antibody would have its most profound
effects on dendritic cells and granulocytes.
The IL2RB and SMAD3 proteins may provide a means
of downregulating inflammation and promoting healing.
SMAD3 is, however, concentrated in the nucleus, and noth-
ing is yet known of its nuclear binding sites and partners.
IL2RB is expressed at the cell surface, and murine knock-
down studies indicate that biologics directed against it may
have therapeutic benefits.
M A N I P U L AT I N G T H E M I C RO B I O M E
The finding of a disordered microbiome in asthma inevi-
tably leads to direct manipulation of the airway bacterial
community. Clinical investigators have supported these
findings with direct interventions with long courses of anti-
biotics.109 A  conclusion from these studies is that chronic
bacterial infections are relevant in a subgroup of preschool
children with persistent wheezing, and such children ben-
efit significantly from antibiotic therapy.109 Persistent bacte-
rial bronchitis (PBB) is a syndrome of childhood wheezing
with a chronic productive cough that is often diagnosed as
asthma.110 It may be relevant that Haemophilus influenzae
and Streptococcus pneumoniae are the most commonly iso-
lated organisms, and that long courses of antibiotics are very
effective in treating the disease.110
Many patients with asthma receive antibiotics at dif-
ferent times without noticeable changes in the course of
their disease. Rational antibiotic therapy would probably
necessitate direct sampling of the airway mucosa with
brushings or lavage, together with assessment of the key
elements of the local inflammatory response. Antibiotics
have profound effects on microbial communities, and
restoration of a normal bacterial flora may be a neces-
sary part of the therapeutic approach. In this context, it
is remarkable that replacement of a normal bacterial com-
munity in the bowel has been very effective in patients
with IBD,111,112 and curative in patients with Clostridium
difficile necrotizing enteroclotitis.113,114
G e n etic s a nd G e no m ic s o f C h ronic I n f l a m m atory D i s ord e r s , I I I   •   4 6 9
C O N C LU S I O N S A N D F U T U R E
P E R S P E C T I VES
GWAS of asthma have provided a rich harvest of knowledge
about the mechanisms of asthma. Next steps in genetics will
include a comprehensive meta-analysis of all GWAS data
internationally, currently being undertaken by the applied
genetics core (TAGC). This exercise may double the num-
ber of loci proven to be influencing asthma.
The contribution of rare mutations to asthma is not
known, and full genomic sequencing of multiple diseased
individuals is the gold standard to be applied to this prob-
lem. Whilst the community waits for sequencing costs to
fall to acceptable levels before undertaking full sequenc-
ing, whole-exome sequencing with enrichment of exonic
sequences by hybridization before ultra-deep sequencing115
has proven to be extremely effective in the identification of
rare mutations in Mendelian diseases,116,117 and merits appli-
cation to complex disorders such as asthma. Another interim
solution may come from the Illumina exome chip, which is
derived from assembled information on 12,000 sequenced
genomes and exomes and catalogues, for each variant that
potentially affects protein structure, the total number of
times it was seen, and the total number of datasets that
included the variant. Non-synonymous variants seen at least
three times across two datasets are to be included on the chip.
It is unlikely that many important genetic effects will
remain undiscovered after completion of global meta-analysis
and identification of any important rare mutations. The
source of the missing asthma heritability is most likely to be
polygenic aggregation on multiple small effects. The large
number of loci identified through candidate gene studies118
that do not show up in GWAS may point to this eventuality.
Molecular genetic studies identify unexpected pathways
and potential targets on the basis of sequence variation in
DNA. The essential components of many disease-related
pathways may, however, not be subject to genetic varia-
tion. Epigenetic variation may be another means to identify
novel disease pathways and therapeutic targets.
Whatever the outcome of future genetic studies, the
biggest advances in understanding and treating asthma
will come from full functional analyses of the genes already
identified, from reagents that manipulate their actions, and
from disentangling their relationship to the microbial part-
ners on the other side of the respiratory mucosa.
R EFE R E N C ES
1. ISAAC. Worldwide variation in prevalence of symptoms of
asthma, allergic rhinoconjunctivitis, and atopic eczema:  ISAAC.
 The International Study of Asthma and Allergies in Childhood
(ISAAC) Steering Committee. Lancet 1998;351:1225–1232.
2. Smith DH, Malone DC, Lawson KA, Okamoto LJ, Battista C,
Saunders WB. A national estimate of the economic costs of asthma.
Am J Respir Crit Care Med 1997;156:787–793.
3. Asthma Agenda. London: National Asthma Campaign; 1998.
4. Vandenplas O, Toren K, Blanc PD. Health and socioeconomic
impact of work-related asthma. Eur Respir J 2003;22:689–697.
5. Duffy DL, Martin NG, Battistutta D, Hopper JL, Mathews JD.
Genetics of asthma and hay fever in Australian twins. Am Rev
Respir Dis 1990;142:1351–1358.
6. Strachan DP. Hay fever, hygiene, and household size. Brit Med J
1989;299:1259–1260.
7. von Mutius E, Fritzsch C, Weiland SK, Roll G, Magnussen
H. Prevalence of asthma and allergic disorders among chil-
dren in united Germany:  a descriptive comparison. Brit Med J
1992;305:1395–1399.
8. Braun-Fahrlander C, Gassner M, Grize L, et  al. Prevalence of hay
fever and allergic sensitization in farmer’s children and their peers
living in the same rural community. SCARPOL team: Swiss Study
on Childhood Allergy and Respiratory Symptoms with Respect to
Air Pollution. Clin Exp Allergy 1999;29:28–34.
9. Sozanska B, Macneill SJ, Kajderowicz-Kowalik M, et al. Atopy and
asthma in rural Poland: a paradigm for the emergence of childhood
respiratory allergies in Europe. Allergy 2007;62:394–400.
10. Ege MJ, Mayer M, Normand AC, et  al. Exposure to environ-
mental microorganisms and childhood asthma. N Engl J Med
2011;364:701–709.
11. Hilty M, Burke C, Pedro H, et al. Disordered microbial communi-
ties in asthmatic airways. PLoS One 2010;5:e8578.
12. Manolio TA. Genomewide association studies and assessment of the
risk of disease. N Engl J Med 2010;363:166–176.
13. Moffatt MF, Kabesch M, Liang L, et al. Genetic variants regulating
ORMDL3 expression contribute to the risk of childhood asthma.
Nature 2007;448:470–473.
14. Bouzigon E, Corda E, Aschard H, et  al. Effect of 17q21 vari-
ants and smoking exposure in early-onset asthma. N Engl J Med
2008;359:1985–1994.
15. Dixon AL, Liang L, Moffatt MF, et al. A genome-wide association
study of global gene expression. Nat Genet 2007;39:1202–1207.
16. Cookson W, Liang L, Abecasis G, Moffatt M, Lathrop M. Mapping
complex disease traits with global gene expression. Nat Rev Genet
2009;10:184–194.
17. Verlaan DJ, Berlivet S, Hunninghake GM, et al. Allele-specific chro-
matin remodeling in the ZPBP2/GSDMB/ORMDL3 locus associ-
ated with the risk of asthma and autoimmune disease. Am J Hum
Genet 2009;85:377–393.
18. Moffatt MF, Gut IG, Demenais F, et  al. A large-scale,
consortium-based genomewide association study of asthma. N Engl
J Med 2010;363:1211–1221.
19. Gudbjartsson DF, Bjornsdottir US, Halapi E, et al. Sequence vari-
ants affecting eosinophil numbers associate with asthma and myo-
cardial infarction. Nat Genet 2009;41:342–347.
20. Himes BE, Hunninghake GM, Baurley JW, et  al. Genome-wide
association analysis identifies PDE4D as an asthma-susceptibility
gene. Am J Hum Genet 2009;84:581–593.
21. Sleiman PM, Flory J, Imielinski M, et  al. Variants of
DENND1B associated with asthma in children. N Engl J Med
2010;362:36–44.
22. Noguchi E, Sakamoto H, Hirota T, et  al. Genome-wide associa-
tion study identifies HLA-DP as a susceptibility gene for pediatric
asthma in Asian populations. PLoS Genet 2011;7:e1002170.
23. Hirota T, Takahashi A, Kubo M, et  al. Genome-wide association
study identifies three new susceptibility loci for adult asthma in the
Japanese population. Nat Genet 2011;43:893–896.
24. Young RP, Dekker JW, Wordsworth BP, Cookson WOCM.
HLA-DR and HLA-DP genotypes and immunoglobulin
4 7 0   •   G e no m ic s in C l inic a l P r actic e
E responses to common major allergens. Clin Exp Allergy
1994;24:431–439.
25.
Choi JH, Lee KW, Kim CW, et  al. The HLA
DRB1*1501-DQB1*0602-DPB1*0501 haplotype is a risk factor
for toluene diisocyanate-induced occupational asthma. Int Arch
Allergy Imm 2009;150:156–163.
26. Torgerson DG, Ampleford EJ, Chiu GY, et  al. Meta-analysis of
genome-wide association studies of asthma in ethnically diverse
North American populations. Nat Genet 2011;43:887–892.
27. Ferreira MA, Matheson MC, Duffy DL, et  al. Identification of
IL6R and chromosome 11q13.5 as risk loci for asthma. Lancet
2011;378:1006–1014.
28. Esparza-Gordillo J, Weidinger S, Folster-Holst R, et al. A common
variant on chromosome 11q13 is associated with atopic dermatitis.
Nat Genet 2009;41:596–601.
29. Marenholz I, Bauerfeind A, Esparza-Gordillo J, et al. The eczema risk
variant on chromosome 11q13 (rs7927894) in the population-based
ALSPAC cohort: a novel susceptibility factor for asthma and hay
fever. Hum Mol Genet 2011;20:2443–2449.
30. Anderson CA, Boucher G, Lees CW, et al. Meta-analysis identifies
29 additional ulcerative colitis risk loci, increasing the number of
confirmed associations to 47. Nat Genet 2011;43:246–252.
31. Cookson W, Moffatt M, Strachan DP. Genetic risks and
childhood-onset asthma. J Allergy Clin Immunol 2011;128:266–
270; quiz 71–72.
32. Weidinger S, Gieger C, Rodriguez E, et al. Genome-wide scan on
total serum IgE levels identifies FCER1A as novel susceptibility
locus. PLoS Genet 2008;4:e1000166.
33. Granada M, Wilk JB, Tuzova M, et  al. A genome-wide associa-
tion study of plasma total IgE concentrations in the Framingham
Heart Study. J Allergy Clin Immunol 2011;129(3):840–845.e21.
doi:10.1016/j.jaci.2011.09.029.
34. Walley AJ, Chavanas S, Moffatt MF, et  al. Gene polymor-
phism in Netherton and common atopic disease. Nat Genet
2001;29:175–178.
35. Palmer CN, Irvine AD, Terron-Kwiatkowski A, et  al. Common
loss-of-function variants of the epidermal barrier protein filaggrin
are a major predisposing factor for atopic dermatitis. Nat Genet
2006;38:441–446.
36. Lees CW, Barrett JC, Parkes M, Satsangi J. New IBD genetics: com-
mon pathways with other diseases. Gut 2011;60:1739–1753.
37. Zhang K, Kaufman RJ. From endoplasmic-reticulum stress to the
inflammatory response. Nature 2008;454:455–462.
38. Sriburi R, Bommiasamy H, Buldak GL, et al. Coordinate regulation
of phospholipid biosynthesis and secretory pathway gene expres-
sion in XBP-1(S)-induced endoplasmic reticulum biogenesis. J Biol
Chem 2007;282:7024–7034.
39. Breslow DK, Collins SR, Bodenmiller B, et al. Orm family proteins
mediate sphingolipid homeostasis. Nature;463:1048–1053.
40. Uhlig S, Gulbins E. Sphingolipids in the lungs. Am J Respir Crit
Care Med 2008;178:1100–1114.
41. Ammit AJ, Hastie AT, Edsall LC, et  al. Sphingosine 1-phosphate
modulates human airway smooth muscle cell functions that pro-
mote inflammation and airway remodeling in asthma. FASEB J
2001;15:1212–1214.
42. Cantero-Recasens G, Fandos C, Rubio-Moscardo F, Valverde MA,
Vicente R. The asthma-associated ORMDL3 gene product regu-
lates endoplasmic reticulum-mediated calcium signaling and cellu-
lar stress. Hum Mol Genet;19:111–121.
43. Lluis A, Schedel M, Liu J, et  al. Asthma-associated polymor-
phisms in 17q21 influence cord blood ORMDL3 and GSDMA
gene expression and IL-17 secretion. J Allergy Clin Immunol
2011;127:1587–1594 e6.
44. Saeki N, Kuwahara Y, Sasaki H, Satoh H, Shiroishi T. Gasdermin
(Gsdm) localizing to mouse Chromosome 11 is predominantly
expressed in upper gastrointestinal tract but significantly suppressed
in human gastric cancer cells. Mamm Genome 2000;11:718–724.
 45. Tamura M, Tanaka S, Fujii T, et al. Members of a novel gene family,
GSDM, are expressed exclusively in the epithelium of the skin and
gastrointestinal tract in a highly tissue-specific manner. Genomics
2007;89:618–629.
46. Saeki N, Usui T, Aoyagi K, et al. Distinctive expression and function
of four GSDM family genes (GSDMA-D) in normal and malig-
nant upper gastrointestinal epithelium. Gene Chromosome Canc
2009;48:261–271.
47. Dunne A, O’Neill LA. The interleukin-1 receptor/Toll-like recep-
tor superfamily: signal transduction during inflammation and host
defense. Science’s STKE [electronic resource] 2003;171:2003, pre3.
48. Baekkevold ES, Roussigne M, Yamanaka T, et al. Molecular charac-
terization of NF-HEV, a nuclear factor preferentially expressed in
human high endothelial venules. Am J Pathol 2003;163:69–79.
49. Schmitz J, Owyang A, Oldham E, et al. IL-33, an interleukin-1-like
cytokine that signals via the IL-1 receptor-related protein ST2
and induces T helper type 2-associated cytokines. Immunity
2005;23:479–490.
50. Carriere V, Roussel L, Ortega N, et al. IL-33, the IL-1-like cytokine
ligand for ST2 receptor, is a chromatin-associated nuclear factor in
vivo. Proc Natl Acad Sci U S A 2007;104:282–287.
51. Kurowska-Stolarska M, Hueber A, Stolarski B, McInnes IB.
Interleukin-33: a novel mediator with a role in distinct disease
pathologies. J Intern Med 2011;269:29–35.
52. Moussion C, Ortega N, Girard JP. The IL-1-like cytokine IL-33 is
constitutively expressed in the nucleus of endothelial cells and epi-
thelial cells in vivo: a novel “alarmin”? PLoS One 2008;3:e3331.
53. Oboki K, Ohno T, Kajiwara N, et  al. IL-33 is a crucial ampli-
fier of innate rather than acquired immunity. Proc Natl Acad Sci
U S A;107:18581–18586.
54. Lloyd CM. IL-33 family members and asthma—bridging
innate and adaptive immune responses. Curr Opin Immunol
2010;22:800–806.
55. Fukao T, Matsuda S, Koyasu S. Synergistic effects of IL-4 and IL-18
on IL-12-dependent IFN-gamma production by dendritic cells. J
Immunol 2000;164:64–71.
56. Reche PA, Soumelis V, Gorman DM, et al. Human thymic stromal
lymphopoietin preferentially stimulates myeloid cells. J Immunol
2001;167:336–343.
57. Ziegler SF, Artis D. Sensing the outside world: TSLP regulates bar-
rier immunity. Nat Immunol 2010;11:289–293.
58. Yang YA, Zhang GM, Feigenbaum L, Zhang YE. Smad3 reduces sus-
ceptibility to hepatocarcinoma by sensitizing hepatocytes to apopto-
sis through downregulation of Bcl-2. Cancer Cell 2006;9:445–457.
59. Tian F, DaCosta Byfield S, Parks WT, et al. Reduction in Smad2/3
signaling enhances tumorigenesis but suppresses metastasis of breast
cancer cell lines. Cancer Res 2003;63:8284–8292.
60. Daly AC, Vizan P, Hill CS. Smad3 protein levels are modulated by
Ras activity and during the cell cycle to dictate transforming growth
factor-beta responses. J Biol Chem 2010;285:6489–6497.
61. Lloyd CM, Hawrylowicz CM. Regulatory T cells in asthma.
Immunity 2009;31:438–449.
62. Kariyawasam HH, Pegorier S, Barkans J, et  al. Activin and trans-
forming growth factor-beta signaling pathways are activated
after allergen challenge in mild asthma. J Allergy Clin Immunol
2009;124:454–462.
63. Ashcroft GS, Yang X, Glick AB, et  al. Mice lacking Smad3 show
accelerated wound healing and an impaired local inflammatory
response. Nat Cell Biol 1999;1:260–266.
64. Anthoni M, Wang G, Leino MS, Lauerma AI, Alenius HT, Wolff
HJ. Smad3—signalling and Th2 cytokines in normal mouse airways
and in a mouse model of asthma. Int J Biol Sci 2007;3:477–485.
65. Ge X, McFarlane C, Vajjala A, et al. Smad3 signaling is required for
satellite cell function and myogenic differentiation of myoblasts.
Cell Res.2011;21(11):1591–1604. doi:10.1038/cr.2011.72.
66. Letourneau S, Krieg C, Pantaleo G, Boyman O. IL-2—
and CD25-dependent immunoregulatory mechanisms in
G e n etic s a nd G e no m ic s o f C h ronic I n f l a m m atory D i s ord e r s , I I I   •   4 7 1
the homeostasis of T-cell subsets. J Allergy Clin Immunol
2009;123:758–762.
67. Setoguchi R, Hori S, Takahashi T, Sakaguchi S. Homeostatic main-
tenance of natural Foxp3(+) CD25(+) CD4(+) regulatory T cells
by interleukin (IL)-2 and induction of autoimmune disease by IL-2
neutralization. JEM 2005;201:723–735.
68. Gaffen SL. Signaling domains of the interleukin 2 receptor.
Cytokine 2001;14:63–77.
69. Montes de Oca P, Malarde V, Proust R, Dautry-Varsat A,
Gesbert F. Ectodomain shedding of interleukin-2 receptor beta
and generation of an intracellular functional fragment. J Biol
Chem;285:22050–22058.
70. Doganci A, Karwot R, Maxeiner JH, et al. IL-2 receptor beta-chain
signaling controls immunosuppressive CD4+ T cells in the draining
lymph nodes and lung during allergic airway inflammation in vivo. J
Immunol 2008;181:1917–1926.
71. Lee GR, Fields PE, Griffin TJ, Flavell RA. Regulation of the Th2 cyto-
kine locus by a locus control region. Immunity 2003;19:145–153.
72. Marsh DG, Meyers DA, Bias WB. The epidemiology and genetics of
atopic allergy. N Engl J Med 1981;305:1551–1559.
73. Marsh DG, Blumenthal MN, Ishikawa T, et  al. HLA and spe-
cific immune respoinsiveness to allergens. In:  Tsuji K, Aizawa
M, Sazazuki T, eds. Proceedings of the Eleventh International
Histocompatability Workshop and Conference. Oxford,
UK: Oxford University Press; 1992:765–767.
74. Moffatt MF, Young A, Schou C, Faux J, et  al. Involvement
of TCR-a/d and HLA-DR in specific allergy. Allergy Suppl
1995;50:164.
75. Moffatt MF, Cookson WO. Tumour necrosis factor haplotypes and
asthma. Hum Mol Genet 1997;6:551–554.
76. Link E, Parish S, Armitage J, et  al. SLCO1B1 variants and
statin-induced myopathy—a genomewide study. N Engl J Med
2008;359:789–799.
77. Tantisira KG, Lasky-Su J, Harada M, et  al. Genomewide associa-
tion between GLCCI1 and response to glucocorticoid therapy in
asthma. N Engl J Med 2011;365:1173–1183.
78. Bird A. DNA methylation patterns and epigenetic memory. Genes
Dev 2002;16:6–21.
79. Bird AP. CpG islands as gene markers in vertebrate nucleus. Trends
Genet 1987;3:342–347.
80. Eckhardt F, Lewin J, Cortese R, et al. DNA methylation profiling of
human chromosomes 6, 20 and 22. Nat Genet 2006;38:1378–1385.
81. Feinberg AP. Phenotypic plasticity and the epigenetics of human
disease. Nature 2007;447:433–440.
82. Liu J, Morgan M, Hutchison K, Calhoun VD. A study of the
influence of sex on genome wide methylation. PLoS One 2010;5:
e10028.
83. Christensen BC, Houseman EA, Marsit CJ, et al. Aging and envi-
ronmental exposures alter tissue-specific DNA methylation depen-
dent upon CpG island context. PLoS Genet 2009;5:e1000602.
84. Boks MP, Derks EM, Weisenberger DJ, et  al. The relationship of
DNA methylation with age, gender and genotype in twins and
healthy controls. PLoS One 2009;4:e6767.
85. Heijmans BT, Kremer D, Tobi EW, Boomsma DI, Slagboom PE.
Heritable rather than age-related environmental and stochastic fac-
tors dominate variation in DNA methylation of the human IGF2/
H19 locus. Hum Mol Genet 2007;16:547–554.
86. Eder W, Ege MJ, von Mutius E. The asthma epidemic. N Engl J Med
2006;355:2226–2235.
87. Huang YJ, Nelson CE, Brodie EL, et  al. Airway microbiota and
bronchial hyperresponsiveness in patients with suboptimally con-
trolled asthma. J Allergy Clin Immunol 2011;127:372–381 e3.
88. Bisgaard H, Hermansen MN, Buchvald F, et al. Childhood asthma
after bacterial colonization of the airway in neonates. N Engl J Med
2007;357:1487–1495.
89. Kraft M. The role of bacterial infections in asthma. Clin Chest Med
2000;21:301–313.
 90. Savage DC. Microbial ecology of the gastrointestinal tract. Annu
Rev Microbiol 1977;31:107–133.
91. Staley JT, Konopka A. Measurement of in situ activities of non-
photosynthetic microorganisms in aquatic and terrestrial habitats.
Annu Rev Microbiol 1985;39:321–146.
92. Turnbaugh PJ, Ley RE, Hamady M, Fraser-Liggett CM,
Knight R, Gordon JI. The human microbiome project. Nature
2007;449:804–810.
93. Ahmed S, Macfarlane GT, Fite A, McBain AJ, Gilbert P, Macfarlane
S. Mucosa-associated bacterial diversity in relation to human ter-
minal ileum and colonic biopsy samples. Appl Environ Microbiol
2007;73:7435–7442.
94. Maslowski KM, Vieira AT, Ng A, et al. Regulation of inflamma-
tory responses by gut microbiota and chemoattractant receptor
GPR43. Nature 2009;461:1282–1286.
95. Noverr MC, Huffnagle GB. The “microflora hypothesis” of allergic
diseases. Clin Exp Allergy 2005;35:1511–1520.
96. Herbst T, Sichelstiel A, Schar C, Yadava K, Burki K, et  al.
Dysregulation of allergic airway inflammation in the absence of
microbial colonization. Am J Respir Crit Care Med 2011;184:
198–205
 97. Nembrini C, Sichelstiel A, Kisielow J, Kurrer M, Kopf M,
Marsland BJ. Bacterial-induced protection against allergic inflam-
mation through a multicomponent immunoregulatory mecha-
nism. Thorax 2011;66:755–763.
98. Noverr MC, Huffnagle GB. Does the microbiota regulate immune
responses outside the gut? Trends Microbiol 2004;12:562–568.
99. Kraft P, Wacholder S, Cornelis MC, et al. Beyond odds ratios—
communicating disease risk based on genetic profiles. Nat Rev
Genet 2009;10:264–269.
100. Hunter DJ, Khoury MJ, Drazen JM. Letting the genome out
of the bottle—will we get our wish? N Engl J Med 2008;358:
105–107.
101. Binia A, Khorasani N, Bhavsar PK, et  al. Chromosome 17q21
SNP and severe asthma. J Hum Genet 2011;56:97–98.
102. Cook DG, Strachan DP. Health effects of passive smoking.
3. Parental smoking and prevalence of respiratory symptoms and
asthma in school age children. Thorax 1997;52:1081–1094.
103. Weedon MN, Lango H, Lindgren CM, et al. Genome-wide asso-
ciation analysis identifies 20 loci that influence adult height. Nat
Genet 2008;40:575–583.
4 7 2   •   G e no m ic s in C l inic a l P r actic e
104. Barrett JC, Hansoul S, Nicolae DL, et al. Genome-wide associa-
tion defines more than 30 distinct susceptibility loci for Crohn’s
disease. Nat Genet 2008;40:955–962.
105. Blakemore AI, Froguel P. Investigation of Mendelian forms of
obesity holds out the prospect of personalized medicine. Ann N Y
Acad Sci 2010;1214:180–189.
106. Maher B. Personal genomes: The case of the missing heritability.
Nature 2008;456:18–21.
107. Risch NJ. Searching for genetic determinants in the new millen-
nium. Nature 2000;405:847–856.
108. Nixon GF. Sphingolipids in inflammation:  pathological impli-
cations and potential therapeutic targets. Br J Pharmacol
2009;158:982–993.
109. Schwerk N, Brinkmann F, Soudah B, Kabesch M, Hansen G.
Wheeze in preschool age is associated with pulmonary bacte-
rial infection and resolves after antibiotic therapy. PLoS One
2011;6:e27913.
110. Donnelly D, Critchlow A, Everard ML. Outcomes in chil-
dren treated for persistent bacterial bronchitis. Thorax
2007;62:80–84.
111. Bennet JD, Brinkman M. Treatment of ulcerative colitis by implan-
tation of normal colonic flora. Lancet 1989;1:164.
112. Borody TJ, Warren EF, Leis S, Surace R, Ashman O. Treatment of
ulcerative colitis using fecal bacteriotherapy. J Clin Gastroenterol
2003;37:42–47.
113. Eiseman B, Silen W, Bascom GS, Kauvar AJ. Fecal enema as an
adjunct in the treatment of pseudomembranous enterocolitis.
Surgery 1958;44:854–859.
114. Hamilton MJ, Weingarden AR, Sadowsky MJ, Khoruts A.
Standardized Frozen Preparation for Transplantation of Fecal
Microbiota for Recurrent Clostridium difficile Infection. Am J
Gastroenterol 2012.; 107(5):761–767. doi:10.1038/ajg.2011.482.
115. Ng SB, Turner EH, Robertson PD, et  al. Targeted capture and
massively parallel sequencing of 12 human exomes. Nature
2009;461:272–276.
116. Ng SB, Buckingham KJ, Lee C, et al. Exome sequencing identifies
the cause of a mendelian disorder. Nat Genet 2010;42:30–35.
117. Biesecker LG. Exome sequencing makes medical genomics a reality.
Nat Genet 2010;42:13–14.
118. Vercelli D. Advances in asthma and allergy genetics in 2007.
J Allergy Clin Immunol 2008;122:267–271.
 31.
GENETICS AND GENOMICS OF NEURO-PSYCHIATRIC
DISEASES, I: SEIZURE DISORDER S
William Owen Pickrell
INTRODUCTION
Epilepsy is one of the most common neurological disorders.
It is defined as an ongoing tendency to experience epileptic
seizures. The prevalence of epilepsy is around 1% (higher
in resource-poor areas),1 and at least 50 million people are
affected worldwide.2 There is significant morbidity associ-
ated with the disease, and although appropriate treatment
can offer freedom from seizures for some people with epi-
lepsy, a significant minority will continue to have seizures
that are refractory to treatment.
There is a clear link between genetics and epi-
lepsy:  first-degree relatives of people with epilepsy are
around two to four times more likely to develop epilepsy
than the general population,3 and monozygotic twins show
a higher rate of concordance than do dizygotic twins.4 This
link has been known for a long time:  Hippocrates wrote
of epilepsy, circa 300 BC: “Its origin is hereditary, like that
of other diseases.”5 The insightful London neurologist
John Reynolds wrote in 1861 that an inherited influence
was present in around 31% of his own cases.6 This figure
is remarkably similar to the estimate that is widely used
today—genetic factors account for around 40% of all
epilepsies.7
It is only during the last 15 years or so that real progress
has been made into understanding some of the complicated
genetics involved with epilepsy. The ability to use powerful
new genomic technology marks an exciting time, promising
to yield further significant advances in our understanding,
with the ultimate aim to improve the treatment and quality
of life of people with the disease.
This chapter briefly describes the classification of sei-
zures and epilepsy before moving on to consider some of
the predominantly monogenetic, genetic, and symptomatic
epilepsies. The study of other genetic epilepsies is then con-
sidered. The chapter concludes by discussing some aspects of
epilepsy treatment that have been influenced by genomics.
473
(For further reference, the reader is directed to the excellent
text by Shorvon et al.,8 and recent reviews.9,10,11)
DEFINING AND CLASSIFYING
S E I ZU R E S A N D E P I L E P SY
An epileptic seizure is a transient occurrence of signs and
symptoms due to abnormal excessive or synchronous neu-
ronal activity in the brain.12
G E N E R A L I Z E D S E I ZU R E S
Seizures that originate from neuronal networks in both
cerebral hemispheres are classified as generalized seizures.
The most common types of generalized seizures are gener-
alized tonic clonic seizures and absence seizures (see Table
31.1 for more types of generalized seizures).
F O C A L S E I ZU R E S
Seizures that originate from neuronal networks limited to
one hemisphere are classified as focal seizures.13 The clini-
cal manifestations of a focal seizure are wide-ranging and
depend on the location in the brain of the abnormal activ-
ity (see Table 31.1 for some features of focal seizures).
Sometimes focal seizures can evolve into generalized sei-
zures as the abnormal neuronal activity spreads to both
hemispheres. This may manifest as an aura (partial sei-
zure); for example, a feeling of déjà-vu, before a general-
ized tonic clonic seizure. (Such a manifestation points to a
temporal-lobe focus for the seizure.)
T H E C L A S S I FI C AT I O N O F E P I L E P SY
Epilepsy is best thought of, not as a single disease, but rather
as a collection of several underlying brain disorders that all
Table 31.1  SOME DIFFERENT TYPES OF EPILEPTIC SEIZURE

GENERALIZED SEIZURES
Tonic-clonic Loss of consciousness with an initial period of stiffness (tonic) followed by rhythmic generalized jerking
(clonic). Gradual recovery of consciousness, with post-ictal confusion lasting minutes to hours.
Absence (typical) Sudden, brief (generally <10 sec) periods of loss of awareness with behavioral arrest (staring episodes) with
rapid recovery, occasional eye movements, automatisms.
Absence (atypical) Longer-than-typical absences and frequently associated with myoclonic or atonic attacks. Start and finish more
gradually; focal features more prominent, and more retained awareness than in typical absences.
Clonic Rare; loss of consciousness with repetitive symmetrical and rhythmical jerking. Rapid onset and cessation.
Tonic Sustained contraction of musculature, generally sudden-onset <1 min in duration. Abrupt onset and rapid
recovery.
Atonic Sudden loss of muscle tone, frequently causing a fall to the ground. Rapid recovery.
Myoclonic Brief muscle contractions with sudden onset and cessation. Single muscle to generalized jerking. Consciousness
generally not impaired.
FOCAL SEIZURES a
Frontal lobe Frequently nocturnal; brief, rapid onset and cessation. Prominent motor features, sometimes with posturing
and head version. Frequent bizarre automatisms/behaviors and vocalization.
Temporal lobe Generally longer in duration than frontal lobe seizures. Variety of sensory disturbances, including psychic
(déjà-vu, jamais-vu, fear), gustatory and olfactory hallucinations. Sensation of epigastric disturbance. Oro-facial
automatisms (e.g., chewing, sucking) or fidgety hand movements. Frequently, altered awareness (complex partial
seizures) and post-ictal confusion. Auditory features may point to lateral temporal lobe involvement.
Occipital lobe Visual disturbance/hallucinations, typically elementary color images in the contralateral eye field.
OTHER SEIZURES
Unknown Sometimes there is insufficient evidence available to classify a seizure as generalized, focal, or both.
a
These are the main types of focal seizure; there are other types. See also Berg et al. 2010.13

cause an increased tendency to experience epileptic seizures.
There are several different ways of classifying the epilepsies.
A useful distinction is that between the symptomatic epi-
lepsies, where the epilepsy occurs with a concurrent under-
lying structural or metabolic problem, such as tuberous
sclerosis; and the idiopathic epilepsies, where the epilepsy
occurs in otherwise normal individuals with no detectable
metabolic or structural problem, such as childhood absence
epilepsy (these were previously known as genetic epilep-
sies).13 There is some overlap, however, between genetic and
symptomatic epilepsy, and in some cases the dividing line
is not so clear. For example, epilepsy due to glucose trans-
porter (GLUT-1) deficiency is a symptomatic epilepsy, as it
is caused by a metabolic problem, but it also can be thought
of as a genetic epilepsy, as it is caused by mutations in the
SLC2A1 gene.
The concept of a seizure threshold can be used to
explain the effect of genetics on the tendency to develop
epilepsy. People with a tendency to develop seizures
spontaneously can be thought of as having a low seizure
threshold, whereas people who develop seizures only
after extremely provoking circumstances; for example,
severe head trauma patients can be thought of as hav-
ing a high seizure threshold. The interplay of genetic and
environmental mechanisms underpins an individual’s
seizure threshold. Figure 31.1 illustrates the spectrum of
some of the causes of epilepsy and the relevant influence
of genetic and environmental factors.
MONOGENIC INHERITED
EPILEPSIES
Monogenic inherited epilepsies are rare, accounting for
a small fraction of all epilepsies, and are predominately
caused by mutations in single genes encoding ion-channels.
Despite their rarity, they are of interest as their study has
led to a better understanding of some of the molecular
mechanisms underpinning epilepsy. Despite being pre-
dominately monogenic, these epilepsies show locus hetero-
geneity (i.e., the same phenotype caused by mutations in
different genes); variable expression (mutations in the same
gene causing different phenotypes); and their causative
mutations are rarely found in the more common sporadic
genetic epilepsies. Table 31.2 lists the main types of mono-
genetic genetic epilepsy, and the next few sections describe
some of the more important of these epilepsies and the
mechanisms involved.

4 7 4   •   G eno m ics in C linical P ractice

 Symptomatic Epilepsies

Trauma
Tumour
Infection
CVD
Metabolic
Neuro-
degenerative
Monogenetic
symptomatic
epilepsies e.g. NCLs
Increasing Increasing
Genetic Environmental
Influences Influences
Genetic focal &
generalised epilepsies
e.g. JME

Monogenetic
genetic epilepsies
e.g. BNFS

Genetic Epilepsies

Figure 31.1 
Some of the causes of epilepsy, with the relative influence of genetic and environmental factors and the “division” between symptomatic
and genetic epilepsies. In some cases the distinction between symptomatic and genetic epilepsy is blurred; for example, disorders of cortical
malformation clearly cause a structural change in the brain (symptomatic), but they may be caused by a mutation in one gene and thus be genetically
predetermined (genetic). The figure is not to scale, and the size of the captions does not reflect the number of patients with epilepsy due to each
cause.  BNFS = benign neonatal familial seizures; CVD = cerebrovascular disease; JME = juvenile myoclonic epilepsy; NCL = neuronal ceroid lipofuscinoses.

N I C OT I N I C AC ET Y L C H O L I N E vitro,25 with variation in this gain-of-function between

CHANNEL EPILEPSIES
Neuronal nicotinic acetylcholine receptors (nAChRs) are
pentameric ligand-gated cationic channels. nAChRs are
encoded by 9α and 3β subunit genes. nAChRs are widely
distributed throughout the central nervous system (CNS),
and different combinations of subunit types in nAChRs
give rise to channels with different functions and locations.14
Autosomal Dominant Nocturnal Frontal
Lobe Epilepsy
Autosomal dominant nocturnal frontal lobe epilepsy
(ADNFLE) is a rare epilepsy consisting of frequent,
brief frontal lobe seizures occurring during sleep. The sei-
zures tend to start suddenly, often with vocalization, and
involve hyperkinetic, sometimes bizarre, movements. They
occasionally will evolve into generalized tonic clonic sei-
zures. A  mutation in the CHRNA4 gene, which encodes
the nAChR α4 subunit, was the first epilepsy-causing
gene mutation to be discovered.26 Mutations in the genes
CHRNB2 and CHRNA4, which encode the nAChR β2
and α4 subunits, respectively, also cause ADNFLE.27,25
Studies have shown that the mutations associated with
ADNFLE increase the sensitivity of nAChR to ACh in
different mutations.30 The intriguing fact that ADNFLE
manifests as seizures that arise exclusively from sleep and
the frontal lobe remains unexplained at present.
P OTA S S I UM C H A N N E L E P I L E P S I E S
There are five channels in the voltage-gated potassium
7 (KV7) family:  KV7.1 channels are mainly expressed in
cardiac muscle, and KCNQ1 mutations cause long QT
syndrome; KV7.2–5 are mainly expressed in the central
nervous system; KV7.1 and KV7.4 are also expressed in the
inner ear, and mutations in KCNQ4 cause a slowly pro-
gressive deafness.15
Mutations in the genes KCNQ2 and KCNQ3, which
encode KV7.2 and KV7.3, respectively, are associated with
benign familial neonatal seizures (BFNS).14,16 In this rare
autosomal dominant syndrome, seizures normally start in
the first week of life and usually remit within a few weeks
to months, although around 10% of patients may develop
further seizures in later life.
KV7 channels act as a “brake” to neuronal firing by
producing a slowly activating K+ current known as the
M-current. Mutations in KCNQ2 and KCNQ3 cause a
reduction in this M-current and hence enhanced neu-
ronal excitability, which is the presumed mechanism for

G enetics and G eno m ics of N euro -P sychiatric D iseases , I :   S eizure D isorders   •   4 75

Table 31.2  SOME OF THE MORE IMPORTANT GENES ASSOCIATED WITH MONOGENETIC GENETIC
EPILEPSIES

GENE PROTEIN PHENOTYPE OMIM REFERENCES

Voltage-gated Ion Channels
KCNQ2 Potassium channel, voltage-gated, Benign familial neonatal seizures; epi- 121200 Singh et al. 1998,65
member 2 leptic encephalopathies Weckhuysen et al. 201266
KCNQ3 Potassium channel, voltage-gated, Benign familial neonatal seizures 121201 Charlier et al. 199867
member 3
SCN1A Sodium channel, voltage-gated Genetic epilepsy with febrile seizures +; 604403; Claes et al. 2001,68 Escayg
­neuronal type 1 α subunit severe myoclonic epilepsy of infancya 607208 et al. 200069
SCN1B Sodium channel, neuronal type 1 β Genetic epilepsy with febrile seizures + 604233 Wallace et al. 199870
subunit
SCN2A Sodium channel, voltage-gated Benign familial neonatal/infantile 607745; Heron et al. 200271
type II α subunit seizures; early infantile epileptic 613721
encephalopathy
Ligand-gated Ion Channels
GABRA1 GABAAb receptor on ligand-gated Autosomal dominant juvenile 611136 Cossette et al 2002,72,Maljevic
chloride channel, α1 subunit ­myoclonic epilepsy; association with et al. 200673
childhood absence epilepsy
GABRG2 GABAAb receptor on ligand-gated Genetic epilepsy with febrile seizures +; 611277; Baulac et al. 2001,74 Wallace
chloride channel, γ2 subunit severe myoclonic epilepsy of infancya; 607208; et al. 200175
childhood absence epilepsy 607681
CHRNA2 Neuronal nicotinic acetylcholine Autosomal dominant nocturnal frontal 610353 Aridon et al. 200676
receptor, α2 subunit lobe epilepsy
CHRNA4 Neuronal nicotinic acetylcholine Autosomal dominant nocturnal frontal 600513 Steinlein et al. 199577
receptor, α4 subunit lobe epilepsy
CHRNB2 Neuronal nicotinic acetylcholine Autosomal dominant nocturnal frontal 605375 De Fusco et al. 200078
receptor, β2 subunit lobe epilepsy
Non-ion channels
LGI1 Leucine-rich glioma-inactivated gene Autosomal dominant lateral temporal 600512 Kalachikov et al 200279
lobe epilepsy
EFHC1 EF-hand domain (c-terminal) con- Familial juvenile myoclonic epilepsy 254770 Suzuki et al. 200480
taining protein 1
OMIM = Online Mendelian Inheritance in Man database: http://www.ncbi.nlm.nih.gov/omim
a
Also known as Dravet’s syndrome.
b
GABAA = Gamma-aminobutyric acid type A.
Adapted with permission from Johnson MR, The genetic contribution to epilepsy: the known and missing heritability. In: The Causes of Epilepsy, eds. Shorvon SD,
et al., pp. 63–67. Cambridge, UK: Cambridge University Press, 2011.

seizures.16 Retigabine, a new anti-epileptic drug used as
an additional therapy for refractory partial-onset sei-
zures, acts by enhancing the activity of KV7.2–5 channels,
increasing M-current and therefore reducing neuronal
excitability.17
A perplexing question is, why do the seizures in BFNS
stop after a few months of life? Heterozygous mutations in
KCNQ2 and KCNQ3 typically only cause a small reduc-
tion of 25–30% in the M-current,32 and it may be that,
after the neonatal period, the additional ion-channels
expressed in the brain mean that this small reduction in
current is less problematic. Alternatively, the fact that
the GABAergic system switches from being an excitatory
to an inhibitory system after the neonatal period might
provide the additional “inhibition” that is required to
prevent seizures in patients with KCNQ2/3 mutations.7
Recently, KCNQ2 mutations have been discovered in
10% of patients with a neonatal/early infantile encepha-
lopathy,15,18 suggesting a role for this gene in other types of
more severe epilepsy.
S O D I UM C H A N N E L E P I L E P S I E S
Voltage-gated sodium channels (VGSC) are well charac-
terized to date and play an important part in initiating and
propagating action potentials in electrically excitable tissue

4 7 6   •   G eno m ics in C linical P ractice

by allowing a flux of inward sodium current in response
to depolarization. Neuronal sodium channels consist of
an α-subunit and two β-subunits. The α-subunit consists
of four domains and forms the central ion-conducting
pore and the channel gate for activation and inactivation.
There are nine different α-subunit subtypes, designated
NaV1.1–1.9.19 The β subunits are thought to affect VGSC
trafficking and gating, and there are four different types of
β subunit.20
Genetic Epilepsy with Febrile Seizures
Genetic epilepsy with febrile seizures+ (GEFS+) is a familial
epilepsy syndrome with a broad phenotype. By definition,
febrile seizures and febrile seizures+ (febrile seizures beyond
the age of 6 years) are the prominent feature. Other, afebrile
seizures also occur, both generalized seizures (absences,
tonic clonic, myoclonic) and, less commonly, focal (fron-
tal and temporal lobe) seizures. Mutations in the SCN1A
gene, which encodes the α1-subunit18 and, less commonly,
the SCN1B gene, which encodes the β1-subunit,19,21 have all
been associated with GEFS+ as well as mutations in genes
encoding the GABA receptor which are discussed in the
next section.
Severe Myoclonic Epilepsy in Infancy
Severe myoclonic epilepsy in infancy (SMEI) or Dravet’s
syndrome is a severe epilepsy syndrome usually starting
at around 6 months of age, usually with a febrile seizure
and progressing to multiple drug-refractory seizures and
developmental delay. Around 80% of all Dravet’s syn-
drome cases are caused by mutations in SCN1A, with
around 95% of these mutations occurring de novo.17,36 The
fact that SCN1A mutations cause the relatively mild phe-
notype of GEFS+ as well as the devastatingly severe SMEI
is partly explained by the fact that the specific, mostly
de novo, nonsense mutations associated with SMEI give
rise to more severe truncated or non-expressed proteins.
This is compared to the mostly missense mutations that
are associated with GEFS+. However, there are probably
additional factors that explain variable expression such as
modifier genes, genetic mosaicism, or other undiscovered
influences.
Both loss- and gain-of-function mutations in SCN1A
have been described in GEFS+, although it generally seems
that loss-of-function mutations are more common. The fact
that NaV1.1 is predominantly expressed in inhibitory inter-
neurons may explain how loss-of-function mutations can
lead to increased seizures.35
G enetics and G eno m ics of N euro -P sychiatric D iseases , I :   S eizure D isorders   •   4 77
Benign Familial Neonatal Infantile Seizures
Benign familial neonatal infantile seizures (BFNIS) are
a rare “mild” autosomal dominant phenotype wherein
patients suffer both partial and general seizures between the
first week and the first few months of life, which remit by
the age of 12 months. BFNIS is associated with mutations
in the SCN2A gene, which encodes the α2-subunit of the
voltage-gated sodium channel.20,22 The mutations in SCN2A
associated with BFNIS seem to increase neuronal excitabil-
ity, which offers a plausible explanation of the cause of the
seizures. Although the expression of the α2-subunit is ini-
tially high in the neonate, it decreases with age. This, together
with the fact that the expression of another subtype—the
α6-subunit—increases and seems to replace the α2-subunit,
gives an explanation of why the seizures in this syndrome
remit.23 Similar explanations might underpin the remission
of seizures in other self-terminating epilepsy syndromes.
A mutation in SCN1A has been associated with a type
of familial hemiplegic migraine.24 Migraine is similar to epi-
lepsy in that it is a common disease of the central nervous
system with symptoms caused by a transient disturbance
of the CNS, and SCN1A mutations are an interesting link
between the two diseases.
G A BA A R EC E P TO R E P I L E P S I E S
Gamma-aminobutyric acid type A  receptors (GABAA)
are ligand-gated ion-channels and are the major inhibitory
ion-channels within the human brain. Binding of GABA to
these receptors causes an influx of chloride, which results in
inhibition of neuronal activity. GABAA receptors have a het-
eropentameric structure formed from at least seven sub-unit
classes, but most GABAA receptors found within the brain
contain two α subunits, two β subunits, and a γ or δ sub-
unit.25 Mutations in the GABRA1, GABRG2, and GABRD
genes, which encode the α1, γ2 and δ subunit, respectively,
have been implicated in a wide variety of epilepsy pheno-
types, including febrile seizures, absence seizures, GEFS+,
and familial juvenile myoclonic epilepsy ( JME).
The majority of mutations seem to cause a loss of
function of the GABAA receptor and a reduction in
GABAA-mediated current. This would then cause a reduc-
tion in inhibition, which seems a plausible mechanism for
seizure generation.
LGI1-A S S O C I AT E D E P I L E P S I E S
The leucine-rich glioma-inactivated 1 (LGI1) gene
derives its name from the leucine-rich regions on the
protein it encodes and the fact that its expression is absent
or significantly downregulated in many high-grade glio-
mas.28 It was the first non–ion channel gene to be associ-
ated with genetic epilepsy in humans, although it seems
likely that its function is closely linked to glutametergic
synapses.
LGI1 encodes a protein with three leucine-rich repeats
in the N-terminal region and seven tandem repeats named
epilepsy-associated repeats (EAR) or epitempin repeats
(EPTP) in its C-terminal region. These repeats are pre-
dicted to form a propeller structure that links presynaptic
ADAM23 and Kv1.1 potassium channels and postsynaptic
ADAM22 and AMPA receptor scaffolds on glutamater-
gic synapses. Mouse models with targeted disruption of
any of these components of the suggested LGI1-associated
synapse bridging complex cause seizures.39 Mice models
suggest that LGI1 has a role in mediating glutamatergic syn-
apse pruning and maturation in the hippocampus and that
LGI1 deficiencies can markedly increase excitatory synapse
transmission.26
Autosomal Dominant Lateral Temporal Lobe
Epilepsy; Autosomal Dominant Partial Epilepsy with
Auditory Features
Autosomal dominant lateral temporal lobe epilepsy
(ADLTE) or autosomal dominant partial epilepsy with
auditory features (ADPEAF) is an inherited epilepsy aris-
ing in early adulthood with predominantly focal seizures
arising from the lateral temporal lobe. Focal seizures consist
mainly of auditory hallucinations, such as a ringing or buzz-
ing, with occasional visual hallucinations. Approximately
50% of patients with ADLTE have mutations in the LGI1
gene.27 Interestingly, antibodies to LGI1 have recently been
described in cases of limbic encephalitis, which causes cog-
nitive disturbance and seizures, offering further evidence of
the role of LGI1 in seizure generation.39
M O N O G E N I C SY M P TO M AT I C
EPILEPSIES
Currently, there are more than 200 Mendelian disorders in
which epilepsy is present as the result of a defect in genes
involved in a wide variety of functions. The genetics and
molecular biology of these disorders continue to be unrav-
elled. A few of these epilepsies are described here; the reader
is advised to consult texts such as Shorvon et al.8 for a more
comprehensive description.
4 7 8   •   G eno m ics in C linical P ractice
P RO G R E S S I VE MYO C L O N I C
EPILEPSIES
The progressive myoclonic epilepsies (PMEs) are a het-
erogeneous group of disorders characterized by epilepsy,
myoclonus, and progressive neurological deterioration,
including ataxia and dementia. The genetic bases of many of
the PMEs have been elicited, and the more common types
are here briefly described.
Unverricht-Lundborg Disease
Unverricht-Lundborg disease (EPM1) is the most com-
mon of the PMEs (Online Mendelian Inheritance in Man
[OMIM] 254800; also known as Baltic myoclonic epilepsy
as it was initially described in Finland) and is an autosomal
recessive disorder caused by mutations in the CSTB gene,28
which encodes cystatin B, a cysteine proteinase inhibitor.
Onset is typically around 6–15  years of age with severe,
progressive, stimulus-sensitive myoclonus; generalized
tonic-clonic seizures; followed by further neurological signs,
including ataxia, dysarthria, and mild dementia. Around
90% of EPM1 cases are caused by a dodecamer repeat in
the CSTB-promoter region. Disease-causing expansions
contain at least 30 repeats, whereas normal alleles typically
contain two or three repeats.29
The exact physiological function of CSTB and the
pathophysiology of mutations associated with EPM1
are currently unknown; however, mice lacking cystatin B
develop progressive myoclonic seizures and ataxia, and cel-
lular changes associated with apoptosis.44
EPM1B (OMIM 612437), which is caused by a muta-
tion in the PRICKLE1 gene, is very similar to EPM1,
although patients may present at a slightly younger age and
may also have an impaired upgaze.30
Lafora Disease
Lafora disease (OMIM 254780) is an autosomal recessive
disorder caused by mutations in the EPM2A31 or NHLRC132
genes. Onset is typically in adolescence with progressive
epilepsy, with several seizure types, including myoclonic
and generalized tonic clonic seizures. Ataxia and cognitive
decline follow, with death typically within 10 years of onset.
Lafora bodies (periodic acid-Schiff-positive intracellular
inclusions), found in various organs, including heart, skel-
etal muscle, and neurons, are pathognomonic.
EPM2A encodes laforin, which is a protein tyrosine
phosphatase, whilst NHLRC1 encodes malin, a ubiqui-
tin ligase. Malin and laforin seem to suppress the cellular
toxicity of misfolded proteins by promoting their degrada- M A L F O R M AT I O N S O F C O RT I C A L
tion through the ubiquitin-proteasome system.33 D EVE L O PM E N T

Malformations of cortical development (MCD), gener-

Neuronal Ceroid Lipofuscinoses
The neuronal ceroid lipofuscinosis (NCL) are the most
common inherited cause of neurodegenerative disease in
childhood and are characterized by visual failure, epilepsy,
progressive dementia, ataxia, and early death (second or
third decade). There is a presence of intracellular lipopig-
ment storage material—ceroid lipofuscin. Currently, muta-
tions in at least 10 genes are known to cause the disease
(see Table 31.3), and inheritance is mostly in an autosomal
recessive form.34
Other Causes of Progressive Myoclonic Epilepsies
Other causes of progressive myoclonic epilepsies
include Sialidoses (OMIM 256550)  caused by muta-
tions in NEU1, which lead to neuraminidase deficiency;
dentate-rubro-pallido-lusian atrophy (DRPLA) (OMIM
125370), caused by an unstable CAG repeat on chromo-
some 12p13.31; and Gaucher disease (OMIM 23100),
caused by mutations in GBA, which lead to deficient
β-glucocerebrosidase activity.
ally readily detectable on magnetic resonance imaging
(MRI), can cause epilepsy, and the genetic basis for many is
being elicited. Knowledge of the causative mutations, with
increasing understanding of gene function, is helping scien-
tists unravel the neurobiology behind brain development,
discover possible mechanisms for epileptogenesis, and also
define criteria for further refining the phenotypes of these
heterogeneous disorders. A  few of the more important
MCDs are described here.
Microcephaly
A group of disorders with the common feature of micro-
cephaly—significantly reduced head circumference
(defined as an occipital-frontal circumference less than two
standard deviations below the mean)—is associated with
epilepsy. Epilepsy is more common in the types of micro-
cephaly with accompanying structural changes and cortical
malformation. Some of the microcephalies with a greater
association with epilepsy and their causative genes are listed
in Table 31.4.

Table 31.3  THE NEURONAL CEROID LIPOFUSCINOSES 50

DISEASE OMIM GENE AND FUNCTION AGE AT ONSET PHENOTYPE

REFERENCE
CLN1 256730 CLN1 encodes lysosomal enzyme PPT1 18 months Seizures (myoclonic jerks prominent), devel-
(palmotoyl protein thoesterase). Decreased opmental delay
PPT1 activity in disease.
CLN2 204500 CLN2 encodes lysosomal enzyme TPP1 2–4 years Seizures dominant feature, progressive
(tripeptidyl peptidase 1). Decreased TPP1 myoclonus
activity in disease.
CLN3 204200 CLN3 encodes membrane protein not trans- 5–10 years Visual failure, learning difficulties, then
ported to membrane in disease. epilepsy
CLN5 256731 CLN5 encodes protein with unclear 4–7 years (Finnish variant): initially impaired motor
function. skills, progressive visual failure, seizures
CLN6 601780 CLN6 encodes membrane protein with 1.5–8 years Motor delay, dysarthria, and ataxia
unclear function.
CLN7 610951 MFSD8 encodes lysosomal transporter pro- 2–7 years Initially severe epileptic seizures, then pro-
tein with unclear function. gressive cognitive and motor deterioration
CLN8 600143 CLN8 encodes membrane protein, exact 5–10 years Generalized tonic clonic seizures, progressive
function unknown. motor problems and learning difficulties
CLN9 609055 Gene unknown. 5–10 years Similar phenotype to CLN3, but distinctive
fibroblasts apparent
CLN10 610127 CTSD encodes lysosomal aspartic protease 0–5 years Early (primary microcephaly, neonatal epi-
cathepsin D. lepsy, death in infancy) and late infantile
forms

G enetics and G eno m ics of N euro -P sychiatric D iseases , I :   S eizure D isorders   •   4 79

Table 31.4  TYPES OF MICROCEPHALY MORE ASSOCIATED WITH EPILEPSY, AND THEIR ASSOCIATED GENES 81

DISEASE OMIM REFERENCE GENE REFERENCE

Microcephaly and periventricular heterotopia 608097 ARFGEF-2 Sheen et al. 200482
Microcephaly and lissencephaly 611603 TUBA1A Morris-Rosendahl et al. 200883
Microcephaly with diffuse polymicrogyria 600118 RAB3GAP Aligianis et al. 200584
Microlissencephaly 300215 ARXa Kitamura et al. 200285
Microcephaly with brainstem and cerebellar hypoplasia 300749 CASK Najm et al. 200886
Microcephaly with asymmetrical polymicrogyria 610031 TUBB2B Jaglin et al. 200987
Microcephaly with calcification 610333, RNASEH2A, Crow et al 200688
610181, RNASEH2B,
610329 RNASEH2C
a
ARX is also linked to early infantile epileptic encephalopathy.

Lissencephaly and Subcortical Band Heterotopia The mutation results in defective transport of glucose across
the blood–brain barrier, depriving the brain of essential
Lissencephaly literally means “smooth brain” and is a neuro-
glucose. The disorder was discovered in 1991 and is char-
nal migration disorder in which there are absent or reduced
acterized by hypoglycorrhachia (low CSF glucose) with
gyra on the surface of the brain. Subcortical band heteroto-
normoglycemia, drug-refractory epilepsy with a variety of
pia (SBH) is a less severe form wherein bands of gray mat-
seizure types, acquired microcephaly, and episodic move-
ter are present within the white matter of the brain.35 Both
ment disorders.40 Although GLUT-1 DS is rare, it is an
cause learning difficulties and epilepsy. Most lissencephaly is
important diagnosis not to miss, as treatment with some
caused by mutations in PAFAH1B1 (LIS1) or DCX.36 DCX
conventional anti-epileptic drugs and other medications
is located on the X-chromosome, and mutations in females
can inhibit GLUT-1 and worsen symptoms, whereas treat-
cause SBH, whereas mutations in males cause lissenceph-
ment with a ketogenic diet can reduce seizure frequency
aly.37 Deletion of PAFAH1B1 as well as neighboring genes
dramatically.41
causes a more severe phenotype with characteristic dysmor-
phic features (Miller-Dieker syndrome). Both PAFAH1B1
and DCX regulate microtubule function, important in the Pyridoxine-Dependent Epilepsy
regulation of migrating neurones.38
Pyridoxine-dependent epilepsy is an autosomal recessive
disease, usually presenting in neonates with seizures, which
Periventricular Nodular Heterotopia respond to treatment with pyridoxine (OMIM 266100).
The disease is caused by homozygous or compound hetero-
Periventricular nodular heterotopia is a neuronal migration
zygous mutations in the ALDH7A1 gene, which encodes
disorder in which grey matter is positioned abnormally within
antiquitin, an aldehyde dehydrogenase. This causes an
the brain, typically along the walls of the lateral ventricles.
elevation in α-aminoadipic semialdehyde (AASA), result-
Mutations in FLN1 on the X chromosome account for most
ing in an intracellular reduction in pyridoxal-5′-phosphate
“classical cases.” Most females with the condition develop
(PLP)—the active vitamin B6 co-factor essential for neuro-
epilepsy, while males tend to die before birth. FLN1 encodes
transmission.42,43 Although rare, this is another example of
a protein with a role in neuronal maturation. There is also an
an important disease to diagnose, as lifelong treatment with
autosomal recessive form of the disease associated with micro-
pyridoxine causes seizure freedom.
cephaly and mutations in the ARFGEF gene (see Table 31.4).39

Other Genetic Conditions

SY M P TO M AT I C E P I L E P S I E S D U E TO
M ETA B O L I C A N D OT H E R C AUS E S
Glucose Transporter Type 1 Deficiency Syndrome
(GLUT-1 DS)
GLUT-1 DS is a rare disorder due to mutations in the
SLC2A1 gene, which encodes GLUT-1 (OMIM 606777).
There are many other genetic conditions in which epilepsy
is a prominent feature. These include:  tuberous sclerosis
(OMIM 191100, autosomal dominant disease with wide-
spread multi-organ hamartomatous lesions, frequently asso-
ciated with TSC1 or TSC2 mutations); neurofibromatosis 1
(OMIM 162200, autosomal dominant neurocutaneous

4 8 0   •   G eno m ics in C linical P ractice

disease caused by mutations in the tumor-suppressor gene
NF1); Rett syndrome (OMIM 312750, X-linked neuro-
developmental disorder with seizures in childhood caused
by mutations in MECP2 gene—mutations in CDKL5 also
cause a similar phenotype).
OT H E R G E N ET I C E P I L E P S I E S
The predominantly monogenetic epilepsies described
above, generally inherited in a Mendelian fashion, account
for only about 1% of all epilepsies. Genetic focal and gen-
eralized epilepsies occurring sporadically and inherited
in a non-Mendelian manner account for about 30% of all
epilepsies. To date, no convincing genetic mechanism has
been discovered to entirely explain these epilepsies, and
they are most likely caused by an interaction of multiple
genetic influences, with the influence of some environmen-
tal factors. Over 100 gene defects have been associated with
these epilepsies, however. The Epilepsy Genetic Association
Database (epiGAD; www.epigad.org) listed 327 epilepsy
gene-association studies involving 120 different genes at the
time of writing and is a useful reference source.44
The next section describes some of the strategies that are
being used to study the complicated genetics of these epi-
lepsies, and some avenues for future study.
G E N O M E -WI D E A S S O C I AT I O N
STUDIES
Genome-wide association studies (GWAS) use a large num-
ber (typically around 0.5 to 1 million) of single-nucleotide
polymorphism (SNP) markers spread throughout the entire
genome of patients and controls to look for any associations
between common genetic variants and disease status (see
Chapters 2 and 6). Typically, a GWAS study requires a large
number of samples to be adequately powered to detect the
relatively weak genetic effects (odds ratios of 1.2–1.3) asso-
ciated with the common disease–common variant hypoth-
esis. From the publication of the first GWAS in 2005 to the
end of 2011, 1617 published associations had been identi-
fied for 249 conditions.45
There have been two GWAS in epilepsy published as
of this writing. The first study included 3445 patients with
genetic and symptomatic focal epilepsies and 6935 controls
of European ancestry, and did not identify any significant
genome-wide associations.46 The second study looked at
1087 patients of Han Chinese descent with genetic and
symptomatic focal epilepsies, and 3444 ethnically matched
controls. One variation on 1q32.1 in the CAMSAP1L1
G enetics and G eno m ics of N euro -P sychiatric D iseases , I :   S eizure D isorders   •   4 81
gene, which encodes a cytoskeletal protein, was significant
on a genome-wide scale.47 Interestingly, neither of these
GWAS seemed to identify any significant variations from
within ion-channel encoding regions.
The lack of positive findings in these studies might be
due in part to the large amount of phenotypical heterogene-
ity. Further GWAS are currently being undertaken in other
forms of epilepsy. However, the lack of success of GWAS
suggests that the complex genetics of epilepsy may be more
due to rare genetic variants than to common variants—the
common disease–rare variant hypothesis.
C O P Y N UM B E R VA R I A N T S
One possible source of rare genetic variants are copy num-
ber variants (CNVs). CNVs are genomic structural variants,
including insertions, deletions, and duplications, which are
typically kilobases to megabases in size (see also Chapters 2
and 10). With the advent of technologies such as compara-
tive genome hybridization (CGH) array platforms, and
their ability to detect CNVs at an increased resolution,
attention has turned to the role of CNVs in diseases with a
complicated genetic basis, like epilepsy.
Several recent studies have published an association
of three microdeletion CNVs with epilepsy. The regions
containing the CNVs all contain possible candidate genes;
for example, 15q13.3 includes the CHRNA7 gene, which
encodes a nAChR. All three microdeletions have also been
associated with other diseases,; for example, the 15q13.3
microdeletion is associated with intellectual disability,48
autism,49 and schizophrenia.50 See Table 31.5 for a summary
of the major CNVs found to date.
Therefore, although CNVs seem a reasonable explana-
tion as a genetic risk factor for epilepsy, there is much that
is unexplained in terms of the mechanisms whereby they
cause disease.
N E X T- G E N E R AT I O N S EQ U E N C I N G
A N D E P I L E P SY
The advent of next-generation sequencing (NGS) technolo-
gies in the last few years has seen a remarkable advance in our
ability to acquire detailed genetic data on individuals (see
also Chapter  10). The main technologies associated with
NGS are whole-exome sequencing (WES), wherein the
entire protein-coding region of the genome is sequenced;
and whole-genome sequencing (WGS), wherein the entire
human genome is sequenced.
Already there have been several published examples
of how NGS has been used to identify causative genes in
Table 31.5  SIGNIFICANT CHROMOSOMAL MICRODELETIONS ASSOCIATED WITH EPILEPSY

MICRODELETION NOTES REFERENCE

15q13.3 Present in 1% of 1223 patients with genetic generalized epilepsy (absent in 3699 controls) Helbig et al. 200989
Present in 1.3% of 539 patients with genetic generalized epilepsy (absent in 3777 controls) Dibbens et al. 200990
Present in 1% of 517 patients with genetic generalized and focal epilepsy (absent in 2493 Mefford et al. 201091
controls)
15q11.2 Present in 1% of 1234 patients with genetic generalized epilepsy (0.1% of 3022 controls) De Kovel et al. 201092
Present in 1% of 517 patients with genetic generalized and focal epilepsy (0.2% of 2493 controls) Mefford et al. 201076
16p13.11 Present in 0.5% of 1234 patients with genetic generalized epilepsy (0.1% of 3022 controls) De Kovel et al. 201077
Present in 1% of 517 patients with genetic generalized and focal epilepsy (absent in 2493 Mefford et al. 201076
controls)

several disorders (Chapter 10). A recently published study
used WGS to identify a mutation in the sodium-channel
gene SCN8A as a cause of a severe epileptic encephalopa-
thy.51 These NGS studies have used small sample sizes
(nuclear family units of three or four, or very small cohorts).
Next-Generation Sequencing Panels
Lemke et  al.52 used NGS technology to test a “panel” of
genes with a known association with epilepsy in 33 patients
with a wide range of sporadic and familial epilepsy pheno-
types (see Chapters 6 and 10). They found, on average, 326
variants per patient from within the panel of 265 genes.
Filtering, using criteria such as including only variants
within coding regions and not present on publically avail-
able databases, reduced the average number of variants to
15 per patient. Software was then used in order to predict
the pathogenicity of the variants. This led to mutations
being found in 16 of the patients, including all eight of the
patients with a well-defined phenotype (some of whom had
already had candidate genes sequenced using traditional
methods with no success).
NGS gene panel methods have the advantage of being
able to offer increased coverage of genes of interest com-
pared to WGS or WES methods. They also produce less
‘noise,’ albeit at the cost of missing variation in genes not
present on the panel. The cost of gene panels continues
to decrease and is already cheaper than sequencing several
genes using traditional methods. Gene panels are beginning
to be used in clinical practice and are likely to be increas-
ingly used in the future.
Large-Scale Next-Generation Sequencing Studies
Heinlein and colleagues have recently published the
results of a study in which 118 unrelated patients with
genetic generalized epilepsy (GGE), and 242 controls were
exome-sequenced.53 No statistically significant variants
were identified. From the list of 87,255 variants identified
at the exome-sequencing stage, 3897 were identified and
genotyped in a larger cohort of 878 patients with IGE and
1830 controls. Again none of these candidate variants met
significance criteria (p <0.05) after Bonferonni correction
for the large number of variants tested. The study had 80%
power to detect single-nucleotide variants (SNVs) with a
frequency of 0.5% and a relative risk of 5.4. Although the
study did not produce any statistically significant findings,
the list of nearly 2000 variants observed exclusively in cases
with IGE is likely to contain several real risk factors and will
be a useful reference for future studies. It is interesting to
note that, as in the GWAS studies described above, none of
the variations detected were in ion channel genes or known
ion-channel modifiers. Several studies are now underway
to use NGS technology with different methodologies in
large-scale cohorts, such as Epi4k.54
EPIGENOMICS
Mechanisms other than changes in nucleotide sequence;
that is, epigenomic changes, may play a part in the etiology
and hereditability of epilepsy (see also Chapter 4).55 Some
examples of epigenetic mechanisms associated with epilepsy
follow.
DNA methylation has a role in regulating gene expres-
sion, and several genes associated with DNA methylation
have been associated with epilepsy; for example, MECP2
and the X-linked Rett syndrome (autism, learning difficul-
ties, and epilepsy—OMIM 312750).56
Histone proteins have a crucial role in gene regu-
lation as well as in organizing and packaging DNA.
Kainate-induced seizures in animals are important models
for research. Kainate administration causes changes in his-
tone proteins; for example, hyperacetylation of histones on
the brain-derived neurotrophic factor (BDNF) promoter.

4 8 2   •   G eno m ics in C linical P ractice

Sodium valproate, an anti-epileptic drug, can inhibit his-
tone deacetylases.57
Another epigenetic mechanism is that of post-
transcriptional RNA editing. For example:  Adenosine to
inosine (A to I) editing of the AMPA receptor subunit
GluR2 was found to be significantly increased in hippocam-
pal tissue from temporal lobe epilepsy patients. This caused
an increase in the permeability of calcium ions in the recep-
tor mediating excitotoxicity.58
G E N O M I C S A N D M A N AG E M E N T
O F E P I L E P SY
G E N ET I C D I AG N O S I S A N D T R E AT M E N T
Currently, there are several examples of a genetic diagno-
sis being clinically useful for a patient with epilepsy. For
instance, a positive genetic diagnosis of benign neonatal
familial seizures in a neonate avoids the need for potentially
more invasive tests; drugs that can potentially worsen sei-
zures can be avoided in a person with a diagnosis of Dravet’s
syndrome; and the ketogenic diet can be initiated in a per-
son with GLUT-1 DS. As our understanding of epilepsy
genomics increases, the role of a genetic diagnosis will play
a greater role for people with epilepsy.
P H A R M AC O G E N ET I C S —T R E AT M E N T
R E S P O NS E
Although surgery is a possible option for some focal epilep-
sies, drug therapy with anti-epileptic drugs (AEDs) remains
the mainstay of epilepsy treatment. There is a wide and often
unpredictable variation in the individual’s response to AED
treatment in terms of AED efficacy and adverse effects. The
epiGAD database (www.epigad.org)67 listed 178 epilepsy
pharmacogenetic association studies at the time of writing
and is a useful reference source (see also Chapter 7).
It has been postulated that the drug transporter PGP,
encoded by the gene ABCB1, has a role in AED response,
and a polymorphism in ABCB1 has been associated with
drug-resistant epilepsy.59 These findings have not always
been consistently replicated, however, and it seems that
variations in ABCB1 do not convincingly explain AED
resistance.60
Several studies have looked at gene-expression patterns
in brain samples from epilepsy surgery patients. Nine stud-
ies using microarray technology to compare patterns of
gene expression in hippocampal sections from patients with
drug-resistant epilepsy and control samples were included
G enetics and G eno m ics of N euro -P sychiatric D iseases , I :   S eizure D isorders   •   4 83
in a recent meta-analysis.61 They identified 233 genes with
altered expression in drug-resistant samples, with signifi-
cant overlap between the studies.
Some AEDs can cause serious adverse reactions; for
example, carbamazepine-induced Stevens-Johnson syn-
drome. One of the more important findings to date has
been the association of HLA-B*1502 with this reaction in
Asian populations.62 No cases of Stevens-Johnson syndrome
occurred in a Taiwanese group that had tested negative for
the HLA-B*1502 allele.63 This finding has led to the recom-
mendation that HLA-B*1502 testing should be performed
individuals of a high-risk ethnicity before carbamazapine
therapy is commenced. The HLA-A*3101 allele has been
associated with carbamazepine-induced Stevens-Johnson
syndrome in Caucasian patients.64
C O N C LU S I O N S
Although the influence of genetics on the etiology of epilepsy
has been understood for several hundred years, it is only dur-
ing the last 15 years or so that significant advances have been
made in our understanding of the mechanisms. Discovering
the genes responsible for rare genetic and symptomatic
Mendelian epilepsies has improved our understanding of the
mechanisms of seizure-generation and of neurophysiology
as a whole, but it has not explained the genetic mechanism
underlying the majority of the epilepsies.
Genome-wide association studies have provided little
evidence for the common disease–common variant hypothesis
in epilepsy to date. The common disease–rare variant hypoth-
esis might provide more of an explanation for the genetics
of epilepsy. Ongoing work using next-generation sequencing
technologies seems likely to yield further results in this area.
Genetic testing as an aid to diagnosis and treatment is begin-
ning to be used in the clinical domain, and we hope that,
as our understanding increases, genomics will begin to have
a greater impact on the treatment of patients with epilepsy.
AC K N OW L E D G E M E N T S
Thanks to Professor Philip Smith, Professor Mark Rees, and
Doctor Seokyung Chung for their support and comments.
REFERENCES
1. Sander JW, Shorvon SD. Epidemiology of the epilepsies. J Neurol
Neurosurg Psychiatry. 61:433–443, 1996.
 2. Ngugi AK, Bmomley C, Kleinschmidt I, Sander JW, Newton
CR. Estimation of the burden of active and life-time epilepsy:  a
meta-analytic approach. Epilepsia. 51:883–890, 2010.
3. Ottman R, Annegers JF, Risch N, Hauser WA, Susser M. Relations
of genetic and environmental factors in the etiology of epilepsy. Ann
Neurol. 39:442–449, 1996.
4. Vadlamudi L, Andermann E, Lombroso CT, et  al. Epilepsy in
twins:  insights from unique historical data of William Lennox.
Neurology. 62:1127–1133, 2004.
5. Hippocrates:  On the sacred disease. Available at http://www.​
humanistictexts.org/hippocrates.htm​. Accessed September 12,
2012.
6. Reynolds JR. Epilepsy:  Its Symptoms, Treatment and Relation to
Other Chronic Convulsive Diseases. London: Churchill, 1861.
7. Guerrini R, Shorvon SD, Andermann F, Andermann E.
Introduction to the concept of genetic epilepsy. In:  Shorvon SD,
Andermann F, Guerrini R (eds.), The Causes of Epilepsy. Cambridge,
UK: Cambridge University Press, 2011, Chapter 5, pp. 43–61.
8. Shorvon SD, Andermann F, Guerrini R. The Causes of Epilepsy. 1st
ed. Cambridge, UK: Cambridge University Press, 2011.
9. Poduri A, Lowenstein D. Epilepsy genetics—past, present, and
future. Curr Opinion Gen Dev. 21:325–332, 2011.
10. Ottman R, Risch R. Genetic epidemiology and gene discovery in
epilepsy. In: Noebels JL, Avoli M, Rogawski MA, et al. (eds.), Jasper’s
Basic Mechanisms of the Epilepsies. 4th ed. Bethesda, MD: National
Center for Biotechnology Information (US); 2012. Available at
http://www.ncbi.nlm.nih.gov/books/​NBK98138/​. Accessed
August 24, 2012.
11. Johnson MR, Robinson RA, Gardiner RM. Molecular genetics of the
epilepsies. From Epilepsy 2011: From Science to Society, A Practical
Guide to Epilepsy, Lecture Notes. Available at http://www.epilepsyso-
ciety.org.uk/FileStorage/​main_content/05-molecular-genetics-of-
the-epilepsies-.pdf​. Accessed August 28, 2012.
12. Fisher RS, van Emde Boas W, Blume W, et al. Epileptic seizures and
epilepsy: definitions proposed by the International League Against
Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE).
Epilepsia. 46:470–472, 2005.
13. Berg AT, Berkovic SF, Brodie MJ, et  al. Revised terminology and
concepts for organization of seizures and epilepsies:  report of the
ILAE Commission on Classification and Terminology, 2005–2009.
Epilepsia. 51:676–685, 2010.
14. Steinlein OK, Kaneko S, Hirose S. Nicotinic acetylcholine recep-
tor mutations. In:  Noebels JL, Avoli M, Rogawski MA, et  al.
(eds.), Jasper’s Basic Mechanisms of the Epilepsies. 4th ed. Bethesda,
MD: National Center for Biotechnology Information (US); 2012.
Available at http://www.ncbi.nlm.nih.​gov/books/NBK98138/​.
Accessed August 24, 2012.
15. Maljevic S, Wuttke TV, Lerche H. Nervous system KV7 disorders:
breakdown of a subthreshold brake. J Physiol. 586:1791–1801, 2008.
16. Schroeder BC, Kubisch C, Stein V, Jentsch TJ. Moderate loss of
function of cyclic-AMP-modulated KCNQ2/KCNQ3 K+ chan-
nels causes epilepsy. Nature. 396:687–690, 1998.
17. Gunthorpe MJ, Large CH, Sankar R. The mechanism of action of
retigabine (ezogabine), a first-in-class K+ channel opener for the
treatment of epilepsy. Epilepsia. 53:412–424, 2012.
18. Saitsu H, Kato M, Koide A, et al. Whole exome sequencing identi-
fies KCNQ2 mutations in Ohtahara syndrome. Ann Neurol. 71:15–
25, 2012.
19. Mantegazza M, Curia G, Biagini G, Ragsdale DS, Avoli M.
Voltage-gated sodium channels as therapeutic targets in epilepsy and
other neurological disorders. Lancet Neurol. 9:413–424, 2012.
20. Oliva M, Berkovic SF, Petrou S. Sodium channels and the neurobi-
ology of epilepsy. Epilepsia [published online ahead of print, 2012].
doi:10.1111/j.1528-1167.2012.03631.x2012
21. Scheffer IE, Harkin LA, Grinton BE, et al. Temporal lobe epilepsy
and GEFS+ phenotypes associated with SCN1B mutations. Brain.
130:100–109, 2007.
4 8 4   •   G eno m ics in C linical P ractice
22. Berkovic SF, Heron SE, Giordano L, et  al. Benign familial
neonatal-infantile seizures: characterization of a new sodium chan-
nelopathy. Ann Neurol. 55:550–557, 2004.
23. Liao Y, Deprez L, Maljevic S, et  al. Molecular correlates of
age-dependent seizures in an inherited neonatal-infantile epilepsy.
Brain. 133:1403–1414, 2010.
24. Dichgans M, Freilinger T, Eckstein G, et al. Mutation in the neu-
ronal voltage-gated sodium channel SCN1A in familial hemiplegic
migraine. Lancet. 366:371–377, 2005.
25. Macdonald RL, Kang JQ, Gallagher MJ. Mutations in GABAA
receptor subunits associated with genetic epilepsies. J Physiol.
588:1861–1869, 2010.
26. Zhou YD, Lee S, Jin Z, Wright M, Smith SE, Anderson MP.
Arrested maturation of excitatory synapses in autosomal dominant
lateral temporal lobe epilepsy. Nat Med. 15:1208–1214, 2009.
27. Ottman R, Winawer MR, Kalachikov S, et  al. LGI1 mutations
in autosomal dominant partial epilepsy with auditory features.
Neurology. 62:1120–1126, 2004.
28. Pennacchio LA, Lehesjoki AE, Stone NE, et  al. Mutations in
the gene encoding cystatin B in progressive myoclonus epilepsy
(EPM1). Science. 271:1731–1734, 1996.
29. Lehtinen MK, Lehesjoki A, Kälviäinen R. Unverricht-Lundborg
disease. In:  Shorvon SD, Andermann F, Guerrini R (eds.), The
Causes of Epilepsy. Cambridge, UK:  Cambridge University Press,
2011, Chapter 16, pp. 135–138.
30. Bassuk AG, Wallace RH, Buhr A, et al. A homozygous mutation in
human PRICKLE1 causes an autosomal recessive progressive myo-
clonic epilepsy-ataxia syndrome. Am J Hum Genet. 83:572–581,
2008.
31. Minassian BA, Lee JR, Herbrick JA, et  al. Mutations in a gene
encoding a novel protein tyrosine phosphatase cause progressive
myoclonus epilepsy. Nat Genet. 20:171–174, 1998.
32. Chan EM, Young EJ, Ianzano L, et al. Mutations in NHLRC1 cause
progressive myoclonus epilepsy. Nat Genet. 35:125–127, 2003.
33. Garyali P, Siwach P, Singh PK, et al. The malin-laforin complex sup-
presses the cellular toxicity of misfolded proteins by promoting their
degradation through the ubiquitin-proteasome system. Hum Mol
Genet. 18:688–670, 2009.
34. Jalanko A, Braulke T. Neuronal ceroid lipofuscinoses. Biochim
Biophys Acta. 1792:697–709, 2009.
35. Mochida GH. Genetics and biology of microcephaly and lissen-
cephaly. Semin Pediatr Neurol. 16:120–126, 2009.
36. Reiner O, Carrozzo R, Shen Y, et  al. Isolation of a Miller-Dieker
lissencephaly gene containing G protein beta-subunit-like repeats.
Nature. 364:717–721, 1993.
37. des Portes V, Pinard JM, Billuart P, et al. A novel CNS gene required
for neuronal migration and involved in X-linked subcortical lami-
nar heterotopia and lissencephaly syndrome. Cell. 92:51–61, 1998.
38. Smith DS, Niethammer M, Ayala R, et al. Regulation of cytoplas-
mic dynein behaviour and microtubule organization by mammalian
Lis1. Nature Cell Biol. 2:767–775, 2000.
39. Lu J, Sheen V. Periventricular heterotopia. Epil Behav. 7:143–149,
2005.
40. De Vivo DC, Trifiletti RR, Jacobson RI, et  al. Defective glucose
transport across the blood–brain barrier as a cause of persistent
hypoglycorrhachia, seizures, and developmental delay. N Engl J
Med. 325:703–709, 1991.
41. Pong AW, Geary BR, Engelstad KM, Matarajan A, Yang H, De Vivo
DC. Glucose transporter type 1 deficiency syndrome: Epilepsy phe-
notypes and outcomes. Epilepsia [published online ahead of print],
2012. doi:10.1111/j.1528-1167.2012.03592
42. Mills PB, Struys E, Jakobs C, et al. Mutations in antiquitin in indi-
viduals with pyridoxine-dependent seizures. Nat Med. 12:307–309,
2006.
43. Gospe SM. Pyridoxine-dependent epilepsy. In:  Shorvon SD,
Andermann F, Guerrini R (eds.), The Causes of Epilepsy. Cambridge,
UK: Cambridge University Press, 2011, Chapter 33, pp. 237–241.
 44. Tan NC, Berkovic SF. The Epilepsy Genetic Association Database
(epiGAD): analysis of 165 genetic association studies, 1996–2008.
Epilepsia. 51:686–689, 2010. Available at www.epigad.org.
45. Hindorff LA, MacArthur J (European Bioinformatics Institute),
Wise A, et al. A Catalog of Published Genome-Wide Association
Studies. Available at www.genome.gov/gwastud​ies​. Accessed
August 20, 2012.
46. Kasperaviciute D, Catarino CB, Heinzen EL, et al. Common genetic
variation and susceptibility to partial epilepsies:  a genome-wide
association study. Brain. 133:2136–2147, 2010.
47. Guo Y, Baum LW, Sham PC, et al. Two-stage genome-wide associa-
tion study identifies variants in CAMSAP1L1 as susceptibility loci
for epilepsy in Chinese. Hum Mol Genet. 21:1184–1189, 2012.
48. Sharp AJ, Mefford HC, Li K, et al. A recurrent 15q13.3 microdele-
tion syndrome associated with mental retardation and seizures. Nat
Genet. 40:322–328, 2008.
49. Pagnamenta AT, Wing K, Akha ES, et al. A 15q13.3 microdeletion
segregating with autism. Eur J Hum Genet. 17:687–692, 2009.
50. Stefansson H, Rujescu D, Cichon S, et al. Large recurrent microde-
letions associated with schizophrenia. Nature 455:232–236, 2008.
51. Veeramah KR, O’Brien JE, Meisler MH, et al. De novo pathogenic
SCN8A mutation identified by whole-genome sequencing of a
family quartet affected by infantile epileptic encephalopathy and
SUDEP. Am J Hum Genet, 90:502–510, 2012.
52. Lemke JR, Riesch E, Scheurenbrand T, et al. Targeted next genera-
tion sequencing as a diagnostic tool in epileptic disorders. Epilepsia.
53:1387–1398, 2012.
53. Heinzen EL, Depondt C, Cavalleri GL, et  al. Exome sequencing
followed by large-scale genotyping fails to identify single rare vari-
ants of large effect in genetic generalized epilepsy. Am J Hum Genet.
91:293–302, 2012.
54. The Epi4K Consortium. Epi4K: gene discovery in 4,000 genomes.
Epilepsia. 53:1457–1467, 2012.
55. Qureshi IA, Mehler MF. Epigenetic mechanisms underlying human
epileptic disorders and the process of epileptogenesis. Neurobiol Dis.
39:53–60, 2010.
56. Amir RE, Van den Veyver IB, Wan M, Tran CQ, Francke U, Zoghbi
HY. Rett syndrome is caused by mutations in X-linked MECP2, encod-
ing methyl-CpG-binding protein 2. Nat Genet. 23:185–188, 1999.
57. Urdinguio RG, Sanchez-Mut JV, Esteller M. Epigenetic mecha-
nisms in neurological diseases:  genes, syndromes, and therapies.
Lancet Neurol. 8:1056–1072, 2009.
58. Vollmar W, Gloger J, Berger E, et al. RNA editing (R/G site) and
flip-flop splicing of the AMPA receptor subunit GluR2 in nervous
tissue of epilepsy patients. Neurobiol Dis. 15:371–379, 2004.
59. Siddiqui A, Kerb R, Weale ME, et al. Association of multidrug resis-
tance in epilepsy with a polymorphism in the drug-transporter gene
ABCB1. N Engl J Med. 348:1442–1448, 2003.
60. Leschziner GD, Andrew T, Pirmohamed M, Johnson MR. ABCB1
genotype and PGP expression, function and therapeutic drug
response: a critical review and recommendations for future research.
Pharmacogenomics J. 7:154–179, 2007.
61. Mirza N, Vasieva O, Marson AG, Pirmohamed M. Exploring the
genomic basis of pharmacoresistance in epilepsy:  an integrative
analysis of large-scale gene expression profiling studies on brain tis-
sue from epilepsy surgery. Hum Mol Genet. 20:4381–4394, 2011.
62. Chung WH, Hung SI, Hong HS, et al. Medical genetics: a marker
for Stevens-Johnson syndrome. Nature. 428:486, 2004.
63. Chen P, Lin JJ, Lu CS, et al. Carbamazepine-induced toxic effects
and HLA-B*1502 screening in Taiwan. N Engl J Med. 364:1126–
1133, 2011.
64. McCormack M, Alfirevic A, Bourgeois S, et al. HLA-A*3101 and
carbamazepine-induced hypersensitivity reactions in Europeans. N
Engl J Med. 364:1134–43, 2011.
65. Singh NA, Charlier C, Stauffer D, et al. A novel potassium channel
gene, KCNQ2, is mutated in an inherited epilepsy of newborns. Nat
Genet. 18;25–29, 1998.
G enetics and G eno m ics of N euro -P sychiatric D iseases , I :   S eizure D isorders   •   4 85
66. Weckhuysen S, Mandelstam S, Suls A, et al. KCNQ2 encephalopa-
thy:  Emerging phenotype of a neonatal epileptic encephalopathy.
Ann Neurol. 71:15–25, 2012.
67. Charlier C, Singh NA, Ryan SG, et al. A pore mutation in a novel
KQT-like potassium channel gene in a genetic epilepsy family. Nat
Genet. 18:53–55, 1998.
68. Claes L, Del-Favero J, Ceulemans B, Lagae L, VanBroeckhoven
C, De Joghe P. De novo mutations in the sodium-channel gene
SCN1A cause severe myoclonic epilepsy of infancy. Am J Hum
Genet. 68:1327–1332, 2001.
69. Escayg A, MacDonald BT, Meisler, MH, et al. Mutations of SCN1A,
encoding a neuronal sodium channel, in two families with GEFS+.
Nat Genet. 24:343–345, 2000.
70. Wallace RH, Wang DW, Singh R, et al. Febrile seizures and general-
ized epilepsy associated with a mutation in the Na+-channel beta1
subunit gene SCN1B. Nat Genet. 19:366–370,1998.
71. Heron SE, Crossland KM, Andermann E, et  al. Sodium-channel
defects in benign familial neonatal-infantile seizures. Lancet.
360:851–852, 2002.
72. Cossette P, Liu L, Brisebois K, et al. Mutation of GABRA1 in an
autosomal dominant form of juvenile myoclonic epilepsy. Nat
Genet. 31:184–189, 2002.
73. Maljevic S, Krampfl K, Cobilanschi J, et  al. A mutation in the
GABAA receptor α1-subunit is associated with absence epilepsy.
Ann Neurol. 59:983–987, 2006.
74. Baulac S, Huberfeld G, Gourfinkel-An I, et al. First genetic evidence
of GABA(A) receptor dysfunction in epilepsy:  a mutation in the
gamma2-subunit gene. Nat Genet. 28:46–48, 2001.
75. Wallace RH, Marini C, Petrou S, et al. Mutant GABA(A) receptor
gamma2-subunit in childhood absence epilepsy and febrile seizures.
Nat Genet. 28:49–52, 2001.
76. Aridon P, Marini C, Di Resta C, et al. Increased sensitivity of the
neuronal nicotinic receptor α2 subunit causes familial epilepsy with
nocturnal wandering and ictal fear. Am J Hum Genet. 79:342–350,
2006.
77. Steinlein OK, Mulley JC, Propping P, et  al. A missense mutation
in the neuronal nicotinic acetylcholine receptor alpha 4 subunit
is associated with autosomal dominant nocturnal frontal lobe epi-
lepsy. Nat Genet. 11:201–203, 1995.
78. De Fusco M, Becchetti A, Patrignani A, et al. The nicotinic recep-
tor beta 2 subunit is mutant in nocturnal frontal lobe epilepsy. Nat
Genet. 26:275–276, 2000.
79. Kalachikov S, Evgrafov O, Ross B, et al. Mutations in LGI1 cause
autosomal-dominant partial epilepsy with auditory features. Nat
Genet. 30:335–41, 2002.
80. Suzuki T, Delgado-Escueta AV, Aguan K, et  al. Mutations in
EFHC1 cause juvenile myoclonic epilepsy. Nat Genet. 36:842–849,
2004.
81. Ross ME. Microcephaly. In: Shorvon SD, Andermann F, Guerrini
R (eds.), The Causes of Epilepsy. Cambridge, UK:  Cambridge
University Press, 2011, Chapter 50, pp. 330–340.
82. Sheen VL, Ganesh VS, Topcu M, et  al. Mutations in ARFGEF2
implicate vesicle trafficking in neural progenitor proliferation and
migration in the human cerebral cortex. Nat Genet. 36:69–76,
2004.
83. Morris-Rosendahl DJ, Najm J, Lachmeijer AMA, et al. Refining the
phenotype of alpha-1a tubulin (TUBA1A) mutation in patients
with classical lissencephaly. Clin Genet. 74: 425–433, 2008.
84. Aligianis IA, Johnson CA, Gissen, P, et al. Mutations of the catalytic
subunit of RAB3GAP cause Warburg Micro syndrome. Nat Genet.
37:221–223, 2005.
85. Kitamura K, Yanazawa M, Sugiyama N, et  al. Mutation of ARX
causes abnormal development of forebrain and testes in mice and
X-linked lissencephaly with abnormal genitalia in humans. Nat
Genet. 32:359–369, 2002.
86. Najm J, Horn D, Wimplinger I, et al. Mutations of CASK cause an
X-linked brain malformation phenotype with microcephaly and
 hypoplasia of the brainstem and cerebellum. Nat Genet. 40:1065–
1067, 2008.
87. Jaglin XH, Poirier K, Saillour Y, et al. Mutations in the beta-tubulin
gene TUBB2B result in asymmetrical polymicrogyria. Nat Genet.
41:746–752, 2009.
88. Crow YJ, Leitch A, Hayward BE, et al. Mutations in genes encod-
ing ribonuclease H2 subunits cause Aicardi-Goutieres syndrome
and mimic congenital viral brain infection. Nat Genet. 38:910–916,
2006.
89. Helbig I, Mefford HC, Sharp AJ, et  al. 15q13.3 microdeletions
increase risk of genetic generalized epilepsy. Nat Genet. 41:160–
162, 2009.
4 8 6   •   G eno m ics in C linical P ractice
90. Dibbens LM, Mullen S, Helbig I, et  al. Familial and sporadic
15q13.3 microdeletions in genetic generalized epilepsy: precedent
for disorders with complex inheritance. Hum Mol Genet. 18:3626–
3631, 2009.
91. Mefford HC, Muhle H, Ostertag P, et al. Genome-wide copy num-
ber variation in epilepsy: novel susceptibility loci in genetic general-
ized and focal epilepsies. PLoS Genetics 6:e1000962, 2010.
92. De Kovel CG, Trucks H, Helbig I, et al. Recurrent microdeletions at
15q11.2 and 16p13.11 predispose to genetic generalized epilepsies.
Brain 133:23–32, 2010.
 32.
GENETICS AND GENOMICS OF NEURO-PSYCHIATRIC
DISEASES, II: MULTIPLE SCLEROSIS
Katharine Harding and Neil Robertson
INTRODUCTION
Multiple sclerosis (MS) is both an inflammatory and
a degenerative disease of the central nervous system
(CNS) and the commonest non-traumatic cause of neu-
rological disability in young adults in Western Caucasian
populations. Its etiology is complex and incompletely
understood, but related to a combination of genetic and
environmental factors.1 Although the disease’s frequency
has increased over time, MS remains relatively rare, with
a variable and geographically dependent prevalence
affecting approximately one per 685 of the population in
Northern Europe.2 Diagnosis can be problematic, clinical
manifestations of disease diverse, and the long-term prog-
noses extremely variable. Improvements in disease man-
agement have led to gradually increasing life expectancy,
which now varies little from that of the general popula-
tion, so patients commonly live with the consequences of
the disease rather than dying from it. However, despite sig-
nificant recent therapeutic advances in MS, there remains
no cure; therefore the impact of the disease on the indi-
vidual is often substantial and prolonged.
A fundamental epidemiological characteristic of MS
is familial recurrence, which was recognized more than
100  years ago and provided an early indication of the
importance of genetics to its frequency, distribution, and
outcome. Subsequently, an understanding of the genetic
contribution to MS has proved valuable in counselling
patients, understanding pathological mechanisms of dis-
ease, identifying novel therapeutic targets, and determining
disease outcome. Genetic studies have also proved infor-
mative in unravelling environmental contributions to the
disease as well as the nature of potential gene–environment
interactions. These combined factors may now offer the
exciting future prospect of a more individualized approach
to patient management and treatment.
487
E P I D E M I O L O GY O F MU LT I P L E
SCLEROSIS
The frequency of MS demonstrates considerable geographic
variation, and a latitudinal gradient has been observed for
some decades1; a less marked longitudinal gradient is also
seen from East to West. In European populations, the preva-
lence increases with distance from the equator, up to approx-
imately 55° north, with the highest worldwide frequency of
MS being reported in Orkney, Scotland, where prevalence
was recently estimated at 402 per 100,000 (95% confidence
interval [CI] 319 to 500).3 However, the frequency of MS
appears to decrease at latitudes further north.4 This pattern
is replicated in the Southern Hemisphere, although the
effect is somewhat attenuated when corrected for age and
sex.5 The cause of these disease gradients remains obscure,
and some observers have questioned their origins,6 sug-
gesting they may occur not as a true environmental effect
but rather as an artifact of historic population movements
over the centuries. Recently, a detailed meta-analysis has
examined these issues in more detail and concluded that
the latitudinal variance in disease prevalence occurs only
in populations of European descent,4 therefore providing
some convincing evidence for an interaction between genes
and environment that contributes to disease susceptibility.
More formal analysis from population migration
studies in people affected by MS has helped shed further
light on this variable geographical distribution of MS and
indicates that the environmental contribution to the risk
of developing disease occurs at a relatively early stage in
life. These studies suggest that migrants moving before
their mid-teenage years appear to adopt the risk profile of
their host country, while migrants moving after this age
retain the risk of their country of origin.7–9 This phenom-
enon is operative for migrants moving from areas of high
prevalence to areas of low prevalence,7 and vice versa.8,9
Although there remains some debate about the underlying
mechanism of this phenomenon and also about the meth-
odology of studies (which may have identified migrants
who were not genetically representative of the indigenous
population10), it is clear that environmental influences
can significantly affect the risk of developing MS and
even override an underlying genetic predisposition to dis-
ease.11 An interesting exception to this is the unexpectedly
low rate of MS in second- and third-generation Japanese
Americans, born in the United States but with Japanese
parents, suggesting that environmental factors may not
be able to override genetic background in all cases, and
that a protective factor or factors may be active in specific
populations.12
Studies of some genetically and geographically isolated
populations which have identified a disease frequency that
does not conform to the more generally accepted pattern
of distribution have also provided valuable insights into
factors important in the development of MS. For example,
the population of Sardinia, Italy, which although situated
in Southern Europe where rates of MS are generally low,
has a relatively high prevalence of MS, estimated at 151.9
per 100,00013 and comparable to prevalence rates more
commonly observed in Northern Europe. Sardinia has a
genetically homogenous population that has been relatively
unaffected by recurrent waves of immigration, principally
because the interior of the island is remote and difficult to
access.13 Consequently, Sardinians have an HLA profile
that is distinct from other European populations’14 but rela-
tively free of population stratification, making them a useful
population in which to study complex genetic traits.14 Due
to the relatively high prevalence of MS within a small popu-
lation and the fact that the Sardinian population has been
well characterized over a number of years, it has proved
a particularly valuable population in which to examine
genetic traits in MS.15–17
Conversely, the Hutterite community of west-
ern and central Canada and the United States has a
lower-than-expected frequency of disease and therefore
also has potential for the study of the effect of rare genes
in MS. The Hutterites are an Anabaptist community that
emigrated from Europe and now comprises around 28,000
individuals in Canada alone.18 Marriage rarely occurs out-
side of these relative socially isolated Hutterite commu-
nities; as a result, there is a high rate of consanguinity, so
all Hutterites are related to each other as second cousins.
However, MS remains rare in this population, despite its
location in an area generally considered to be a high risk for
the development of MS,19 with a prevalence of only 21 per
100,000 (although absolute numbers are small). Attempts
4 8 8   •   G enomics in C linic a l P r actice
to identify protective factors relevant in this population
have so far been unsuccessful.18
Finally, a small number of large families with more
cases of MS than would be expected by chance alone have
been identified, including a Canadian pedigree containing
15 affected individuals over four generations, which sug-
gested a pattern of autosomal dominant inheritance (albeit
with incomplete penetrance) leading to MS in this kin-
dred20; and a Swedish kindred with 22 cases (albeit related
more distantly).21 However, detailed genetic analysis has
so far failed to identify any such genes,22 and it now seems
unlikely that an autosomal dominant form of MS exists.
C L I N I C A L C H A R AC T E R I S T I C S
O F   MU LT I P L E S C L E R O S I S
As in other autoimmune diseases, MS is more common in
females, with a ratio of approximately 1 male to 2.5 females,
which appears to be increasing over time,23 and a peak age
of onset in the third and fourth decades.24 However, the age
range of presentation is broad, and patients have been known
to present in almost every decade of life (Figure 32.1).
In the majority of patients, the initial disease course fol-
lows a relapsing pattern, where episodes of neurological dys-
function are interspersed with periods of clinical stability25
at a rate of approximately one significant relapse per year in
the early stages. However, in around 10–20% of patients,
there is progressive neurological deterioration from the
onset of the disease,26–28 commonly termed primary pro-
gressive disease.25 Of patients whose initial disease course
is relapsing, at least 65% will go on to develop progressive
neurological deterioration at a later date, which is known
as the secondary progressive phase of disease1,25 (Figure 32.2).
MS relapses manifest as a subacute onset of neurologi-
cal dysfunction in the absence of intercurrent illness (such
as infection), evolving over days or weeks, and may affect
any part of the nervous system. Common symptoms fre-
quently relate to involvement of clinically eloquent areas
of the CNS and include visual loss, sensory disturbance, or
motor weakness, either alone or in combination. Less com-
monly, brainstem or cerebellar dysfunction may be seen1,29,30
(Table 32.1). Younger patients are more likely to experience
bouts of optic neuritis than are older patients, while the fre-
quency of certain other symptoms, such as lower limb motor
disturbance, sphincter disturbance, and facial weakness,
increases with age.31 Although older patients have fewer
relapses32 and are less likely to experience a severe relapse,33
they are more likely to be left with residual disability as a
consequence of the relapse, while younger patients make a
 350

300

250 All
Number of patients

Female
200 Male

150

100

50

0
1–5 6–10 11–15 16–20 21–25 26–30 31–35 36–40 41–45 46–50 51–55 56–60 61–65 66–70 71–75
Age at onset (years)

Figure 32.1  Age at onset of multiple sclerosis.

more complete recovery.31 In most patients, recovery from
relapse occurs within two months, although improvements
may continue for up to a year following relapse.34 Overall
relapse rates within the MS population occur at approxi-
mately 0.3 relapses per patient per year32 and decrease grad-
ually with age and disease duration, so that even in a group
selected for highly active disease, it is unusual to find more
than two significant relapses per year.35
Two other commonly reported symptoms that are char-
acteristic of MS and also provide interesting clinical insights
into the pathology of disease, and in particular the effects
of demyelination, are Lhermitte’s phenomenon (distal tran-
sient sensory symptoms on neck flexion, thought to occur
as the result of ephaptic transmission), and Uhtoff ’s phe-
nomenon (transient neurological deterioration experienced
in sites of prior demyelination symptoms when core body

KEY
Initial relapsing course Current relapsing course

Initial progressive course Current progressive course

Benign MS (10%) Relapsing Secondary

remitting MS progressive MS
(40%) (40%)
Disability

Disability

Time Time

Relapsing with sequelae Primary progressive MS (10%)

Disability

Disability

Time Time

Figure 32.2  Changes in level of disability over time for each disease course group.

G enetics a nd G enomics of N euro -P sychi atric D ise a ses , I I :   Multiple S clerosis   •   489
 Table 32.1  FREQUENCY OF NEUROLOGICAL SYMPTOMS AT ONSET OF MS
SITE FREQUENCY
Long tracts 45–50% 21,28,33
Motor weakness, sensory disturbance, autonomic symptoms
Optic nerve 17–22% 21,28,33
Loss of visual acuity and color vision, painful eye on movement
Brainstem 9–13%21,28,33 Diplopia, vertigo, speech or swallow dysfunction, internuclear ophthalmoplegia
Cerebellum 7–15% 21,28
Ataxia, nystagmus, dysarthria
temperature is raised, such as following a hot bath or shower,
or after exercise1). Despite the widespread nature of pathol-
ogy in the MS-affected brain, frank dementia is seen only
rarely, and usually in the latter stages of disease. However,
certain subtler cognitive deficits are frequently observed in
earlier stages, and these are particularly related to speed of
processing and executive dysfunction, although verbal and
visuospatial memory may also be impaired. In contrast to
Alzheimer’s disease or vascular dementia, language function
is often well preserved. However, overall cognitive dysfunc-
tion does not correlate well with levels of physical disability
and may be confounded by other common symptomatic
manifestations of MS, including fatigue and depression.37
D I AG N O S I S O F MU LT I P L E S C L E RO S I S
From the earliest description of the disease, establishing a
diagnosis of MS has necessitated a demonstration of dis-
semination of disease activity in both space and time38
and remains a cornerstone of initial disease assessment,
although this is now usually augmented by the results of
paraclinical tests. On occasion the diagnosis remains chal-
lenging, despite the technological advances now available to
clinicians, principally because of the wide range of diseases
that can mimic clinical features of MS (Table 32.2), but also
because of the particular challenges of establishing diagnos-
tic certainty in patients with primary progressive disease.
Diagnostic criteria have gradually been refined over time
as the understanding of information provided by magnetic
resonance imaging (MRI) in MS has improved,38–41 along
with the contribution of informative paraclinical tests.
At present, a diagnosis can still be made on clinical
grounds alone if a patient has had two attacks, at least one
of which can be corroborated with objective clinical signs,
and in the absence of evidence for an alternative diagnosis.
However, in practice, most clinicians still depend on sup-
port from the additional investigations described below,
and in the absence of clinical evidence, MRI is the test that
is perhaps most commonly employed to supplement clini-
cal information38 (Figure 32.3; Table 32.3). However, the
lack of a single specific and sensitive test for the diagnosis
4 9 0   •   G enomics in C linic a l P r actice
SYMPTOMS
of MS, and the syndromic nature of the diagnostic process,
inevitably lead to some significant difficulties in managing
patients in the very earliest stages of disease. In particular,
the need to establish evidence for clinical or radiologi-
cal change in order to demonstrate dissemination in space
and time commonly leads to a delay from the onset of first
symptoms to a firm diagnosis. Prior to the widespread use
of MRI in diagnosis, the mean time from disease-onset to
diagnosis was approximately eight years, but more recently
this has been reduced to four years27 with more widespread
availability of imaging technology. However, even with
these advances, interval review of historical cohorts from
specialist centers consistently reveals a small proportion of
patients in whom a previously established diagnosis of MS
has been reversed.27,42 Whilst absolute numbers of patients
with initially erroneous diagnoses are small, it does result
in a degree of fallibility in diagnosing MS, which may also
in turn inject some inconsistency into the interpretation of
genetic studies of MS.
P RO G N O S I S O F MU LT I P L E S C L E RO S I S
Although relapses are a characteristic feature of early dis-
ease, the evidence from natural history studies suggests that
relapse characteristics and frequency do not contribute sig-
nificantly to longer-term outcome.43 Prognosis is therefore
best assessed using fixed disability, the most commonly used
measure of which is the Expanded Disability Status Scale
(EDSS),44 a 20-point scale ranging from “no disability” to
“death from MS” (Table 32.4). Commonly used disability
milestones include EDSS 4 (inability to walk 500 meters
unaided), EDSS 6 (requiring unilateral support to mobi-
lize), and EDSS 8 (wheelchair-bound). Alternative mea-
sures of disability exist, although these are either derived
from the EDSS—such as the MS Severity Score (MSSS),
which incorporates EDSS and disease duration,45 or its pre-
decessor the Progression Index46—or are not widely used,
making them less helpful for direct comparisons; for exam-
ple, the MS Functional Composite (MSFC).47
Most patients with MS will eventually develop progres-
sion of disability even when the initial disease course has
Table 32.2  DIFFERENTIAL DIAGNOSES FOR Table 32.2 CONTINUED
MULTIPLE SCLEROSIS
PATHOLOGY DIAGNOSIS
PATHOLOGY DIAGNOSIS
Cerebral autosomal dominant arteriopa-
Systemic inflammatory thy with subcortical infarcts and leukoen-
diseases: cephalopathy (CADASIL)
Primary systemic Wegener’s granulomatosis Malignancy: Gliomatosis
vasculitides Malignancy-related angiitis
Microscopic polyangiitis Primary lymphoma
Classical polyarteritis nodosa Angiocentric B cell lymphoma
Churg-Strauss Metastases
Small vessel vasculitis (and Leukemia
Henoch-Schönlein purpura)
Non-Hodgkin’s lymphoma
Kawasaki disease
Infections:
Giant cell arteritis
Viral Herpes simplex (HSV)/Herpes zos-
Takayasu’s arteritis ter (HZV)/Epstein-Barr (EBV)/
Cytomegalovirus (CMV)
Non-vasculitic systemic Systemic lupus erythematosus
disorders Human immunodeficiency virus (HIV)
Primary anti-phospholipid syndrome
Hepatitis B
Rheumatoid disease
Progressive multifocal leukoencephalopa-
Systemic sclerosis thy (PML)
Sjögren’s syndrome Bacterial Tuberculosis
Mixed connective tissue disease Meningovascular syphilis
Cryoglobulinemia Lyme disease
Sarcoidosis Fungal Aspergillosis
CNS inflammatory Acute disseminated encephalomyelopathy Cryptococcus
disease Histoplasmosis
Progressive multifocal leukoencephalopa- Other Malaria
thy (PML)
Whipples
Antibody mediated encephalitis
Miscellaneous Mitochondrial disease
Tissue-specific CNS Primary angiitis of the central nervous
vasculitides system
Eales disease
been relapsing. In one large early cohort, 41.3% of patients
Susac’s syndrome
had converted to a progressive phase by 10 years, and 57.6%
Cogan’s syndrome did so after 15  years,24 and it may be that the vast major-
Non-vasculitic Atheroma ity if not all patients would convert to progressive disease
vasculopathies
Migraine if followed up for a long enough period of time. Although
Cryoglobulinemia there is huge variability in the trajectories of disability accu-
Cardiac emboli mulation—ranging from benign disease with little or no
Bacterial endocarditis accumulation of disability over more than 20 years, to the
catastrophic Marberg variant resulting in death or severe
Cardiac myxoma
disability within two years—accumulation of disability is
Hypertension
generally slow, and occurs over decades. Median time from
Drug induced (cocaine) onset of disease to EDSS 4 in large population studies has
Radiation been reported as between 11.4  years36 and 14.6  years48;
Posterior reversible encephalopathy median time to EDSS 6 between 15.024 and 23.3 years26,48;
syndrome
median time to EDSS 7, 33.1 years36; and median time to
Cholesterol emboli EDSS 8, 46.4 years24 (Figure 32.4). Younger patients have
Fibromuscular dysplasia been shown to accumulate disability more slowly,29,49,50
Moya Moya and older patients to reach fixed levels of disability more

G enetics a nd G enomics of N euro -P sychi atric D ise a ses , I I :   Multiple S clerosis   •   491
 (A) (B) (C)

(D) (E) (F)

(G) (H)

Figure 32.3 
MRI of brain in multiple sclerosis: (a) Axial T2 image showing periventricular and juxtacortical hyperintense lesions. (b) Axial T2
image showing juxtacortical hyperintense lesions. (c) Axial T1 image post-contrast showing gadolinium-enhancing lesions.(d) Axial T1 image
post-contrast showing gadolinium-enhancing lesions. (e) Axial T1 image showing periventricular and juxtacortical black holes. (f ) Axial T1 image
showing juxtacortical black holes (g) Sagittal T2 image of spinal cord showing single lesion at C4/C5. (h) Coronal T2 image showing brainstem
hyperintense lesions (as well as periventricular and juxtacortical lesions).

rapidly.30,51 However, patients who are younger at disease PAT H O L O GY O F MU LT I P L E

onset are also younger at reaching disability milestones,50 S C L E RO S I S
whilst patients who are older at disease onset attain disabil-
The inflammatory component of MS pathology is largely
ity milestones at an older age,30 suggesting that a younger
T-cell–driven.1 Th17 cells (a subset of Th1 cells) have been
age at onset of disease could be considered to confer a worse
identified as an essential component in the development of
long-term prognosis.52
a number of autoimmune diseases, including rheumatoid

4 9 2   •   G enomics in C linic a l P r actice

Table 32.3 McDONALD 2010 CRITERIA FOR DIAGNOSIS OF MULTIPLE SCLEROSIS 38

CLINICAL PRESENTATION ADDITIONAL DATA NEEDED TO DIAGNOSE MS

1 attack with objective clinical evidence of ≥2 lesions Dissemination in time on MRI criteriab
1 attack with objective clinical evidence of 1 lesion Dissemination in spacea and timeb on MRI criteria
Insidious onset suggestive of primary progressive MS ≥1 year of progression, with ≥2 of:
•E vidence for dissemination in space in the brain on MRIa
•E vidence for dissemination in space in the spinal cord on MRIa
• Positive oligoclonal bands on CSF
a
Dissemination in space is demonstrated by lesions in at least 2 of the following characteristic areas: periventricular, juxtacortical, infratentorial, and spinal cord.
b
Dissemination in time is demonstrated either by simultaneous presence of gadolinium-enhancing and non-enhancing lesions, or by new T2 lesions or
gadolinium-enhancing lesions on a follow-up scan after an interval of at least a month.

arthritis, psoriasis, and inflammatory bowel disease, as well
as MS.53 Under the influence of IL-23, Th0 cells differenti-
ate into Th17 cells, which are then responsible for producing
a number of pro-inflammatory cytokines, including IL-6,
IL-17, IL-22, TNF-α, and IFN-γ,54 which are key factors
in the development of pathology. Interestingly, IL-23–defi-
cient mice have been shown to be resistant to developing
mouse models of MS and inflammatory arthropathies.54
Th17 cells are activated in the periphery and cross the
blood–brain barrier into the CNS, where a local inflam-
matory reaction results in damage to axons and oligo-
dendrocytes (which form the myelin sheath surrounding
axons, enabling rapid propagation of neural signaling).
Subsequently, proinflammatory cytokines amplify the
response, recruit activated macrophages and microglia,
and compromise the integrity of the blood–brain barrier,
allowing ingress of further activated T cells.55 Propagation
of these effects leads to continuing tissue damage, and
oligodendrocytes are significantly depleted in active MS
lesions and in chronic stable lesions,56 as well as in clini-
cally silent lesions.57 However, more recently it has become
apparent that B cells also have an important role to play,
which may have considerable relevance for the develop-
ment of novel treatments.1

Table 32.4  THE EXPANDED DISABILITY STATUS SCALE 44

EDSS SCORE
0 Normal neurological examination
1 No disability, signs present in one functional neurological system
1.5 No disability, signs present in more than one functional system
2.0 Minimal disability in 1 functional system
2.5 Minimal disability in 2 functional systems
3.0 Moderate disability in one functional system, or mild disability in 3 or 4 functional systems
3.5 Fully ambulatory but with moderate disability in one functional system and mild disability in one or two functional systems, or
moderate disability in two functional systems, or mild disability in five functional systems
4.0 Fully ambulatory without aid, up and about 12 hours a day, significant disability in one functional system, or combinations of
lesser grades exceeding previous steps. Able to walk 500 meters without aid or rest
4.5 Able to walk 300 meters without aid or rest
5.0 Able to walk 200 meters without aid or rest
5.5 Able to walk 100 meters without aid or rest
6.0 Intermittent or unilateral assistance required to walk 100 meters with or without rest
6.5 Constant bilateral assistance required to walk about 20 meters without resting
7.0 Unable to walk more than 5 meters even with aid, essentially restricted to wheelchair, but transfers alone
7.5 Unable to take more than a few steps, essentially restricted to wheelchair, may need help to transfer
8.0 Restricted to chair but up and about most of the day, has effective use of arms
8.5 Restricted to bed or chair, some effective use of arms
9.0 Helpless bed patient, can communicate and eat
9.5 Helpless bed patient, unable to communicate or eat
10 Death due to MS

G enetics a nd G enomics of N euro -P sychi atric D ise a ses , I I :   Multiple S clerosis   •   493
Proportion of patients not yet reaching disability milestone 1.0 a neurodegenerative process that is independent of neuroin-
EDSS 4 flammation.62 However, recent histopathological evidence
EDSS 6
EDSS 8
has suggested that neurodegeneration and neuroinflam-
0.8
mation may coexist, with neuroinflammation driving the
process so that active lesions in relapsing, acute, and pro-
0.6 gressive MS show ongoing inflammation and axonal loss,
with no observable difference between clinical subgroups
of MS.61 As will be discussed below, the recent advances in
0.4 genetics in MS further implicate the immune system in MS
pathogenesis and do not provide any additional evidence
that neurodegeneration is a separate, or indeed the primary
0.2
underlying, process (Figure 32.5).
Finally, radiological correlates of both neuroinflam-
0.0 mation and neurodegeneration have been demonstrated
with gadolinium-enhancement of T2 lesions, considered
0 10 20 30 40 50 60
to reflect the level of inflammation and breakdown of the
Years since onset of disease
blood–brain barrier; T1 black holes representing areas of
Figure 32.4 
Kaplan-Maier survival curves showing time to Expanded neurodegeneration and atrophy; and in vivo magnetic reso-
Disability Status Scales 4, 6, and 8 in a UK cohort of patients nance spectroscopy that is able to measure axonal damage
(unpublished data).
using levels of N-acetyl aspartate.63

Neurodegeneration, which commonly follows the

CU R R E N T T R E AT M E N TS F O R
inflammatory phase of the disease, is less well understood,
MU LT I P L E S C L E RO S I S
although it is thought to be the pathological substrate for
the progressive clinical phase of disease58 and is an impor- Patients with MS are usually managed by specialist MS ser-
tant histopathological feature in later phases of the dis- vices, which commonly include a wide range of paraclinical
ease. It is characterized by axonal loss within inflammatory specialists such as physiotherapists, occupational therapists,
lesions as well as the surrounding normal-appearing white speech therapists, and psychologists, who provide a range
matter (NAWM).58–60 It remains a matter of controversy of supportive care to help patients manage their disabilities
as to whether the neurodegeneration is a consequence of and the wider impact of their disease. Acutely, significant
the inflammation, a distinct but coexisting phenomenon, relapses that result in functional disability are treated with
or whether the neurodegeneration is in fact the primary a short course of oral or intravenous steroids together with
mechanism and causes a secondary inflammatory pro- supportive care and management of related symptomatic
cess.1 Evidence for separate processes includes the fact that issues, such as pain, spasticity, and sphincter dysfunction.
disease-modifying treatments have not been shown to have Steroids are often effective in shortening the time of recov-
a major impact on the accumulation of disability, and that ery from the acute event, but they do not influence the even-
MRI markers for inflammation and degeneration show tual outcome.64 Specific symptomatic treatments include
only modest correlation. Additionally, axonal deterioration drugs such as amitriptyline and gabapentin for neuropathic
has been demonstrated in NAWM at some distance from pain; baclofen, tizanidine, and the recently licensed canna-
inflammatory lesions.61 binoids65 for spasticity; anticholinergic drugs to treat blad-
Given the nature of the clinical diagnosis, it is perhaps der dysfunction; and amantadine for fatigue.1
not surprising that some observers support the concept However, the last two decades have seen the evolution
of pathological heterogeneity in MS. Four different pat- of a range of treatments that are effective in altering the
terns of histological change have been proposed for MS natural history of disease. These so called disease-modifying
lesions:  T-cell–mediated demyelination, antibody and treatments (DMTs) are universally directed at minimizing
complement–mediated demyelination, oligodendrocyte the neuroinflammatory component of disease to reduce
dystrophy and apoptosis, and primary oligodendrocyte or prevent clinical relapse and radiologically evident, new
demyelination. The first three are seen in both relapsing and inflammatory lesions.35,66–71 The mainstay of treatment cur-
progressive MS, but the last pattern has only been observed rently remains interferon-β (IFNβ),66,72 which has been in
in primary progressive MS, which suggests the possibility of widespread use since the 1990s, and has been shown to

4 9 4   •   G enomics in C linic a l P r actice

 (A) (B)
GM P

WM

PP

(C) (D)

WM

BV

Figure 32.5 
Histopathological changes in multiple sclerosis: (a) Histopathological specimen stained for myelin with Luxol Fast Blue, showing
loss of myelin within a slowly expanding plaque. 4x magnification. (WM = white matter, GM = gray matter, P = plaque) (b) High-power view
of peri-plaque area, showing demyelination in the plaque compared to surrounding white matter. 40x magnification. (PP = peri-plaque) (c)
Histopathological specimen stained for HLA, showing activated macrophages within the peri-plaque rim (arrows) of a slowly expanding MS
plaque. 4x magnification. (BV = blood vessel) (d) High-power view within the plaque showing few HLA-positive macrophages (arrows) in
relatively acellular, demyelinated plaque. 40x magnification.

reduce relapse rates by approximately 30%,66,72 although
a commonly used alternative is glatiramer acetate, which
has a similar efficacy.73 More recently, several new treat-
ments have been developed, including natalizumab,68 fin-
golimod,69 alemtuzumab,35 laquinimod,70 BG-12,71 and
teriflunomide74 (Table 32.5). Nevertheless, none of these
treatments has conclusively been shown to have a convinc-
ing impact on long-term progression of disability,75–77 and
indeed there are currently no effective treatments for the
neurodegenerative component of disease.
G E N ET I C S A N D MU LT I P L E S C L E R O S I S
FA M I LY S T U D I E S
For nearly a century, a genetic role in the etiology of MS
has been suggested by epidemiological studies and observed
patterns of familial recurrence. Relatives of MS probands
have an increased risk of disease compared to the general
population, proportional to their degree of relatedness to
the proband.78–80 In general, male relatives have a lower risk
than female relatives with the same degree of kinship78,79
(Figure 32.6). Similar patterns of inheritance are observed
in both high- and low-prevalence populations.80 The sibling
recurrence risk measures the risk to siblings relative to that
for the general population and provides a measure of genetic
risk in that population. It lies between 20 and 30 for MS in
both Europe17 and Australia,80 suggesting that genetic risk
remains similar, and observed variation in prevalence may
be more likely to relate to environmental factors affecting
the rate of MS within different populations with the same
genetic backgrounds.
Classical twin studies have also been useful in fur-
ther quantifying the genetic contribution to MS etiology.
Despite some early controversy,81 a substantially increased
risk to monozygotic twins of around 25% has been con-
firmed as compared to dizygotic twins, who have the same

G enetics a nd G enomics of N euro -P sychi atric D ise a ses , I I :   Multiple S clerosis   •   495
Table 32.5  CURRENT TREATMENTS FOR MULTIPLE SCLEROSIS AND THEIR MECHANISMS OF ACTION
DRUG MECHANISM OF ACTION
Interferon-β66 Immune system modulation, mechanism not clear
Glatiramer Cross-reactivity with and suppression of immune
acetate73 response against myelin basic protein
Natalizumab68 Blocks migration of lymphocytes into the CSF by
binding to cell surface receptors, thereby preventing
their binding to endothelial cell receptors and cross-
ing the blood–brain barrier
Fingolimod69 Sphingosine-1-phosphate receptor antagonist, which
prevents egress of T cells from lymph nodes, seques-
tering autoactive T cells in lymph tissue
Alemtuzumab35 Monoclonal antibody targeted at CD52, resulting in
prolonged T-cell depletion and modulation of the
lymphocyte repertoire
Laquinimod70 Reduction of inflammatory cell infiltration to central
nervous system
BG-1271 Activation of cellular defense mechanisms against
oxidative stress
Teriflunomide74 Reversible inhibition of mitochondrial enzymes
risk as non-twin siblings.82–84 The increased but imperfect
concordance between monozygotic twins adds weight to
the hypothesis that there is both a genetic and an environ-
mental contribution to MS risk.
Studies of half-siblings have proven to be another useful
method for understanding the genetic contribution to MS.
There have been two large studies of half-siblings in Canada,
which showed an increased risk to half-siblings of patients
with MS as compared to the general population, but less
than that for full siblings85,86 (Figure 32.6). Additionally,
there is also evidence for a parent-of-origin effect, with
half-siblings who have maternal genes in common having a
greater risk of developing MS than paternal half-siblings,86
raising the possibility of parental imprinting and epistatic
effects, although evidence at the genetic level for such
mechanisms has not been available until very recently.87,88
Furthermore, epidemiological studies of MS probands who
were either adopted themselves, or who have adopted rela-
tives, show no increase in risk of MS in the non-biological
relatives of MS patients despite cohabitation,89 although
risk for biological relatives of the same patients is increased
in a pattern similar to that of non-adopted probands, sug-
gesting that familial recurrence in MS is purely a genetic
phenomenon.89 Furthermore, there is no increased risk
to spouses90 or step-siblings91 of MS probands, making it
unlikely that a shared environmental factor or transmissible
agent is responsible for disease ignition.
4 9 6   •   G enomics in C linic a l P r actice
MODE OF REDUCTION IN RELAPSE RATES
ADMINISTRATION
Subcutaneous/ 20% (compared to placebo)70
intramuscular
Subcutaneous 29% (compared to placebo)65
Intravenous 59% (compared to placebo)66
Oral 45% (compared to placebo)67
Intravenous 74% (compared to IFN-β)32
Oral 23% (compared to placebo)68
Oral 47% (compared to placebo)69
Oral 31.5% (compared to placebo)74
It is apparent from these studies that the genetic con-
tribution to MS disease susceptibility is complex and likely
to involve multiple genes and epistatic effects. More formal
modelling using the familial risks as outlined above has sug-
gested that at least six genes (but likely to be more, with no
upper limit to the model) contribute to disease susceptibil-
ity, and that an autosomal dominant pattern of inheritance
fits the observed patterns much more closely than recessive
models.92 Finally, it is possible and indeed likely that genetic
heterogeneity exists, such that the risk alleles in one patient
may be different from those in another.92
T H E M A J O R H I S TO C O M PAT I B I L IT Y
COMPLEX
The association of the HLA region with susceptibility to
MS was first identified in 1972,93 and has since been con-
firmed using a variety of genetic analysis methods and
in a wide range of populations. The HLA DRB1*1501–
HLA-DQA1*0102–HLA-DQB1*0602 extended haplo-
type has been shown to be most strongly associated with
MS, increasing risk of MS three-fold, or by 17-fold in homo-
zygous patients.94 Disentangling the effect of each individ-
ual allele has proven to be difficult, as HLA-DRB1*1501,
HLA-DQA1*0102, and HLA-DQB1*0602 are in strong
linkage disequilibrium in European populations where MS
 1.9%
2.4% 3.2%
MZ: monozygotic
DZ: dizygotic
0.6%
Figure 32.6 Pedigree diagram summarizing lifetime recurrence risk to siblings, half-siblings, and offspring of MS patients, according to published
data.78,79,81–86
is most prevalent, and it was not clear for many years whether
the effect was due to one allele or a combined effect.94 In
African-American populations, the HLA-DRB1*1501
allele is more frequently inherited alone, which allowed
confirmation that it is this allele that is responsible for the
impact on disease susceptibility; African-American patients
carrying this allele were at similar increased risk as European
patients carrying the HLA DRB1*1501–DQA1*0102–
DQB1*0602 extended haplotype.95 The HLA-DRB1*1501
allele has now been shown to be associated with risk of
MS in a wide range of populations, including Northern
European and European-descended populations,96–100
Sardinian,16 Turkish,101 Mexican,102 Ashkenazi Jewish,103
and African-American populations,95 although association
is more difficult to demonstrate in populations where the
frequency of HLA-DRB1*1501 is low.16,95
A recent meta-regression confirmed a significant
positive association of disease prevalence with latitude
in populations of European descent, and when corrected
for HLA-DRB1 carriage, the association is enhanced.4
However, although HLA-DRB1 is clearly an important
variable in the European population, it does not account
for all observed variations in prevalence, suggesting a signif-
icant and possibly essential role for environmental factors in
disease evolution.4
The modest effect on MS susceptibility conferred by sin-
gle genes, including HLA-DRB1*1501, makes the study of
the genetic contribution to MS difficult. Prior to more wide-
spread availability of the technology for genome-wide analy-
sis, the most powerful methods for analysis were those using
G enetics a nd G enomics of N euro -P sychi atric D ise a ses , I I :   Multiple S clerosis   •   497
2.1%
MZ: 25% 4.4% 1.3%
DZ: 3.8%
3.2%
nuclear families with at least one affected member, and studies
involving this familial construct were initially the only ones
able to detect an effect for HLA-DRB1*1501,104–107 although
they remained largely underpowered. However, transmis-
sion disequilibrium analysis (TDT) of MS families was also
able to confirm that HLA-DRB1*1501 is over-transmitted
to MS patients as compared to their unaffected siblings,16,96,98
even in populations with a low prevalent frequency of
HLA-DRB1*1501.16 These early linkage studies also sug-
gested that association with MS might be present within other
genetic regions,104–107 albeit with a smaller effect size.106
With the advent of the publication of the Human
Genome Project,108,109 effective international collaborations,
and availability of low-cost genome-wide typing, explora-
tion of the complexities of the association of the MHC with
MS susceptibility has become a more realistic proposition,
and in recent years substantial advances have been made.
The association of HLA-DRB1*1501 has been unequivo-
cally confirmed in genome-wide association studies
(GWAS) in MS in large-scale studies of unrelated MS cases
and controls.100 The genome-wide approach has also been
able to identify other MHC susceptibility loci in addition
to the HLA-DRB1*1501 locus; in particular HLA-A*0201
has been shown to exert a protective effect,97,100 and if pres-
ent in conjunction with HLA-A*0201, it reduces the risk
from an odds ratio (OR) of 3.6 to 1.9.97 Independent effects
of other MHC loci have also been observed, including a
protective effect from HLA-C*05,110 and increased risk
associated with HLA-A*0301,97 HLA-DRB1*0301, and
HLA-DQB1*0201.100
 Large-scale genome-wide studies have also allowed
more detailed dissection of the MS risk associated with
the HLA-DRB1 locus, and it has become apparent that
trans-epistatic effects between HLA-DRB1 alleles modu-
late the overall risk. The presence of HLA-DRB1*1501 in
combination with HLA-DRB1* 17 confers a risk for MS
in excess of that seen from HLA-DRB1*1501 in conjunc-
tion with alternative HLA-DRB1 alleles.111 HLA-DRB1*08
is independently associated with a small increased risk of
MS, but if present together with HLA-DRB1*1501, the
risk of MS doubles.111,112 In contrast, HLA-DRB1*14 exerts
a protective effect,111,112 such that if inherited in conjunc-
tion with HLA-DRB1*1501, it almost entirely negates the
risk associated with HLA-DRB1*1501.112 Epistatic effects
have also been demonstrated between HLA-DRB1*1501,
HLA-DQA1*0102, which increases the risk of MS if
inherited together, and between HLA-DRB1*1501 and
HLA-DQB1*0602, thus also implicating HLA-DQ in MS
susceptibility.113
In concordance with the observation of a parent-of-
origin effect in epidemiological studies,86 over-transmission
of HLA-DRB1*1501 from mothers to affected offspring
has also been observed, and this effect is most marked
for daughters (OR 3.31, 95% CI 2.59–4.24), although
maternal transmission is also increased to sons (OR 2.13,
95% CI 5 1.44–3.14), in contrast to paternal transmis-
sion of HLA-DRB1*1501 (OR for daughter 2.14, 95%
CI 1.64–2.78; OR for son 2.16, 95% CI 1.37–3.39).87
Additionally, HLA-DRB1*08 is over-transmitted
from HLA-DRB1*1501-negative mothers to
HLA-DRB1*1501-positive daughters (OR 4.80, 95% CI
1.43–16.06), and maternal transmission of HLA-DRB1*14
to HLA-DRB1*1501-negative is also distorted (transmis-
sion to daughters: OR 0.40, 95% CI 0.11–1.43; sons: OR
0.14, 95% CI 0.02–0.84).87
N O N-M H C G E N E S
Early candidate-gene approaches exploring the association
of genes outside the MHC region with risk for MS were
unrewarding, and genome-linkage screens were underpow-
ered to detect the effect of other genes.104–106 Conflicting
results were obtained from studies of genes that seemed
likely candidates on a theoretical or functional basis to
be relevant to MS, such as the T-cell surface molecule
CTLA-4,114 or APOE-ε4,115 and later studies using more
powerful methods of analysis have failed to confirm a role
for these genes in MS susceptibility.100
However, a putative association with MS for a region on
chromosome 5p13 was suggested in one of the early genome
4 9 8   •   G enomics in C linic a l P r actice
linkage screens.104 Subsequently, IL-7R on chromosome
5p13 was confirmed as a susceptibility locus for MS in both
candidate-gene analyses116,117 and the first genome-wide
study in MS,110 thus unequivocally establishing IL-7R as an
MS susceptibility locus. This first genome-wide study also
identified two single-nucleotide polymorphisms (SNPs)
within the IL-2RA gene on chromosome 10p15 as a second
non-MHC MS susceptibility locus.110 Both these genes have
a much smaller effect size than HLA-DRB1*1501: IL-2RA
has an odds ratio of 1.25 (95% CI 1.16–1.36), while IL-7R
has an odds ratio of 1.18 (95% CI 1.11–1.26).110
More recently, an international collaborative GWAS
analyzed 9,772 MS cases and 17,376 controls. Nearly all pre-
vious associations were replicated, and further SNPs associ-
ated with MS were identified, so that almost 50 SNPs outside
the MHC complex have now been shown to be associated
with MS,100 including SNPs within genes for cytokines
(CXCR5, IL2RA, IL7R, IL7, IL12RB1, IL22RA2, IL12A,
IL12B, IRF8, TNFRSF1A, TNFRSF14, TNFSF14), sig-
naling molecules (CD37, CD40, CD58, CD80, CD86,
CLECL1), signal transduction (CBLB, GPR65, MALT1,
RGS1, STAT3, TAGAP, TYK2), as well as genes relevant
for vitamin D (CYP27B1, CYP24A1), and response to MS
therapies (VCAM1 and IL2RA).100 Furthermore, analysis
of known genetic roles using a gene ontology approach has
demonstrated that genes involved in immune system pro-
cesses, particularly lymphocyte function and specifically
T-cell activation and proliferation are over-represented in
MS, confirming our previous understanding of MS as a pri-
mary immunological disorder100 (Table 32.6).
Replication studies have since confirmed the asso-
ciation of MS susceptibility with CD6,118–120, CD58,121
CLEC16A,118 IL12A,122 IL12B,123 IRF8,119,120 KIF21B,124
MPHOSPH9/CDK2AP1,122 RGS1,122 STAT3,121
TMEM39A,123,124 TNFRSF1A,119,120 and TYK2.125 The
importance of replication studies to confirm genetic asso-
ciation with MS was emphasized by a small GWAS that
initially suggested association with KIF1B,126 which a
larger GWAS approach was unable to replicate.127 A further
study combining several independent datasets and utilizing
non-equivalence statistical methods subsequently demon-
strated that this association was a false positive.128
In addition, association of MS with a number of
other genetic loci has been suggested, including chro-
mosome 5p13.1,129 chromosome 3p24.1, the MLANA
gene on chromosome 9p24.1, and chromosome 2p21.121
However, these have not been replicated in other cohorts
and therefore remain unsubstantiated. Functional stud-
ies have now started to unravel the possible mechanisms
underlying the association of IL-7R and IL-2RA with
Table 32.6  CONFIRMED GENES ASSOCIATED WITH MS 100

CHROMOSOME RS NUMBER GENE P VALUE ODDS RATIO (95% CI)

6 HLA-DRB1*1501   <1x10–320
3.1
1 rs4648356 MMEL1 1.00x10 –14
1.14 (1.12–1.16)
1 rs11810217 EVI5 5.80x10–15 1.15 (1.13–1.16)
1 rs1335532 CD58 3.20x10 –16
1.22 (1.19–1.24)
1 rs1323292 RGS1 2.30x10–08 1.12 (1.1–1.14)
1 rs7522462 C1orf106 1.90x10 –09
1.11 (1.1–1.13)
2 rs12466022 no gene 6.20x10 –10
1.11 (1.1–1.13)
2 rs7595037 PLEK 5.10x10–11 1.11 (1.1–1.12)
2 rs17174870 MERTK 1.30x10 –08
1.11 (1.09–1.13)
2 rs10201872 SP140 1.80x10 –10
1.14 (1.12–1.16)
3 rs11129295 EOMES 1.20x10–09 1.11 (1.09–1.12)
3 rs2293370 TMEM39A 2.70x10 –09
1.13 (1.11–1.15)
3 rs9282641 CD86 1.00x10 –11
1.21 (1.18–1.24)
3 rs2243123 IL12A 7.20x10–06 1.08 (1.06–1.1)
4 rs228614 NFKB1 1.40x10 –07
1.09 (1.07–1.1)
5 rs6897932 IL7R 1.70x10–08 1.11 (1.09–1.13)
5 rs4613763 PTGER4 2.50x10 –16
1.2 (1.18–1.22)
5 rs2546890 IL12B 1.20x10 –11
1.11 (1.1–1.13)
6 rs12212193 BACH2 3.80x10–08 1.09 (1.08–1.1)
6 rs802734 THEMIS 5.50x10 –09
1.1 (1.09–1.12)
6 rs11154801 AHI1 1.00x10 –13
1.13 (1.11–1.15)
6 rs17066096 IL22RA2 6.00x10–13 1.14 (1.12–1.15)
6 rs13192841 no gene 1.30x10 –08
1.1 (1.09–1.12)
6 rs1738074 TAGAP 6.80x10–15 1.13 (1.12–1.15)
7 rs354033 ZNF767 4.70x10 –09
1.11 (1.1–1.13)
8 rs1520333 IL7 1.60x10 –07
1.1 (1.08–1.11)
8 rs4410871 MYC 7.70x10–09 1.11 (1.09–1.12)
8 rs2019960 PVT1 5.20x10 –09
1.12 (1.1–1.13)
10 rs3118470 IL2RA 3.20x10 –11
1.12 (1.1–1.13)
10 rs1250550 ZMIZ1 6.30x10–09 1.1 (1.09–1.12)
10 rs7923837 HHEX 4.90x10 –09
1.1 (1.08–1.11)
11 rs650258 CD6 2.00x10–11 1.12 (1.1–1.13)
12 rs1800693 TNFRSF1A 4.10x10 –14
1.12 (1.11–1.14)
12 rs10466829 CLECL1 1.40x10 –08
1.09 (1.08–1.11)
12 rs12368653 CYP27B1 1.70x10–09 1.1 (1.09–1.12)
14 rs4902647 ZFP36L1 9.30x10 –12
1.11 (1.1–1.13)
14 rs2300603 BATF 2.00x10 –08
1.11 (1.09–1.12)
14 rs2119704 GPR65 2.20x10–10 1.22 (1.19–1.25)
16 rs2744148 SOX8 8.40x10 –08
1.12 (1.1–1.14)
16 rs7200786 CLEC16A 8.50x10 –17
1.15 (1.13–1.16)
16 rs13333054 IRF8 1.30x10–08 1.11 (1.1–1.13)
17 rs9891119 STAT3 1.80x10 –10
1.11 (1.09–1.12)
17 rs180515 RPS6KB1 8.80x10–08 1.09 (1.08–1.11)

(continued)

G enetics a nd G enomics of N euro -P sychi atric D ise a ses , I I :   Multiple S clerosis   •   499
Table 32.6 CONTINUED

CHROMOSOME RS NUMBER GENE P VALUE ODDS RATIO (95% CI)

18 rs7238078 MALT1 2.50x10 –09
1.12 (1.1–1.14)
19 rs1077667 TNFSF14 9.40x10 –14
1.16 (1.14–1.18)
19 rs8112449 TYK2 1.20x10–06 1.08 (1.07–1.1)
19 rs874628 MPV17L2 1.30x10 –08
1.11 (1.09–1.12)
19 rs2303759 DKKL1 5.20x10 –09
1.11 (1.09–1.13)
20 rs2425752 CD40 5.10x10–10 1.11 (1.1–1.13)
20 rs2248359 CYP24A1 2.50x10 –11
1.12 (1.1–1.13)
20 rs6062314 ZBTB46 1.30x10–07 1.16 (1.14–1.19)
22 rs2283792 MAPK1 4.70x10 –09
1.1 (1.08–1.11)
22 rs140522 SCO2 1.70x10 –08
1.1 (1.09–1.12)

MS susceptibility. IL-7R is found in two forms: if exon
6 is included in the transcript, this encodes a membrane-
bound form of IL-7R, while if exon 6 is not included, a
soluble form of IL-7R is produced. When the risk allele
is present, it appears to modify exon splicing such that
exon 6 is skipped twice as often, thus increasing the pro-
duction of soluble IL-7R as compared to membrane-
bound IL-7R. This is likely to affect IL-7 signaling and
in turn T-cell proliferation and maintenance.117 Similarly,
IL-2RA is also found in soluble and membrane bound
forms. The two IL-2RA SNPs associated with MS have
been shown to be associated with variations in soluble
IL-2RA levels in healthy controls and MS patients.130
Soluble IL-2RA can inhibit T-cell signaling, but can also
enhance T-cell proliferation and expansion.130 Patterns of
cell surface expression of IL-2RA have also been shown
to vary depending on IL-2RA genotype.131 Thus, genetic
variation may lead to alterations in T-cell development,
functioning, and signaling that may in turn lead to the
development of MS.
G E N OT Y P E –P H E N OT Y P E
C O R R E L AT I O NS
A further question that is distinct from those relating to dis-
ease susceptibility is whether there are genetic influences on
phenotypes. Some epidemiological studies of familial dis-
ease have provided some evidence for concordance within
affected families, suggesting a genetic contribution to dis-
ease expression in addition to disease susceptibility.132–134 In
particular, concordance within families has been noted for
disease course132,134 and age at onset,134 but not for neuro-
logical symptoms at onset of disease.132 In addition there is
also some limited evidence for familial concordance in the
gender of affected patients.132
Early studies to further characterize potential pheno-
typical associations were limited by power,135 but associa-
tion between HLA-DRB1*1501 and a younger age at onset
of disease has now been confirmed by several independent
studies136–139; although it remains possible that this asso-
ciation is simply a consequence of the strong association
of HLA-DRB1*1501 with disease susceptibility, mak-
ing HLA-DRB1*1501-positive individuals more likely
to manifest disease earlier.138 Less robustly, an association
between HLA-DRB1*1501 and female gender has also
been observed,136,138,139 although other studies have failed to
confirm this.137 No association of HLA-DRB1*1501 with
disease course has so far been identified.24,136–138,140
Association of HLA with a more rapid progression of
disability was recognized shortly after its association with
disease susceptibility was confirmed,141 and HLA-DQ poly-
morphisms have been shown to influence the progression
of demyelination in viral mouse models of MS.142 However,
deeper analysis of the contribution of HLA-DRB1*1501
to accumulation of disability has been limited by power135
and by the difficulties inherent in the collection of dis-
ability data, including the slow rate of progression and the
high rate of variability. Consequently, different measures
of disability have been used in previous studies, making
direct comparisons difficult. However, no association of
HLA-DRB1*1501 with progression index135,138 or MSSS139
has been found. Association between HLA-DRB1*01,
HLA-DRB1*04, and more rapid time to EDSS 6 has been
observed using survival analysis, although no effect was
seen for time to EDSS 3.140 Finally, HLA-DRB1*01 was
over-represented in benign cases (EDSS ≤3 after ≥20 years
of disease) as compared to malignant cases (EDSS >6
within five years of disease onset)143 in a larger study uti-
lizing an extremes-of-outcome approach (a more powerful
statistical method144).

5 0 0   •   G enomics in C linic a l P r actice

 With the identification of other genes that are associ-
ated with MS susceptibility, interest has now focused on
the possible impact of these genes on disability accumula-
tion. Although once again different measures of disability
have been employed, making comparisons problematic,
early results show some promise. Findings from studies of
single SNPs have been conflicting: MGAT5, a glycosylat-
ing enzyme important for T-cell regulation, has been shown
to be associated with MSSS,145 but this was not detected
in a large GWAS.100 The immunoglobulin heavy chain
gene IGHV4-39 is thought to be important in the B-cell
response in MS,146 but an extremes-of-outcome approach
failed to detect an association.
However, approaches that analyze only one SNP at a
time may be unable to detect the combined effect of sev-
eral SNPs that could be more relevant for accumulation of
disability. Multivariate modelling suggests that, although
the contribution of individual SNPs to MSSS is minimal,
incorporation of multiple SNPs into such a model improves
its predictive power, and that IL-2 in particular is relevant
to this.147 Finally, a weak association of TYK2 with time to
EDSS 4 and 6, and KIF21B with time to EDSS 8 has also
been suggested,148 suggesting that genes may not take effect
across the whole disease trajectory, but may be relevant at
particular stages of disease. If this is true, then approaches
that combine all disability data into one measure such as
the MSSS may be unable to dissect temporal genetic effects.
Therefore, it will be important to clarify appropriate meth-
ods of phenotypical definition in order to adequately assess
the genetic contribution to progression of disability over
time.
Physical disability is not the only measurable pheno-
typical aspect of MS, and a limited number of studies have
been performed using alternative phenotypical outcomes.
Investigation of genetic influence on cognitive impair-
ment in MS has primarily focused on the contribution
of the APOE ε4 allele, and although no association has
been demonstrated to date,149–151 at least two studies may
have been underpowered. HLA-DRB1*1501 carriage has
been reported to be associated with a significantly poorer
performance on the PASAT test in a large cohort of MS
patients,152 suggesting that in addition to its effects on dis-
ease susceptibility and progression of disability, it may also
influence cognitive outcomes.
Genetic associations with radiological appearances in
MS have also been investigated. HLA-DRB1*01 carriage
was noted to be associated with a larger number of T2 lesions
and a smaller brain volume than in HLA-DRB1*01-negative
patients.152 More recent studies utilizing novel methods of
genetic analysis identified an association of smaller brain
G enetics a nd G enomics of N euro -P sychi atric D ise a ses , I I :   Multiple S clerosis   •   501
volume as measured radiologically with glutamate153 and
histone deacetylase gene polymorphisms,154 both of which
are implicated in neurodegenerative processes.
Although genotype–phenotype correlation analysis
remains in its infancy, there are several promising lines of
investigation, and it seems likely that, although the contri-
bution of individual genes to MS disease phenotype is mod-
est, genetic factors do play a role in the changes in disease
manifestation over time and its eventual prognosis.
G E N E –E N VI RO N M E N T
I N T E R AC T I O N S
The strong evidence for a significant environmental contri-
bution to disease etiology has led to a wide-ranging search
for relevant factors over many years. A number of associa-
tions have been reported, but only a limited number have
stood the test of time:  these include latitudinal gradient,
smoking, and prior infection with the Epstein-Barr virus.
Recent genetic associations have prompted investigators to
explore how these genes may interact with these environ-
mental factors to increase disease risk. The observation of
geographical variation in MS prevalence1 and a season-of-
birth effect for MS risk, with increased risk for those born
in May and a reduced risk for those born in November,155
hints at an association of disease susceptibility with the
number of sunshine hours in the day. Some have hypoth-
esized that this could possibly be mediated via vitamin D
levels, since exposure to sunlight increases the production
of vitamin D in the skin, which is then activated within
the liver.156 Vitamin D is involved in a number of different
regulatory mechanisms; primarily calcium homeostasis, but
it is also important in immune-system modulation and shift
in T-cell regulation156 and has been of interest as a possible
etiological agent and therapeutic target in MS for many
years.156 Activated vitamin D is a member of the steroid
receptor super-family, modifying transcription of down-
stream genes by binding a vitamin D–responsive element
(VDRE) in promoter regions.94,156 A  strongly conserved
VDRE has been shown to be present in the promoter
region of HLA-DRB1*1501, and functional studies have
shown that the addition of vitamin D modifies the expres-
sion of HLA-DRB1*1501 on the surface of lymphoblastoid
cells.94 Analysis of inheritance of rare genetic variants in MS
families has identified one such variant in the CYP27B1
gene that causes complete loss of gene function and is
associated with MS.157 Finally, MS-associated genetic loci
have been shown to be enriched for VDREs, particularly
in enhancer and promoter regions, and furthermore that
these VRDEs are more accessible to transcription factors in
lymphoblastoid cell lines are than non-immune cell types,88
such that vitamin D preferentially enhances transcription
of MS-associated genes within immunologically active cells.
Smoking is more common in patients with MS than the
general population,158,159 and it has also been linked with
earlier conversion to secondary progression.158,160,161 The
incidence and prevalence of MS in women has been shown
to have significantly increased over the last century, while
rates in men have remained relatively constant.23 One possi-
ble explanation for this may be the changing smoking habits
of women during the twentieth century.159 Tobacco taken in
other forms, such as the chewing tobacco that is popular in
Sweden, is not associated with an increase in the frequency
of MS, suggesting that risk is specific to the inhalational
mode of administration of tobacco and not tobacco per
se.162 It has been postulated that cigarette smoke has direct
cytotoxic effects in the lungs, leading to local immune
reactions and sensitization-to-self antigens, which in turn
increases the likelihood of developing autoimmune dis-
eases, including MS.159 Interaction of smoking with genetic
risk factors has been observed, such that smokers with two
genetic risk factors (presence of HLA-DRB1*1501 and
absence of the protective factor HLA-A*02) have an odds
ratio of 13.5 for developing MS, compared to nonsmokers
without either genetic risk factor, suggesting that priming
the immune system in the lungs by cigarette smoking may
tip the balance towards development of MS in genetically
susceptible people.163 However, a study comparing rates of
smoking and MS in MS patients and their healthy siblings
found no difference.164 This study did not take into account
HLA profiles of participants, which might be expected to
be more similar than in the general population, therefore
further replication of these findings in other populations is
required in order to understand the interactions between
smoking and genetic profile in MS.
The clear association of MS with underlying autoim-
mune dysfunction has led to interest in possible infec-
tious triggers of disease, although the only infectious
agent with consistent findings of possible association with
MS is the Epstein-Barr virus (EBV).159 EBV seropositiv-
ity is significantly higher in MS patients than in the gen-
eral population, but there appears to be a relationship
between seroconversion in the teenage years and increased
risk of MS,165,166 contrasting with populations with a low
frequency of MS where EBV is acquired much earlier
in life.159 It seems unlikely that EBV is a direct cause of
MS, as viral loads are variable and not always increased
in patients with newly diagnosed MS,159 and prospective
studies have found a prolonged lag period of several years
between acquisition of EBV and development of MS.166
5 0 2   •   G enomics in C linic a l P r actice
EBV is currently thought to act by stimulating overactivity
of the immune system, promoting a proinflammatory state,
and allowing survival of a T-cell population with broad-
ened specificity and higher likelihood of self-reactivity—
the “hygiene hypothesis,” where a more hygienic early life
leads to late exposure to childhood viruses and a less bal-
anced immune response as a consequence.159 Interaction
between EBV and genetic profile has been shown to affect
risk of MS.167,168 HLA-DRB1*1501-positive patients with
high titers of antibody against the EBV nuclear antigen
(EBNA) have a nine-fold increased risk for MS com-
pared to HLA-DRB1*1501-negative patients with low
anti-EBNA antibody titres.167 Additionally, the interaction
between high anti-EBNA titers and the presence of genetic
risk factors (presence of HLA-DRB1*1501 or absence of
HLA-A*02) has been shown to confer extra risk in addi-
tion to the risk conferred by each individual factor.168
It is becoming apparent that the interaction between
environmental factors and underlying genetic susceptibility
may be as important as, or more important than the role
of individual genes in predisposing individuals to MS. It is
likely that multiple factors will be relevant in any individual
patient, and disentangling the relative contribution of each
factor will require well-designed, large epidemiological
studies that incorporate genetic information into analyses.
There has been no investigation as of this writing of the con-
tribution of gene–environment interactions to the clinical
variability of disease, and this may also be a fruitful line of
investigation.
C O N C LU S I O N S
Evidence for a polygenic etiology for MS is strong. HLA-
DRB1*1501 undoubtedly carries the largest effect, con-
ferring increased risk of MS and affecting the clinical
manifestations of MS, including age at onset, gender ratios,
and accumulation of disability. More recently, other genes
have been shown to be associated with MS, and there is early
evidence that these may also play a role in altering the dis-
ease course, although the contribution of individual genes is
likely to be modest. Additionally, epistasis and gene–envi-
ronment interactions are important in MS disease suscep-
tibility. We are only at the very early stages of unravelling
the complexities of the role of genetics in MS, but impor-
tant advances have been made in recent years, particularly
after the publication of the Human Genome Project. A
deeper understanding of the mechanisms underlying the
onset of MS and the subsequent accumulation of disability
may allow development of new therapeutic targets with the
ultimate aim of preventing disability and perhaps finding a
cure for the disease.
REFERENCES
1. Compston A and Coles A, 2008. Multiple sclerosis. Lancet,
372(9648); 1502–1517.
2. Hirst C, Ingram G, Pickersgill T, Swingler R, Compston DAS, and
Robertson NP, 2009. Increasing prevalence and incidence of mul-
tiple sclerosis in South East Wales. J Neurol Neurosurg Psychiatry,
80(4); 386–391.
3. Visser EM, Wilde K, Wilson JF, Yong KK, and Counsell CE, 2012.
A new prevalence study of multiple sclerosis in Orkney, Shetland
and Aberdeen city. J Neurol Neurosurg Psychiatry, 83(7); 719–724.
4. Simpson S, Blizzard L, Otahal P, der Mei IV, and Taylor B, 2011.
Latitude is significantly associated with the prevalence of multiple
sclerosis:  a meta-analysis. J Neurol Neurosurg Psychiatry, 82(10);
1132–1141.
5. Zivadinov R, Iona L, Monti-Bragadin L, Bosco A, Jurjevic A, Taus C,
et al., 2003. The use of standardized incidence and prevalence rates
in epidemiological studies on multiple sclerosis. A meta-analysis
study. Neuroepidemiology, 22(1); 65–74.
6. Koch-Henriksen N and Sorensen PS, 2011. Why does the
north-south gradient of incidence of multiple sclerosis seem to have
disappeared on the northern hemisphere? J Neurol Sci, 311(1-2);
58–63.
7. Dean G and Kurtzke JF, 1971. On the risk of multiple sclerosis
according to age at immigration to South Africa. Br Med J, 3(5777);
725–729.
8. Kurtzke JF, Delasnerie-Lauprêtre N, and Wallin MT, 1998. Multiple
sclerosis in North African migrants to France. Acta Neurol Scand,
98(5); 302–309.
9. Elian M, Nightingale S, and Dean G, 1990. Multiple sclerosis among
United Kingdom-born children of immigrants from the Indian
subcontinent, Africa and the West Indies. J Neurol Neurosurg
Psychiatry, 53(10); 906–911.
10. Kurtzke JF and Bui-Quoc-Huong, 1980. Multiple sclerosis in a
migrant population:  2. Half-Orientals immigrating in childhood.
Ann Neurol, 8(3); 256–260.
11. Compston A, 1997. Genetic epidemiology of multiple sclerosis. J
Neurol Neurosurg Psychiatry, 62(6); 553–561.
12. Detels R, Visscher BR, Malmgren RM, Coulson AH, Lucia MV,
and Dudley JP, 1977. Evidence for lower susceptibility to multiple
sclerosis in Japanese-Americans. Am J Epidemiol, 105(4); 303–310.
13. Granieri E, Casetta I, Govoni V, Tola MR, Marchi D, Murgia SB,
et  al., 2000. The increasing incidence and prevalence of MS in a
Sardinian province. Neurology, 55(6); 842–848.
14. Lampis R, Morelli L, Congia M, Macis MD, Mulargia A, Loddo
M, et  al., 2000. The inter-regional distribution of HLA class  II
haplotypes indicates the suitability of the Sardinian population
for case-control association studies in complex diseases. Hum Mol
Genet, 9(20); 2959–2965.
15. Marrosu MG, Murru MR, Costa G, Murru R, Muntoni F, and
Cucca F, 1998. DRB1-DQA1-DQB1 loci and multiple sclerosis
predisposition in the Sardinian population. Hum Mol Genet, 7(8);
1235–1237.
16. Marrosu MG, Murru R, Murru MR, Costa G, Zavattari P, Whalen
M, et  al., 2001. Dissection of the HLA association with multiple
sclerosis in the founder isolated population of Sardinia. Hum Mol
Genet, 10(25); 2907–2916.
17. Montomoli C, Prokopenko I, Caria A, Ferrai R, Mander A, Seaman
S, et al., 2002. Multiple sclerosis recurrence risk for siblings in an iso-
lated population of Central Sardinia, Italy. Genet Epidemiol, 22(3);
265–271.
G enetics a nd G enomics of N euro -P sychi atric D ise a ses , I I :   Multiple S clerosis   •   503
18. Dyment DA, Cader MZ, Datta A, Broxholme SJ, Cherny SS, Willer
CJ, et al., 2008. A first stage genome-wide screen for regions shared
identical-by-descent in Hutterite families with multiple sclerosis.
Am J Med Genet B Neuropsychiatr Genet, 147B(4); 467–472.
19. Hader WJ, Seland TP, Hader MB, Harris CJ, and Dietrich DW,
1996. The occurrence of multiple sclerosis in the Hutterites of
North America. Can J Neurol Sci, 23(4); 291–295.
20. Dyment DA, Cader MZ, Willer CJ, Risch N, Sadovnick AD, and
Ebers GC, 2002. A multigenerational family with multiple sclerosis.
Brain, 125(7); 1474–1482.
21. Binzer M, Forsgren L, Holmgren G, Drugge U, and Fredrikson S,
1994. Familial clustering of multiple sclerosis in a northern Swedish
rural district. J Neurol Neurosurg Psychiatry, 57(4); 497–499.
22. Dyment DA, Cader MZ, Herrera BM, Ramagopalan SV, Orton
SM, Chao M, et al., 2008. A genome scan in a single pedigree with a
high prevalence of multiple sclerosis. J Neurol Neurosurg Psychiatry,
79(2); 158–162.
23. Orton SM, Herrera BM, Yee IM, Valdar W, Ramagopalan SV,
Sadovnick AD, et  al., 2006. Sex ratio of multiple sclerosis in
Canada: a longitudinal study. Lancet Neurol, 5(11); 932–936.
24. Weinshenker BG, Bass B, Rice GP, Noseworthy J, Carriere W,
Baskerville J, et al., 1989. The natural history of multiple sclerosis: a
geographically based study. I. Clinical course and disability. Brain,
112(1); 133–146.
25. Lublin FD and Reingold SC, 1996. Defining the clinical course of
multiple sclerosis: results of an international survey. National Multiple
Sclerosis Society (USA) Advisory Committee on Clinical Trials of
New Agents in Multiple Sclerosis. Neurology, 46(4); 907–911.
26. Runmarker B and Andersen O, 1993. Prognostic factors in a mul-
tiple sclerosis incidence cohort with twenty-five years of follow-up.
Brain, 116(Pt 1); 117–134.
27. Cottrell DA, Kremenchutzky M, Rice GP, Koopman WJ, Hader
W, Baskerville J, et al., 1999. The natural history of multiple sclero-
sis: a geographically based study. 5. The clinical features and natu-
ral history of primary progressive multiple sclerosis. Brain, 122(4);
625–639.
28. Koch M, Kingwell E, Rieckmann P, and Tremlett H, 2009. The
natural history of primary progressive multiple sclerosis. Neurology,
73(23); 1996–2002.
29. Boiko A, Vorobeychik G, Paty D, Devonshire V, Sadovnick D, and
University of British Columbia MS Clinic Neurologists, 2002. Early
onset multiple sclerosis:  a longitudinal study. Neurology, 59(7);
1006–1010.
30. Tremlett H and Devonshire V, 2006. Is late-onset multiple sclerosis
associated with a worse outcome? Neurology, 67(6); 954–959.
31. Cossburn M, Ingram G, Hirst C, Ben-Shlomo Y, Pickersgill TP, and
Robertson NP, 2012. Age at onset as a determinant of presenting
phenotype and initial relapse recovery in multiple sclerosis. Mult
Scler, 18(1); 45–54.
32. Tremlett H, Zhao Y, Joseph J, Devonshire V, and Neurologists UC,
2008. Relapses in multiple sclerosis are age- and time-dependent. J
Neurol Neurosurg Psychiatry, 79(12); 1368–1374.
33. Vercellino M, Romagnolo A, Mattioda A, Masera S, Piacentino C,
Merola A, et  al., 2009. Multiple sclerosis relapses:  a multivariable
analysis of residual disability determinants. Acta Neurol Scand,
119(2); 126–130.
34. Hirst CL, Ingram G, Pickersgill TP, and Robertson NP, 2012.
Temporal evolution of remission following multiple sclerosis relapse
and predictors of outcome. Mult Scler, 18(8); 1152–1158.
35. Campath 1-H in Multiple Sclerosis (CAMMS223) Trial
Investigators, Coles AJ, Compston DAS, Selmaj KW, Lake SL,
Moran S, et al., 2008. Alemtuzumab vs. interferon beta-1a in early
multiple sclerosis. N Engl J Med, 359(17); 1786–1801.
36. Confavreux C and Vukusic S, 2006. Natural history of multiple scle-
rosis: a unifying concept. Brain, 129(3); 606–616.
37. Langdon DW, 2011. Cognition in multiple sclerosis. Curr Opin
Neurol, 24(3); 244–249.
 38. Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Filippi
M, et al., 2011. Diagnostic criteria for multiple sclerosis: 2010 revi-
sions to the McDonald criteria. Ann Neurol, 69(2); 292–302.
39. Poser CM, Paty DW, Scheinberg L, McDonald WI, Davis FA,
Ebers GC, et  al., 1983. New diagnostic criteria for multiple
sclerosis:  guidelines for research protocols. Ann Neurol, 13(3);
227–231.
40. McDonald WI, Compston A, Edan G, Goodkin D, Hartung HP,
Lublin FD, et  al., 2001. Recommended diagnostic criteria for
multiple sclerosis:  guidelines from the International Panel on the
Diagnosis of Multiple Sclerosis. Ann Neurol, 50(1); 121–127.
41. Polman CH, Reingold SC, Edan G, Filippi M, Hartung HP, Kappos
L, et al., 2005. Diagnostic criteria for multiple sclerosis: 2005 revi-
sions to the “McDonald Criteria.” Ann Neurol, 58(6); 840–846.
42. Hirst C, Swingler R, Compston DAS, Ben-Shlomo Y, and
Robertson NP, 2008. Survival and cause of death in multiple scle-
rosis:  a prospective population-based study. J Neurol Neurosurg
Psychiatry, 79(9); 1016–1021.
43. Scalfari A, Neuhaus A, Degenhardt A, Rice GP, Muraro PA,
Daumer M, et al., 2010. The natural history of multiple sclerosis: a
geographically based study 10:  relapses and long-term disability.
Brain, 133(7); 1914–1929.
44. Kurtzke JF, 1983. Rating neurologic impairment in multiple sclero-
sis: an expanded disability status scale (EDSS). Neurology, 33(11);
1444–1452.
45.. Roxburgh RHSR, Seaman SR, Masterman T, Hensiek AE, Sawcer
SJ, Vukusic S, et al., 2005. Multiple Sclerosis Severity Score: using
disability and disease duration to rate disease severity. Neurology,
64(7); 1144–1151.
46. Poser S, Ritter G, Bauer HJ, Grosse-Wilde H, Kuwert EK, and Raun
NE, 1981. HLA-antigens and the prognosis of multiple sclerosis.
J Neurol, 225(3); 219–221.
47. Fischer JS, Rudick RA, Cutter GR, and Reingold SC, 1999. The
Multiple Sclerosis Functional Composite Measure (MSFC):  an
integrated approach to MS clinical outcome assessment. Mult Scler,
5(4); 244–250.
48. Debouverie M, 2009. Gender as a prognostic factor and its impact
on the incidence of multiple sclerosis in Lorraine, France. J Neurol
Sci, 286(1–2); 14–17.
49. Simone IL, Carrara D, Tortorella C, Liguori M, Lepore V, Pellegrini
F, et al., 2002. Course and prognosis in early-onset MS: comparison
with adult-onset forms. Neurology, 59(12); 1922–1928.
50. Renoux C, Vukusic S, Mikaeloff Y, Edan G, Clanet M, Dubois B,
et  al., 2007. Natural history of multiple sclerosis with childhood
onset. N Engl J Med, 356(25); 2603–2613.
51. Qiu W, Wu JS, Castley A, James I, Joseph J, Christiansen FT, et al.,
2010. Clinical profile and HLA-DRB1 genotype of late onset
multiple sclerosis in Western Australia. J Clin Neurosci, 17(8);
1009–1013.
52. Confavreux C and Vukusic S, 2006. Age at disability milestones in
multiple sclerosis. Brain, 129(3); 595–605.
53. Korn T, 2008. Pathophysiology of multiple sclerosis. J Neurol,
255(Suppl 6); 2–6.
54. Boniface K, Blom B, Liu YJ, and de Waal Malefyt R, 2008. From
interleukin-23 to T-helper 17 cells: human T-helper cell differentia-
tion revisited. Immunol Rev, 226; 132–146.
55. Lindquist S, Hassinger S, Lindquist JA, and Sailer M, 2011. The bal-
ance of pro-inflammatory and trophic factors in multiple sclerosis
patients: effects of acute relapse and immunomodulatory treatment.
Mult Scler, 17(7); 851–866.
56. Lucchinetti C, Brück W, Parisi J, Scheithauer B, Rodriguez M, and
Lassmann H, 1999. A quantitative analysis of oligodendrocytes
in multiple sclerosis lesions. A study of 113 cases. Brain, 122(12);
2279–2295.
57. Mews I, Bergmann M, Bunkowski S, Gullotta F, and Brück W, 1998.
Oligodendrocyte and axon pathology in clinically silent multiple
sclerosis lesions. Mult Scler, 4(2); 55–62.
5 0 4   •   G enomics in C linic a l P r actice
58. Bjartmar C, Kidd G, Mörk S, Rudick R, and Trapp BD, 2000.
Neurological disability correlates with spinal cord axonal loss and
reduced N-acetyl aspartate in chronic multiple sclerosis patients.
Ann Neurol, 48(6); 893–901.
59. Bitsch A, Schuchardt J, Bunkowski S, Kuhlmann T, and Brück W,
2000. Acute axonal injury in multiple sclerosis. Correlation with
demyelination and inflammation. Brain, 123(6); 1174–1183.
60. Lassmann H, 2011. Review:  the architecture of inflammatory
demyelinating lesions:  implications for studies on pathogenesis.
Neuropathol Appl Neurobiol, 37(7); 698–710.
61. Frischer JM, Bramow S, Dal-Bianco A, Lucchinetti CF, Rauschka
H, Schmidbauer M, et al., 2009. The relation between inflammation
and neurodegeneration in multiple sclerosis brains. Brain, 132(5);
1175–1189.
62. Lucchinetti C, Brück W, Parisi J, Scheithauer B, Rodriguez M,
and Lassmann H, 2000. Heterogeneity of multiple sclerosis
lesions:  implications for the pathogenesis of demyelination. Ann
Neurol, 47(6); 707–717.
63. Lucchinetti C and Brück W, 2004. The pathology of primary pro-
gressive multiple sclerosis. Mult Scler, 10(Suppl 1); S23–S30.
64. Brusaferri F and Candelise L, 2000. Steroids for multiple sclerosis
and optic neuritis: a meta-analysis of randomized controlled clinical
trials. J Neurol, 247(6); 435–442.
65. Zajicek JP, Hobart JC, Slade A, Barnes D, Mattison PG, and on
behalf of the MUSEC Research Group, 2012. MUltiple Sclerosis
and Extract of Cannabis:  results of the MUSEC trial. J Neurol
Neurosurg Psychiatry, 83(11); 1125–1132.
66. PRISMS Study Group, 1998. Randomised double-blind
placebo-controlled study of interferon beta-1a in relapsing/
remitting multiple sclerosis. PRISMS (Prevention of Relapses
and Disability by Interferon beta-1a Subcutaneously in Multiple
Sclerosis) Study Group. Lancet, 352(9139); 1498–1504.
67. Johnson KP, Brooks BR, Cohen JA, Ford CC, Goldstein J, Lisak
RP, et  al., 1995. Copolymer 1 reduces relapse rate and improves
disability in relapsing-remitting multiple sclerosis:  results of a
phase III multicenter, double-blind placebo-controlled trial. The
Copolymer 1 Multiple Sclerosis Study Group. Neurology, 45(7);
1268–1276.
68. Polman CH, O’Connor PW, Havrdova E, Hutchinson M, Kappos
L, Miller DH, et  al., 2006. A randomized, placebo-controlled
trial of natalizumab for relapsing multiple sclerosis. N Engl J Med,
354(9); 899–910.
69. Kappos L, Radue EW, O’Connor P, Polman C, Hohlfeld R,
Calabresi P, et al., 2010. A placebo-controlled trial of oral fingoli-
mod in relapsing multiple sclerosis. N Engl J Med, 362(5); 387–401.
70. Comi G, Jeffery D, Kappos L, Montalban X, Boyko A, Rocca MA,
et al., 2012. Placebo-controlled trial of oral laquinimod for multiple
sclerosis. N Engl J Med, 366(11); 1000–1009.
71. Gold R, Kappos L, Arnold DL, Bar-Or A, Giovannoni G, Selmaj
K, et al., 2012. Placebo-controlled phase 3 study of oral BG-12 for
relapsing multiple sclerosis. N Engl J Med, 367(12); 1098–1107.
72. Rice GP, Incorvaia B, Munari L, Ebers G, Polman C, D’Amico R,
et  al., 2001. Interferon in relapsing-remitting multiple sclerosis.
Cochrane Database Syst Rev, (4); CD002002.
73. Mikol DD, Barkhof F, Chang P, Coyle PK, Jeffery DR, Schwid SR,
et al., 2008. Comparison of subcutaneous interferon beta-1a with
glatiramer acetate in patients with relapsing multiple sclerosis (the
REbif vs Glatiramer Acetate in Relapsing MS Disease [REGARD]
study): a multicentre, randomised, parallel, open-label trial. Lancet
Neurol, 7(10); 903–914.
74. O’Connor P, Wolinsky JS, Confavreux C, Comi G, Kappos L,
Olsson TP, et al., 2011. Randomized trial of oral teriflunomide for
relapsing multiple sclerosis. N Engl J Med, 365(14); 1293–1303.
75. Wolinsky JS, Narayana PA, O’Connor P, Coyle PK, Ford C, Johnson
K, et al., 2007. Glatiramer acetate in primary progressive multiple
sclerosis:  results of a multinational, multicenter, double-blind,
placebo-controlled trial. Ann Neurol, 61(1); 14–24.
 76. Tur C, Montalban X, Tintore M, Nos C, Ro J, Aymerich FX, et al.,
2011. Interferon β-1b for the treatment of primary progressive mul-
tiple sclerosis: five-year clinical trial follow-up. Arch Neurol, 68(11);
1421–1427.
77. La Mantia L, Vacchi L, Pietrantonj CD, Ebers G, Rovaris M,
Fredrikson S, et  al., 2012. Interferon beta for secondary pro-
gressive multiple sclerosis. Cochrane Database Syst Rev, 1;
CD005181.
78. Sadovnick AD and Baird PA, 1988. The familial nature of multiple
sclerosis: age-corrected empiric recurrence risks for children and sib-
lings of patients. Neurology, 38(6); 990–991.
79. Robertson NP, Fraser M, Deans J, Clayton D, Walker N, and
Compston DA, 1996. Age-adjusted recurrence risks for relatives of
patients with multiple sclerosis. Brain, 119(2); 449–455.
80. O’Gorman C, Freeman S, Taylor BV, Butzkueven H, Australian and
New Zealand MS Genetics Consortium (ANZgene), et al., 2011.
Familial recurrence risks for multiple sclerosis in Australia. J Neurol
Neurosurg Psychiatry, 82(12); 1351–1354.
81. French Research Group on MS, 1992. Multiple sclerosis in 54 twin-
ships:  concordance rate is independent of zygosity. Ann Neurol,
32(6); 724–727.
82. Ebers GC, Bulman DE, Sadovnick AD, Paty DW, Warren S, Hader
W, et  al., 1986. A population-based study of multiple sclerosis in
twins. N Engl J Med, 315(26); 1638–1642.
83. Mumford CJ, Wood NW, Kellar-Wood H, Thorpe JW, Miller DH,
and Compston DA, 1994. The British Isles survey of multiple scle-
rosis in twins. Neurology, 44(1); 11–15.
84. Willer CJ, Dyment DA, Risch NJ, Sadovnick AD, Ebers GC,
and Group CCS, 2003. Twin concordance and sibling recurrence
rates in multiple sclerosis. Proc Natl Acad Sci U S A, 100(22);
12877–12882.
85. Sadovnick AD, Ebers GC, Dyment DA, and Risch NJ, 1996.
Evidence for genetic basis of multiple sclerosis. The Canadian
Collaborative Study Group. Lancet, 347(9017); 1728–1730.
86. Ebers GC, Sadovnick AD, Dyment DA, Yee IML, Willer CJ, and
Risch N, 2004. Parent-of-origin effect in multiple sclerosis: observa-
tions in half-siblings. Lancet, 363(9423); 1773–1774.
87. Chao MJ, Herrera BM, Ramagopalan SV, Deluca G, Handunetthi
L, Orton SM, et al., 2010. Parent-of-origin effects at the major his-
tocompatibility complex in multiple sclerosis. Hum Mol Genet,
19(18); 3679–3689.
88. Disanto G, Sandve GK, Berlanga-Taylor AJ, Ragnedda G, Morahan
JM, Watson CT, et al., 2012. Vitamin D receptor binding, chroma-
tin states and association with multiple sclerosis. Hum Mol Genet,
21(16); 3575–3586.
89. Ebers GC, Sadovnick AD, and Risch NJ, 1995. A genetic basis for
familial aggregation in multiple sclerosis. Canadian Collaborative
Study Group. Nature, 377(6545); 150–151.
90. Robertson NP, O’Riordan JI, Chataway J, Kingsley DP, Miller
DH, Clayton D, et al., 1997. Offspring recurrence rates and clini-
cal characteristics of conjugal multiple sclerosis. Lancet, 349(9065);
1587–1590.
91. Dyment DA, Yee IML, Ebers GC, Sadovnick AD, and Canadian
Collaborative Study Group, 2006. Multiple sclerosis in steps-
iblings:  recurrence risk and ascertainment. J Neurol Neurosurg
Psychiatry, 77(2); 258–259.
92. Lindsey JW, 2005. Familial recurrence rates and genetic models of
multiple sclerosis. Am J Med Genet A, 135(1); 53–58.
93. Jersild C, Svejgaard A, and Fog T, 1972. HL-A antigens and mul-
tiple sclerosis. Lancet, 1(7762); 1240–1241.
94. Ramagopalan SV, Knight JC, and Ebers GC, 2009. Multiple scle-
rosis and the major histocompatibility complex. Curr Opin Neurol,
22(3); 219–225.
95. Oksenberg JR, Barcellos LF, Cree BAC, Baranzini SE, Bugawan TL,
Khan O, et  al., 2004. Mapping multiple sclerosis susceptibility to
the HLA-DR locus in African Americans. Am J Hum Genet, 74(1);
160–167.
G enetics a nd G enomics of N euro -P sychi atric D ise a ses , I I :   Multiple S clerosis   •   505
96. Yaouanq J, Semana G, Eichenbaum S, Quelvennec E, Roth MP,
Clanet M, et al., 1997. Evidence for linkage disequilibrium between
HLA-DRB1 gene and multiple sclerosis. Science, 276(5313);
664–665.
97. Fogdell-Hahn A, Ligers A, Gronning M, Hillert J, and Olerup
O, 2000. Multiple sclerosis:  a modifying influence of HLA
class  I  genes in an HLA class  II associated autoimmune disease.
Tissue Antigens, 55(2); 140–148.
98. Harbo HF, Lie BA, Sawcer S, Celius EG, Dai KZ, Oturai A, et al.,
2004. Genes in the HLA class  I  region may contribute to the
HLA class II-associated genetic susceptibility to multiple sclerosis.
Tissue Antigens, 63(3); 237–247.
99. Cree BAC, Rioux JD, McCauley JL, Gourraud PAFD, Goyette P,
McElroy J, et al., 2010. A major histocompatibility Class I locus
contributes to multiple sclerosis susceptibility independently from
HLA-DRB1*15:01. PLoS One, 5(6); e11296.
100. IMSGC, 2011. Genetic risk and a primary role for cell-mediated
immune mechanisms in multiple sclerosis. Nature, 476(7359);
214–219.
101. Saruhan-Direskeneli G, Esin S, Baykan-Kurt B, Ornek I, Vaughan
R, and Eraksoy M, 1997. HLA-DR and -DQ associations with
multiple sclerosis in Turkey. Hum Immunol, 55(1); 59–65.
102. Alvarado-de la Barrera C, Zuniga-Ramos J, Ruiz-Morales JA,
Estanol B, Granados J, and Llorente L, 2000. HLA class II geno-
types in Mexican mestizos with familial and nonfamilial multiple
sclerosis. Neurology, 55(12); 1897–1900.
103. Kwon OJ, Karni A, Israel S, Brautbar C, Amar A, Meiner Z, et al.,
1999. HLA class  II susceptibility to multiple sclerosis among
Ashkenazi and non-Ashkenazi Jews. Arch Neurol, 56(5); 555–560.
104. Ebers GC, Kukay K, Bulman DE, Sadovnick AD, Rice G,
Anderson C, et al., 1996. A full genome search in multiple sclero-
sis. Nat Genet, 13(4); 472–476.
105. Haines JL, Ter-Minassian M, Bazyk A, Gusella JF, Kim DJ,
Terwedow H, et al., 1996. A complete genomic screen for multiple
sclerosis underscores a role for the major histocompatibility com-
plex. The Multiple Sclerosis Genetics Group. Nat Genet, 13(4);
469–471.
106. Sawcer S, Jones HB, Feakes R, Gray J, Smaldon N, Chataway J,
et al., 1996. A genome screen in multiple sclerosis reveals suscep-
tibility loci on chromosome 6p21 and 17q22. Nat Genet, 13(4);
464–468.
107. Sawcer S, Ban M, Maranian M, Yeo TW, Compston A, Kirby A,
et al., 2005. A high-density screen for linkage in multiple sclerosis.
Am J Hum Genet, 77(3); 454–467.
108. Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Baldwin
J, et al., 2001. Initial sequencing and analysis of the human genome.
Nature, 409(6822); 860–921.
109. Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton
GG, et  al., 2001. The sequence of the human genome. Science,
291(5507); 1304–1351.
110. IMSGC, Hafler DA, Compston A, Sawcer S, Lander ES, Daly
MJ, et al., 2007. Risk alleles for multiple sclerosis identified by a
genomewide study. N Engl J Med, 357(9); 851–862.
111. Dyment DA, Herrera BM, Cader MZ, Willer CJ, Lincoln MR,
Sadovnick AD, et al., 2005. Complex interactions among MHC
haplotypes in multiple sclerosis: susceptibility and resistance. Hum
Mol Genet, 14(14); 2019–2026.
112. Barcellos LF, Sawcer S, Ramsay PP, Baranzini SE, Thomson
G, Briggs F, et  al., 2006. Heterogeneity at the HLA-DRB1
locus and risk for multiple sclerosis. Hum Mol Genet, 15(18);
2813–2824.
113. Lincoln MR, Ramagopalan SV, Chao MJ, Herrera BM, Deluca
GC, Orton SM, et  al., 2009. Epistasis among HLA-DRB1,
HLA-DQA1, and HLA-DQB1 loci determines multiple sclerosis
susceptibility. Proc Natl Acad Sci U S A, 106(18); 7542–7547.
114. van Veen T, Crusius JBA, van Winsen L, Xia B, Barkhof F, Pena
AS, et  al., 2003. CTLA-4 and CD28 gene polymorphisms in
 susceptibility, clinical course and progression of multiple sclerosis.
J Neuroimmunol, 140(1-2); 188–193.
115. Burwick RM, Ramsay PP, Haines JL, Hauser SL, Oksenberg JR,
Pericak-Vance MA, et al., 2006. APOE epsilon variation in mul-
tiple sclerosis susceptibility and disease severity:  some answers.
Neurology, 66(9); 1373–1383.
116. Lundmark F, Duvefelt K, Iacobaeus E, Kockum I, Wallstrom E,
Khademi M, et al., 2007. Variation in interleukin 7 receptor alpha
chain (IL7R) influences risk of multiple sclerosis. Nat Genet,
39(9); 1108–1113.
117. Gregory SG, Schmidt S, Seth P, Oksenberg JR, Hart J, Prokop A,
et al., 2007. Interleukin 7 receptor alpha chain (IL7R) shows allelic
and functional association with multiple sclerosis. Nat Genet,
39(9); 1083–1091.
118. Hoppenbrouwers IA, Aulchenko YS, Janssens AC, Ramagopalan
SV, Broer L, Kayser M, et  al., 2009. Replication of CD58 and
CLEC16A as genome-wide significant risk genes for multiple scle-
rosis. J Hum Genet, 54(11); 676–680.
119. De Jager PL, Jia X, Wang J, de Bakker PIW, Ottoboni L, Aggarwal
NT, et  al., 2009. Meta-analysis of genome scans and replication
identify CD6, IRF8 and TNFRSF1A as new multiple sclerosis
susceptibility loci. Nat Genet, 41(7); 776–782.
120.. International MS Genetics Consortium (IMSGC), 2011. The
genetic association of variants in CD6, TNFRSF1A and IRF8 to
multiple sclerosis:  a multicenter case-control study. PLoS One,
6(4); e18813.
121. Patsopoulos NA, Bayer Pharma MS Genetics Working Group
Steering Committee of Studies Evaluating IFNβ-1band a
CCR1-Antagonist, Australia and New Zealand Genetics
(ANZgene) Consortium, GeneMSA, et  al., 2011. Genome-wide
meta-analysis identifies novel multiple sclerosis susceptibility loci.
Ann Neurol, 70(6); 897–912.
122. International Multiple Sclerosis Genetics Consortium (IMSGC),
2010. IL12A, MPHOSPH9/CDK2AP1 and RGS1 are novel mul-
tiple sclerosis susceptibility loci. Genes Immun, 11(5); 397–405.
123. Varade J, Comabella M, Ortiz MA, Arroyo R, Fernandez O,
Pinto-Medel MJ, et al., 2012. Replication study of 10 genes show-
ing evidence for association with multiple sclerosis: validation of
TMEM39A, IL12B and CLBL genes. Mult Scler, 18(7); 959–965.
124.. International Multiple Sclerosis Genetics Consortium
(IMSGC), 2010. Comprehensive follow-up of the first
genome-wide association study of multiple sclerosis identifies
KIF21B and TMEM39A as susceptibility loci. Hum Mol Genet,
19(5); 953–962.
125. Ban M, Goris A, Lorentzen AR, Baker A, Mihalova T, Ingram G,
et al., 2009. Replication analysis identifies TYK2 as a multiple scle-
rosis susceptibility factor. Eur J Hum Genet, 17(10); 1309–1313.
126. Aulchenko YS, Hoppenbrouwers IA, Ramagopalan SV, Broer L,
Jafari N, Hillert J, et al., 2008. Genetic variation in the KIF1B locus
influences susceptibility to multiple sclerosis. Nat Genet, 40(12);
1402–1403.
127. International Multiple Sclerosis Genetics Consortium (IMSGC),
Booth DR, Heard RN, Stewart GJ, Cox M, Scott RJ, et al., 2010.
Lack of support for association between the KIF1B rs10492972[C]
variant and multiple sclerosis. Nat Genet, 42(6); 469–470; author
reply, 470–471.
128. Gourraud PA and (IMSGC) IMSGC, 2011. When is the absence
of evidence, evidence of absence? Use of equivalence-based anal-
yses in genetic epidemiology and a conclusion for the KIF1B
rs10492972*C allelic association in multiple sclerosis. Genet
Epidemiol, 35(6); 568–571.
129. Matesanz F, Gonzalez-Perez A, Lucas M, Sanna S, Gayan J, Urcelay
E, et al., 2012. Genome-wide association study of multiple sclerosis
confirms a novel locus at 5p13.1. PLoS One, 7(5); e36140.
130. Maier LM, Anderson DE, Severson CA, Baecher-Allan C, Healy
B, Liu DV, et al., 2009. Soluble IL-2RA levels in multiple sclerosis
5 0 6   •   G enomics in C linic a l P r actice
subjects and the effect of soluble IL-2RA on immune responses.
J Immunol, 182(3); 1541–1547.
131. Dendrou CA, Plagnol V, Fung E, Yang JHM, Downes K, Cooper
JD, et al., 2009. Cell-specific protein phenotypes for the autoim-
mune locus IL2RA using a genotype-selectable human biore-
source. Nat Genet, 41(9); 1011–1015.
132. Robertson NP, Clayton D, Fraser M, Deans J, and Compston DA,
1996. Clinical concordance in sibling pairs with multiple sclerosis.
Neurology, 47(2); 347–352.
133. Brassat D, Azais-Vuillemin C, Yaouanq J, Semana G, Reboul J,
Cournu I, et al., 1999. Familial factors influence disability in MS
multiplex families. French Multiple Sclerosis Genetics Group.
Neurology, 52(8); 1632–1636.
134. Hensiek AE, Seaman SR, Barcellos LF, Oturai A, Eraksoi M,
Cocco E, et al., 2007. Familial effects on the clinical course of mul-
tiple sclerosis. Neurology, 68(5); 376–383.
135. Weinshenker BG, Santrach P, Bissonet AS, McDonnell SK,
Schaid D, Moore SB, et  al., 1998. Major histocompatibility
complex class  II alleles and the course and outcome of MS:  a
population-based study. Neurology, 51(3); 742–747.
136. Celius EG, Harbo HF, Egeland T, Vartdal F, Vandvik B, and
Spurkiand A, 2000. Sex and age at diagnosis are correlated with
the HLA-DR2, DQ6 haplotype in multiple sclerosis. J Neurol Sci,
178(2); 132–135.
137. Masterman T, Ligers A, Olsson T, Andersson M, Olerup O, and
Hillert J, 2000. HLA-DR15 is associated with lower age at onset in
multiple sclerosis. Ann Neurol, 48(2); 211–219.
138. Hensiek AE, Sawcer SJ, Feakes R, Deans J, Mander A, Akesson
E, et  al., 2002. HLA-DR 15 is associated with female sex and
younger age at diagnosis in multiple sclerosis. J Neurol Neurosurg
Psychiatry, 72(2); 184–187.
139. Smestad C, Brynedal B, Jonasdottir G, Lorentzen AR, Masterman
T, Akesson E, et al., 2007. The impact of HLA-A and -DRB1 on
age at onset, disease course and severity in Scandinavian multiple
sclerosis patients. Eur J Neurol, 14(8); 835–840.
140. Romero-Pinel L, Pujal JM, Martinez-Yelamos S, Gubieras L, Matas
E, Bau L, et al., 2011. HLA-DRB1: genetic susceptibility and dis-
ability progression in a Spanish multiple sclerosis population. Eur J
Neurol, 18(2); 337–342.
141. Jersild C, Fog T, Hansen GS, Thomsen M, Svejgaard A, and
Dupont B, 1973. Histocompatibility determinants in multiple
sclerosis, with special reference to clinical course. Lancet, 2(7840);
1221–1225.
142. Pavelko KD, Drescher KM, McGavern DB, David CS, and
Rodriguez M, 2000. HLA-DQ polymorphism influences progres-
sion of demyelination and neurologic deficits in a viral model of
multiple sclerosis. Mol Cell Neurosci, 15(6); 495–509.
143. DeLuca GC, Ramagopalan SV, Herrera BM, Dyment DA,
Lincoln MR, Montpetit A, et al., 2007. An extremes of outcome
strategy provides evidence that multiple sclerosis severity is deter-
mined by alleles at the HLA-DRB1 locus. Proc Natl Acad Sci U S
A, 104(52); 20896–20901.
144. Risch N and Zhang H, 1995. Extreme discordant sib pairs for
mapping quantitative trait loci in humans. Science, 268(5217);
1584–1589.
145. Brynedal B, Wojcik J, Esposito F, Debailleul V, Yaouanq J,
Martinelli-Boneschi F, et al., 2010. MGAT5 alters the severity of
multiple sclerosis. J Neuroimmunol, 220(1-2); 120–124.
146. Watson CT, Ramagopalan SV, Morrison KM, Ebers GC, and
Breden F, 2010. IGHV4-39 deletion polymorphism does not asso-
ciate with risk or outcome of multiple sclerosis. J Neuroimmunol,
225(1-2); 164–166.
147. Sombekke MH, Arteta D, van de Wiel MA, Crusius JBA, Tejedor
D, Killestein J, et  al., 2010. Analysis of multiple candidate genes
in association with phenotypes of multiple sclerosis. Mult Scler,
16(6); 652–659.
 148. Harding K, Ingram G, Cossburn M, Hirst C, Pickersgill T,
Ben-Shlomo Y, et  al., 2012. Genotype-phenotype correlation
for non-HLA disease associated risk alleles in multiple sclerosis.
Neurosci Lett, 526(1); 15–19.
149. van der Walt A, Stankovich J, Bahlo M, Taylor BV, van der Mei
IAF, Foote SJ, et  al., 2009. Apolipoprotein genotype does not
influence MS severity, cognition, or brain atrophy. Neurology,
73(13); 1018–1025.
150. Ghaffar O, Reis M, Pennell N, O’Connor P, and Feinstein A, 2010.
APOE epsilon4 and the cognitive genetics of multiple sclerosis.
Neurology, 74(20); 1611–1618.
151. Carmona O, Masuet C, Santiago O, Alia P, Moral E, Alonso-
Magdalena L, et al., 2011. Multiple sclerosis and cognitive decline: is
ApoE-4 a surrogate marker? Acta Neurol Scand, 124(4); 258–263.
152. Okuda DT, Srinivasan R, Oksenberg JR, Goodin DS, Baranzini
SE, Beheshtian A, et al., 2009. Genotype-phenotype correlations in
multiple sclerosis: HLA genes influence disease severity inferred by
1HMR spectroscopy and MRI measures. Brain, 132(1); 250–259.
153. Strijbis EM, Inkster B, Vounou M, Naegelin Y, Kappos L, Radue
EW, et al., 2013. Glutamate gene polymorphisms predict brain vol-
umes in multiple sclerosis. Mult Scler. 19(3); 281–288.
154. Inkster B, Strijbis EMM, Vounou M, Kappos L, Radue EW,
Matthews PM, et al., 2013. Histone deacetylase gene variants pre-
dict brain volume changes in multiple sclerosis. Neurobiol Aging.
34(1):238–247.
155. Willer CJ, Dyment DA, Sadovnick AD, Rothwell PM, Murray TJ,
Ebers GC, et al., 2005. Timing of birth and risk of multiple sclero-
sis: population based study. BMJ, 330(7483); 120.
156. Smolders J, Damoiseaux J, Menheere P, and Hupperts R, 2008.
Vitamin D as an immune modulator in multiple sclerosis, a review.
J Neuroimmunol, 194(1–2); 7–17.
157. Ramagopalan SV, Dyment DA, Cader MZ, Morrison KM,
Disanto G, Morahan JM, et  al., 2011. Rare variants in the
CYP27B1 gene are associated with multiple sclerosis. Ann Neurol,
70(6); 881–886.
G enetics a nd G enomics of N euro -P sychi atric D ise a ses , I I :   Multiple S clerosis   •   507
158. Hernán MA, Jick SS, Logroscino G, Olek MJ, Ascherio A, and Jick
H, 2005. Cigarette smoking and the progression of multiple sclero-
sis. Brain, 128(6); 1461–1465.
159. O’Gorman C, Lucas R, and Taylor B, 2012. Environmental risk
factors for multiple sclerosis: a review with a focus on molecular
mechanisms. Int J Mol Sci, 13(9); 11718–11752.
160. Healy BC, Ali EN, Guttmann CRG, Chitnis T, Glanz BI, Buckle
G, et al., 2009. Smoking and disease progression in multiple sclero-
sis. Arch Neurol, 66(7); 858–864.
161. Sundström P and Nyström L, 2008. Smoking worsens the progno-
sis in multiple sclerosis. Mult Scler, 14(8); 1031–1035.
162. Hedström AK, Bäarnhielm M, Olsson T, and Alfredsson L, 2009.
Tobacco smoking, but not Swedish snuff use, increases the risk of
multiple sclerosis. Neurology, 73(9); 696–701.
163. Hedström AK, Sundqvist E, Bäarnhielm M, Nordin N, Hillert J,
Kockum I, et al., 2011. Smoking and two human leukocyte anti-
gen genes interact to increase the risk for multiple sclerosis. Brain,
134(3); 653–664.
164. Jafari N, Hoppenbrouwers IA, Hop WCJ, Breteler MMB, and
Hintzen RQ, 2009. Cigarette smoking and risk of MS in multiplex
families. Mult Scler, 15(11); 1363–1367.
165. Haahr S, Koch-Henriksen N, Moller-Larsen A, Eriksen LS, and
Andersen HM, 1995. Increased risk of multiple sclerosis after late
Epstein-Barr virus infection:  a historical prospective study. Mult
Scler, 1(2); 73–77.
166. Haahr S, Plesner AM, Vestergaard BF, and Höllsberg P, 2004. A
role of late Epstein-Barr virus infection in multiple sclerosis. Acta
Neurol Scand, 109(4); 270–275.
167. De Jager PL, Simon KC, Munger KL, Rioux JD, Hafler DA, and
Ascherio A, 2008. Integrating risk factors: HLA-DRB1* 1501 and
Epstein-Barr virus in multiple sclerosis. Neurology, 70(13 Pt 2);
1113–1118.
168 Sundqvist E, Sundström P, Lindén M, Hedström AK, Aloisi
F, Hillert J, et  al., 2012. Epstein-Barr virus and multiple sclero-
sis: interaction with HLA. Genes Immun, 13(1); 14–20.
 33.
GENETICS AND GENOMICS OF NEURO-PSYCHIATRIC
DISEASES, III: THE COMMON DEMENTIAS
Amy Gerrish, Rebecca Sims, and Julie Williams
INTRODUCTION
Dementia is a collective syndrome of numerous diseases and
conditions that are commonly defined by a global decline
in cognitive functioning, including memory problems, dif-
ficulties with language use, as well as issues with executive
functioning and other activities of daily living. The World
Alzheimer Report estimates the global figure of dementia
sufferers to be 35.6 million[1]‌, with this number predicted
to almost double every 20 years. This increasing incidence
makes dementia a major health concern, with 115.4  mil-
lion people predicted to be diagnosed with dementia by
2050. However, there is hope for future therapeutic inter-
vention through recent advances in the molecular genetics
of the dementias, which are pinpointing potential patho-
genic mechanisms. In this chapter, we focus on our current
understanding of the molecular genetics of the most com-
mon forms of dementia: Alzheimer’s disease (AD), vascular
dementia (VaD), and frontotemporal dementia (FTD).
A L Z H E I M E R’ S D I S E A S E
AD is a devastating, progressive neurodegenerative disor-
der that is the most common form of dementia[2]‌, account-
ing for an estimated 50–60% of all dementia cases. AD is
characterized by deficits in memory and decline in at least
one of these areas of cognition: aphasia, apraxia, agnosia,
and executive functioning. Sufferers may also experience
a range of behavioral symptoms such as psychosis, agita-
tion, and depression. The mean duration of disease after
symptom-onset is 10.3 years, but it may range from two to
over 20 years[3]. Diagnosis of AD was standardized in 1984[4],
with the criteria revised in 2011[5]. Neuropathologically, the
disease is characterized by extracellular senile plaques (SPs)
and intracellular neurofibrillary tangles (NFTs). SPs pre-
dominantly consist of extracellular deposits of β-amyloid
508
(Aβ) peptides, which are proteolytic derivatives of the amy-
loid precursor protein (APP). Two species of fibrillar Aβ,
which are 40 and 42 amino acids long and known as Aβ40
and Aβ42, respectively, are found in SPs. Aβ42, the slightly
more hydrophobic form, is particularly prone to aggrega-
tion. NFTs are relatively insoluble, intracellular aggregates
composed of paired helical filaments that occupy the cell
body and may extend into the dendrites. These tangles con-
tain abnormally phosphorylated microtubule-associated
tau (τ) protein. NFTs are not unique to AD and occur in
other neurodegenerative disorders, including frontotem-
poral dementia, corticobasal degeneration, progressive
supranuclear palsy, and dementia pugilistica. A small num-
ber are also found in normal aging.
SPs and NFTs are region-specific, occurring predomi-
nantly in the hippocampus, entorhinal cortex, and associa-
tion areas of the neocortex, with the cognitive phenotype of
AD reflecting the location of the lesions. Memory impair-
ment suffered by AD patients has been related to hippocam-
pal dysfunction, while aphasia is produced by dysfunction
of the left posterior association cortex and visuospatial dys-
function is linked to the right posterior association cortex[6]‌.
There is significant neuronal loss in AD. Degeneration of
the basal forebrain cholinergic system is a classic feature
and has been correlated with the severity of dementia in the
later stages of disease. In addition, neuronal loss or atrophy
in the locus ceruleus and raphe nuclei of the brainstem leads
to deficits in noradrenergic and serotonergic transmitters,
respectively[7].
The genetics of AD is traditionally divided into sporadic
late-onset AD (LOAD) with an age at onset greater than
60–65 years, and rare, early-onset AD (EOAD), with age
at onset below 60 years. EOAD accounts for fewer than 5%
of all AD cases[8]‌, with a small proportion segregating in an
autosomal dominant manner. The prevalence of EOAD
has been estimated to be 41.2 per 100,000 individuals at
risk, with the prevalence of autosomal dominant EOAD
calculated to be 5.3 per 100,000 individuals at risk (12.9% the first pathway, APP is cleaved by α-secretase, which cuts
of EOAD)[9]. within the Aβ domain. This precludes the generation of the
Aβ peptide and therefore the pathway involving α-secretase
cleavage of the APP protein is termed “non-amyloidogenic.”
G E N ET I C S O F E A R LY- O N S ET
In the alternative amyloidogenic pathway, APP is cleaved by
A L Z H E I M E R’ S D I S E A S E
β-secretase to generate soluble extracellular β-N-terminal
fragment (β-APPs) and the C-terminal APP fragment,
Many of the initial studies of the molecular genetics of AD
C99. Subsequent cleavage of C99 by γ-secretase results in
focused on EOAD in rare families where the disease is usu-
the formation of the Aβ peptide.
ally transmitted in an autosomal dominant fashion. The
Autosomal dominant mutations in the APP and pre-
identification of rare mutations in three genes that cause
senilin genes appear to cause AD through alterations in
EOAD has had a major impact on our understanding of the
APP processing. In APP, pathogenic mutations are found
pathogenesis of the disease.
in exons 16 and 17, clustered around the β- and γ-secretase
The first success in the search for AD genes was the
sites. Presenilin 1 and 2 (PSEN1 and PSEN2) have been
discovery of a positive genetic linkage between a locus on
implicated in γ-secretase function[21;  22]. Mutations in
chromosome 21q and autosomal dominant EOAD[10]. The
PSEN1 and PSEN2 are predominantly located in the
amyloid precursor protein (APP) gene was cloned and local-
highly conserved transmembrane domains and are pre-
ized to the long arm of chromosome 21, and a point mutation
sumed to distort the precise conformation of the molecule
was discovered in a family showing linkage on chromosome
within the membrane.
21[11]. Several pathogenic mutations have since been identi-
The discovery of the aforementioned mutations and
fied in autosomal dominant EOAD families, and there are
their mechanism of action led to the formulation of the
currently 32 APP mutations listed in the Alzheimer Disease
amyloid cascade hypothesis, which postulates that elevated
and Frontotemporal Dementia Mutation Database (http://
levels of Aβ result in the oligomerisation, fibrillogenesis,
www.molgen.ua.ac.be/ADMutations/default.cfm). These
and aggregation of the peptide. This is thought to initi-
act to increase production of Aß, specifically Aß42, the most
ate a cascade of injurious events, including direct neuronal
amyloidogenic form of the peptide[12; 13].
damage plus free radical production and glial activation.
It was noted that the APP mutations did not account for a
Soluble Aβ oligomers are themselves a neurotoxic species
substantial proportion of autosomal dominant EOAD cases,
prior to deposition in SPs[23], as they can lead to the genera-
meaning that additional pathogenic loci were still to be identi-
tion of hydroxyl ions and oxidative injury of phospholipid
fied. In 1992, an AD linkage study identified a candidate locus
membranes, and this loss of membrane integrity leads to
on the long arm of chromosome 14, and, a few years later, the
cell death[24]. Aβ may also exert toxicity through influ-
presenilin 1 (PSEN1) gene was isolated by a positional clon-
ences on calcium channels, exaggerating calcium influx and
ing[14; 15]. Similarly, evidence for a locus on chromosome 1q was
initiating cell death. The aggregation of Aβ in plaques is
reported in several related kindreds of German-Russian origin
also thought to trigger the inflammatory response and the
with multiple cases of autosomal dominant EOAD[16]. The
activation of microglia that surround extracellular lesions.
presenilin 2 (PSEN2) gene was subsequently identified based
These reactive changes may further compromise brain
on its homology to PSEN1[17; 18]. The AD & FTD Mutation
function.
Database lists 185 PSEN1 pathogenic mutations and 13
PSEN2 pathogenic mutations. The effect of these mutations
again seems to be the enhanced production of Aß42[19; 20]. It
G E N ET I C S O F L AT E - O N S ET
should be noted that around one-third of dominantly inher-
A L Z H E I M E R’ S D I S E A S E
ited AD cases are not associated with known mutations in
either the APP or PSEN genes, which implies the existence of
While much progress has been made in determining the
further disease loci[9]‌.
causes of EOAD, the inheritance pattern of the more com-
mon, late-onset form of the disease appears to be much
APP PROCESSING AND THE AMYLOID more complex. Considerable evidence shows that familial
C A S C A D E H Y P OT H E S I S factors play an important role in the etiology of LOAD,
with the heritability estimated to be between 58% and
APP is a ubiquitously expressed trans-membrane protein 79%[25]. Numerous studies have investigated the role of
that is processed via one of two proteolytic pathways. In the early-onset genes in the more common late-onset form

G enetics and G enomics o f N euro -P sychiatric D iseases , I I I :   T he C ommon D ementias   •   509

of the disease. However, the results of these studies are far
from compelling, with a number of conflicting publications
for APP[25a–25e] and the PSENs[26; 27]. Most convincingly, the
recent powerful genome-wide association studies show no
evidence for a role of APP or the PSENs in the etiology of
LOAD[28–32].
C A N D I DAT E - G E N E S T U D I E S
A N D A P O L I P O P ROT E I N E (A P O E)
Successes in linkage analyses have not been restricted to
EOAD families. Late-onset families have also been exam-
ined, leading to the identification of a linkage region for
LOAD on chromosome 19[33]. Subsequent studies identi-
fied an association between variation at the apolipoprotein
E (APOE) gene on chromosome 19q13.2 with LOAD[34].
There are three major isoforms of apoE:  E2, E3 and E4,
which differ from each other by one or two amino acids.
These isoforms are coded for by alleles ε2, ε3 and ε4,
respectively. Numerous studies have consistently found an
increased risk of LOAD in carriers of the ε4 allele, whereas
the ε2 allele appears to have a protective effect.
In the subsequent 15  years, little progress was made
in the field of AD genetics despite over 1300 association
studies examining more than 650 genes[26]. Although many
significant associations with AD were reported, most have
failed to replicate in subsequent studies. Recently, thanks
to our increased knowledge of human genetic variation
provided by the Human Genome and HapMap projects,
and improvements in high-throughput genotyping, it has
become possible to take a genome-wide hypothesis-free
approach to association studies. Since the first genome-wide
association studies (GWAS) in 2005, GWAS have proved
enormously successful in identifying susceptibility loci in
complex disease[35].
G E N O M E -WI D E A S S O C I AT I O N
S T U D I E S O F A L Z H E I M E R’S D I S E A S E
Since 2007, multiple GWAS of AD have been performed.
The early GWAS studies were not sufficiently powered
to detect the modest effect sizes indicative of common
susceptibility alleles[35a]. In 2009, two AD GWAS stud-
ies were published that were much larger and therefore
more powerful than previous studies. The first study was
performed by the Genetic and Environmental Risk in
Alzheimer’s Disease (GERAD) consortium[29]. The second
was performed independently by the European Alzheimer’s
Disease Initiative (EADI) consortium[30]. The most signifi-
cant single-nucleotide polymorphism (SNP) identified in
5 1 0   •   G enomics in C linical P ractice
both studies was a proxy for the APOE ε4 SNP (rs429358)
rs2075650 (GERAD P  =  2  × 10–157; EADI P  =  1  × 10–
130
). GERAD also reported genome-wide significant evi-
dence for two novel susceptibility loci: rs11136000 in the
CLU gene (P = 8.5 × 10–10, odds ratio [OR] = 0.86) and
rs3851179 and rs541458, two SNPs 5′ to the PICALM
gene (P  =  1.3  × 10–9, OR  =  0.86 and P  =  8.3  × 10–10,
OR  =  0.86 respectively). Encouragingly, EADI also iden-
tified genome-wide significant association with the same
allele of rs11136000 in CLU (P = 7.5 × 10–9, OR = 0.86)
and found support for the PICALM locus (P = 0.03 and
P  =  3  × 10–3 for rs3851179 and rs541458, respectively).
In addition, EADI identified genome-wide significant
association with rs6656401 in the CR1 gene (P  =  3.7  ×
10–9, OR  =  1.21). A  proxy of this SNP rs3818361 (pair-
wise r2 = 0.83), also in the CR1 gene, showed association in
the GERAD dataset (GERAD P = 9.2 × 10−6, OR = 1.17;
EADI P = 8.9 × 10−8, OR = 1.19).
Additionally, the GERAD study identified a significant
excess of SNPs that, although not meeting stringent criteria
for genome-wide significance, still showed strong evidence
for association with AD (P < 1 × 10–5), indicating that addi-
tional susceptibility genes remained to be identified. SNPs
within this “sub-threshold” range included variants 5′ to
the BIN1 gene (e.g., rs744373, P = 3.2 × 10−6, OR = 1.17)
and SNPs at the MS4A gene cluster (e.g., rs610932, P = 1.4
x10−6, OR = 0.87). The BIN1 locus received further support
from a subsequent GWAS by the Cohorts for Heart and
Aging Research in Genomic Epidemiology (CHARGE)
consortium[32] (P  =  4.9  × 10–4, OR  =  1.13). When the
CHARGE, GERAD, and EADI data were combined, the
BIN1 SNP surpassed the threshold for genome-wide signif-
icance (P = 1.59 × 10–11, OR = 1.15). Taken together, these
studies provide compelling evidence that CLU, PICALM,
CR1, and BIN1 are genuine susceptibility genes for AD.
These findings have been subsequently replicated in several
independent datasets[31; 36–38].
Recently, the GERAD consortium reported find-
ings from an extended study (GERAD+), which identi-
fied genome-wide significant evidence for association at
the ABCA7 locus (P  =  5.0×10−21) and the MS4A gene
cluster (P  =  1.2×10−16)[37]. The American Alzheimer’s
Disease Genetic Consortium (ADGC) have also
reported genome-wide significant evidence for associa-
tion at the MS4A gene cluster and provided further sup-
port for ABCA7. When combining data from ADGC
and GERAD+, SNPs at CD33 (P  =  1.6×10−9), CD2AP
(P = 8.6×10−9), and EPHA1 (P = 6.0×10−10) also exceed
criteria for genome-wide significant association with
AD[31; 37].
 PAT H O G E N I C M E C H A N I S M S I N L AT E -
O N S ET A L Z H E I M E R’ S D I S E A S E
The pathogenic actions of the early-onset autosomal domi-
nant mutations in APP, PSEN1, and PSEN2 implicated Aβ
aggregation/clearance as a key process in AD. Interestingly,
the genetic risk factors identified in LOAD may well have
an impact on these processes. APOE and CLU have both
been shown to modify Aβ clearance at the blood–brain bar-
rier[39], while CR1 has been shown to act as a receptor for
Aβ, implicating CR1 in Aβ clearance from the brain and cir-
culatory system[40]. Furthermore, ABCA7 has been shown
to regulate APP processing and inhibit Aβ secretion in cell
cultures over-expressing APP[41].
Pathway analysis of GWAS data has implicated the
immune system and cholesterol metabolism in the etiol-
ogy of AD[42]. The implication that the immune system,
specifically the inflammatory response, plays a role in AD
is not new[43;  44]; however, this has generally been consid-
ered a secondary event. Both CLU and CR1 play signifi-
cant roles in inflammation as well as in innate and adaptive
immunity. CLU is a multifunctional molecule whose roles
include inhibition of complement activation[45]. Alterations
in CLU structure and/or expression may alter the rate of
Aβ clearance via the complement system. CR1 is a poly-
morphic protein that has numerous functions within the
immune system. It acts as a negative regulator inhibiting
both the classical and alternative pathways, and on eryth-
rocytes is a vehicle for the clearance of C3b-coated immune
complexes[46]. It is hypothesized that inhibition of the
complement system by the longer forms of CR1, which
are a risk factor for AD[47], confers disease risk by exces-
sively decreasing C3b opsonisation and clearance of Aβ.
This theory is supported by AD mouse models wherein
inhibition of complement activation has been shown to
correlate with enhanced Aβ plaque deposition and loss of
neuronal integrity[48;  49]. ABCA7, CD33, and EPHA1 all
have putative functions in the immune system. ABCA7 is
known to modulate phagocytosis and clearance of apop-
totic cells[50], CD33 has the potential to mediate both cell–
cell interactions and signaling functions in both the innate
and adaptive immune systems[51], and EPHA1 has putative
functions in both apoptosis[52] and inflammation[53]. BIN1
and CD2AP may also have potential functions in immu-
nity as BIN1-knockout mosaic mice have been reported to
show reduced inflammation with aging[54], and CD2AP is
hypothesized to be critical to the formation of an effective
T-cell-antigen-presenting cell junction[55].
Cholesterol metabolism has long been implicated in
AD. Studies have shown that high levels of cholesterol
G enetics and G enomics o f N euro -P sychiatric D iseases , I I I :   T he C ommon D ementias   •   511
during middle age are correlated with the subsequent devel-
opment of dementia, and that statins, which lower choles-
terol levels, may have a protective effect against the onset
of dementia[56]. ApoE is the main cholesterol transporter
within the brain[57], with the E4 isoform shown to deliver
cholesterol to neurons with decreased efficiency[58]. CLU
is the second major apolipoprotein within the brain and
parallels many of ApoE’s properties in lipid metabolism
and trafficking[59;  60]. ABCA7, which is part of the ABC
transporter superfamily[61], also plays a role in lipid metabo-
lism. It is expressed throughout the brain[62] and is known
to be involved in the release of cholesterol and phospho-
lipids from cells to lipoprotein particles[63]. Through this
role, ABCA7 may interact with both APOE and CLU.
Furthermore, the closest homologue to ABCA7, ABCA1,
plays a critical role in peripheral lipid transport and regu-
lates cholesterol efflux from the plasma membrane to the
apolipoproteins A-I[61].
One of the earliest changes detected in Alzheimer’s
pathology is within the clathrin-mediated endocytic
pathway[64], and receptor-mediated endocytosis has fur-
ther been implicated by the recent GWAS findings. Two
AD risk genes, PICALM and BIN1, have a direct role in
clathrin-mediated endocytosis. PICALM is involved in
the initial assembly of the clathrin-coated vesicle, whereas
BIN1 binds lipid membranes and detects and induces
membrane curvature[65]. Two other AD susceptibility genes
have also been linked to endocytosis. Most members of the
Siglec family, which includes CD33, act as endocytic recep-
tors, mediating endocytosis through a mechanism indepen-
dent of clathrin[66]. CD2AP is a scaffold adaptor protein[55]
that associates with cortactin, a protein also involved in the
regulation of receptor-mediated endocytosis[67]. There are
several hypotheses about how defects in endocytosis and
intracellular trafficking could cause AD, including disrupt-
ing APP processing[68; 69] and/or neurotransmitter release[70].
VA S C U L A R D E M E N T I A
VaD accounts for approximately 20% of dementia, mak-
ing it the second most common form[71]. The disease often
manifests through impairments in judgement and executive
functioning and is characterized by a progressive deterio-
ration in cognitive function. VaD reflects a heterogeneous
grouping of disorders with subtypes including stroke-related
dementia (multi- and single-infarct dementia), subcorti-
cal VaD, VaD caused by inflammatory disease, and mixed
AD. The heterogeneous nature of VaD presents a consider-
able obstacle to the identification of genetic susceptibility
factors and is further confounded by the limitations of the
diagnostic criteria for VaD[72;  73]. This is reflected in the
modest number of genetic studies published on VaD com-
pared with AD. However, advances have been made in
understanding the Mendelian forms of the disorder. While
the genetic architecture of sporadic VaD has received less
attention, a number of studies implicate the APOE gene;
these are also discussed below.
G E N ET I C S O F M E N D E L I A N
VA S C U L A R D E M E N T I A
Cerebral autosomal dominant arteriopathy with subcor-
tical infarcts and leucoencephalopathy (CADASIL) is
a subcortical small-vessel disease accompanied by recur-
rent transient ischemic attacks (TIAs), lacunar strokes,
migraine, neuropsychiatric complications, and dementia[74].
Despite being rare, affecting around one in 100,000 people,
CADASIL is the most common form of hereditary recur-
rent stroke. The disease can affect individuals of both sexes
from their twenties upwards, with the mean age at onset of
first stroke calculated at 46 years. Sufferers display promi-
nent signal abnormalities on brain MRI, including white
matter abnormalities and small subcortical infarcts, which
can often be observed prior to the first stroke. Genetic link-
age analysis mapped the CADASIL locus to chromosome
19q12[74]. Subsequent studies refined the region of inter-
est to a 2cM interval on chromosome 19p13.1[75;  76]. The
NOTCH3 gene, located within the linkage region, was sub-
sequently identified to contain mutations that segregated
with CADASIL within families[77; 78]. Over 190 pathogenic
variations in the NOTCH3 gene have been identified to
correlate with CADASIL[79]. These mutations may affect
receptor trafficking, processing, ligand binding, or signal
transduction. However, the exact mechanisms underlying
the pathogenesis of CADASIL remain unclear.
Cerebral autosomal recessive arteriopathy with sub-
cortical infarcts and leucoencephalopathy (CARASIL)
is an autosomal recessive disease similar to CADASIL.
CARASIL is characterized by non-hypertensive leucoen-
cephalopathy associated with alopecia and spondylosis. As
in CADASIL, recurrent strokes lead to progressive cogni-
tive impairment, ultimately resulting in dementia in most
individuals. CARASIL is extremely rare, with only one case
described in a Caucasian population and approximately
52 cases documented in Asian populations. Genome-wide
linkage analysis and fine-mapping identified homozygous
mutations in the HTRA1 gene that co-segregated with dis-
ease[80]. The HTRA1 gene on chromosome 10q25 encodes
5 1 2   •   G enomics in C linical P ractice
a serine protease that is known to influence different pro-
cesses, including transforming growth factor-β (TGF-β)
signaling. Mutated forms of the HTRA1 protein have been
shown to exhibit decreased protease activity; resulting in
elevated TGF-β signaling in CARASIL[80].
Hereditary cerebral hemorrhage with amyloidosis
(HCHWA) is one of a number of rare forms of cerebral
amyloid angiopathy (CAA). HCHWA is characterized pri-
marily by hemorrhagic strokes and dementia[81; 82], which are
often accompanied by migraines and psychiatric problems.
Several subtypes of HCHWA have been documented: the
majority result from mutations in the APP gene, including
the “Dutch,” “Italian,” “Flemish,” “Piedmont,” and “Arctic”
types[83; 84]. These forms of HCHWA have a common age of
onset between 40 and 50 years and are caused by mutations
that lead to abnormal deposition of amyloid in the walls
of the leptomeningeal arteries and cortical arterioles[83; 85].
As discussed previously, distinct mutations in APP have
been shown to result in early-onset Alzheimer’s disease,
highlighting the overlap that exists between AD and VaD.
Another subtype of HCHWA, the “Icelandic type,” results
from a L68Q mutation[86;  87] in the CST3 gene, which
encodes the cysteine protease inhibitor cystatin C.  The
L68Q mutation leads to the deposition of an N-terminal
degradation product of the mutated cystatin C protein as
vascular amyloid in the leptomeninges, cerebral cortex,
basal ganglia, brainstem, and cerebellum. The majority of
“Icelandic type” HCHWA patients suffer their first stroke
in their mid-twenties. Two other rare, hereditary forms of
CAA are caused by two mutations of the ITM2B gene (also
known as BRI2), namely familial British dementia (FBD)
and familial Danish dementia (FDD). ITM2B encodes a
transmembrane protein that is processed at the C-terminus
to produce a secreted peptide that inhibits the deposition
of Aβ. FBD is a result of a point mutation in the normal
stop codon of the ITM2B gene[88], whereas FDD is associ-
ated with a decamer duplication between codons 265 and
266[89]. Both mutations result in an extended precursor pro-
tein through abolition of the normal stop codon.
G E N ET I C S O F S P O R A D I C
VA S C U L A R D E M E N T I A
Sporadic and common types of VaD are thought to result
from a combination of genetic and environmental factors.
Two twin studies have sought to estimate the heritability of
VaD. One study supported a genetic component to disease,
using discharge records from a Finnish hospital to calculate
probandwise concordance rates as 31% in monozygotic
versus 12.5% in dizygotic twins[90]. In contrast, Bergem
and colleagues used a Norwegian register–based study of
elderly twins that combined interview and clinical data to
estimate identical probandwise concordance rates of 29%
in monozygotic and dizygotic twins[91]. Despite the latter
study’s being the least powerful of the two, it is perhaps the
most methodologically comprehensive and suggests that
environmental influences dominate in VaD.
Studies investigating the genetic architecture of sporadic
VaD are limited and have mostly focused on specific candi-
date genes. A  number of associations have been reported
(see[92] for a review); however, these are largely based on
small sample sizes (typically fewer than 250 cases) and can-
not be considered robust associations by current standards.
The AD genes, including APP, the PSENs, and APOE, have
strong connections to VaD disease pathology. As discussed,
mutations in APP lead to the development of autosomal
dominant forms of AD as well as subtypes of HCHWA.
The PSENs, mutations of which cause AD, interact directly
with NOTCH3[93] which has been shown to segregate with
CADASIL. There is also substantial evidence connect-
ing APOE to VaD. APOE genotype has been associated
with a number of factors linked to VaD, including choles-
terol metabolism[94], hypertension[95], intracerebral hemor-
rhage[96; 97], ischemic heart disease[98], and cerebral amyloid
angiopathy[99]. Additionally, APOE has shown association
with the risk of stroke. Meta-analysis of data from 31 stud-
ies, including 5,961 stroke sufferers and 17,965 healthy con-
trols, observed modest associations between the APOE ε4
allele and ischemic stroke and subarachnoid hemorrhage,
but not with intracerebral hemorrhage, which was associated
with ε2 genotypes[100]. Direct association between APOE
and VaD has also been investigated, with perhaps the most
compelling evidence coming from large population-based
studies. A study of the Framingham cohort, involving over
1,000 individuals aged between 70 and 100 years old, iden-
tified associations between the APOE ε4 allele and both
dementia following stroke and multi-infarct dementia[101].
Similarly, Slooter and colleagues found that the rate of VaD
and mixed VaD and AD were increased among APOE ε4
carriers[102]. However given the well-established and strong
association between APOE and AD[103], investigations of
APOE in relation to VaD are liable to be complicated by
issues of diagnosis, as dementia sufferers often exhibit mixed
AD and vascular pathology[104]. Indeed, the Gothenburg
longitudinal study found that the APOE ε4 allele was asso-
ciated with an increased risk of AD and mixed dementia,
but not pure VaD[105]. In an attempt to dissect this complex
diagnostic issue, Polvikoski and colleagues sought to estab-
lish the prevalence of neuropathologically confirmed AD
G enetics and G enomics o f N euro -P sychiatric D iseases , I I I :   T he C ommon D ementias   •   513
and other dementias, together with APOE genotype[106].
They reported a strong relationship between APOE and
both clinically diagnosed VaD and AD. However, neuro-
pathological analyses showed that over half of those with
clinically diagnosed VaD demonstrated AD pathology. This
study suggests that the association between APOE ε4 and
VaD probably results from an association with underlying
and undiagnosed AD. Indeed, other post-mortem stud-
ies of VaD have either found no association with VaD and
APOE, or an association only in cases of mixed VaD and
AD[107; 108].
A recent genome-wide association study of VaD used
a large prospective population-based cohort from the
Netherlands, known as the Rotterdam Study, to identify
genome-wide significant association with a SNP close to
the androgen receptor (AR) gene on the X chromosome[109].
The association seen in the discovery sample of 67 VaD cases
and 5,633 controls (P = 1.3 × 10–8) replicated in two inde-
pendent samples (combined sample of 256 cases and 5,839
controls; combined-P  =  0.024). However, as the authors
note, the study included small numbers of VaD cases, and
additional replication is necessary before this can be consid-
ered an established association.
F R O N TOT E M P O R A L D E M E N T I A
FTD accounts for 5–15% of dementia cases and, after AD,
is the second most common early-onset dementia. It results
from degeneration of the frontal and temporal lobes, often
in conjunction with the degeneration of subcortical brain
regions. FTD manifests clinically as three distinct syn-
dromes; behavioral variant FTD (bvFTD), which accounts
for about half of all FTD cases[110]; plus two language
clinical variants, semantic dementia (SD) and progressive
non-fluent aphasia (PNFA) in which selective language
difficulties are present[111]. There is also an overlap of FTD
with motor-neuron disease/amyotrophic lateral sclerosis
(FTD-MND/ALS), as well as Parkinsonian syndromes,
progressive supranucular palsy (PSP), and corticobasal syn-
drome (CBS)[112].
FTD is a pathologically heterogeneous disease, and its
associated pathology, termed frontotemporal lobar degen-
eration (FTLD), can be subdivided into groups based on
the presence of either τ or ubiquitin protein inclusions in
the brain[113]. Approximately 50% of FTD patients show
τ-positive pathology at post-mortem[114] and are collectively
known as FTLD-tau[113]. FTLD-tau includes patients with
Pick’s disease, corticobasal degeneration (CBD), PSP, argyr-
ophilic grain disease (AGD), multiple system tauopathy
with dementia (MSTD), and frontotemporal dementia and
Parkinsonism linked to chromosome 17 (FTDP-17). The
remaining pathological subtypes of FTD possess τ-negative
but ubiquitin-positive inclusions. In the majority of these
cases, and about 40% of all FTD patients[114], the major
protein constituent of the ubiquitin inclusions is the TAR
DNA-binding protein 43(TDP-43)[115;  116]. This patho-
logical subtype is therefore known as FTLD-TDP. A small
number of FTD cases (~5–10%) show ubiquitin-positive
but TDP-43-negative pathology[114]. These aggregates have
recently been shown to be immunoreactive to the fused-in-
sarcoma (FUS) antibody and have therefore been termed
FTLD-FUS[117]. The pathological heterogeneity of FTD is
further emphasized by a group of cases known as fronto-
temporal lobar degeneration-ubiquitin proteasome system
(FTLD-UPS), with ubiquitin-positive but TDP-43- and
FUS-negative inclusions[111].
G E N ET I C S O F
F R O N TOT E M P O R A L D E M E N T I A
Most cases of FTD are sporadic; however, studies suggest
that 25–50% of disease sufferers have a first-degree rela-
tive with FTD[118;  119]. While a substantial proportion of
patients reporting a family history of disease shows consis-
tencies with autosomal dominant inheritance, the molecu-
lar genetic basis of FTD is heterogeneous, and autosomal
dominant FTD may be caused by mutations in several genes,
including MAPT[120–122;  123], GRN[123a];[124], VCP[125;  126],
CHMP2B[127;  128], TMEM106B[129–131], C90RF72[132;  133],
TARDBP[134–136], and FUS[137].
F R O N TOT E M P O R A L D E M E N T I A
A N D M A PT
Around 30% of familial FTD cases are caused by mutations
in the microtubule-associated protein tau gene (MAPT),
which encodes the τ protein. These cases are characterized
by τ-pathology with abnormal accumulation of hypophos-
phorylated tau protein in neurons and/or glial cells[119; 138].
FTDP-17 is a subtype of FTLD-tau that can largely be
categorized into two major types:  predominant-dementia
or predominant-Parkinsonism[139]. The dementia phe-
notype is commonly characterized by cognitive difficul-
ties, including problems with memory and language; and
marked personality changes, including disinhibition and
apathy[140]. Those with the Parkinsonism-predominant
subtype usually experience gait impairment, rigidity,
5 1 4   •   G enomics in C linical P ractice
bradykinesia, and resting tremor[141]. However, both
dementia- and Parkinsonism-predominant FTD can occur
in families carrying the same mutations[142]. In contrast to
AD, those with FTDP-17 have relatively intact episodic
memory and have fewer difficulties with orientation.
Rather, they present with deficits in verbal fluency, abstract
thinking, attention, and executive function (see[143] for a
neuropsychology review).
Spillantini and colleagues[144] first reported link-
age to chromosome 17q21-22 in families with FTD and
Parkinsonism. Further analysis of this region revealed muta-
tions in MAPT as the cause of the symptoms of FTD[120–
123]
. To date, 44 MAPT disease mutations have been
reported in 132 families (see the Alzheimer Disease and
Frontotemporal Dementia Mutation Database: www.mol-
gen.ua.ac.be/FTDMutations/). The τ protein is involved
in microtubule assembly and stabilization. MAPT muta-
tions can be grouped largely according to their position in
the gene. This defines their effect on MAPT mRNA and
protein, which in turn influences the resultant pathology.
MAPT mutations are found in 10–30% of FTD patients
with a positive family history of the disease and up to 70%
of cases from families exhibiting an autosomal dominant
mode of disease transmission[118; 145; 146].
MA PT A N D A L Z H E I M E R’S D I S E A S E
Despite the robust association between MAPT mutations,
τ pathology, and FTDP-17, mutations in MAPT do not
appear to be a common cause of general dementia. Studies
have generally reported no, or only a weak, association
with AD[28; 147–149], Parkinson’s disease[150; 151], and sporadic
FTD[152;  153]. These findings are of particular interest as a
number of other dementias, including AD, are associated
with extensive τ pathology[154].
N O N - M A PT F R O N TOT E M P O R A L
DEMENTIA
While a family history of neurodegenerative disease may be
present in up to 50% of individuals with FTD, a substantial
proportion of hereditary FTD does not result from muta-
tions in MAPT[155]. Furthermore, approximately 50% of
FTD patients do not present with τ pathology but appear
with pathological ubiquitin inclusions (FTLD-TDP,
FTLD-FUS, and FTLD-UPS)[111; 156].
Two independent studies reported linkage to chromo-
some 17q21-22, in patients with FTLD-TDP pathology,
which was not attributable to known mutations in the MAPT
gene[157;  158]. Several mutations in the progranulin (GRN)
gene, located 1.7 Mb centromeric of MAPT on chromo-
some 17q21.3, were identified to segregate with FTLD-TDP
in eight families[124]. Additionally, pathological mutations
in GRN were identified in a Belgian family with autosomal
dominant FTLD-TDP[159]. The GRN mutations identi-
fied were over three times more frequent than mutations in
MAPT, emphasizing a major role for the gene in FTD.
Mutations within the valosin-containing protein (VCP)
gene have been shown to cause a rare familial syndrome with
FTLD-TDP pathology[125; 126], and a total of 14 different mis-
sense mutations have been observed across 30 families[160].
Four families have been identified with muta-
tions within the charged multivesicular body protein 2B
(CHMP2B) gene[127; 128]. These families have been shown to
have FTLD-UPS pathology[161], which presents with pre-
dominantly frontal lobe syndrome; however, there is also
evidence of global cognitive impairment[162].
The wide clinical and pathological differences in FTD,
including those with known mutations, raise concern for
the application of GWAS approaches, which assume etio-
logical homogeneity. Despite this, the first GWAS study
of FTD compared an international sample of 515 indi-
viduals with FTLD-TDP to 2,509 control individuals[129]
and identified association with several SNPs mapping to
a single linkage disequilibrium (LD) block on chromo-
some 7p21, which contains the transmembrane protein
106B (TMEM106B) gene (rs1990622, P  =  1.08  × 10–11,
OR = 0.61). This finding has been replicated in indepen-
dent samples comprising those with FTLD-TDP[130] and
those clinically diagnosed with FTD[131]. The association is
particularly strong in GRN mutation carriers, and associa-
tion between the TMEM106B risk allele and lower plasma
granulin protein levels has been identified[163; 164].
A number of interesting sub-threshold hits were
reported in the GWAS and have been the focus of a subse-
quent study[165]. Although none of the putative associations
were replicated in a clinical cohort of 470 patients, convinc-
ing evidence for association was found on chromosome 9 in
a subgroup of 84 patients with Frontotemporal dementia-
amyotrophic lateral sclerosis (FTD-ALS). Studies dating
back to 2006 have identified linkage to chromosome 9p21
in families with FTD and/or ALS[166–168], although the
causative mutation had proven difficult to pinpoint. Using
next-generation sequencing strategies, two groups recently
identified a GGGGCC hexanucleotide expansion in an
intron of the C90RF72 gene[132; 133]. Little is known about
the function of the C9ORF72 protein, but the hexanucleo-
tide expansion leads to the loss of an alternatively spliced
transcript and the formation of nuclear RNA foci, which
G enetics and G enomics o f N euro -P sychiatric D iseases , I I I :   T he C ommon D ementias   •   515
may be toxic. A  recent screening study utilized a cohort
from the Mayo Clinic with FTD and/or ALS to exam-
ine the frequency of the C90RF72 expansion[169]. The fre-
quency was highest in those with FTD and ALS (21.6%);
when restricting to cases with a positive family history, the
expansion frequency rose to 47.6%. The expansion was also
detected in FTD-only cases (14.7% of cases with a positive
family history and 3.7% of sporadic cases), and in ALS-only
cases (24% of cases with a positive family history and 4.1%
of sporadic cases). The C90RF72 expansion is therefore a
major cause of FTD/ALS.
Additionally, rare mutations within the TARDBP gene,
which encodes the TDP-43 protein, have been reported
in patients with FTD and FTD-ALS[134–136]. Further rare
mutations within the FUS gene have been observed within
FTD and FTD-ALS patients[137]; however, so far no FUS
mutations have been detected in patients with FTLD-FUS
pathology[170].
C O N C LU S I O N S
The identification of genes that contribute to disease risk is
focusing research on biological pathways that correlate with
the development of different dementias. Until recently, most
success had been achieved in identifying mutations causing
rare, Mendelian forms of dementia, which do not influ-
ence susceptibility to the more common, sporadic forms
of disease. However, the advent of GWAS has enabled the
identification of a number of novel susceptibility loci. For
example, nine genes influencing AD have been identified
in the past three years. These findings have refined previous
ideas and defined new putative disease mechanisms, pro-
viding impetus for focused studies aimed at understanding
AD pathogenesis. In addition, the recent genetic advances
highlight the potential for gene discovery, both when large
cohorts are analyzed and when more sophisticated statis-
tical methods are employed. Furthermore, these datasets
provide a rich resource for interrogating novel hypotheses,
allowing the analysis of refined disease phenotypes and the
examination of genetic loci contributing to multiple phe-
notypes. It is clear that the coming years will be an exciting
time for dementia genetics as we move closer to unravelling
the genetic architecture of these diseases, using more pow-
erful GWAS and next-generation sequencing technologies.
REFERENCES
1. Alzheimer’s Disease International. 2010. World Alzheimer Report
2010 (http://www.alz.co.uk/research/files/World Alzheimer​
Report2010.pdf ).
 2. Nussbaum, R.  L., and C. E.  Ellis. 2003. Alzheimer’s disease and
Parkinson’s disease. N Engl J Med 348 (14):1356–1364.
3. Mann, U.  M., E. Mohr, M. Gearing, and T. N.  Chase. 1992.
Heterogeneity in Alzheimer’s disease:  progression rate segregated
by distinct neuropsychological and cerebral metabolic profiles. J
Neurol Neurosurg Psychiatry 55 (10):956–959.
4. McKhann, G., D. Drachman, M. Folstein, R. Katzman, D. Price,
and E. M.  Stadlan. 1984. Clinical diagnosis of Alzheimer’s dis-
ease: report of the NINCDS-ADRDA Work Group under the aus-
pices of Department of Health and Human Services Task Force on
Alzheimer’s Disease. Neurology 34 (7):939–944.
5. McKhann, G. M., D. S. Knopman, H. Chertkow, et al. 2011. The
diagnosis of dementia due to Alzheimer’s disease: recommendations
from the National Institute on Aging–Alzheimer’s Association
workgroups on diagnostic guidelines for Alzheimer’s disease.
Alzheimers Dement 7 (3):263–269.
6. Cummings, J. L. 2003. Alzheimer’s disease: from molecular biology
to neuropsychiatry. Semin Clin Neuropsychiatry 8 (1):31–36.
7. Lantos, P., and N. Cairns. 2000. The neuropathology of Alzheimer’s
disease. In Dementia, edited by J. O’Brien, D. Arnes, and A. Burns.
London: Arnold. 443–459.
8. Shastry, B. S., and F. J. Giblin. 1999. Genes and susceptible loci of
Alzheimer’s disease. Brain Res Bull 48 (2):121–127.
9. Campion, D., C. Dumanchin, D. Hannequin, et  al. 1999.
Early-onset autosomal dominant Alzheimer disease:  prevalence,
genetic heterogeneity, and mutation spectrum. Am J Hum Genet
65 (3):664–670.
10. St. George-Hyslop, P. H., R. E. Tanzi, R. J. Polinsky, et al. 1987. The
genetic defect causing familial Alzheimer’s disease maps on chromo-
some 21. Science 235 (4791):885–890.
11. Goate, A., M. C.  Chartier-Harlin, M. Mullan, et  al. 1991.
Segregation of a missense mutation in the amyloid precursor pro-
tein gene with familial Alzheimer’s disease. Nature 349 (6311):
704–706.
12. Citron, M., T. Oltersdorf, C. Haass, et  al. 1992. Mutation of the
beta-amyloid precursor protein in familial Alzheimer’s disease
increases beta-protein production. Nature 360 (6405):672–674.
13. Suzuki, N., T. T.  Cheung, X. D.  Cai, et  al. 1994. An increased
percentage of long amyloid beta protein secreted by familial amy-
loid beta protein precursor (beta APP717) mutants. Science 264
(5163):1336–1340.
14. Sherrington, R., E. I. Rogaev, Y. Liang, et al. 1995. Cloning of a gene
bearing missense mutations in early-onset familial Alzheimer’s dis-
ease. Nature 375 (6534):754–760.
15. Van Broeckhoven, C., H. Backhovens, M. Cruts, et  al. 1992.
Mapping of a gene predisposing to early-onset Alzheimer’s disease
to chromosome 14q24.3. Nat Genet 2 (4):335–339.
16. Levy-Lahad, E., E. M.  Wijsman, E. Nemens, et  al. 1995. A famil-
ial Alzheimer’s disease locus on chromosome 1. Science 269
(5226):970–973.
17. Levy-Lahad, E., W. Wasco, P. Poorkaj, et al. 1995. Candidate gene
for the chromosome 1 familial Alzheimer’s disease locus. Science 269
(5226):973–977.
18. Rogaev, E. I., R. Sherrington, E. A. Rogaeva, et al. 1995. Familial
Alzheimer’s disease in kindreds with missense mutations in a gene
on chromosome 1 related to the Alzheimer’s disease type 3 gene.
Nature 376 (6543):775–8.
19. Borchelt, D. R., G. Thinakaran, C. B. Eckman, et al. 1996. Familial
Alzheimer’s disease-linked presenilin 1 variants elevate Abeta1–
42/1–40 ratio in vitro and in vivo. Neuron 17 (5):1005–1013.
20. Wolfe, M. S., W. Xia, B. L. Ostaszewski, T. S. Diehl, W. T. Kimberly,
and D. J. Selkoe. 1999. Two transmembrane aspartates in presenilin-1
required for presenilin endoproteolysis and gamma-secretase activ-
ity. Nature 398 (6727):513–517.
21. Esler, W.  P., W. T.  Kimberly, B. L.  Ostaszewski, et  al. 2000.
Transition-state analogue inhibitors of gamma-secretase bind
directly to presenilin-1. Nat Cell Biol 2 (7):428–434.
5 1 6   •   G enomics in C linical P ractice
22. Li, Y. M., M. T. Lai, M. Xu, et al. 2000. Presenilin 1 is linked with
gamma-secretase activity in the detergent solubilized state. Proc Natl
Acad Sci U S A 97 (11):6138–6143.
23. Walsh, D. M., I. Klyubin, J. V. Fadeeva, M. J. Rowan, and D. J. Selkoe.
2002. Amyloid-beta oligomers: their production, toxicity and thera-
peutic inhibition. Biochem Soc Trans 30 (4):552–557.
24. Pappolla, M.  A., Y. J.  Chyan, R. A.  Omar, et  al. 1998. Evidence
of oxidative stress and in vivo neurotoxicity of beta-amyloid in a
transgenic mouse model of Alzheimer’s disease: a chronic oxidative
paradigm for testing antioxidant therapies in vivo. Am J Pathol 152
(4):871–877.
25. Gatz, M., C. A. Reynolds, L. Fratiglioni, et al. 2006. Role of genes
and environments for explaining Alzheimer disease. Arch Gen
Psychiatry 63 (2):168–174.
25a. Olson, J. M., K. A. Goddard, and D. M. Dudek. 2001. The amyloid
precursor protein locus and very-late-onset Alzheimer disease. Am J
Hum Genet 69 (4):895–899.
25b. Athan, E.  S., J. H.  Lee, A. Arriaga, et  al. 2002. Polymorphisms
in the promoter of the human APP gene:  functional evalua-
tion and allele frequencies in Alzheimer disease. Arch Neurol 59
(11):1793–1799.
25c. Guyant-Marechal, L., A. Rovelet-Lecrux, L. Goumidi, et al. 2007.
Variations in the APP gene promoter region and risk of Alzheimer
disease. Neurology 68 (9):684–687.
25d. Nowotny, P., X. Simcock, S. Bertelsen, et al. 2007. Association stud-
ies testing for risk for late-onset Alzheimer's disease with common
variants in the beta-amyloid precursor protein (APP). Am J Med
Genet B Neuropsychiatr Genet 144B (4):469–474.
25e. Theuns, J., E. Marjaux, M. Vandenbulcke, et  al. 2006. Alzheimer
dementia caused by a novel mutation located in the APP C-terminal
intracytosolic fragment. Hum Mutat 27 (9):888–896.
26. Bertram, L., M. B.  McQueen, K. Mullin, D. Blacker, and R.
E.  Tanzi. 2007. Systematic meta-analyses of Alzheimer disease
genetic association studies:  the AlzGene database. Nat Genet 39
(1):17–23.
27. Wragg, M., M. Hutton, and C. Talbot. 1996. Genetic association
between intronic polymorphism in presenilin-1 gene and late-onset
Alzheimer’s disease. Alzheimer’s Disease Collaborative Group.
Lancet 347 (9000):509–512.
28. Gerrish, A., G. Russo, A. Richards, et al. 2012. The role of variation
at AbetaPP, PSEN1, PSEN2, and MAPT in late onset Alzheimer’s
disease. J Alzheimers Dis 28 (2):377–387.
29. Harold, D., R. Abraham, P. Hollingworth, et al. 2009. Genome-wide
association study identifies variants at CLU and PICALM associ-
ated with Alzheimer’s disease. Nat Genet 41 (10):1088–1093.
30. Lambert, J. C., S. Heath, G. Even, et al. 2009. Genome-wide asso-
ciation study identifies variants at CLU and CR1 associated with
Alzheimer’s disease. Nat Genet 41 (10):1094–1099.
31. Naj, A. C., G. Jun, G. W. Beecham, et al. 2011. Common variants at
MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are associated with
late-onset Alzheimer’s disease. Nat Genet 43 (5):436–441.
32. Seshadri, S., A. L. Fitzpatrick, M. A. Ikram, et al. 2010. Genome-wide
analysis of genetic loci associated with Alzheimer disease. JAMA
303 (18):1832–1840.
33. Pericak-Vance, M.  A., J. L.  Bebout, P. C.  Gaskell, Jr., et  al. 1991.
Linkage studies in familial Alzheimer disease: evidence for chromo-
some 19 linkage. Am J Hum Genet 48 (6):1034–1050.
34. Strittmatter, W.  J., A. M.  Saunders, D. Schmechel, et  al. 1993.
Apolipoprotein E:  high-avidity binding to beta-amyloid and
increased frequency of type 4 allele in late-onset familial Alzheimer
disease. Proc Natl Acad Sci U S A 90 (5):1977–1981.
35. Hindorff, L. A., J. MacArthur, A. Wise, et al. A Catalog of Published
Genome-Wide Association Studies [cited March 5, 2012]. Available
from www.genome.gov/gwastudies.
35a. Colhoun, H.  M., P. M.  McKeigue, and G. Davey Smith. 2003.
Problems of reporting genetic associations with complex outcomes.
Lancet 361 (9360):865–872.
 36. Carrasquillo, M. M., O. Belbin, T. A. Hunter, et al. 2010. Replication
of CLU, CR1, and PICALM associations with Alzheimer disease.
Arch Neurol 67 (8):961–964.
37. Hollingworth, P., D. Harold, R. Sims, et al. 2011. Common variants
at ABCA7, MS4A6A/MS4A4E, EPHA1, CD33 and CD2AP are
associated with Alzheimer’s disease. Nat Genet 43 (5):429–435.
38. Hu, X., E. Pickering, Y. C.  Liu, et  al. 2011. Meta-analysis for
genome-wide association study identifies multiple variants at the
BIN1 locus associated with late-onset Alzheimer’s disease. PLoS
One 6 (2):e16616.
39. Bell, R. D., A. P. Sagare, A. E. Friedman, et al. 2007. Transport path-
ways for clearance of human Alzheimer’s amyloid beta-peptide and
apolipoproteins E and J in the mouse central nervous system. J Cereb
Blood Flow Metab 27 (5):909–918.
40. Rogers, J., R. Li, D. Mastroeni, et al. 2006. Peripheral clearance of
amyloid beta peptide by complement C3-dependent adherence to
erythrocytes. Neurobiol Aging 27 (12):1733–1739.
41. Chan, S. L., W. S. Kim, J. B. Kwok, et al. 2008. ATP-binding cassette
transporter A7 regulates processing of amyloid precursor protein in
vitro. J Neurochem 106 (2):793–804.
42. Jones, L., P. A. Holmans, M. L. Hamshere, et al. 2010. Genetic evi-
dence implicates the immune system and cholesterol metabolism in
the aetiology of Alzheimer’s disease. PLoS One 5 (11):e13950.
43. Zotova, E., J. A. Nicoll, R. Kalaria, C. Holmes, and D. Boche. 2010.
Inflammation in Alzheimer’s disease: relevance to pathogenesis and
therapy. Alzheimers Res Ther 2 (1):1.
44. McGeer, E.  G., and P. L.  McGeer. 2001. Innate immunity in
Alzheimer’s disease: a model for local inflammatory reactions. Mol
Interv 1 (1):22–29.
45. Jones, S. E., and C. Jomary. 2002. Clusterin. Int J Biochem Cell Biol
34 (5):427–431.
46. Birmingham, D.  J., W. Chen, G. Liang, H. C.  Schmitt, K. Gavit,
and H. N. Nagaraja. 2003. A CR1 polymorphism associated with
constitutive erythrocyte CR1 levels affects binding to C4b but not
C3b. Immunology 108 (4):531–538.
47. Brouwers, N., C. Van Cauwenberghe, S. Engelborghs, et al. 2011.
Alzheimer risk associated with a copy number variation in the
complement receptor 1 increasing C3b/C4b binding sites. Mol
Psychiatry 17 (2):223–233.
48. Maier, M., Y. Peng, L. Jiang, T. J. Seabrook, M. C. Carroll, and C.
A. Lemere. 2008. Complement C3 deficiency leads to accelerated
amyloid beta plaque deposition and neurodegeneration and modu-
lation of the microglia/macrophage phenotype in amyloid precur-
sor protein transgenic mice. J Neurosci 28 (25):6333–6341.
49. Wyss-Coray, T., F. Yan, A. H.  Lin, et  al. 2002. Prominent
neurodegeneration and increased plaque formation in
complement-inhibited Alzheimer’s mice. Proc Natl Acad Sci U S
A 99 (16):10837–10842.
50. Jehle, A.  W., S. J.  Gardai, S. Li, et  al. 2006. ATP-binding cassette
transporter A7 enhances phagocytosis of apoptotic cells and associ-
ated ERK signaling in macrophages. J Cell Biol 174 (4):547–556.
51. Crocker, P. R., J. C. Paulson, and A. Varki. 2007. Siglecs and their
roles in the immune system. Nat Rev Immunol 7 (4):255–266.
52. Depaepe, V., N. Suarez-Gonzalez, A. Dufour, et  al. 2005. Ephrin
signalling controls brain size by regulating apoptosis of neural pro-
genitors. Nature 435 (7046):1244–1250.
53. Ivanov, A.  I., and A. A.  Romanovsky. 2006. Putative dual role of
ephrin-Eph receptor interactions in inflammation. IUBMB Life 58
(7):389–394.
54. Chang, M.  Y., J. Boulden, J. B.  Katz, et  al. 2007. Bin1 ablation
increases susceptibility to cancer during aging, particularly lung can-
cer. Cancer Res 67 (16):7605–7612.
55. Dustin, M. L., M. W. Olszowy, A. D. Holdorf, et al. 1998. A novel
adaptor protein orchestrates receptor patterning and cytoskeletal
polarity in T-cell contacts. Cell 94 (5):667–677.
56. Duron, E., and O. Hanon. 2008. Vascular risk factors, cognitive
decline, and dementia. Vasc Health Risk Manag 4 (2):363–381.
G enetics and G enomics o f N euro -P sychiatric D iseases , I I I :   T he C ommon D ementias   •   517
57. Beffert, U., M. Danik, P. Krzywkowski, C. Ramassamy, F. Berrada,
and J. Poirier. 1998. The neurobiology of apolipoproteins and their
receptors in the CNS and Alzheimer’s disease. Brain Res Brain Res
Rev 27 (2):119–142.
58. Gong, J. S., M. Kobayashi, H. Hayashi, et al. 2002. Apolipoprotein
E (ApoE) isoform-dependent lipid release from astrocytes prepared
from human ApoE3 and ApoE4 knock-in mice. J Biol Chem 277
(33):29919–29926.
59. Stuart, W.  D., B. Krol, S. H.  Jenkins, and J. A.  Harmony. 1992.
Structure and stability of apolipoprotein J–containing high-density
lipoproteins. Biochemistry 31 (36):8552–8559.
60. Calero, M., T. Tokuda, A. Rostagno, et  al. 1999. Functional and
structural properties of lipid-associated apolipoprotein J (clusterin).
Biochem J 344 Pt 2:375–383.
61. Kaminski, W. E., E. Orso, W. Diederich, J. Klucken, W. Drobnik, and
G. Schmitz. 2000. Identification of a novel human sterol-sensitive
ATP-binding cassette transporter (ABCA7). Biochem Biophys Res
Commun 273 (2):532–538.
62. Kim, W. S., M. L. Fitzgerald, K. Kang, et al. 2005. Abca7 null mice
retain normal macrophage phosphatidylcholine and cholesterol
efflux activity despite alterations in adipose mass and serum choles-
terol levels. J Biol Chem 280 (5):3989–3995.
63. Abe-Dohmae, S., Y. Ikeda, M. Matsuo, et al. 2004. Human ABCA7
supports apolipoprotein-mediated release of cellular cholesterol and
phospholipid to generate high density lipoprotein. J Biol Chem 279
(1):604–611.
64. Cataldo, A. M., D. J. Hamilton, J. L. Barnett, P. A. Paskevich, and
R. A. Nixon. 1996. Properties of the endosomal-lysosomal system
in the human central nervous system: disturbances mark most neu-
rons in populations at risk to degenerate in Alzheimer’s disease. J
Neurosci 16 (1):186–199.
65. Dawson, J. C., J. A. Legg, and L. M. Machesky. 2006. Bar domain
proteins:  a role in tubulation, scission and actin assembly in
clathrin-mediated endocytosis. Trends Cell Biol 16 (10):493–498.
66. Tateno, H., H. Li, M. J. Schur, et al. 2007. Distinct endocytic mech-
anisms of CD22 (Siglec-2) and Siglec-F reflect roles in cell signaling
and innate immunity. Mol Cell Biol 27 (16):5699–5710.
67. Lynch, D.  K., S. C.  Winata, R. J.  Lyons, et  al. 2003. A
Cortactin-CD2-associated protein (CD2AP) complex provides a
novel link between epidermal growth factor receptor endocytosis
and the actin cytoskeleton. J Biol Chem 278 (24):21805–21813.
68. Koo, E. H., and S. L. Squazzo. 1994. Evidence that production and
release of amyloid beta-protein involves the endocytic pathway. J
Biol Chem 269 (26):17386–17389.
69. Nordstedt, C., G. L.  Caporaso, J. Thyberg, S. E.  Gandy, and P.
Greengard. 1993. Identification of the Alzheimer beta/A4 amyloid
precursor protein in clathrin-coated vesicles purified from PC12
cells. J Biol Chem 268 (1):608–612.
70. Harel, A., F. Wu, M. P. Mattson, C. M. Morris, and P. J. Yao. 2008.
Evidence for CALM in directing VAMP2 trafficking. Traffic 9
(3):417–429.
71. Leblanc, G. G., J. F. Meschia, D. T. Stuss, and V. Hachinski. 2006.
Genetics of vascular cognitive impairment: the opportunity and the
challenges. Stroke 37 (1):248–255.
72. Pohjasvaara, T., R. Mantyla, R. Ylikoski, M. Kaste, and T. Erkinjuntti.
2000. Comparison of different clinical criteria (DSM-III, ADDTC,
ICD-10, NINDS-AIREN, DSM-IV) for the diagnosis of vascular
dementia. National Institute of Neurological Disorders and Stroke—
Association Internationale pour la Recherche et l’Enseignement en
Neurosciences. Stroke 31 (12):2952–2957.
73. Wetterling, T., R. D. Kanitz, and K. J. Borgis. 1996. Comparison
of different diagnostic criteria for vascular dementia (ADDTC,
DSM-IV, ICD-10, NINDS-AIREN). Stroke 27 (1):30–36.
74. Tournier-Lasserve, E., A. Joutel, J. Melki, et  al. 1993. Cerebral
autosomal dominant arteriopathy with subcortical infarcts and
leukoencephalopathy maps to chromosome 19q12. Nat Genet 3
(3):256–259.
 75. Ducros, A., T. Nagy, S. Alamowitch, et al. 1996. Cerebral autosomal
dominant arteriopathy with subcortical infarcts and leukoencepha-
lopathy, genetic homogeneity, and mapping of the locus within a
2-cM interval. Am J Hum Genet 58 (1):171–181.
76. Sabbadini, G., A. Francia, L. Calandriello, et  al. 1995. Cerebral
autosomal dominant arteriopathy with subcortical infarcts and
leucoencephalopathy (CADASIL). Clinical, neuroimaging, path-
ological and genetic study of a large Italian family. Brain 118 (Pt
1):207–215.
77. Joutel, A., C. Corpechot, A. Ducros, et al. 1996. Notch3 mutations
in CADASIL, a hereditary adult-onset condition causing stroke
and dementia. Nature 383 (6602):707–710.
78. Joutel, A., K. Vahedi, C. Corpechot, et al. 1997. Strong clustering
and stereotyped nature of Notch3 mutations in CADASIL patients.
Lancet 350 (9090):1511–1515.
79. Yamamoto, Y., L. Craggs, M. Baumann, H. Kalimo, and R.
N.  Kalaria. 2011. Review:  molecular genetics and pathology of
hereditary small vessel diseases of the brain. Neuropathol Appl
Neurobiol 37 (1):94–113.
80. Hara, K., A. Shiga, T. Fukutake, et al. 2009. Association of HTRA1
mutations and familial ischemic cerebral small-vessel disease. N Engl
J Med 360 (17):1729–1739.
81. Bornebroek, M., J. Haan, M. L. Maat-Schieman, S. G. Van Duinen,
and R. A.  Roos. 1996. Hereditary cerebral hemorrhage with
amyloidosis-Dutch type (HCHWA-D):  I—A review of clinical,
radiologic and genetic aspects. Brain Pathol 6 (2):111–114.
82. Maat-Schieman, M. L., S. G. van Duinen, M. Bornebroek, J. Haan,
and R. A. Roos. 1996. Hereditary cerebral hemorrhage with amyloi-
dosis–Dutch type (HCHWA-D): II—A review of histopathologi-
cal aspects. Brain Pathol 6 (2):115–120.
83. Levy, E., M. D.  Carman, I. J.  Fernandez-Madrid, et  al. 1990.
Mutation of the Alzheimer’s disease amyloid gene in hereditary
cerebral hemorrhage, Dutch type. Science 248 (4959):1124–1126.
84. Van Broeckhoven, C., J. Haan, E. Bakker, et al. 1990. Amyloid beta
protein precursor gene and hereditary cerebral hemorrhage with
amyloidosis (Dutch). Science 248 (4959):1120–1122.
85. Herzig, M. C., D. T. Winkler, P. Burgermeister, et al. 2004. Abeta is
targeted to the vasculature in a mouse model of hereditary cerebral
hemorrhage with amyloidosis. Nat Neurosci 7 (9):954–960.
86. Abrahamson, M., A. Grubb, I. Olafsson, and A. Lundwall. 1987.
Molecular cloning and sequence analysis of cDNA coding for the
precursor of the human cysteine proteinase inhibitor cystatin C.
FEBS Lett 216 (2):229–233.
87. Palsdottir, A., M. Abrahamson, L. Thorsteinsson, et  al. 1988.
Mutation in cystatin C gene causes hereditary brain haemorrhage.
Lancet 2 (8611):603–604.
88. Vidal, R., B. Frangione, A. Rostagno, et  al. 1999. A stop-codon
mutation in the BRI gene associated with familial British dementia.
Nature 399 (6738):776–781.
89. Vidal, R., T. Revesz, A. Rostagno, et al. 2000. A decamer duplica-
tion in the 3′ region of the BRI gene originates an amyloid peptide
that is associated with dementia in a Danish kindred. Proc Natl Acad
Sci U S A 97 (9):4920–4925.
90. Raiha, I., J. Kaprio, M. Koskenvuo, T. Rajala, and L. Sourander. 1996.
Alzheimer’s disease in Finnish twins. Lancet 347 (9001):573–578.
91. Bergem, A. L., K. Engedal, and E. Kringlen. 1997. The role of hered-
ity in late-onset Alzheimer disease and vascular dementia. A twin
study. Arch Gen Psychiatry 54 (3):264–270.
92. Kim, Y., M. Kong, J. An, J. Ryu, and C. Lee. 2011. Genetic dissection
of susceptibility to vascular dementia. Psychiatr Genet 21 (2):69–76.
93. Sisodia, S. S., and P. H. St. George-Hyslop. 2002. Gamma-secretase,
Notch, Abeta and Alzheimer’s disease: where do the presenilins fit
in? Nat Rev Neurosci 3 (4):281–290.
94. Eichner, J.  E., S. T.  Dunn, G. Perveen, D. M.  Thompson, K.
E.  Stewart, and B. C.  Stroehla. 2002. Apolipoprotein E polymor-
phism and cardiovascular disease: a HuGE review. Am J Epidemiol
155 (6):487–495.
5 1 8   •   G enomics in C linical P ractice
95. Hirono, N., M. Yasuda, S. Tanimukai, H. Kitagaki, and E. Mori.
2000. Effect of the apolipoprotein E epsilon4 allele on white mat-
ter hyperintensities in dementia. Stroke 31 (6):1263–1268.
96. Rosand, J., E. M. Hylek, H. C. O’Donnell, and S. M. Greenberg.
2000. Warfarin-associated hemorrhage and cerebral amyloid angi-
opathy: a genetic and pathologic study. Neurology 55 (7):947–951.
97. Woo, D., L. R. Sauerbeck, B. M. Kissela, et al. 2002. Genetic and
environmental risk factors for intracerebral hemorrhage: prelimi-
nary results of a population-based study. Stroke 33 (5):1190–1195.
98. Song, Y., M. J. Stampfer, and S. Liu. 2004. Meta-analysis: apolipo-
protein E genotypes and risk for coronary heart disease. Ann Intern
Med 141 (2):137–147.
99. McCarron, M. O., K. W. Muir, J. A. Nicoll, et al. 2000. Prospective
study of apolipoprotein E genotype and functional outcome fol-
lowing ischemic stroke. Arch Neurol 57 (10):1480–1484.
100. Sudlow, C., N. A. Martinez Gonzalez, J. Kim, and C. Clark. 2006.
Does apolipoprotein E genotype influence the risk of ischemic
stroke, intracerebral hemorrhage, or subarachnoid hemorrhage?
Systematic review and meta-analyses of 31 studies among 5961
cases and 17,965 controls. Stroke 37 (2):364–370.
101. Myers, R. H., E. J. Schaefer, P. W. Wilson, et al. 1996. Apolipoprotein
E epsilon4 association with dementia in a population-based
study: The Framingham Study. Neurology 46 (3):673–677.
102. Slooter, A.  J., M. X.  Tang, C. M.  van Duijn, et  al. 1997.
Apolipoprotein E epsilon4 and the risk of dementia with stroke.
A population-based investigation. JAMA 277 (10):818–821.
103. Farrer, L.  A., L. A.  Cupples, J. L.  Haines, et  al. 1997. Effects of
age, sex, and ethnicity on the association between apolipopro-
tein E genotype and Alzheimer disease. A  meta-analysis. APOE
and Alzheimer Disease Meta Analysis Consortium. JAMA 278
(16):1349–1356.
104. Blacker, D., and S. Lovestone. 2006. Genetics and dementia nosol-
ogy. J Geriatr Psychiatry Neurol 19 (3):186–191.
105. Skoog, I., C. Hesse, O. Aevarsson, et al. 1998. A population study
of apoE genotype at the age of 85: relation to dementia, cerebro-
vascular disease, and mortality. J Neurol Neurosurg Psychiatry 64
(1):37–43.
106. Polvikoski, T., R. Sulkava, L. Myllykangas, et al. 2001. Prevalence
of Alzheimer’s disease in very elderly people: a prospective neuro-
pathological study. Neurology 56 (12):1690–1696.
107. Betard, C., Y. Robitaille, M. Gee, et al. 1994. Apo E allele frequen-
cies in Alzheimer’s disease, Lewy body dementia, Alzheimer’s
disease with cerebrovascular disease and vascular dementia.
Neuroreport 5 (15):1893–1896.
108. Saunders, A.  M., W. J.  Strittmatter, D. Schmechel, et  al. 1993.
Association of apolipoprotein E allele epsilon 4 with late-onset
familial and sporadic Alzheimer’s disease. Neurology 43
(8):1467–1472.
109. Schrijvers, E.  M., B. Schurmann, P. J.  Koudstaal, et  al. 2012.
Genome-wide association study of vascular dementia. Stroke 43
(2):315–319.
110. Seelaar, H., W. Kamphorst, S. M. Rosso, et al. 2008. Distinct genetic
forms of frontotemporal dementia. Neurology 71 (16):1220–1226.
111. Seelaar, H., J. D. Rohrer, Y. A. Pijnenburg, N. C. Fox, and J. C. van
Swieten. 2011. Clinical, genetic and pathological heterogeneity of
frontotemporal dementia: a review. J Neurol Neurosurg Psychiatry
82 (5):476–486.
112. Kertesz, A., P. McMonagle, M. Blair, W. Davidson, and D.
G. Munoz. 2005. The evolution and pathology of frontotemporal
dementia. Brain 128 (Pt 9):1996–2005.
113. Neumann, M., M. Tolnay, and I. R. Mackenzie. 2009. The molecu-
lar basis of frontotemporal dementia. Expert Rev Mol Med 11:e23.
114. Goedert, M., B. Ghetti, and M. G. Spillantini. 2012. Frontotemporal
dementia: implications for understanding Alzheimer disease. Cold
Spring Harb Perspect Med 2 (2):a006254.
115. Arai, T., M. Hasegawa, H. Akiyama, et  al. 2006. TDP-43 is
a component of ubiquitin-positive tau-negative inclusions in
 frontotemporal lobar degeneration and amyotrophic lateral sclero-
sis. Biochem Biophys Res Commun 351 (3):602–611.
116. Neumann, M., D. M.  Sampathu, L. K.  Kwong, et  al. 2006.
Ubiquitinated TDP-43 in frontotemporal lobar degeneration and
amyotrophic lateral sclerosis. Science 314 (5796):130–133.
117. Urwin, H., K. A. Josephs, J. D. Rohrer, et al. 2010. FUS pathology
defines the majority of tau- and TDP-43-negative frontotemporal
lobar degeneration. Acta Neuropathol 120 (1):33–41.
118. Poorkaj, P., M. Grossman, E. Steinbart, et al. 2001. Frequency of
tau gene mutations in familial and sporadic cases of non-Alzheimer
dementia. Arch Neurol 58 (3):383–387.
119. Rosso, S. M., L. Donker Kaat, T. Baks, et al. 2003. Frontotemporal
dementia in The Netherlands:  patient characteristics and preva-
lence estimates from a population-based study. Brain 126 (Pt
9):2016–2022.
120. Clark, L.  N., P. Poorkaj, Z. Wszolek, et  al. 1998. Pathogenic
implications of mutations in the tau gene in pallido-ponto-nigral
degeneration and related neurodegenerative disorders linked to
chromosome 17. Proc Natl Acad Sci U S A 95 (22):13103–13107.
121. Goedert, M., M. G. Spillantini, R. A. Crowther, et al. 1999. Tau
gene mutation in familial progressive subcortical gliosis. Nat Med
5 (4):454–457.
122. Hutton, M., C. L. Lendon, P. Rizzu, et al. 1998. Association of mis-
sense and 5′-splice-site mutations in tau with the inherited demen-
tia FTDP-17. Nature 393 (6686):702–705.
123. Poorkaj, P., T. D. Bird, E. Wijsman, et al. 1998. Tau is a candidate
gene for chromosome 17 frontotemporal dementia. Ann Neurol 43
(6):815–825.
123a. Cruts, M., I. Gijselinck, J. van der Zee, et al. 2006. Null mutations
in progranulin cause ubiquitin-positive frontotemporal dementia
linked to chromosome 17q21. Nature 442 (7105):920–924.
124. Mackenzie, I.  R., M. Baker, G. West, et  al. 2006. A family with
tau-negative frontotemporal dementia and neuronal intranuclear
inclusions linked to chromosome 17. Brain 129 (Pt 4):853–867.
125. Schroder, R., G. D.  Watts, S. G.  Mehta, et  al. 2005. Mutant
valosin-containing protein causes a novel type of frontotemporal
dementia. Ann Neurol 57 (3):457–461.
126. Watts, G. D., J. Wymer, M. J. Kovach, et al. 2004. Inclusion body
myopathy associated with Paget disease of bone and frontotempo-
ral dementia is caused by mutant valosin-containing protein. Nat
Genet 36 (4):377–381.
127. Parkinson, N., P. G. Ince, M. O. Smith, et al. 2006. ALS pheno-
types with mutations in CHMP2B (charged multivesicular body
protein 2B). Neurology 67 (6):1074–1077.
128. Skibinski, G., N. J. Parkinson, J. M. Brown, et al. 2005. Mutations
in the endosomal ESCRTIII-complex subunit CHMP2B in fron-
totemporal dementia. Nat Genet 37 (8):806–808.
129. Van Deerlin, V. M., P. M. Sleiman, M. Martinez-Lage, et al. 2010.
Common variants at 7p21 are associated with frontotemporal lobar
degeneration with TDP-43 inclusions. Nat Genet 42 (3):234–239.
130. van der Zee, J., and C. Van Broeckhoven. 2011. TMEM106B a
novel risk factor for frontotemporal lobar degeneration. J Mol
Neurosci 45 (3):516–521.
131. van der Zee, J., T. Van Langenhove, G. Kleinberger, et al. 2011.
TMEM106B is associated with frontotemporal lobar degen-
eration in a clinically diagnosed patient cohort. Brain 134 (Pt
3):808–815.
132. DeJesus-Hernandez, M., I. R. Mackenzie, B. F. Boeve, et al. 2011.
Expanded GGGGCC hexanucleotide repeat in noncoding region
of C9ORF72 causes chromosome 9p-linked FTD and ALS.
Neuron 72 (2):245–256.
133. Renton, A. E., E. Majounie, A. Waite, et al. 2011. A hexanucleo-
tide repeat expansion in C9ORF72 is the cause of chromosome
9p21-linked ALS-FTD. Neuron 72 (2):257–268.
134. Benajiba, L., I. Le Ber, A. Camuzat, et al. 2009. TARDBP muta-
tions in motoneuron disease with frontotemporal lobar degenera-
tion. Ann Neurol 65 (4):470–473.
G enetics and G enomics o f N euro -P sychiatric D iseases , I I I :   T he C ommon D ementias   •   519
135. Borroni, B., C. Bonvicini, A. Alberici, et al. 2009. Mutation within
TARDBP leads to frontotemporal dementia without motor neu-
ron disease. Hum Mutat 30 (11):E974–E983.
136. Kovacs, G.  G., J. R.  Murrell, S. Horvath, et  al. 2009. TARDBP
variation associated with frontotemporal dementia, supranuclear
gaze palsy, and chorea. Mov Disord 24 (12):1843–1847.
137. Kwiatkowski, T.  J., Jr., D. A.  Bosco, A. L.  Leclerc, et  al. 2009.
Mutations in the FUS/TLS gene on chromosome 16 cause familial
amyotrophic lateral sclerosis. Science 323 (5918):1205–1208.
138. Morris, H. R., M. N. Khan, J. C. Janssen, et al. 2001. The genetic
and pathological classification of familial frontotemporal demen-
tia. Arch Neurol 58 (11):1813–1816.
139. van Swieten, J., and M. G.  Spillantini. 2007. Hereditary fronto-
temporal dementia caused by Tau gene mutations. Brain Pathol 17
(1):63–73.
140. Neary, D., J. Snowden, and D. Mann. 2005. Frontotemporal
dementia. Lancet Neurol 4 (11):771–780.
141. Arima, K., A. Kowalska, M. Hasegawa, et al. 2000. Two brothers
with frontotemporal dementia and Parkinsonism with an N279K
mutation of the tau gene. Neurology 54 (9):1787–1795.
142. Yasuda, M., T. Kawamata, O. Komure, et al. 1999. A mutation in
the microtubule-associated protein tau in pallido-nigro-luysian
degeneration. Neurology 53 (4):864–868.
143. Hodges, J.  R., and B. Miller. 2001. The neuropsychology of
frontal variant frontotemporal dementia and semantic demen-
tia. Introduction to the special topic papers: Part II. Neurocase 7
(2):113–121.
144. Spillantini, M. G., T. D. Bird, and B. Ghetti. 1998. Frontotemporal
dementia and Parkinsonism linked to chromosome 17:  a new
group of tauopathies. Brain Pathol 8 (2):387–402.
145. Bird, T., D. Knopman, J. VanSwieten, et  al. 2003. Epidemiology
and genetics of frontotemporal dementia/Pick’s disease. Ann
Neurol 54 Suppl 5:S29–S31.
146. Rizzu, P., J. C. Van Swieten, M. Joosse, et al. 1999. High prevalence
of mutations in the microtubule-associated protein tau in a popu-
lation study of frontotemporal dementia in the Netherlands. Am J
Hum Genet 64 (2):414–421.
147. Mukherjee, O., J. S. Kauwe, K. Mayo, J. C. Morris, and A. M. Goate.
2007. Haplotype-based association analysis of the MAPT locus in
late onset Alzheimer’s disease. BMC Genet 8:3.
148. Myers, A.  J., M. Kaleem, L. Marlowe, et  al. 2005. The H1c hap-
lotype at the MAPT locus is associated with Alzheimer’s disease.
Hum Mol Genet 14 (16):2399–2404.
149. Roks, G., B. Dermaut, P. Heutink, et  al. 1999. Mutation screen-
ing of the tau gene in patients with early-onset Alzheimer’s disease.
Neurosci Lett 277 (2):137–139.
150. Zabetian, C. P., C. M. Hutter, S. A. Factor, et al. 2007. Association
analysis of MAPT H1 haplotype and subhaplotypes in Parkinson’s
disease. Ann Neurol 62 (2):137–144.
151. Zhang, J., Y. Song, H. Chen, and D. Fan. 2005. The tau gene hap-
lotype h1 confers a susceptibility to Parkinson’s disease. Eur Neurol
53 (1):15–21.
152. Bernardi, L., R. G. Maletta, C. Tomaino, et al. 2006. The effects of
APOE and tau gene variability on risk of frontotemporal demen-
tia. Neurobiol Aging 27 (5):702–709.
153. Verpillat, P., A. Camuzat, D. Hannequin, et al. 2002. Association
between the extended tau haplotype and frontotemporal demen-
tia. Arch Neurol 59 (6):935–939.
154. Lee, V.  M., M. Goedert, and J. Q.  Trojanowski. 2001.
Neurodegenerative tauopathies. Annu Rev Neurosci 24:1121–1159.
155. Tolnay, M., and A. Probst. 2002. Frontotemporal lobar degenera-
tion—tau as a Pied Piper? Neurogenetics 4 (2):63–75.
156. Forman, M. S., J. Farmer, J. K. Johnson, et al. 2006. Frontotemporal
dementia:  clinicopathological correlations. Ann Neurol 59
(6):952–962.
157. Rademakers, R., M. Cruts, B. Dermaut, et  al. 2002. Tau nega-
tive frontal lobe dementia at 17q21:  significant finemapping
 of the candidate region to a 4.8 cM interval. Mol Psychiatry 7
(10):1064–1074.
158. Rosso, S. M., W. Kamphorst, B. de Graaf, et al. 2001. Familial
frontotemporal dementia with ubiquitin-positive inclu-
sions is linked to chromosome 17q21–22. Brain 124 (Pt 10):
1948–1957.
159. Cruts, M., I. Gijselinck, J. van der Zee, et al. 2006. Null mutations
in progranulin cause ubiquitin-positive frontotemporal dementia
linked to chromosome 17q21. Nature 442 (7105):920–924.
160. Kimonis, V. E., E. Fulchiero, J. Vesa, and G. Watts. 2008. VCP dis-
ease associated with myopathy, Paget disease of bone and fronto-
temporal dementia: review of a unique disorder. Biochim Biophys
Acta 1782 (12):744–748.
161. Holm, I.  E., E. Englund, I. R.  Mackenzie, P. Johannsen, and A.
M. Isaacs. 2007. A reassessment of the neuropathology of fronto-
temporal dementia linked to chromosome 3. J Neuropathol Exp
Neurol 66 (10):884–891.
162. Gydesen, S., J. M.  Brown, A. Brun, et  al. 2002. Chromosome 3
linked frontotemporal dementia (FTD-3). Neurology 59 (10):
1585–1594.
163. Cruchaga, C., C. Graff, H. H. Chiang, et al. 2011. Association of
TMEM106B gene polymorphism with age at onset in granulin
mutation carriers and plasma granulin protein levels. Arch Neurol
68 (5):581–586.
5 2 0   •   G enomics in C linical P ractice
164. Finch, N., M. M. Carrasquillo, M. Baker, et al. 2011. TMEM106B
regulates progranulin levels and the penetrance of FTLD in GRN
mutation carriers. Neurology 76 (5):467–474.
165. Rollinson, S., S. Mead, J. Snowden, et  al. 2011. Frontotemporal
lobar degeneration genome wide association study replication
confirms a risk locus shared with amyotrophic lateral sclerosis.
Neurobiol Aging 32 (4):758 e1–7.
166. Luty, A. A., J. B. Kwok, E. M. Thompson, et al. 2008. Pedigree with
frontotemporal lobar degeneration—motor neuron disease and
Tar DNA binding protein-43 positive neuropathology:  genetic
linkage to chromosome 9. BMC Neurol 8:32.
167. Morita, M., A. Al-Chalabi, P. M. Andersen, et al. 2006. A locus on
chromosome 9p confers susceptibility to ALS and frontotemporal
dementia. Neurology 66 (6):839–844.
168. Vance, C., A. Al-Chalabi, D. Ruddy, et  al. 2006. Familial amyo-
trophic lateral sclerosis with frontotemporal dementia is linked to
a locus on chromosome 9p13.2–21.3. Brain 129 (Pt 4):868–876.
169. Boeve, B.  F., K. B.  Boylan, N. R.  Graff-Radford, et  al. 2012.
Characterization of frontotemporal dementia and/or amyotrophic
lateral sclerosis associated with the GGGGCC repeat expansion in
C9ORF72. Brain 135 (Pt 3):765–783.
170. Mackenzie, I.  R., R. Rademakers, and M. Neumann. 2010.
TDP-43 and FUS in amyotrophic lateral sclerosis and frontotem-
poral dementia. Lancet Neurol 9 (10):995–1007.
 34.
GENETICS AND GENOMICS OF NEURO-PSYCHIATRIC
DISEASES, IV: SCHIZOPHRENIA
AND BIPOLAR DISORDER
Jinbo Fan
INTRODUCTION
Neuro-psychiatric disorders occur globally and constitute
several clinically distinct conditions. Typically, the whole
group is represented by schizophrenia and bipolar disor-
der. Individually, any one of the two conditions may lead
to lifelong morbidity with huge health, personal, family,
and social implications. Financial and socio-economic
costs are enormous in most cases. There are several under-
lying factors that are etiologically related to either disorder.
Genetic or hereditary factors are considered by all clini-
cians, psychiatrists, and other health professionals caring
for these conditions. Genetic factors are reflected in the
family history (excess of first-degree affected close rela-
tives) and higher concordance among monozygotic twins
(identical) compared to dizygotic (non-identical) twins.
Large scale genetic studies of all parameters have helped in
the accumulation of large amounts of pertinent data that
provide good grounds for discussion of the genetic aspects
of schizophrenia and bipolar disorders. This has led to the
creation of a distinct discipline of neuro-psychiatric genet-
ics. With the advent of genomic laboratory techniques and
the increasing power of bioinformatic tools, large-scale
genomic studies have been possible, generating massive
genome-sequencing data for both disorders. This chapter
provides an introduction to the field of neuro-psychiatric
genetics and genomics focusing on schizophrenia and
bipolar disorders. Relevant clinical applications are high-
lighted. However, detailed description and interpreta-
tion are not possible due to limited space and scope. The
interested reader is advised to consult other dedicated
resources.1
521
THE BURDEN OF SCHIZOPHRENIA
AND BIPOL AR DISORDER
Schizophrenia and bipolar disorder are two major mental
illnesses. The fundamental distinction between dementia
praecox (schizophrenia) and manic-depressive insanity was
first proposed by a German psychiatrist, Emil Kraepelin,
nearly 100  years ago.2 According to the Diagnostic and
Statistical Manual of Mental Disorders, Fourth Edition,
Text Revision (DSM-IV-TR),3 schizophrenia is a chronic
and severe psychotic disorder characterized by the presence
of hallucinations, delusions, disorganized speech or behav-
ior, and the so-called negative symptoms. The lifetime prev-
alence of schizophrenia is usually estimated to be between
0.5% and 1%. Bipolar disorder (BD), previously known
as manic-depressive illness, is a lifelong, highly recurrent
mood disorder characterized by periods of mania (bipolar
I) or hypomania (bipolar II) that alternate with episodes
of major depression (DSM-IV-TR). It is estimated that 1%
of the general adult population has bipolar I disorder (clas-
sic manic-depressive illness),4 but that the full spectrum
of the disorder increases its lifetime prevalence to 4.5%
(1.0% bipolar I, 1.1% bipolar II, and 2.4% sub-threshold
BD).5 Despite the clear clinical distinctions (clinical char-
acteristics and distinct treatment regimens) between them,
schizophrenia and bipolar disorders are similar in a few
epidemiological respects, including population prevalence
rate, worldwide distribution, age at onset, risk for suicide,
and so forth.6
Schizophrenia and bipolar disorder can interfere with
cognition and behavior, severely impacting relationships
with family, friends, and employers, and have devastating
and usually lifelong impacts on an individual’s capacity to
function in society. Schizophrenia and bipolar disorder are
ranked twelfth and fourteenth, respectively, among the 20
leading causes of disability worldwide.7 The burden of ill-
ness associated with schizophrenia and bipolar disorder is
substantial and may include unemployment, marital dys-
function, suicide attempts, and increased health services
usage. Treatments are not uniformly effective, and the bio-
logical basis for the diseases is obscure, although several
neurotransmitter, neuroendocrine, and second-messenger
systems appear to be involved. Thus, there is intense interest
in determining the etiology of schizophrenia and bipolar
disorder, with the goal of developing more effective treat-
ment and prevention strategies.
T H E G E N ET I C E P I D E M I O L O GY O F
SCHIZOPHRENIA AND BIPOLAR
DISORDER
The pathophysiological mechanism underlying schizophre-
nia and bipolar disorder is not well understood yet. Family,
twin, and adoption epidemiological studies have unequiv-
ocally demonstrated a strong contribution of inherited
genetic variation to the risk for both diseases. In the case
of schizophrenia, the combined data from a large number
of European family and twin studies showed that the risk
to first-degree relatives of schizophrenia patients is approxi-
mately 12-fold higher than in the general population,8 and
the estimated heritability of schizophrenia is between 70%
and 90%, based on meta-analyses of twin studies.9 In the
case of bipolar disorder, previous family studies have found
that first-degree relatives of bipolar disorder patients are at
least 5 to 10 times more likely to develop the illness than
relatives of control subjects, and the estimated heritabil-
ity of bipolar disorder is between 70% and 90%, based on
twin studies.10 Schizophrenia and bipolar disorder are both
considered to be among the most heritable major mental
illnesses.
A more recent, landmark, large-scale epidemiological
study of nearly 2  million nuclear Swedish families clearly
demonstrated that first-degree relatives of patients with
either schizophrenia or bipolar disorder are at increased
risk of these diseases, while half-siblings had a significantly
increased risk, which is nevertheless substantially lower than
that of the full-siblings.11 Heritability for schizophrenia and
bipolar disorder was estimated at 64% and 59%, respec-
tively. These estimates are lower than previous estimates
(~80%) based on twin studies. Nonetheless, both types
of studies and results show that schizophrenia and bipolar
5 2 2   •   G eno m ics in C linical P ractice
disease are heritable (estimated heritability of 60–80% of
the liabilities). This study also found a significantly elevated
risk of schizophrenia in first-degree relatives of probands
with bipolar disorder and vice versa, which indicates that
schizophrenia and bipolar disorder partly share a common
genetic cause.11
GENOMIC STUDIES IN
N E U R O -P SYC H I AT R I C G E N ET I C
CONDITIONS
T H E P R E – G E N O M E -WI D E
A S S O C I AT I O N S T U D I E S (GWA S ) E R A
In the past two decades, researchers have invested a large
amount of effort in finding risk gene(s) for psychiatric dis-
orders. Thousands of linkage studies and candidate-gene
association studies were performed to find the genetic
underpinnings of schizophrenia and bipolar disorder.
Genome-wide linkage studies have identified multiple
linked chromosomal regions, and association studies have
implicated many candidate genes, but the results have been
inconsistent.
G E N O M E -WI D E L I N K AG E S T U D I E S
Linkage analysis is a powerful method of mapping the loca-
tion of disease-causing loci by identifying genetic markers
that are co-inherited with a phenotype of interest (e.g.,
disease status) in families. Evidence of genetic linkage is
widely regarded as the most powerful statistical evidence
for the existence of an underlying genetic basis for a dis-
ease. To that end, many genome-wide linkage studies have
been conducted since the 1990s. These studies differ not
only with respect to ascertainment criteria, but also in their
marker density and method of analysis. Genome-wide sig-
nificant evidence at various chromosomal regions has been
reported, although none of those regions has been consis-
tently replicated across independent studies. Nevertheless,
researchers have reported linkage signals in many partially
overlapping regions. The lack of consistent significant find-
ings may reflect insufficient power, differing designs, the
presence of genetic heterogeneity, or the absence of a sus-
ceptibility locus in the region. Since most individual studies
were conducted in small numbers of families, a clearer pic-
ture may emerge following combined analysis of individual
studies. In the case of schizophrenia, the largest genome
scan meta-analysis (GSMA) was carried out on 32 inde-
pendent genome-wide linkage scan analyses that included
3,255 pedigrees with 7,413 genotyped cases affected
with schizophrenia or related disorders.12 This combined
analysis revealed genome-wide suggestive evidence for
linkage on chromosomes 5q and 2q. Separate analysis of
European-ancestry samples found linkage evidence fall-
ing just short of the genome-wide significance on 8p. In
the case of bipolar disorder, a similar combined analysis of
genome-wide linkage studies was performed using the orig-
inal genotype data from 11 bipolar disorder genome-wide
linkage scans comprising 5,179 individuals from 1,067
families: this study found genome-wide significant linkage
to bipolar disorder on chromosomes 6q and 8q.13 However,
due to the broad chromosome regions identified in previous
linkage studies and subsequent combined analysis, attempts
to identify susceptibility genes located in those identified
chromosomal regions have so far had only limited success.
C A N D I DAT E - G E N E A S S O C I AT I O N S T U D I E S
Although linkage analysis has a long history of success in
Mendelian disorders, the absence of consistently signifi-
cant findings in schizophrenia and bipolar disorder indi-
cates that there must be multiple genes involved—either
many genes with strong alleles of low frequency, or many
genes with weak alleles of high frequency. Under these
circumstances, association analysis could be more power-
ful than linkage analysis. Association studies have several
other advantageous features. First, large pedigrees are not
required. Second, these studies can detect genes of moder-
ate effects that result from alleles that are common in the
population.14
To date, hundreds of candidate genes have been tested,
but a causal role has not been established for any of them. The
most widely studied candidates have been genes related to
neurotransmitter, neuroendocrine, and second-messenger
systems. Because the pathogenesis of schizophrenia and
bipolar disorder is still unknown, virtually the entire set of
brain-expressed genes could be considered candidates. The
more compelling candidate gene(s) include both positional
(based on the results of linkage studies) and biological
candidates (based on their connection to relevant neuro-
biological pathways). Thousands of candidate-gene associa-
tion studies have been published, with largely inconsistent
findings, and positive associations have often been followed
by non-replications. Inconsistent results may be related to
variations in ascertainment, phenotype definition, control
selection, confounding by population substructure, and
limited power (an individual study has typically included a
few hundred individuals). It is widely known that small sam-
ple size leads to low statistical power in individual studies,
G enetics and G eno m ics of N euro -P sychiatric D iseases , I V   •   5 2 3
with a concomitant increased rate of both false positive and
false negative results and subsequent difficulty replicating
or refuting previous findings. To deal with these ambigui-
ties, meta-analysis of individual studies can be performed,
thereby increasing their power and allowing a more global
interpretation of all available data.
In schizophrenia, a comprehensive meta-analysis of
all genetic polymorphisms having genotype data avail-
able in at least four independent case-control samples has
been performed. Using the proposed criteria for the assess-
ment of cumulative evidence in genetic association stud-
ies, four candidate genes (dopamine receptor D1 [DRD1],
dystrobrevin-binding protein 1 [DTNBP1], methylene-
tetrahydrofolate reductase [MTHFR], and tryptophan
hydroxylase 1 [TPH1]) have been characterized as showing
“strong” epidemiological credibility.15 In bipolar disorder, a
systematic meta-analysis to catalogue and test all candidate
genes has not been published yet. However, a comprehen-
sive review of published linkage and association studies
pointed out several promising “positional or functional can-
didate genes” that seem to be involved in bipolar disorder,
including the Disrupted in schizophrenia 1 (DISC1), neu-
regulin 1 (NRG1), serotonin transporter gene (SLC6A4),
and the brain-derived neurotrophic factor gene (BDNF).16
I N T H E GWA S E R A
Genome-wide association studies that enable the simulta-
neous and cost-effective analysis of hundreds of thousands
of single-nucleotide polymorphisms (SNPs) for association
with a trait in hundreds or thousands of persons have revo-
lutionized the genetic analysis of complex human diseases.17
GWA S F O R S C H I Z O P H R E N I A
In schizophrenia, the first GWAS using DNA pools and
chip-based mass spectrometry was published in 2006.18 In
this study, over 25,000 SNPs located within approximately
14,000 genes were tested in 320 patients with schizophre-
nia of European descent and 325 matched controls. The
most consistent finding was observed with a SNP rs752016
(odds ratio [OR]  =  1.49, P  =  0.006) near the PLXNA2
gene locus on chromosome 1q32.18 The initial wave of
schizophrenia GWAS failed to find risk loci surpassing
the widely acknowledged threshold for genome-wide sig-
nificance (P value < 5.0E-8).19–24 As GWAS tests hundreds
of thousands of SNP individually for an association with
the trait, to account for the large number of statistical tests
carried out, a very stringent P value (< 5.0E-8) is adopted
as the threshold of significance. This stringent threshold
reduces the occurrence of false positives, but it may cause
many real associations to be missed, especially for variants
that have a small effect on the disease. Thus, early GWAS
with limited sample sizes (several hundred case and control
subjects) most likely lacked power to detect the moderate
risk conferred by the common variants in major psychiatric
disorders. Subsequent GWAS with a few thousand case sub-
jects plus a comparable number of control subjects began to
find several association signals surpassing the genome-wide
significance level (P value < 5.0E-8).25–32 These observa-
tions led to subsequent meta-analysis of GWAS in collab-
orative samples, such as that carried out by the Psychiatric
GWAS Consortium,33 which have proven to be remarkably
successful for analyzing psychiatric disorders.34 To date,
17 schizophrenia GWAS or meta-analyses have been pub-
lished. Although there has been only limited consistency
across studies regarding the top associated genomic regions,
the recent GWAS and meta-analysis of some of the GWAS
revealed 21 association signals (P value < 5.0E-8) surpass-
ing genome-wide significance level. Among these, a few
chromosomal loci, including 2q32, 6p21–p22.1, 10q24,
11p11, 11q24, and 18q21.2, were implicated in different
studies (Table 34.1). On the basis of overall evidence from
large-scale GWAS or meta-analysis, the most notable find-
ing was genetic variants across the major histocompatibility
complex (MHC) region at 6p21.32–p22.1, a chromosomal
locus of high linkage disequilibrium (LD) previously
implicated in schizophrenia, which is supported by several
GWAS or meta-analyses.25,27,29,32,33
GWA S F O R B I P O L A R D I S O R D E R
The first GWAS of bipolar disorder was published in 2007,35
in which approximately 469,000 SNPs across the human
genome were tested in 1,868 patients of British descent with
bipolar disorder and approximately 3,000 shared controls.
The strongest signal (P value = 6.3E-8) in bipolar disorder
was with SNP rs420259 at chromosome 16p12, which falls
just short of the genome-wide significance level.35 To date,
13 bipolar disorder GWAS or meta-analyses have been
published.25,36–47 As is the case in schizophrenia GWAS,
there has been only limited consistency across BD studies

Table 34.1  DETAILS OF GENE REGIONS IDENTIFIED IN GWAS FOR SCHIZOPHRENIA

CHROMOSOME EXAMPLE CLOSEST GENE OMIM OVERALLP ODDS MAF ANCESTRY REFERENCE
LOCUS VARIANT VALUE RATIO
1p21.3 rs1625579 MIR137 614304 1.6E-11 1.12 0.20 European 34
1q24.2 rs10489202 BRP44 614737 9.5E-9 1.19 0.14 Chinese 28
2p15.1 rs2312147 VRK2 602169 1.9E-9 1.09 0.39 European 30
2q32.1 rs1344706 ZNF804A 612282 2.5E-11 1.10 0.41 European 31
2q32.3 rs17662626 PCGEM1 605443 4.3E-11 1.18 0.09 European 34
6p21-p22.1 rs13194053 HIST1H2AH 615013 9.5E-9 1.21 0.18 European 25, 27
6p21-p22.1 rs2021722 TRIM26 600830 2.2E-12 1.15 0.22 European 34
6p21-p22.1 rs1635 NKAPL n/a 6.9E-12 1.28 0.33 Chinese 32
6p21-p22.1 rs6932590 PRSS16 607169 1.4E-12 1.16 0.22 European 29
8p12 rs16887244 LSM1 607281 1.3E-10 1.20 0.32 Chinese 28
8p23.2 rs10503253 CSMD1 608397 1.5E-8 1.16 0.19 European 34
8q21.3 rs7004633 MMP16 602262 2.8E-8 1.10 0.18 European 34
10q24.32 rs7914558 CNNM2 607803 1.8E-9 1.10 0.41 European 34
10q24.33 rs11191580 NT5C2 600417 1.1E-8 1.15 0.09 European 34
11p11.2 rs11038167 TSPAN18 n/a 1.1E-11 1.29 0.40 Chinese 32
11p11.2 rs11819869 AMBRA1 611359 3.9E-9 1.25 0.15 European 26
11q24.2 rs548181 STT3A 601134 2.9E-8 1.20 0.12 European 34
11q24.2 rs12807809 NRGN 602350 2.4E-9 1.15 0.17 European 29
18q21.2 rs9960767 TCF4 602228 4.1E-9 1.23 0.06 European 29
18q21.2 rs12966547 CCDC68 n/a 2.6E-10 1.09 0.42 European 34
18q21.2 rs4309482 TCF4 602228 7.8E-9 1.09 0.42 European 30
Abbreviations: OMIM: Online Mendelian Inheritance in Man database; MAF: minor allele frequency

5 2 4   •   G eno m ics in C linical P ractice

Table 34.2  DETAILS OF GENE REGIONS IDENTIFIED IN GWAS FOR BIPOLAR DISORDER

CHROMOSOME EXAMPLE CLOSEST OMIM OVERALLP ODDS MAF ANCESTRY REFERENCE

LOCUS VARIANT GENE VALUE RATIO
1p31.1 rs4650608 PTGFR 600563 1.6E-8 1.14 0.32 European 38
2q11.2 rs2271893 LMAN2L 609552 1.3E-10 1.16 0.33 European 38
3p22.2 rs9834970 TRANK1 n/a 3.1E-11 1.18 0.48 European 38
6q25.2 rs9371601 SYNE1 608441 4.3E-9 1.15 0.36 European 36
10q21.2 rs10994336 ANK3 600465 9.1E-9 1.45 0.05 European 41
10q21.2 rs10994397 ANK3 600465 5.5E-10 1.35 0.06 European 36
10q21.2 rs4948418 ANK3 600465 3.7E-10 n/a n/a European 38
11q14.1 rs12576775 ODZ4 610084 2.7E-8 1.18 0.18 European 36
11q14.1 rs12576775 ODZ4 610084 6.2E-9 1.14 n/a European 42
12p13.33 rs4765913 CACNA1C 114205 1.5E-8 1.14 0.32 European 36
12p13.33 rs4765913 CACNA1C 114205 9.8E-10 1.14 n/a European 42
12q13.12 rs7296288 DHH 605423 9.4E-9 1.15 0.48 European 36
12q13.12 rs7296288 DHH 605423 9.0E-9 1.11 n/a European 42
13q14.11 rs1012053 DGKH 604071 1.5E-8 1.59 0.16 European 37
19p13.11 rs1064395 NCAN 600826 2.1E-9 1.17 0.14 European 39
20q11.22 rs3818253 TRPC4AP 608430 3.9E-8 1.16 n/a European 42
Abbreviations: OMIM: Online Mendelian Inheritance in Man; MAF: minor allele frequency

regarding the top associated genomic regions; however,
meta-analysis of some of the GWAS revealed 16 association
signals surpassing genome-wide significance level (Table
34.2)36–39,41,42—among them, only four loci (10q21, 11q14,
12p13, and 12q13) were implicated by different studies.
The largest study to date, a meta-analysis of ~750,000
markers on a combined sample of ~18,000 subjects of
European and Asian ancestry, identified association signals
in four chromosome regions, three of which were novel
(1p31.1, 2q11.2, and 3p22.2).38 On the basis of overall evi-
dence from large-scale GWAS or meta-analyses, the most
notable findings were with genetic variants located in or
near four genes (CACNA1C, ANK3, ODZ4, and DHH).
CACNA1C, located on chromosome 12p13, encodes an
alpha subunit of the L-type, voltage-gated calcium channel,
which mediates a variety of calcium-dependent processes in
neurons.41,46 ANK3, located on chromosome 10q21, encodes
ankyrin G, which is an adaptor protein found at axonal ini-
tial segment and has been shown to regulate the assembly of
voltage-gated sodium channels.41 ODZ4, located on chromo-
some 11, encodes a member of a family of cell-surface pro-
teins, the teneurins, and is related to the Drosophila pair-rule
gene ten-m (ODZ). These genes are probably involved in
cell-surface signaling and neuronal pathfinding.36 DHH,
located on chromosome 12q13, encodes desert hedgehog.
The hedgehog gene family encodes signaling molecules that
play an important role in regulating morphogenesis.42
Most of the chromosomal loci identified by schizophre-
nia and bipolar disorder GWAS have been reported in pre-
vious linkage studies. Most of the GWAS-implicated genes
(Tables 34.1 and 34.2) could be considered as positional or
functional candidates according to gene locations or puta-
tive functions of their encoded proteins; however, none of
those genes has been tested in a previous candidate-gene
association study, except one gene—CACNA1C, which
was previously reported in a large-scale candidate-gene
association study.48 Although the risk conferred by each
individual locus is small (odds ratios that are conferred by
each additional risk allele are no greater than 1.10–1.30), it
is still notable that a small impact on risk does not detract
from the most important aspect of GWAS findings—novel
associations, provided that they are replicable, shed light
on the etiology of schizophrenia and bipolar disorder. For
example, the results of schizophrenia GWAS implicated
MIR137-mediated dysregulation as a new etiological mech-
anism in schizophrenia,34 and the results of bipolar disorder
GWAS indicate that ion channelopathies may be involved
in the pathogenesis of bipolar disorder.41
C O P Y N UM B E R VA R I AT I O N S
As in many traits and diseases, the risk conferred by each
common risk variant is small (odds ratios <1.5), and these
GWAS “hits” (surpassing genome-wide significance level)

G enetics and G eno m ics of N euro -P sychiatric D iseases , I V   •   5 2 5

fail to explain the vast majority of genetic heritability for
schizophrenia and bipolar disorder, either individually or
collectively.17,49 Many reasons for the missing heritability
have been proposed. One plausible explanation is that rare
variants, which are not captured by GWAS platforms, may
make significant contributions to the heritability of schizo-
phrenia and bipolar disorder.
Copy number variations (CNVs) including submicro-
scopic deletions and duplications of segments of DNA
are a major component of the difference between human
genomes. CNVs may range from about 1 kilo-base to sev-
eral mega-bases in size. It is estimated that approximately
0.4% of the genomes of unrelated people typically differ
with respect to CNVs.50 CNVs can be detected by various
GWAS platforms, and analyses of GWAS data have identi-
fied many rare CNVs in schizophrenia and bipolar disorder
patients. In contrast to common risk variants identified by
GWAS, these rare CNVs have fairly large effect sizes (OR
> 5) on disease risk.51,52 Deletions at 1q21.1, 15q13.3, and
22q11.2 have been associated with schizophrenia,53,54 a
few studies reported that the burden of rare, large CNVs
is increased in schizophrenia.54–56 In the case of bipolar
disorder, a genome-wide survey reported that singleton
deletions (more than 100 kb in length) are more com-
mon in bipolar disorder patients than in control subjects.57
Microduplications on 10q11 and 6q27 have been associ-
ated with early-onset bipolar disorder.58 Some studies failed
to find a significant association between CNVs and diseases
status, and it has been suggested that that common CNVs
based on current GWAS platforms are unlikely to play a
major role in the genetic basis of common diseases.49,59
SHARED RISK GENES
A recent landmark, large-scale epidemiological study of
nearly 2  million nuclear Swedish families definitively
showed how these two illnesses actually represent one
Table 34.3  DETAILS OF GENE REGIONS IDENTIFIED IN COMBINED GWAS FOR SCHIZOPHRENIA AND
BIPOLAR DISORDER
CHROMOSOME EXAMPLE CLOSEST GENE OMIM
LOCUS VARIANT
2q32.1 rs1344706 ZNF804A 612282
3p21.1 rs2239547 ITIH4 600564
10q21.2 rs10994359 ANK3 600465
12p13.33 rs4765905 CACNA1C 114205
12p13.33 rs1024582 CACNA1C 114205
Abbreviations: OMIM: Online Mendelian Inheritance in Man; MAF: minor allele frequency
5 2 6   •   G eno m ics in C linical P ractice
molecular signature, by reporting a significantly elevated risk
of schizophrenia in first-degree relatives of probands with
bipolar disorder, and vice versa.11 Previous genome-wide
linkage studies have also shown overlapping chromosomal
regions of susceptibility, including 18p11, 10p14, and
22q11-13.6 It is notable that some of the risk gene regions
identified by bipolar disorder GWAS have been reported as
associated with schizophrenia in follow-up replication stud-
ies, and vice versa. More recently, the combined analyses of
schizophrenia and bipolar disorder GWAS data identified
several shared risk loci, including the CACNA1C, ANK3,
ITIH3-ITIH4, and ZNF804A gene regions, each of which
reached genome-wide significance (Table 34.3).23,34,60 Taken
together, findings from recent large-scale epidemiological
study and GWAS further challenge the current nosological
dichotomy between these disorders, and are consistent with
a reappraisal of these diseases as distinct diagnostic entities.11
T H E P O S T- GWA S E R A : WH AT N E X T ?
The over 20 GWAS for schizophrenia or bipolar disorder
make it clear that psychiatric genetic studies are enter-
ing a new era. Researchers have gone from being able to
analyze a small number of polymorphisms in a few genes,
to testing most of the common variants across the entire
genome simultaneously. GWAS for schizophrenia and
bipolar disorder together have recorded more than 20
disease-associated loci surpassing the genome-wide sig-
nificance level. An important next step will be performing
cellular, animal, and molecular biology studies to eluci-
date the pathophysiological mechanisms underlying the
observed associations. But does this mean that geneticists
have elucidated the genetic basis of schizophrenia and
bipolar disorder? Far from it: there are many areas that still
need to be addressed.
First, it will be important to conduct further in-depth
studies of risk loci identified by GWAS to discover causal
OVERALL ODDS MAF ANCESTRY REFERENCE
P VALUE RATIO
9.9E-9 1.12 0.41 European 23
7.8E-9 1.12 0.28 European 34
2.5E-8 1.22 0.06 European 34
7.0E-9 1.11 0.33 European 34
1.9E-8 1.07 0.34 European 60
variants underlying the association signals. Although
GWAS is a powerful methodology to rapidly and system-
atically uncover risk loci, it does not circumvent the process
of refining the implicated chromosomal regions to find the
precise causal variant(s). The correlations (linkage disequi-
librium) between tested and untested variants across the
human genome indicate that the most significant variant(s)
identified by GWAS might not represent the causal
variant(s). One of the first steps will involve further rounds
of genotyping to create a full picture of all the possible com-
mon variants that might explain the association signals.
This should include deep-sequencing efforts to identify rare
variants and copy number variants in different populations
as well, and should also attempt to define independent asso-
ciation signals in the disease-associated regions.
Second, assessments of the predictive value of the iden-
tified risk variants are needed. Give the current observations
of the small effect size of common risk variants and the low
frequency of rare risk variants with stronger effect size, indi-
vidual variants will not be useful for predicting the disease
risk. However, more extensive studies are still needed to
assess the usefulness of combining information from both
common and rare risk variants. In order to address this
question, future collaborative studies with large-scale sam-
ples with well-defined phenotypes are needed.
Third, more risk gene(s) are still waiting to be found.
The current GWAS “hits” are only the tips of the iceberg
and explain only a small proportion of the excess familial
risk. Each individual GWAS and meta-analysis only focused
on their top findings (surpassing genome-wide significance
level)—there must be many more risk variants to discover.
Extensive international collaboration to assemble sample
sizes of an order of magnitude of tens of thousands of case
and control subjects will be critical in order to detecting the
risk variants with odds ratios of approximately 1.10 or less,
but these variants with small effect size are still critical from
an etiological point of view.
C O N C LU S I O N S
The recent schizophrenia and bipolar disorder GWAS
have provided convincing evidence for more than 20 gene
regions involved in disease susceptibility. In almost every
case, the implicated genes within these loci are not identi-
fied by prior linkage and candidate-gene association stud-
ies. Thus, GWAS uncovered new etiological pathways to be
explored in future studies. The risk conferred by common
variants identified by GWAS is small. As in other diseases,
it is clear that these GWAS “hits” fail to explain the vast
G enetics and G eno m ics of N euro -P sychiatric D iseases , I V   •   5 2 7
majority of genetic heritability for schizophrenia and bipo-
lar disorder, either individually or collectively. Many more
risk genes for schizophrenia and bipolar disorder, probably
of small effect, remain to be found. Recent developments in
molecular genomic tools and statistical approaches enable
investigators to capture almost all genetic information from
the genome to find rare and common variants of modest-to-
large effect. Identifying susceptibility variants/genes of
schizophrenia and bipolar disorder would ultimately lead to
improved prevention and treatment for patients with major
psychiatric disorders.
REFERENCES
1. McGuffin P, Owen MJ, and Gottesman II (2004). Psychiatric
Genetics and Genomics. New York: Oxford University Press.
2. Kraepelin E (1991). Psychiatry:  A  Textbook for Students and
Physicians. Science History Publications. Sagamore Beach, MA,
USA.
3. American Psychiatric Association (2000). Diagnostic and
Statistical Manual of Mental Disorders, 4th Edition, Text Revision
(DSM-IV-TR). American Psychiatric Association. Arlington, VA,
USA.
4. Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, and
Walters EE (2005). Lifetime prevalence and age-of-onset distribu-
tions of DSM-IV disorders in the National Comorbidity Survey
Replication. Arch Gen Psychiatry 62 (6):593–602.
5. Merikangas KR, Akiskal HS, Angst J, et  al. (2007). Lifetime
and 12-month prevalence of bipolar spectrum disorder in the
National Comorbidity Survey replication. Arch Gen Psychiatry 64
(5):543–552.
6. Berrettini W (2003). Evidence for shared susceptibility in bipolar
disorder and schizophrenia. Am J Med Genet C Semin Med Genet
123C (1):59–64.
7. WHO (2008). The Global Burden of Disease:  2004 Update.
Geneva: WHO.
8. Gottesman I (1990). Schizophrenia Genesis: The Origins of Madness,
1st edition. W. H. Freeman. San Francisco, CA, USA.
9. Sullivan PF, Kendler KS, and Neale MC (2003). Schizophrenia as a
complex trait: evidence from a meta-analysis of twin studies. Arch
Gen Psychiatry 60 (12):1187–1192.
10. McGuffin P, Rijsdijk F, Andrew M, Sham P, Katz R, and Cardno
A (2003). The heritability of bipolar affective disorder and the
genetic relationship to unipolar depression. Arch Gen Psychiatry 60
(5):497–502.
11. Lichtenstein P, Yip BH, Bjork C, et  al. (2009). Common genetic
determinants of schizophrenia and bipolar disorder in Swedish fam-
ilies: a population-based study. Lancet 373 (9659):234–239.
12. Ng MY, Levinson DF, Faraone SV, et al. (2009). Meta-analysis of 32
genome-wide linkage studies of schizophrenia. Mol Psychiatry 14
(8):774–785.
13. McQueen MB, Devlin B, Faraone SV, et al. (2005). Combined anal-
ysis from eleven linkage studies of bipolar disorder provides strong
evidence of susceptibility loci on chromosomes 6q and 8q. Am J
Hum Genet 77 (4):582–595.
14. Risch N and Merikangas K (1996). The future of genetic studies of
complex human diseases. Science 273 (5281):1516–1517.
15. Allen NC, Bagade S, McQueen MB, et  al. (2008). Systematic
meta-analyses and field synopsis of genetic association stud-
ies in schizophrenia:  the SzGene database. Nat Genet 40
(7):827–834.
 16. Serretti A and Mandelli L (2008). The genetics of bipolar disor-
der:  genome “hot regions,” genes, new potential candidates and
future directions. Mol Psychiatry 13 (8):742–771.
17. Manolio TA (2010). Genomewide association studies and assess-
ment of the risk of disease. N Engl J Med 363 (2):166–176.
18. Mah S, Nelson MR, Delisi LE, et al. (2006). Identification of the
semaphorin receptor PLXNA2 as a candidate for susceptibility to
schizophrenia. Mol Psychiatry 11 (5):471–478.
19. Kirov G, Zaharieva I, Georgieva L, et  al. (2009). A genome-wide
association study in 574 schizophrenia trios using DNA pooling.
Mol Psychiatry 14 (8):796–803.
20. Lencz T, Morgan TV, Athanasiou M, et al. (2007). Converging evi-
dence for a pseudoautosomal cytokine receptor gene locus in schizo-
phrenia. Mol Psychiatry 12 (6):572–580.
21. Liu Y, Chen G, Norton N, et al. (2009). Whole genome association
study in a homogenous population in Shandong peninsula of China
reveals JARID2 as a susceptibility gene for schizophrenia. J Biomed
Biotechnol 2009:536918.
22. Need AC, Ge D, Weale ME, et al. (2009). A genome-wide inves-
tigation of SNPs and CNVs in schizophrenia. PLoS Genet 5
(2):e1000373.
23. O’Donovan MC, Craddock N, Norton N, et  al. (2008).
Identification of loci associated with schizophrenia by genome-wide
association and follow-up. Nat Genet 40 (9):1053–1055.
24. Sullivan PF, Lin D, Tzeng JY, et  al. (2008). Genomewide associa-
tion for schizophrenia in the CATIE study: results of stage 1. Mol
Psychiatry 13 (6):570–584.
25. Purcell SM, Wray NR, Stone JL, et al. (2009). Common polygenic
variation contributes to risk of schizophrenia and bipolar disorder.
Nature 460 (7256):748–752.
26. Rietschel M, Mattheisen M, Degenhardt F, et al. (2012). Association
between genetic variation in a region on chromosome 11 and
schizophrenia in large samples from Europe. Mol Psychiatry 17
(9):906–917.
27. Shi J, Levinson DF, Duan J, et  al. (2009). Common variants on
chromosome 6p22.1 are associated with schizophrenia. Nature 460
(7256):753–757.
28. Shi Y, Li Z, Xu Q, et  al. (2011). Common variants on 8p12
and 1q24.2 confer risk of schizophrenia. Nat Genet 43
(12):1224–1227.
29. Stefansson H, Ophoff RA, Steinberg S, et  al. (2009). Common
variants conferring risk of schizophrenia. Nature 460
(7256):744–747.
30. Steinberg S, de Jong S, Andreassen OA, et  al. (2011). Common
variants at VRK2 and TCF4 conferring risk of schizophrenia. Hum
Mol Genet 20 (20):4076–4081.
31. Williams HJ, Norton N, Dwyer S, et al. (2011). Fine mapping of
ZNF804A and genome-wide significant evidence for its involve-
ment in schizophrenia and bipolar disorder. Mol Psychiatry 16
(4):429–441.
32. Yue WH, Wang HF, Sun LD, et al. (2011). Genome-wide associa-
tion study identifies a susceptibility locus for schizophrenia in Han
Chinese at 11p11.2. Nat Genet 43 (12):1228–1231.
33. Cichon S, Craddock N, Daly M, et al. (2009). Genomewide associa-
tion studies: history, rationale, and prospects for psychiatric disor-
ders. Am J Psychiatry 166 (5):540–556.
34. Ripke S, Sanders AR, Kendler KS, et al. (2011). Genome-wide asso-
ciation study identifies five new schizophrenia loci. Nat Genet 43
(10):969–976.
35. Burton PR, Clayton DG, Cardon LR, et al. (2007). Genome-wide
association study of 14,000 cases of seven common diseases and
3,000 shared controls. Nature 447 (7145):661–678.
36. Sklar P, Ripke S, Scott LJ, et  al. (2011). Large-scale genome-wide
association analysis of bipolar disorder identifies a new susceptibility
locus near ODZ4. Nat Genet 43 (10):977–983.
37. Baum AE, Akula N, Cabanero M, et  al. (2008). A genome-wide
association study implicates diacylglycerol kinase eta (DGKH)
5 2 8   •   G eno m ics in C linical P ractice
and several other genes in the etiology of bipolar disorder. Mol
Psychiatry 13 (2):197–207.
38. Chen DT, Jiang X, Akula N, et al. (2013). Genome-wide association
study meta-analysis of European and Asian-ancestry samples identi-
fies three novel loci associated with bipolar disorder. Mol Psychiatry
18 (2):195–205.
39. Cichon S, Muhleisen TW, Degenhardt FA, et  al. (2011).
Genome-wide association study identifies genetic variation in
neurocan as a susceptibility factor for bipolar disorder. Am J Hum
Genet 88 (3):372–381.
40. Djurovic S, Gustafsson O, Mattingsdal M, et  al. (2010). A
genome-wide association study of bipolar disorder in Norwegian
individuals, followed by replication in Icelandic sample. J Affect
Disord 126 (1–2):312–316.
41. Ferreira MA, O’Donovan MC, Meng YA, et  al. (2008).
Collaborative genome-wide association analysis supports a role
for ANK3 and CACNA1C in bipolar disorder. Nat Genet 40
(9):1056–1058.
42. Green EK, Hamshere M, Forty L, et al. (2013). Replication of bipo-
lar disorder susceptibility alleles and identification of two novel
genome-wide significant associations in a new bipolar disorder
case-control sample. Mol Psychiatry 18(12):1302–1307.
43. Hattori E, Toyota T, Ishitsuka Y, et  al. (2009). Preliminary
genome-wide association study of bipolar disorder in the Japanese
population. Am J Med Genet B Neuropsychiatr Genet 150B
(8):1110–1117.
44. Lee MT, Chen CH, Lee CS, et al. (2011). Genome-wide associa-
tion study of bipolar I disorder in the Han Chinese population. Mol
Psychiatry 16 (5):548–556.
45. Scott LJ, Muglia P, Kong XQ, et al. (2009). Genome-wide associa-
tion and meta-analysis of bipolar disorder in individuals of European
ancestry. Proc Natl Acad Sci U S A 106 (18):7501–7506.
46. Sklar P, Smoller JW, Fan J, et al. (2008). Whole-genome association
study of bipolar disorder. Mol Psychiatry 13 (6):558–569.
47. Smith EN, Bloss CS, Badner JA, et al. (2009). Genome-wide asso-
ciation study of bipolar disorder in European American and African
American individuals. Mol Psychiatry 14 (8):755–763.
48. Sklar P, Gabriel SB, McInnis MG, et al. (2002). Family-based asso-
ciation study of 76 candidate genes in bipolar disorder: BDNF is a
potential risk locus. Mol Psychiatry 7 (6):579–593.
49. Visscher PM, Goddard ME, Derks EM, and Wray NR (2012).
Evidence-based psychiatric genetics, a.k.a. the false dichotomy
between common and rare variant hypotheses. Mol Psychiatry 17
(5):474–485.
50. Kidd JM, Cooper GM, Donahue WF, et al. (2008). Mapping and
sequencing of structural variation from eight human genomes.
Nature 453 (7191):56–64.
51. Lee KW, Woon PS, Teo YY, and Sim K (2012). Genome wide asso-
ciation studies (GWAS) and copy number variation (CNV) stud-
ies of the major psychoses: what have we learnt? Neurosci Biobehav
Rev 36 (1):556–571.
52. Mowry BJ and Gratten J (2013). The emerging spectrum of allelic
variation in schizophrenia: current evidence and strategies for the
identification and functional characterization of common and rare
variants. Mol Psychiatry 18 (1):38–52.
53. Stefansson H, Rujescu D, Cichon S, et  al. (2008). Large recur-
rent microdeletions associated with schizophrenia. Nature 455
(7210):232–236.
54. Stone JL, O'Donovan MC, Gurling H, et  al. (2008). Rare chro-
mosomal deletions and duplications increase risk of schizophrenia.
Nature 455 (7210):237–241.
55. Kirov G, Grozeva D, Norton N, et  al. (2009). Support for the
involvement of large copy number variants in the pathogenesis of
schizophrenia. Hum Mol Genet 18 (8):1497–1503.
56. Walsh T, McClellan JM, McCarthy SE, et al. (2008). Rare structural
variants disrupt multiple genes in neurodevelopmental pathways in
schizophrenia. Science 320 (5875):539–543.
 57. Zhang D, Cheng L, Qian Y, et  al. (2009). Singleton deletions
throughout the genome increase risk of bipolar disorder. Mol
Psychiatry 14 (4):376–380.
58. Priebe L, Degenhardt FA, Herms S, et  al. (2012). Genome-wide
survey implicates the influence of copy number variants (CNVs) in
the development of early-onset bipolar disorder. Mol Psychiatry 17
(4):421–432.
59. Craddock N, Hurles ME, Cardin N, et  al. (2010). Genome-wide
association study of CNVs in 16,000 cases of eight common diseases
and 3,000 shared controls. Nature 464 (7289):713–720.
60. Smoller JW, Craddock N, Kendler K, et al. (2013). Identification
of risk loci with shared effects on five major psychiatric disorders: a
genome-wide analysis. Lancet 381 (9875):1371–1379.

G enetics and G eno m ics of N euro -P sychiatric D iseases , I V   •   5 2 9

 35.
GENETICS AND GENOMICS OF NEURO-PSYCHIATRIC
DISEASES, V: LEARNING AND BEHAVIORAL DISORDER S
F. Lucy Raymond
INTRODUCTION
Compared to many areas of medicine, relatively little is
known about the underlying cellular mechanisms that
lead to learning and behavioral disorders in humans. The
reasons for this are many:  social, political, medical, and
practical. The advent of the Human Genome Project and
developments in molecular genetics over the last 20 years
now mean that from a practical point of view, this area
can be considered in detail, which was not possible
previously.
Why should one even consider trying to understand the
genetic basis of disorders of learning or behavior? The over-
whelming reason is that families who are caring for children
with a severe disability wish to understand how and why
this has happened to their child. If a mechanism of disease
can be identified, there is an opportunity to understand
and to clarify the cause of the problems. Knowledge of the
disease can also lead to a greater understanding of the dis-
ease and provide prognostic information and, potentially,
therapeutic interventions. In many cases, having a diagnosis
means that the lengthy search for the cause of the disease can
now cease, and further unnecessary investigations can be
avoided. Also, the identification of a specific diagnosis may
mean that other medical systems such as cardiac or nutri-
tional status now need closer surveillance than would have
been done if the diagnosis were not made. This can prevent
further morbidity and avoid early mortality. Increasingly,
as the genetic basis of these diseases is understood, there is
more potential to provide therapeutic trials that will ame-
liorate the underlying biological defect.
This chapter focuses on the identification of novel
genes and abnormalities of these genes that cause intel-
lectual disability and autism. However, close analysis of a
few conditions with striking behavioral phenotypes gives
us astonishing insights into the workings of the human
brain.
530
I N T E L L E C T UA L D I S A B I L I T Y
The definition of intellectual disability (ID) requires there
to be significantly sub-average general intellectual function-
ing (Criterion A), which is accompanied by limitations in
adaptive functioning in at least two of the following skill
areas: communication, self-care, home living, social/inter-
personal skills, use of community resources, self-direction,
functional academic skills, work, leisure, health, and safety
(Criterion B). The onset must occur before the age of
18 years (Criterion C). General intellectual functioning is
defined by the intelligence quotient, IQ. “Adaptive func-
tioning” refers to how effectively individuals cope with com-
mon life demands. These are less objective measures and rely
on information gathered from independent sources, such as
teacher evaluations and educational, developmental, and
medical history; nevertheless, these data are extremely use-
ful. In the United Kingdom, the ICD-10 Classification of
Mental and Behavioural Disorders (WHO, Geneva 1992) is
used, whilst in the United States, the Diagnostic and
Statistical Manual of Mental Disorders’ diagnostic classifica-
tion is used, which is broadly similar to the WHO classifi-
cation[1,2]. IQ across the population is normally distributed
and is set at 100, and an IQ of less than 70 is classified as
indicating intellectual impairment or disability.
C AUS E S O F I N T E L L EC T UA L
D I S A B I L IT Y: E P I D E M I O L O GY
The underlying causes of ID are extremely heterogeneous,
as the developing brain is susceptible to a wide variety of
environmental and genetic insults[3,4]. In 1995, a consensus
conference held under the auspices of the American College
of Medical Genetics undertook a literature review of nine
surveys of the intellectually disabled that were carried out
between 1977 and 1994[5]. This review enabled a general
distribution of the causes of mental retardation to be deter- defects. Chromosomal abnormalities are collectively the
mined. Table 35.1 summarizes their findings, together single most frequent identifiable cause of mental retarda-
with a more recent survey of 429 individuals with mental tion (Table 35.1). The most common chromosomal cause
retardation carried out as part of the Australian Child and of ID is trisomy of chromosome 21, which underlies Down
Adolescent (ACAD) study[6]. Overall, intellectual disability syndrome. Other abnormalities associated with intellectual
(IQ <70] is a common condition that affects approximately disability include autosomal and sex chromosome aneu-
1–1.5% of the population[7]. Moderate to profound ID (IQ ploidies; partial chromosomal deletions and duplications;
<50) is estimated to affect 0.3–0.5% of the population. The translocations; insertions; inversions; ring chromosomes;
causes of ID divide into environmental and genetic fac- and uniparental disomies. Single-gene disorders, both as
tors, although environmental and genetic causes frequently recognized syndromes and known monogenic disorders,
coexist in an individual. Environmental exposures can be collectively include a significant proportion of causes of ID
subdivided into prenatal, perinatal, and postnatal expo- [7–21%).
sure. The long-term consequences of extreme prematurity A striking statistic from Table 35.1 is that, in many cases
and the associated medical complications account for an (up to 40%), the underlying etiology remains unknown, but
increasingly significant proportion of children with ID, with improved technology and understanding this figure is
while the proportion of children suffering from postnatal gradually reducing.
infections is diminishing[4,8]. Perinatal injury due to birth The relative excess of males in the population with
problems remains common, as does the prenatal exposure severe ID (1.3–1.7:1) suggested that X-linked disease genes
to teratogens during pregnancy. These include fetal expo- are significant contributors to the overall genetic etiol-
sure to sodium valproate and other anticoagulants, alco- ogy[4,11]. Studies using modern cytogenetic techniques, to
hol, and high levels of blood glucose in diabetic mothers. exclude chromosome abnormalities, suggest that where
A  small proportion of children also suffer from accidents two male siblings are severely retarded, in the absence of
or infections beyond the postnatal period that result in a chromosome abnormality, the likelihood of this being
intellectual disability. The genetic contribution to ID is due to an X-linked gene abnormality is as high as 80%[12].
high, as empirical recurrence risks for siblings of severely However, the contribution of X-linked disease genes to the
affected individuals are 5–8% if a single case is observed, male excess of cases with ID is probably overestimated, as
and 12–15% if two siblings are affected[9,10]. These figures surveys of singleton male cases shows a ten-fold reduction
include both chromosomal abnormalities and single-gene in expected monogenic cases[13].
Although a diagnosis (genetic or environmental) can-
Table 35.1  COMPARISON OF THE CAUSES OF not be reached in about 40% of patients with an IQ less
INTELLECTUAL DISABILITY REPORTED BY THE than 50, this figure is declining as genome-sequence analy-
CONSENSUS CONFERENCE (1997) AND THE ACAD sis is improving. It is likely that there will be a significant
STUDY (2000) (ADAPTED FROM PARTINGTON ET AL., increase in our ability to attribute causation in the near
2000) future for the severe cases. However, approximately 76%
CATEGORY/GROUP CONSENSUS ACAD
of cases with mild mental retardation (IQ 50–70) remain
CONFERENCE (%) STUDY (%) undefined, largely due to the multifactorial causation and
Chromosomal abnormalities 4–28 21 limits in our knowledge of the contribution of genes to the
Down syndrome 12.9–16.1 15 normal distribution of IQ in an assumed polygenic model
Recognizable syndromes 3–7 2 of disability[4].
Provisional unique syndromes 1–5 3
Known monogenic conditions 3–9 7 G E N ET I C C AUS E S
Structural CNS abnormalities 7–17 6 O F I N T E L L EC T UA L D I S A B I L I T Y
Complications of prematurity 2–10 8
Many chromosome abnormalities are detected by the
Environmental/teratogenic 5–13 8 visual inspection of the whole genome at the resolution
Metabolic/endocrine 1–5 0 of approximately 5–10 Mb. This will detect large gains or
Unknown 30–50 46 losses of chromosome material and rearrangements, which
Male/female ratio (all categories) 1.35–1.4 1.38 are almost always of clinical significance. Although the
Male/female ratio (unknown – 1.77 technique of karyotyping is ultimately limited by the reso-
diagnosis subgroup) lution of the microscope, the quality of the chromosome

G enetics and G eno m ics o f N euro -P syc h iatric D iseases , V   •   5 3 1

preparations has gradually improved over the last 20 years,
and reevaluation of a patient’s chromosomes where a chro-
mosome abnormality is suspected is well worthwhile.
G E N O M I C M I C RO D E L ET I O NS A N D
D U P L I C AT I O NS
There are recurrent small microdeletions of the genome
that are associated with characteristic syndromes. Routine
chromosome analysis would appear normal, as these micro-
deletions are too small to be detected by G-banding and
light microscopy and would only be detected using specific
genomic DNA fluorescent probes, which fail to bind where
there is deletion. The history of detecting microdeletion
syndromes started with the recognition of discrete syn-
dromic phenotypes associated with ID, such as Prader-Willi,
Miller-Dieker, Angelman, and Williams syndromes. Patients
with similar phenotypes were then found to have similar
submicroscopic deletions in discrete genomic regions that
then defined their condition[14–17]. Since then, a large num-
ber of microdeletion syndromes have been described, each
with a discrete phenotype and microdeletion: for example,
Rubinstein-Taybi, Smith-Magenis, Williams, DiGeorge,
or 22q11 deletion syndrome. It is notable that most of the
microdeletion syndromes are associated with a degree of ID.
The spectrum of disability depends on the specific deletion
syndrome and the extent of the deletion. The intellectual
disability in 22q11 deletion syndrome is frequently mild
and may be unrecognized clinically, whereas the degree of
disability in Rubinstein-Taybi and Angelman syndromes
is usually very significant. Larger deletions in any one area
tend to be associated with more severe intellectual disability
than small deletions, as more genes are deleted, but the cor-
relation entirely depends on the nature of the genes within
the deleted region. In Williams syndrome, although there
is a general reduction in intellectual ability, affected indi-
viduals are frequently described as being excessively sociable
and have relatively preserved expressive speech and lan-
guage. In contrast, a patient with a 1.5 Mb duplication that
includes the minimum Williams syndrome–critical region
has recently been reported in whom expressive speech and
language is severely delayed due to the global intellectual
problems, suggesting that within 7q11.23, a locus exists that
is responsible for speech and language development that is
exquisitely sensitive to gene dosage[18].
The observation that microdeletions within the genome
result in ID as a nearly universal feature led Flint and col-
leagues in 1995 to develop a more systematic screen of the
genome for microdeletions in a cohort of patients with men-
tal retardation[19]. This initial experiment did not screen all
5 3 2   •   G eno m ics in C linical P ractice
ends of the chromosomes—telomeres—as not all were eas-
ily studied; however, the technology was developed using
FISH (fluorescent in situ hybridization) with a diagnostic
yield of 3.5 to 11%[20–22].
As a result of more systematic screening of the human
genome in patients with ID, new syndromes have emerged
in which the identification of the deletion initially defined
the condition—from genotype to phenotype. Where many
cases are reported, the clinical features of the condition
emerge post-genotyping and gradually assume the status of
recognizable syndromes such as 1p36 or 2q37.3 microdele-
tion syndromes, as common clinical features have emerged
once enough patients with the same deletion have been
collected[23–25]. However, several of the subtelomeric dele-
tions have been reported only rarely, and the delineation of
the associated clinical phenotype in such cases is therefore
slower to emerge[26].
The principle is therefore well established that small
deletions or duplications of chromosome material can lead
to ID. The advances in array-comparative genomic hybrid-
ization using the bacterial artificial chromosome (BAC)
probes 1 Mb apart initially and then single-nucleotide
polymorphisms (SNP) or oligomer probes have gradually
replaced routine karyotype analysis in many diagnostic lab-
oratories, with a resulting steady increase in diagnostic yield
to 10–15% of all cases[27–32].
The resolution has now increased so that single exons
or single-gene deletions or duplications are now identifi-
able using the array comparative genomic hybridization
(aCGH)[33–35]. Not only has the identification of new
syndromes been elucidated by aCGH, but the increas-
ing knowledge of the sequence has led to a furthering of
our understanding of the molecular mechanism by which
duplications and deletions occur. The presence of low
copy-repeat sequences at and near the location of dele-
tion breakpoints is thought to account for recurrent dele-
tion events by non-homologous recombination[26,36]. This
is emerging as a common feature of deletion and duplica-
tion syndromes on autosomes, although for non-recurrence
deletions, non-homologous endjoining is more frequently
observed, especially on the X chromosome[33–35, 37].
It has emerged that copy number variation (CNV)
throughout the genome is frequent and has no patho-
logical significance in most cases; it is a feature of normal
human variation[38–41]. In 2004, the DECIPHER database
(http://decipher.sanger.ac.uk/) was launched to collate
and aid interpretation of CNVs that are identified in indi-
viduals[42]. As many of the CNVs identified in individu-
als are rare, the identification of other cases with a similar
rare CNV enables phenotypical interpretation of these
findings. Similarly, the documentation of more common
CNVs prevents over-interpretation of the clinical signifi-
cance of deletions or duplications found in an individual.
Large-scale analysis of multiple cases of developmental
delay by aCGH reveals that the de novo mutation rate for
CNVs varies across the genome[43], and that the previously
simplistic view of causation with one abnormality leading
to ID is inaccurate. Girirajan and colleagues showed that,
for some CNVs—for example, 22q11 deletion—a single
genomic deletion is sufficient to lead to intellectual disabil-
ity itself, whilst for other microdeletions—such as 16p11.2
deletion—this is relatively common in the normal popula-
tion, and additional genomic deletions or duplication may
be required in the genome before the child or adult presents
to medical attention. This leads to the concept of a burden
of genomic damage needing to reach a threshold before pre-
senting as a disease, and is the beginning of the revelation of
data underlining the normal IQ continuum in the general
population[44].
It is also emerging that CNV variation within the
genome may not in itself be pathological, but individual
genomic architectures may predispose to deletions in sub-
sequent generations. In patients with Williams syndrome,
although this usually arises de novo in the individual, it is
observed that the chromosome carrying the de novo dele-
tion is frequently associated with a small genomic inver-
sion, suggesting that the presence of one relatively benign
genomic variation in a parent may predispose to an addi-
tional pathological variation in a subsequent generation[45].
I D E N T I FI C AT I O N O F S I N G L E G E N E S
T H AT C AUS E I N T E L L EC T UA L
D I S A B I L IT Y
The Human Genome Project has had a significant impact
on the rate of identifying genes that cause intellectual dis-
ability. The focus was initially on the identification of
novel genes on the X chromosome, as this was thought to
be where many such genes lie, based on the observation
of a male excess and the identification of disease genes on
the X.  In addition, the clinical imperative to identify the
genetic cause of X-linked disease was high, due to the major
clinical implications for other family members.
A consistent finding from surveys of the ID popula-
tion is an excess of males over females in a proportion of
about three to two (male/female ratio of 1.38)[6, 11, 46]. This
male excess is very apparent (male/female ratio of 1.77) in
the group of patients in which a diagnosis cannot be made
(Table 35.1). Historically, a variety of explanations has
been proposed for the excess of males observed in the ID
G enetics and G eno m ics o f N euro -P syc h iatric D iseases , V   •   5 3 3
population, such as (i)  “the greater size of the male head
exposing the infant to greater difficulties and injuries dur-
ing labor”; (ii) bias of ascertainment, with more males likely
to come to the attention of the authorities because affected
males “are more difficult to manage in the house than
affected females” or because parents seek assistance more
frequently for boys than for girls because of different expec-
tations for males; (iii) greater mortality among females with
mental retardation; and (iv) hormonal contributions to the
causation of ID[4, 6,47–49]. However, none of these explana-
tions has been substantiated.
In 1972, it was proposed by Lehrke that major genes
related to intellectual functioning were located on the X
chromosome and that the male excess was due to differ-
ences in the sex chromosome constitution between males
and females[50]. This hypothesis was initially supported by
reports of large families in which ID was segregating in
an X-linked pedigree and then by surveys that described
a marked excess number of brothers with mental retarda-
tion compared to affected sisters[50–52]. The concept of
X-linked intellectual disability (XLID) has subsequently
been reinforced by the mapping of several well-recognized
intellectual disability syndromes to the X chromosome
and by the cloning of numerous genes for syndromic and
non-syndromic XLID.
X-L I N K E D I N T E L L EC T UA L
D I S A B I L IT Y
X-linked intellectual disability is currently recognized to
represent up to 5% of all ID, although if XLID were to
explain the approximate 30% excess of males, then XLID
would account for ~14% of all ID[4].
Historically, XLID has been subdivided into syn-
dromic and non-syndromic forms. Syndromic X-linked
ID (MRXS) is defined as ID associated with one or more
distinguishing somatic, neurological, behavioral, or meta-
bolic manifestations for which there is good evidence of an
X-linked pattern of inheritance, or MRX (non-syndromic
ID), which indicates that intellectual disability was the pre-
dominant feature without additional diagnostic clues.
The distinction between syndromic and non-syndromic
forms of X-linked ID is not always clear-cut, but it was valu-
able in the early days of gene discovery[4, 53]. For example,
fragile X syndrome was initially classified as non-syndromal,
as no distinguishing features in addition to ID were identi-
fied in 11 affected males from two generations of one fam-
ily[54]. However, when the family was restudied several years
later, the affected males were found to have prognathism,
large ears, and macro-orchidism, clinical features typical of
fragile X syndrome[55]. Hence, once a disease-causing gene
is identified, then genotype–phenotype correlations in
families that initially appear to have a non-syndromal form
of XLID may actually show features of syndromic XLID
upon clinical reevaluation.
The recent progress in the identification of abnormali-
ties in several XLID genes has shown that mutations in the
same gene may be responsible for both non-syndromic (NS)
and syndromic forms of XLMR. In the Aristaless-related
paired-class homeobox (ARX) gene, the same 24-bp inser-
tion duplication within exon 2 is found in patients with
NS-XLMR, X-linked West syndrome (characterized by
infantile spasms, hypsarrythmias, and ID), and Partington
syndrome (an extrapyramidal neurological disorder charac-
terized by dystonic movements of the hands, dysarthria, and
ID)[56, 57]. It has been suggested that the phenotypical het-
erogeneity caused by the same mutation in different families
could be due to differences in genetic and environmental
backgrounds that are specific to each family[56]. Further
examples of allelism include mutations within MECP2
(Rett syndrome), RPS6KA3 (Coffin-Lowry syndrome),
and ATRX (α-thalassemia mental retardation syndrome,
X-linked)[58–63]. The mutations in these genes in these fami-
lies are presumed to cause only a partial loss of function of
the encoded proteins, which could explain the absence of
syndromal features[3].
MO R E T H A N 100 I D G E N E S O N T H E
X C H RO MO S O M E
The rate of identification of genes that lead to moderate to
severe intellectual disability reflects the gradual improve-
ments in human genetics and DNA-sequencing technology.
The first gene to be identified was fragile X, by link-
age and positional cloning of genes on the X chromosome,
however, the first family was reported in 1943[64]. Positional
cloning, deletion mapping of multiple X-linked families,
and mapping of X-autosomal translocation then identified
a number of disease-causing genes. More recently, the sys-
tematic analysis of the whole of the coding sequence of the
X chromosome in a large cohort of families with X-linked
disease made a step change in the rate of identification of
disease-causing genes, bringing the total count of syn-
dromic and non-syndromic genes on the X chromosome to
over 100 genes[65]. Since then, a few more genes have been
identified using systematic sequence analysis combined
with functional analysis or high-resolution aCGH identify
single-exon deletions[35, 66, 67].
Although not all families from the 250-family cohort
were resolved by sequence analysis of the coding region of
5 3 4   •   G eno m ics in C linical P ractice
the X chromosome, 50–70% were after detailed sequence
analysis, CNV assessment, and detailed analysis of variants.
This suggests that it is unlikely that there are many more
genes of significant frequency to be found on the X chro-
mosome. Not all X-linked families have been resolved, and
the shortfall may in part be due to limitations in sequence
coverage and undue caution in ascribing pathogenicity,
but it may also suggest that coding-sequence anomalies are
not the only cause of disease. Recently, two cases have been
reported where the presence of non-coding abnormalities
upstream of CUL4B and HCFC1 cause disease[35, 68], sug-
gesting that this may prove to be a significant, albeit small,
additional contribution to the origins of disease.
High-throughput DNA sequence analysis of probands
identified two previously unsuspected observations:
Up to 10% of genes on the X chromosome are potentially
redundant in males. It was noted that in a large
normal control group, or in families where segregation
analysis was possible, non-pathogenic loss-of-function
mutations were observed without deleterious effects;
and
Four to five rare coding variants per individual are
observed.
This challenges some of the previous analyses of variants
in smaller family cohorts and underscores the need for con-
stant reevaluation of causative variants in the literature[69].
AU TO S O M A L C AUS E S O F
I N T E L L EC T UA L D I S A B I L I T Y
Although there will probably be a significant number of
genes on autosomes that cause intellectual disability, the
identification of these genes has been slow. As adults with
significant intellectual disability have reduced reproductive
fitness and rarely have children of their own, there a few auto-
somal dominant pedigrees in the literature. Nevertheless,
there are significant numbers of children with ID, sug-
gesting that the new-mutation rate or autosomal-recessive
causes of disease may be high.
Autosomal-recessive disease is likely to be more fre-
quent in cultures where consanguineous (inbred) marriage
is more common, as this increases the chance of homozy-
gous abnormalities’ being inherited from related heterozy-
gous parents. Three genes, PRSS12 on chromosome 4q26,
CRBN on chromosome 3p26, and CC2D1A on chromo-
some 19p13.12, were the first to be identified using auto-
zygosity mapping of highly consanguineous families[70–72].
This method is powerful for detecting autosomal recessive
genes that cause disease, but it requires the identification
of rare families with the condition. Recently, a significant
contribution has been the systematic sequence analysis of
regions of linkage in 50 Iranian families who were highly
consanguineous[73]. The convincing identification of novel
genes that cause disease from this cohort has been more
complex, as each family is rare and carries a number of rare
private variants. Large-scale analysis of DNA from consan-
guineous families and comparing variants against ethnically
matched controls is still needed before multiple new genes
that cause disease can be convincingly identified.
The identification of autosomal dominant genes
that cause intellectual disability is being carried out in
two primary ways. First, most microdeletion syndromes
are associated with de novo copy number variant evens,
but there are many candidate genes within the can-
didate region. Several groups working on syndromic
causes of intellectual disability have taken the approach
of deletion-mapping the minimal critical region that
causes the disease or selected patients who clinically
have the phenotype and yet do not apparently have the
classic microdeletion syndrome. Through this method,
for example, the single gene that is responsible for the
clinical features of Smith-Magenis syndrome has been
identified[74]. This required the delineation of patients
with smaller and smaller deletions and the sequencing of
the few genes within the minimum deletion interval in
a series of patients who were clinically affected but did
not have a deletion. Mutations in RAI1 were identified in
this syndrome. Of note is that the behavioral phenotype
of self-hugging, head-banging, onychotillomania, and
polyembolokoilomania, characteristic of Smith-Magenis
syndrome, was also seen in the patients with intragenic
mutations in RAI1, suggesting that single-gene disorders
are sufficient to alter behavior quite significantly.
Similarly, a common microdeletion is associated with
Rubinstein-Taybi syndrome, and subsequent mutation
analyses of the gene CREBBP are found to be sufficient
to develop the full phenotype[75]. Also, the severe phe-
notype associated with a de novo 9q34 deletion has since
been defined as “Kleefstra syndrome,” where mutations or
disruption of one allele of EHMT1 are sufficient to cause
disease[76]; and that of Pitt-Hopkins syndrome is associated
with mutations in TCF4[77].
Although this strategy is powerful where individuals
have striking syndromic features, it is less powerful for
non-syndromic ID, where the location of an abnormality
in the genome is less apparent. The recent development of
next-generation sequence technology has now permitted
G enetics and G eno m ics o f N euro -P syc h iatric D iseases , V   •   5 3 5
this area of research to expand our knowledge. Taking a
trio analysis approach, in which the model of disease is
assumed to be de novo, systematic sequence analysis of
unaffected parents and an affected child has shown that
10–30% of severe ID can be identified by this method[78].
Since the initial report of 10 trios, many further de novo
variants have been identified[79]. For any one individual,
there are 5–6 rare variants, and the challenge is to identify
which variant is the cause of disease. The description of
2–3 pathogenic de novo variants within a gene is likely
to be necessary before confidently assigning the gene to
non-syndromic ID. Worldwide coordination of rare gene
variant reporting in ID is likely to be the way forward
to safely describe the number of genes that remain to be
identified. Based on knowledge of the X chromosome
where over 100 genes cause disease from a target of 700
genes on the X, it is likely that there are at least 2,000
autosomal genes that can cause this phenotype.
A L L E L I S M :  T H E S P E C T RU M
O F N E U R O D E VE L O PM E N TA L
DISORDER S
Research to understand the genetic basis of intellectual dis-
ability, autism, and neuropsychiatric disorders like schizo-
phrenia has traditionally worked on the assumption that
the genetic basis of each of these disorders was sufficiently
distinct to merit a different experimental approach. In the
most recent few years, however, there has been increasing
evidence of both clinical and etiological overlap between
these conditions.
AU T I S M
Autism is a common childhood disorder, and the estimated
prevalence in the population is about 4 in 10,000. The crite-
ria for diagnosis include impairments in social interactions
and communication, restricted and stereotyped patterns
of interests and activities, and the presence of develop-
mental abnormalities by three years of age[1]. Males are
more commonly affected, and the sibling recurrence risk
is 3–10%. Twin studies show much higher concordance
rates in monozygotic than in dizygotic twins, suggesting
a major genetic component to the underlying etiology[80].
Identifying the genetic basis of autism has been an area of
sustained work for the last 5 to 10 years, and to date, we do
not have any clear answers, as the genetics appears to be sig-
nificantly more complex than first thought. Although cases
of autism cluster within families, the models of inheritance
make it unlikely that it is frequently due to highly penetrant
single-gene defects as found in X-linked intellectual disabil-
ity. Mutations in a rare X-linked gene, NLGN4, have been
described that cause autism and associated severe mental
retardation, but this is very uncommon[81, 82].
The disease models suggested that a polygenic mode of
inheritance was more likely, wherein disease only occurs in
an individual if alterations or variants are present in sev-
eral genes within the same individual. In 1998, the first
genome-wide linkage was performed on 99 families with
affected sib pairs with disease. This identified six suscep-
tibility loci on chromosomes 4, 7q, 10, 16p, 19p, and 22,
with a maximum LOD (log of the odds) score >1.0[83].
The locus on 7q was the most significant and has since
been replicated with additional sample sets[84]. Additional
whole-genome-wide scans of different cohorts have identi-
fied numerous other loci, including 3q, 15q, 4q, 5p, 6q, 10q,
18q, and Xp[85–90]. Whilst each study identifies different
areas of the genome that carry susceptibility loci, two par-
ticular regions on chromosome 7q22-31 and 2q32 appear
to be consistently identified as regions of the genome where
a common susceptibility gene lies.
The remaining many susceptibility loci for autism iden-
tified indicate either that the studies have been underpow-
ered to detect truly significant linkage, or that the genes that
underlie autism are very numerous. Furthermore, it is likely
that the clinical entity of autism is highly heterogeneous in
its etiology and reflects a final common pathway of a series
of complex brain-developmental abnormalities.
Although autism alone is rarely associated with cytoge-
netically visible chromosomal deletions (except a recurrent
maternally inherited 15q11-13 duplication and 16p11.2
deletion), once fine-scale genomic assessments using
aCGH- or SNP-based analysis emerged, it was noted that
de novo copy number variation makes a significant contri-
bution to disease causation[91,92]. Since then, association
studies, CNV and mutation analysis of candidate genes
have implicated several of the synaptic genes as contribu-
tors to autistic spectrum disorders:  NRXN1, SHANK3,
SEMA5A, and CNTNAP2[93,94,95]. Many of the genes that
are disrupted that are associated with autistic spectrum
disorders are also reported in individuals with intellectual
disability, or are genes that form biological interactions and
are functionally similar. Many are critical to the structural
integrity of the synapse or are key components of the neu-
rodevelopment of the synapse.
As with the identification of new genes that cause
intellectual disability, the progression to next-generation
sequencing technology has enabled the start of comprehen-
sive surveys of gene abnormalities in the genome. A de novo
5 3 6   •   G eno m ics in C linical P ractice
model of disease has been reported, and like that of intel-
lectual disability genetics, the identification of passenger
variants from deleterious mutations is a major challenge
for the field[96–98,99]. Notwithstanding the wealth of data, it
is clear that the functional convergence of ID and autism
on synaptic function is important. At this stage, it is not
clear at a neurodevelopmental level how defects in the same
gene, such as PTCHD1, can lead to autism in one family
but ID in another[100]. It remains to be elucidated whether
this is determined by other genetic modifications within the
genome of the individual, or whether brain development
when faced with an abnormal repertoire of proteins only
has a limited capacity to express a phenotype. The clinically
distinct phenotypes we experience and thus describe that
have historically been categorized separately are perhaps
now hindering our understanding of early neurodevelop-
ment and its associated abnormalities.
SCHIZOPHRENIA
The genetic basis of schizophrenia has been actively inves-
tigated for many years and is the subject of a separate chap-
ter in this book. However, recent data suggest that many of
the same genes that are important for intellectual develop-
ment and predispose to autism are also involved in a pre-
disposition to severe mental illness and schizophrenia. The
identification of a microdeletion on 22q11 was the first
significant finding to suggest this association[101,102], and
since then, large-scale CNV analysis has identified several
regions of the genome that are strongly associated with dis-
ease[103–105]. In the future, understanding the biological basis
of ID, autism, and schizophrenia, with their similarities and
differences, is the next great and interesting challenge for
the field of human neurodevelopment.
C O N C LU S I O N
The first clear epidemiological studies that showed that
intellectual disability has a genetic basis were conducted in
1938[11]. However, not until 1959 was the basis for Down
syndrome clearly identified as due to additional chromo-
some material in every cell[106]. Over the next 50 years, and
especially in the last 10  years, an astonishing discovery
of many single genes that cause intellectual impairment
when abnormal has taken place. This avalanche is likely to
continue at an equally astonishing pace for the next few
years, when mutations in very many genes are going to be
identified and found to result in intellectual disabilities in
humans.
 What is emerging in the genetics of intellectual dis-
ability is that there are more than 2,000 genes that cause
disease, and each gene abnormality is a rare cause of an
abnormality. It is likely that this pattern will be followed
for other phenotypes, and that there is a large number of
genes that, when abnormal, give rise to unusual behavior;
but each gene only rarely causes of problems. Similarly, in
the intellectual disability field, the assumption that much
of the male excess of affected individuals is due to X-linked
single-gene abnormalities is likely to be incorrect, and
a polygenic disease or increased burden of abnormality
model may be more accurate interpretations of the data.
The observation that males are more susceptible to autism
and intellectual disability or that fewer genomic insults
are needed to present to health care professions is still
unexplained.
The power of contemporary genetics is its ability to
analyze large numbers of genes in a great many clinical
samples. This rate of sequence analysis has increased annu-
ally and exponentially, and the availability of affordable
whole-genome sequence analysis will herald a new era
when the identification of sequence variants ceases to be the
rate-limiting step. For behavioral genetics and the study of
intellectual development, this has been a major issue, due
to the small number of affected individuals with the same
condition. The next challenge is to understand the cel-
lular interactions and mechanisms by which the disorders
are manifest and to understand the complex phenotypical
variation that we see in abnormalities within the same gene
in different individuals.
AC K N OW L E D G E M E N T S
F. Lucy Raymond is supported by research grants from the
Wellcome Trust, Birth Defects Foundation, Action Medical
Research, and National Institute of Health Research
(NIHR) Biomedical Research Centre, Cambridge, United
Kingdom.
REFERENCES
1. World Health Organization. The ICD-10 Classification of Mental
and Behavioural Disorders. Geneva: WHO; 1992.
2. American Psychiatric Association. Diagnostic and Statistical
Manual of Mental Disorders, 4th edition (DSM-IV). Washington,
DC: American Psychiatric Association; 1994.
3. Chelly J, Mandel JL. Monogenic causes of X-linked mental retarda-
tion. Nat Rev Genet. 2001;2(9):669–680.
4. Stevenson RE, Schwartz CE, Schroer RJ. X-Linked Mental
Retardation. Oxford, UK: Oxford University Press; 2000.
G enetics and G eno m ics o f N euro -P syc h iatric D iseases , V   •   5 3 7
5. Curry CJ, Stevenson RE, Aughton D, Byrne J, Carey JC, Cassidy
S, et  al. Evaluation of mental retardation:  recommendations of a
Consensus Conference:  American College of Medical Genetics.
Am J Med Genet. Nov 12, 1997;72(4):468–477. PubMed
PMID: 9375733.
6. Partington M, Mowat D, Einfeld S, Tonge B, Turner G. Genes on
the X chromosome are important in undiagnosed mental retarda-
tion. Am J Med Genet. 2000;92(1):57–61.
7. McLaren J, Bryson SE. Review of recent epidemiological studies of
mental retardation:  prevalence, associated disorders, and etiology.
Am J Ment Retard. 1987;92(3):243–254.
8. MacKay DF, Smith GC, Dobbie R, Pell JP. Gestational age at
delivery and special educational need:  retrospective cohort study
of 407,503 schoolchildren. PLoS Med. 2010 Jun;7(6):e1000289.
PubMed PMID: 20543995. Pubmed Central PMCID: 2882432.
9. Crow YJ, Tolmie JL. Recurrence risks in mental retardation. J Med
Genet. 1998;35(3):177–182.
10. Bundey S, Webb TP, Thake A, Todd J. A community study of
severe mental retardation in the West Midlands and the impor-
tance of the fragile X chromosome in its aetiology. J Med Genet.
1985;22(4):258–266.
11. Penrose LS. A Clinical and Genetic Study of 1280 Cases of Mental
Defect. London: His Majesty’s Stationery Office; 1938.
12. Turner G, Partington M. Recurrence risks in undiagnosed mental
retardation. J Med Genet. 2000;37(12):E45.
13. Mandel JL, Chelly J. Monogenic X-linked mental retarda-
tion:  is it as frequent as currently estimated? The paradox of
the ARX (Aristaless X) mutations. Eur J Hum Genet. 2004
Sep;12(9):689–693. PubMed PMID: 15319782.
14. Schwartz CE, Johnson JP, Holycross B, et al. Detection of submicro-
scopic deletions in band 17p13 in patients with the Miller-Dieker
syndrome. Am J Hum Genet. 1988 Nov;43(5):597–604. PubMed
PMID: 2903661.
15. Knoll JH, Nicholls RD, Magenis RE, Graham JM, Jr., Lalande
M, Latt SA. Angelman and Prader-Willi syndromes share a com-
mon chromosome 15 deletion but differ in parental origin of the
deletion. Am J Med Genet. 1989 Feb;32(2):285–290. PubMed
PMID: 2564739.
16. Ewart AK, Morris CA, Atkinson D, et  al. Hemizygosity at the
elastin locus in a developmental disorder, Williams syndrome. Nat
Genet. 1993 Sep;5(1):11–16. PubMed PMID: 7693128.
17. Ledbetter DH, Riccardi VM, Airhart SD, Strobel RJ, Keenan
BS, Crawford JD. Deletions of chromosome 15 as a cause of
the Prader-Willi syndrome. N Engl J Med. Feb 5, 1981;304(6):
325–329. PubMed PMID: 7442771.
18. Somerville MJ, Mervis CB, Young EJ, et al. Severe expressive-language
delay related to duplication of the Williams-Beuren locus.
N Engl J Med. Oct 20, 2005;353(16):1694–1701. PubMed
PMID: 16236740.
19. Flint J, Wilkie AO, Buckle VJ, Winter RM, Holland AJ,
McDermid HE. The detection of subtelomeric chromosomal
rearrangements in idiopathic mental retardation. Nat Genet.
1995;9(2):132–140.
20. Knight SJ, Regan R, Nicod A, et al. Subtle chromosomal rearrange-
ments in children with unexplained mental retardation. Lancet.
1999;354(9191):1676–1681.
21. Slavotinek A, Rosenberg M, Knight S, et al. Screening for submi-
croscopic chromosome rearrangements in children with idiopathic
mental retardation using microsatellite markers for the chromo-
some telomeres. J Med Genet. 1999 May;36(5):405–411. PubMed
PMID: 10353788.
22. Flint J, Knight S. The use of telomere probes to investigate sub-
microscopic rearrangements associated with mental retarda-
tion. Curr Opin Genet Dev. 2003 Jun;13(3):310–316. PubMed
PMID: 12787795.
23. Shapira SK, McCaskill C, Northrup H, et al. Chromosome 1p36
deletions:  the clinical phenotype and molecular characterization
 of a common newly delineated syndrome. Am J Hum Genet. 1997
Sep;61(3):642–650. PubMed PMID: 9326330.
24. Heilstedt HA, Ballif BC, Howard LA, et al. Physical map of 1p36,
placement of breakpoints in monosomy 1p36, and clinical character-
ization of the syndrome. Am J Hum Genet. 2003 May;72(5):1200–
1212. PubMed PMID: 12687501.
25. Aldred MA, Sanford RO, Thomas NS, et al. Molecular analysis of 20
patients with 2q37.3 monosomy: definition of minimum deletion
intervals for key phenotypes. J Med Genet. 2004 Jun;41(6):433–
439. PubMed PMID: 15173228.
26. Willatt L, Cox J, Barber J, et al. 3q29 microdeletion syndrome: clin-
ical and molecular characterization of a new syndrome. Am J Hum
Genet. 2005 Jul;77(1):154–160. PubMed PMID: 15918153.
27. Veltman JA, Schoenmakers EF, Eussen BH, et al. High-throughput
analysis of subtelomeric chromosome rearrangements by use of
array-based comparative genomic hybridization. Am J Hum Genet.
2002 May;70(5):1269–1276. PubMed PMID: 11951177.
28. Vissers LE, de Vries BB, Osoegawa K, et al. Array-based compara-
tive genomic hybridization for the genomewide detection of sub-
microscopic chromosomal abnormalities. Am J Hum Genet. 2003
Dec;73(6):1261–1270. PubMed PMID: 14628292.
29. Shaw-Smith C, Redon R, Rickman L, et al. Microarray based com-
parative genomic hybridisation (array-CGH) detects submicro-
scopic chromosomal deletions and duplications in patients with
learning disability/mental retardation and dysmorphic features. J
Med Genet. 2004 Apr;41(4):241–248. PubMed PMID: 15060094.
30. de Vries BB, Pfundt R, Leisink M, et al. Diagnostic genome profiling
in mental retardation. Am J Hum Genet. 2005 Oct;77(4):606–616.
PubMed PMID: 16175506.
31. Lugtenberg D, de Brouwer AP, Kleefstra T, et  al. Chromosomal
copy number changes in patients with non-syndromic X-linked
mental retardation detected by array CGH. J Med Genet. 2005
Sep 16. PubMed PMID: 16169931.
32. Veltman JA, Yntema HG, Lugtenberg D, et al. High resolution pro-
filing of X chromosomal aberrations by array comparative genomic
hybridisation. J Med Genet. 2004 Jun;41(6):425–432. PubMed
PMID: 15173227.
33. Froyen G, Corbett M, Vandewalle J, et  al. Submicroscopic dupli-
cations of the hydroxysteroid dehydrogenase HSD17B10 and the
E3 ubiquitin ligase HUWE1 are associated with mental retar-
dation. Am J Hum Genet. 2008 Feb;82(2):432–443. PubMed
PMID: 18252223. Pubmed Central PMCID: 2426915.
34. Froyen G, Belet S, Martinez F, et  al. Copy-number gains of
HUWE1 due to replication- and recombination-based rearrange-
ments. Am J Hum Genet. Aug 10, 2012;91(2):252–264. PubMed
PMID: 22840365. Pubmed Central PMCID: 3415555.
35. Whibley AC, Plagnol V, Tarpey PS, et  al. Fine-scale survey of X
chromosome copy number variants and indels underlying intellec-
tual disability. Am J Hum Genet. Aug 13, 2010;87(2):173–188.
PubMed PMID: 20655035. Pubmed Central PMCID: 2917707.
36. Shaw CJ, Withers MA, Lupski JR. Uncommon deletions of the
Smith-Magenis syndrome region can be recurrent when alternate
low-copy repeats act as homologous recombination substrates. Am J
Hum Genet. 2004 Jul;75(1):75–81. PubMed PMID: 15148657.
37. Liu P, Carvalho CM, Hastings PJ, Lupski JR. Mechanisms for
recurrent and complex human genomic rearrangements. Curr Opin
Genet Dev. 2012 Jun;22(3):211–220. PubMed PMID: 22440479.
Pubmed Central PMCID: 3378805.
38. Sebat J, Lakshmi B, Troge J, et al. Large-scale copy number polymor-
phism in the human genome. Science. Jul 23, 2004;305(5683):525–
528. PubMed PMID: 15273396.
39. Sharp AJ, Locke DP, McGrath SD, et al. Segmental duplications and
copy-number variation in the human genome. Am J Hum Genet.
2005 Jul;77(1):78–88. PubMed PMID: 15918152.
40. Conrad DF, Andrews TD, Carter NP, Hurles ME, Pritchard
JK. A high-resolution survey of deletion polymorphism in the
5 3 8   •   G eno m ics in C linical P ractice
human genome. Nat Genet. 2006 Jan;38(1):75–81. PubMed
PMID: 16327808.
41. Hinds DA, Kloek AP, Jen M, Chen X, Frazer KA. Common dele-
tions and SNPs are in linkage disequilibrium in the human genome.
Nat Genet. 2006 Jan;38(1):82–85. PubMed PMID: 16327809.
42. Firth HV, Richards SM, Bevan AP, et al. DECIPHER: Database of
chromosomal imbalance and phenotype in humans using Ensembl
resources. Am J Hum Genet. 2009 Apr;84(4):524–533. PubMed
PMID: 19344873. Pubmed Central PMCID: 2667985.
43. Cooper GM, Coe BP, Girirajan S, et  al. A copy number varia-
tion morbidity map of developmental delay. Nat Genet. 2011
Sep;43(9):838–846. PubMed PMID: 21841781. Pubmed Central
PMCID: 3171215.
44. Girirajan S, Rosenfeld JA, Coe BP, et al. Phenotypic heterogeneity
of genomic disorders and rare copy-number variants. N Engl J Med.
Oct 4, 2012;367(14):1321–1331. PubMed PMID:  22970919.
Pubmed Central PMCID: 3494411.
45. Osborne LR, Li M, Pober B, et al. A 1.5 million-base pair inversion
polymorphism in families with Williams-Beuren syndrome. Nat
Genet. 2001;29(3):321–325.
46. Penrose LS. The Biology of Mental Defect:  Anchor Press Ltd;
London, 1949.
47. Ireland WW. The Mental Affections of Children:  Idiocy, Imbecility
and Insanity. London: JA Churchill; 1900.
48. Dewey WJ, Barrai I, Morton NE, Mi MP. Recessive genes in severe
mental defect. Am J Hum Genet. 1965 May;17:237–256. PubMed
PMID: 14295493.
49. Nance WE, Engel E. One X and four hypotheses:  response to
Lehrke’s “A theory of X-linkage of major intellectual traits.” Am J
Ment Defic. 1972 May;76(6):623–625. PubMed PMID: 5031082.
50. Lehrke R. Theory of X-linkage of major intellectual traits. Am J
Ment Defic. 1972 May;76(6):611–619. PubMed PMID: 5031080.
51. Turner G, Turner B. X-linked mental retardation. J Med Genet.
1974 Jun;11(2):109–113. PubMed PMID: 4841078.
52. Herbst DS, Miller JR. Nonspecific X-linked mental retarda-
tion II:  the frequency in British Columbia. Am J Med Genet.
1980;7(4):461–469.
53. Frints SG, Froyen G, Marynen P, Fryns JP. X-linked mental retar-
dation:  vanishing boundaries between non-specific (MRX) and
syndromic (MRXS) forms. Clin Genet. 2002 Dec;62(6):423–432.
PubMed PMID: 12485186.
54. Martin JP, Bell J. A pedigree of mental defect showing sex-linkage. J
Neurol Psychiatry. 1943;6:154.
55. Richards BW, Sylvester PE, Brooker C. Fragile X-linked mental
retardation:  the Martin-Bell syndrome. J Ment Defic Res. 1981
Dec;25 Pt 4:253–256. PubMed PMID: 7328634.
56. Bienvenu T, Poirier K, Friocourt G, et  al. ARX, a novel
Prd-class-homeobox gene highly expressed in the telencepha-
lon, is mutated in X-linked mental retardation. Hum Mol Genet.
2002;11(8):981–991.
57. Stromme P, Mangelsdorf ME, Shaw MA, et  al. Mutations in the
human ortholog of Aristaless cause X-linked mental retardation and
epilepsy. Nat Genet. 2002;30(4):441–445.
58. Couvert P, Bienvenu T, Aquaviva C, et  al. MECP2 is highly
mutated in X-linked mental retardation. Hum Mol Genet.
2001;10(9):941–946.
59. Gibbons RJ, Picketts DJ, Villard L, Higgs DR. Mutations in a puta-
tive global transcriptional regulator cause X-linked mental retar-
dation with alpha-thalassemia (ATR-X syndrome). Cell. Mar 24,
1995;80(6):837–845. PubMed PMID: 7697714.
60. Merienne K, Jacquot S, Pannetier S, et al. A missense mutation in
RPS6KA3 (RSK2) responsible for non-specific mental retardation.
Nat Genet. 1999;22(1):13–14.
61. Trivier E, De Cesare D, Jacquot S, et  al. Mutations in the
kinase RSK-2 associated with Coffin-Lowry syndrome. Nature.
1996;384(6609):567–570.
 62. Amir RE, Van den Veyver IB, Wan M, Tran CQ, Francke U,
Zoghbi HY. Rett syndrome is caused by mutations in X-linked
MECP2, encoding methyl- CpG-binding protein 2. Nat Genet.
1999;23(2):185–188. PubMed PMID: 10508514.
63. Guerrini R, Shanahan JL, Carrozzo R, Bonanni P, Higgs DR,
Gibbons RJ. A nonsense mutation of the ATRX gene causing mild
mental retardation and epilepsy. Ann Neurol. 2000;47(1):117–121.
PubMed PMID: 10632111.
64. Oberle I, Rousseau F, Heitz D, et al. Instability of a 550-base pair
DNA segment and abnormal methylation in fragile X syndrome.
Science. 1991;252(5010):1097–1102.
65. Tarpey P, Smith R, Pleasance E, et al. A systematic, large scale rese-
quencing screen of the X chromosome coding exons in mental retar-
dation Nat Genet. 2009; 41(5):535-43. doi:10.1038/ng.367.
66. Shoubridge C, Tarpey PS, Abidi F, et al. Mutations in the guanine
nucleotide exchange factor gene IQSEC2 cause nonsyndromic intel-
lectual disability. Nat Genet. 2010 Jun;42(6):486–488. PubMed
PMID: 20473311. Pubmed Central PMCID: 3632837.
67. McLarren KW, Severson TM, du Souich C, et  al. Hypomorphic
temperature-sensitive alleles of NSDHL cause CK syndrome.
Am J Hum Genet. Dec 10, 2010;87(6):905–914. PubMed
PMID: 21129721. Pubmed Central PMCID: 2997364.
68. Huang L, Jolly LA, Willis-Owen S, et al. A noncoding, regulatory
mutation implicates HCFC1 in nonsyndromic intellectual dis-
ability. Am J Hum Genet. Oct 5, 2012;91(4):694–702. PubMed
PMID: 23000143. Pubmed Central PMCID: 3484651.
69. Raymond FL, Tarpey P. The genetics of mental retardation.
Hum Mol Genet. Oct 15, 2006;15 Spec No 2:R110-6. PubMed
PMID: 16987873.
70. Basel-Vanagaite L, Attia R, Yahav M, et al. The CC2D1A, a mem-
ber of a new gene family with C2 domains, is involved in autosomal
recessive nonsyndromic mental retardation. J Med Genet. 2005
Jul 20. PubMed PMID: 16033914.
71. Molinari F, Rio M, Meskenaite V, et  al. Truncating neurotrypsin
mutation in autosomal recessive nonsyndromic mental retarda-
tion. Science. Nov 29, 2002;298(5599):1779–1781. PubMed
PMID: 12459588.
72. Higgins JJ, Pucilowska J, Lombardi RQ, Rooney JP. A mutation in
a novel ATP-dependent Lon protease gene in a kindred with mild
mental retardation. Neurology. Nov 23, 2004;63(10):1927–1931.
PubMed PMID: 15557513.
73. Najmabadi H, Hu H, Garshasbi M, et al. Deep sequencing reveals
50 novel genes for recessive cognitive disorders. Nature. Oct 6,
2011;478(7367):57–63. PubMed PMID: 21937992.
74. Slager RE, Newton TL, Vlangos CN, Finucane B, Elsea SH.
Mutations in RAI1 associated with Smith-Magenis syndrome. Nat
Genet. 2003 Apr;33(4):466–468. PubMed PMID: 12652298.
75. Petrij F, Giles RH, Dauwerse HG, et al. Rubinstein-Taybi syndrome
caused by mutations in the transcriptional co-activator CBP. Nature.
Jul 27, 1995;376(6538):348–351. PubMed PMID: 7630403.
76. Kleefstra T, Smidt M, Banning MJ, et  al. Disruption of the gene
euchromatin histone methyltransferase1 (Eu-HMTase1) is associ-
ated with the 9q34 subtelomeric deletion syndrome. J Med Genet.
2005 Apr;42(4):299–306. PubMed PMID:  15805155. Pubmed
Central PMCID: 1736026.
77. Brockschmidt A, Todt U, Ryu S, et  al. Severe mental retardation
with breathing abnormalities (Pitt-Hopkins syndrome) is caused
by haploinsufficiency of the neuronal bHLH transcription fac-
tor TCF4. Hum Mol Genet. Jun 15, 2007;16(12):1488–1494.
PubMed PMID: 17478476.
78. Vissers LE, de Ligt J, Gilissen C, et al. A de novo paradigm for men-
tal retardation. Nat Genet. 2010 Dec;42(12):1109–1112. PubMed
PMID: 21076407.
79. Rauch A, Wieczorek D, Graf E, et  al. Range of genetic muta-
tions associated with severe non-syndromic sporadic intellec-
tual disability:  an exome sequencing study. Lancet. Nov 10,
2012;380(9854):1674–1682. PubMed PMID: 23020937.
G enetics and G eno m ics o f N euro -P syc h iatric D iseases , V   •   5 3 9
80. Bailey A, Le Couteur A, Gottesman I, et  al. Autism as a strongly
genetic disorder: evidence from a British twin study. Psychol Med.
1995 Jan;25(1):63–77. PubMed PMID: 7792363.
81. Jamain S, Quach H, Betancur C, et al. Mutations of the X-linked
genes encoding neuroligins NLGN3 and NLGN4 are associ-
ated with autism. Nat Genet. 2003;34(1):27–29. PubMed
PMID: 12669065.
82. Gauthier J, Bonnel A, St-Onge J, et  al. NLGN3/NLGN4 gene
mutations are not responsible for autism in the Quebec population.
Am J Med Genet B Neuropsychiatr Genet. Jan 5, 2005;132(1):74–
75. PubMed PMID: 15389766.
83. International Molecular Genetic Study of Autism Consortium. A
full genome screen for autism with evidence for linkage to a region
on chromosome 7q. Hum Mol Genet. 1998 Mar;7(3):571–578.
PubMed PMID: 9546821.
84. International Molecular Genetic Study of Autism Consortium. A
genomewide screen for autism: strong evidence for linkage to chro-
mosomes 2q, 7q, and 16p. Am J Hum Genet. 2001 Sep;69(3):570–
581. PubMed PMID: 11481586.
85. Philippe A, Martinez M, Guilloud-Bataille M, et al. Genome-wide
scan for autism susceptibility genes. Paris Autism Research
International Sib Pair Study. Hum Mol Genet. 1999 May;8(5):805–
812. PubMed PMID: 10196369.
86. Yonan AL, Alarcon M, Cheng R, et  al. A genomewide screen of
345 families for autism-susceptibility loci. Am J Hum Genet. 2003
Oct;73(4):886–897. PubMed PMID: 13680528.
87. Auranen M, Vanhala R, Varilo T, et  al. A genomewide screen for
autism-spectrum disorders: evidence for a major susceptibility locus
on chromosome 3q25-27. Am J Hum Genet. 2002 Oct;71(4):777–
790. PubMed PMID: 12192642.
88. Alarcon M, Cantor RM, Liu J, Gilliam TC, Geschwind DH.
Evidence for a language quantitative trait locus on chromosome 7q
in multiplex autism families. Am J Hum Genet. 2002 Jan;70(1):60–
71. PubMed PMID: 11741194.
89. Liu J, Nyholt DR, Magnussen P, et  al. A genomewide screen for
autism susceptibility loci. Am J Hum Genet. 2001 Aug;69(2):327–
340. PubMed PMID: 11452361.
90. Risch N, Spiker D, Lotspeich L, et  al. A genomic screen of
autism: evidence for a multilocus etiology. Am J Hum Genet. 1999
Aug;65(2):493–507. PubMed PMID: 10417292.
91. Durand CM, Betancur C, Boeckers TM, et  al. Mutations in the
gene encoding the synaptic scaffolding protein SHANK3 are
associated with autism spectrum disorders. Nat Genet. 2007
Jan;39(1):25–27. PubMed PMID:  17173049. Pubmed Central
PMCID: 2082049.
92. Sebat J, Lakshmi B, Malhotra D, et  al. Strong association of de
novo copy number mutations with autism. Science. Apr 20,
2007;316(5823):445–449. PubMed PMID:  17363630. Pubmed
Central PMCID: 2993504.
93. Kim HG, Kishikawa S, Higgins AW, et al. Disruption of neurexin 1
associated with autism spectrum disorder. Am J Hum Genet. 2008
Jan;82(1):199–207. PubMed PMID: 18179900. Pubmed Central
PMCID: 2253961.
94. Alarcon M, Abrahams BS, Stone JL, et  al. Linkage, asso-
ciation, and gene-expression analyses identify CNTNAP2
as an autism-susceptibility gene. Am J Hum Genet. 2008
Jan;82(1):150–159. PubMed PMID: 18179893. Pubmed Central
PMCID: 2253955.
95. Weiss LA, Arking DE, Gene Discovery Project of Johns Hopkins,
the Autism Consortium, Daly MJ, Chakravarti A. A genome-wide
linkage and association scan reveals novel loci for autism. Nature.
Oct 8, 2009;461(7265):802–808. PubMed PMID:  19812673.
Pubmed Central PMCID: 2772655.
96. O’Roak BJ, Deriziotis P, Lee C, et  al. Exome sequencing in spo-
radic autism spectrum disorders identifies severe de novo mutations.
Nat Genet. 2011 Jun;43(6):585–589. PubMed PMID: 21572417.
Pubmed Central PMCID: 3115696.
 97. Sanders SJ, Murtha MT, Gupta AR, et  al. De novo mutations
revealed by whole-exome sequencing are strongly associated with
autism. Nature. May 10, 2012;485(7397):237–241. PubMed
PMID: 22495306. Pubmed Central PMCID: 3667984.
98. Neale BM, Kou Y, Liu L, et  al. Patterns and rates of exonic de
novo mutations in autism spectrum disorders. Nature. May 10,
2012;485(7397):242–245. PubMed PMID:  22495311. Pubmed
Central PMCID: 3613847.
99. Yu TW, Chahrour MH, Coulter ME, et  al. Using whole-exome
sequencing to identify inherited causes of autism. Neuron. Jan 23,
2013;77(2):259–273. PubMed PMID: 23352163.
100. Noor A, Whibley A, Marshall CR, et  al. Disruption at the
PTCHD1 locus on Xp22.11 in autism spectrum disor-
der and intellectual disability. Science Trans Med. Sep 15,
2010;2(49):49ra68. PubMed PMID:  20844286. Pubmed
Central PMCID: 2987731.
101. Lindsay EA, Morris MA, Gos A, et  al. Schizophrenia and chro-
mosomal deletions within 22q11.2. Am J Hum Genet. 1995
Jun;56(6):1502–1503. PubMed PMID:  7762575. Pubmed
Central PMCID: 1801107.
5 4 0   •   G eno m ics in C linical P ractice
102. Karayiorgou M, Morris MA, Morrow B, et al. Schizophrenia sus-
ceptibility associated with interstitial deletions of chromosome
22q11. Proc Natl Acad Sci U S A. Aug 15, 1995;92(17):7612–7616.
PubMed PMID: 7644464. Pubmed Central PMCID: 41195.
103. Moreno-De-Luca D, Consortium S, Mulle JG, Simons Simplex
Collection Genetics C, et al. Deletion 17q12 is a recurrent copy
number variant that confers high risk of autism and schizophre-
nia. Am J Hum Genet. Nov 12, 2010;87(5):618–630. PubMed
PMID: 21055719. Pubmed Central PMCID: 2978962.
104. Mulle JG, Dodd AF, McGrath JA, et al. Microdeletions of 3q29
confer high risk for schizophrenia. Am J Hum Genet. Aug 13,
2010;87(2):229–236. PubMed PMID:  20691406. Pubmed
Central PMCID: 2917706.
105. Glessner JT, Reilly MP, Kim CE, et al. Strong synaptic transmis-
sion impact by copy number variations in schizophrenia. Proc Natl
Acad Sci U S A. Jun 8, 2010;107(23):10584–10589. PubMed
PMID: 20489179. Pubmed Central PMCID: 2890845.
106. Lejeune J, Turpin R, Gautier M. Chromosomic diagnosis of
mongolism. Arch Fr Pediatr. 1959;16:962–963. PubMed
PMID: 14415503.
 36.
CLINICAL CANCER GENOMICS
Joanne Ngeow and Charis Eng
INTRODUCTION
From the vantage point of 2014, it is difficult to imagine a
time when cancer was not widely accepted as a genetic dis-
ease, in the most basic sense of being caused by alterations in
the structure and function of genes. Indeed, it was not until
the second half of the twentieth century that the heritable
nature of common cancers started to be widely accepted.1,2
Over the decades since then, genealogical and epidemiolog-
ical observations set the stage for a burst of scientific activ-
ity in the 1990s, when developments in positional cloning
facilitated the discovery of genes responsible for the most
common cancer predisposition syndromes, and for several
less common ones. These discoveries provided invaluable
insights in the biology of not only inherited cancer, but
also sporadic disease, and led the way for cancer genomics
research and gene-directed management and treatment.
This is reflected in the U.S. Department of Health and
Human Services’ recently released Healthy People 2020
benchmarks, aimed at improving the health care of all
Americans. These healthcare objectives are released every
decade using firm evidence-based information regarding
cost-effective clinical benefits at the population level. The
2020 objectives focused on new health-promotion areas
to concentrate on and for the first time included genomic
medicine in the list of priorities. The genomic objectives of
Healthy People 2020 emphasize the importance of obtain-
ing a family and genetic history as a potentially powerful
guide for clinical and public health initiatives.
The first genomic recommendation is that women with
a family history of breast or ovarian cancer should receive
genetic counseling. In 2005, 23% of women with a family
history of breast or ovarian cancer received genetic counsel-
ing. Healthy People 2020 would like to see at least a 10%
improvement.3 The second recommendation is to increase
the number of patients newly diagnosed with colorectal can-
cer who obtain genetic testing to rule out Lynch syndrome.
These genomic recommendations are based on the thought
541
that knowing this information will lead to gene-enabled
management and improve the health of involved patients.
By increasing their knowledge through genetic counseling
and appropriate testing, the risk of developing cancer in
high-risk individuals can be significantly reduced.
I N H E R I T E D S US C E P T I B I L IT Y TO C A N C E R
Currently, over 200 hereditary cancer susceptibility syn-
dromes have been described, the majority of which are
inherited in an autosomal dominant manner. Although
many of these are rare syndromes, they are thought to
account for at least 5–10% of all cancer, amounting to
a substantial burden of morbidity and mortality in the
human population (Figure 36.1). An inherited cancer
susceptibility is suspected in families with the following
characteristics: two or more relatives with the same type of
cancer on the same side of the family; several generations
affected; earlier ages of cancer diagnosis than what is typi-
cally seen for that cancer type; individuals with multiple
primary cancers; the occurrence of cancers in one family,
which are known to be genetically related (such as breast
and ovarian cancer, or colon and uterine cancer); and the
occurrence of nonmalignant conditions and cancer in the
same person and/or family (e.g., Marfanoid habitus and
medullary thyroid cancer in multiple endocrine neoplasia
type 2B [MEN 2B]) (Figure 36.2). Because of phenotypical
variability, age-related penetrance, and gender-specific can-
cer risks, however, many families with an inherited cancer
syndrome will not meet these criteria.
Cancers that arise as a result of a strong germline pre-
disposition are typically managed differently from those
that arise sporadically. Patients may undergo more extensive
local therapy if they are at increased risk for metachronous
malignancy. Obvious examples include the consideration
of bilateral mastectomy instead of breast conservation in
patients with breast cancer carrying a germline BRCA1
or BRCA2 mutation, and subtotal colectomy instead of
 Inherited Although highly penetrant genes are the ones most
5– likely to be clinically relevant, they only explain a small
10% Familial portion of the hereditability of cancer. As cancer is rela-
15–20%
tively common in the general population, it is possible
to have a chance clustering of the same or related can-
cers within a family (Figure 36.1). These familial cluster-
ings are most likely due to low-penetrance alleles with
risks that do not clearly exceed an action threshold.
70–80% Genome-wide association studies (GWAS) have further
Sporadic
identified large numbers of genetic variants that are,
individually, associated with very limited increases in
risk, and the clinical utility of identifying these variants
is completely unclear.6
Figure 36.1 
Estimated frequency distribution of heritable, “familial,” For the reader’s reference, a table is included with the
and sporadic cancers. The majority of common cancers are believed more commonly encountered inherited cancer syndromes
to be sporadic; 5–10% are heritable and arise due to highly penetrant (Table 36.1). From this longer list, we have chosen four
germline mutations in Mendelian genes. An additional 10–15% are
referred to as “familial” and may be caused by low-penetrance genes,
syndromes to highlight how understanding the underlying
gene–environment interactions, or both. genetic cause of a disease allows for gene-enabled manage-
ment. We discuss in detail hereditary breast and ovarian
cancer syndrome (HBOC) and hereditary non-polyposis
limited resection in patients with colorectal cancer who
colorectal cancer syndrome (HNPCC), as targeted by
have Lynch syndrome. Germline changes associated with
Healthy People 2020, to illustrate the clinical utility of
susceptibility have also recently been predictive of differ-
germline genetic testing and risk-reduction strategies. We
ential response to treatment approaches. Recent success
highlight Cowden syndrome (CS), a condition associated
of inhibitors of poly(ADP-ribose) polymerase inhibitors
with a spectrum of protean malignant and benign clinical
in BRCA-mutated breast, ovarian, and pancreatic cancers
features, to illustrate the clinical diagnostic difficulty in
indicates a possible role for this in the future,4 as does the
identifying patients at risk. Genotype-specific recommen-
suggestion of sensitivity to cis-platinum in the neoadjuvant
dations are available for component neoplasias where there
treatment of BRCA1-mutated breast cancer.5 Perhaps the
are clear genotype–phenotype correlations, and we high-
most important application of genetic testing is in individ-
light multiple endocrine neoplasia 2 (MEN 2) to illustrate
ual risk assessment for developing cancer, allowing for suc-
how genotype–phenotype correlation fine-tunes clinical
cessful risk-reduction strategies for the at-risk patient whilst
management.
avoiding unwarranted surveillance in unaffected family
members.
H E R E D I TA RY B R E A S T C A N C E R
Sporadic Inherited
Increasing numbers of genes have been implicated in the
development of breast cancer. Although 5–10% of breast
cancers are believed to be caused by mutations in single
genes, including such mutations predispose carriers to a
very high lifetime risk of developing breast, ovarian and
other cancers. In addition to BRCA1 and BRCA 2, heredi-
tary cancer syndromes that include breast cancer risk are
caused by germline mutations in PTEN, TP53, STK11,
• Later ages of onset (60s–70s) • Early ages of onset (under 50s) CDH1, NF1, and SDHB/D, and germline promoter hyper-
• Little or no family history of • Multiple generations with methylation of KLLN (Figure 36.3). Importantly, for many
cancer cancer
• Single or unilateral cancers • Multiple or bilateral cancers of these genes, women who carry these germline mutations
• Clustering of certain cancers (e.g. are at a higher risk of developing breast cancer at a young
breast and ovarian)
age than the recommended age of mammogram screening
Figure 36.2  Clinical “red flags” suggestive of heritable cancers. for the general population.

5 4 2   •   G e no m ic s in C l inic a l P r actic e
Table 36.1  HEREDITARY CANCER SYNDROMES ACCORDING TO ORGAN SYSTEMS

SYNDROME (OMIM ENTRY ) COMPONENT TUMORS MODE OF GENES

INHERITANCE
Hereditary Breast Cancer Syndromes
Hereditary breast cancer and ovarian cancer Breast cancer Dominant BRCA1
­syndrome (113705, 600185) Ovarian cancer Recessive BRCA2
Prostate cancer BRCA2
Pancreatic cancer
Melanoma
Fanconi anemia/medulloblastoma
Cowden syndrome (158350) Breast cancer Dominant PTEN
Thyroid cancer SDHB, C, D
Endometrial cancer KLLN
Renal cancer AKT1
Colon cancer PIK3CA
Li-Fraumeni syndrome (151623) Soft-tissue sarcoma Dominant TP53
Breast cancer
Osteosarcoma
Leukemia
Brain tumors
Adrenocortical carcinoma
Bannayan-Riley-Ruvalcaba syndrome (153480) Breast cancer Dominant PTEN
Thyroid cancer
Hereditary lobular breast cancer with diffuse Lobular breast cancer Dominant CDH1
gastric cancer (192090) Gastric cancer
Androgen insensitivity syndrome (300068); Male breast cancer Dominant AR
Reifenstein syndrome (312300) Prostate cancer
Klinefelter syndrome Male breast cancer – XXY
Extragonadal germ cell tumor
Werner syndrome (277700) Sarcoma/osteosarcoma Recessive WRN
Meningioma
Breast cancer
Ataxia telangiectasia (208900) Breast cancer Recessive ATM
Leukemia
Lymphoma
Hereditary Gastrointestinal Malignancies
HNPCC, including “Lynch II” syndrome Colon cancer Dominant MLH1
(120435, 120436, 114500, 114400) Endometrial cancer MSH2
Ovarian cancer MSH6
Renal pelvis cancers PMS2
Ureteral cancers
Pancreatic cancer
Stomach and small-bowel cancers
Hepatobiliary cancers
Familial polyposis, including attenuated Colon cancer Dominant APC
phenotype (175100) Medulloblastoma
Desmoid tumors
Hepatoblastoma
Thyroid cancer (papillary)
Small-bowel cancer (duodenum,
periampullary)
Biliary cancer
Stomach cancer
Pancreatic cancer
MUTYH-associated polyposis (604933) Colon cancer Recessive MUTYH
Hereditary mixed polyposis syndrome (601228) Colon cancer Dominant GREM1
Hereditary gastric cancer (137215) Diffuse gastric cancers Dominant CDH1

(continued)
Table 36.1 CONTINUED

SYNDROME (OMIM ENTRY ) COMPONENT TUMORS MODE OF GENES

INHERITANCE
Juvenile polyposis (174900) Gastrointestinal cancers Dominant SMAD4/DPC4
Pancreatic cancer BMPR1A
Peutz-Jeghers syndrome (175200) Colon cancer Dominant STK11
Small-bowel cancer
Breast cancer
Ovarian cancer (granulosa cell)
Pancreatic cancer
Hereditary melanoma pancreatic cancer syn- Pancreatic cancer Dominant CDKN2A/p16
drome (606719) Melanoma
Hereditary pancreatitis (167800) Pancreatic cancer Dominant PRSS1
Turcot syndrome (276300) Colon cancer Dominant APC
Basal-cell carcinoma MLH1
Ependymoma MSH2
Medulloblastoma (APC typically) MSH6
Glioblastoma (Lynch typically) PMS2
Familial Barrett’s esophagus (614266) Esophageal adenocarcinoma Dominant MSR1
CTHRC1
ASCC1
Familial gastrointestinal stromal tumor Gastrointestinal stromal tumors Dominant KIT
(606764) PDGFR
Endocrine Cancer Predisposition Syndromes
MEN 1 (131100) Pancreatic islet cell tumors Dominant MEN1
Pituitary adenomas
Parathyroid adenomas
MEN 2 (171400) Medullary thyroid cancers Dominant RET
Pheochromocytoma
Parathyroid hyperplasia
Familial medullary thyroid cancer (155240) Medullary thyroid cancers Dominant RET
Familial papillary thyroid cancer (188500) Papillary thyroid cancer Dominant Multiple loci
Hereditary paraganglioma-pheochromocytoma Paragangliomas Dominant SDHA
syndromes (171300, 16800, 115310, 605373, Pheochromocytomas SDHB
601650) SDHC
SDHD
SDHAF2
VHL
TMEM127
MAX
Hereditary Genodermatoses with Cancer Predisposition
Melanoma syndrome (155600, 155601, Malignant melanoma Dominant CDKN2 (p16)
609448, 608035) CDK4
CMM
Basal-cell cancers, Gorlin syndrome (109400) Basal-cell cancers Dominant PTCH
Brain tumors
Cowden syndrome (158350) See above Dominant See above
Neurofibromatosis type 1 (162200) Neurofibrosarcoma Dominant NF1
Pheochromocytomas
Optic gliomas
Meningiomas
Breast cancers
Malignant peripheral nerve-sheath
tumors
Neurofibromatosis type 2 (101000) Vestibular schwannomas Dominant NF2

(continued)

5 4 4   •   G e no m ic s in C l inic a l P r actic e
Table 36.1 CONTINUED

SYNDROME (OMIM ENTRY ) COMPONENT TUMORS MODE OF GENES

INHERITANCE
Tuberous sclerosis complex (191100) Myocardial rhabdomyoma Dominant TSC1
Multiple bilateral renal TSC2
angiomyolipoma
Ependymoma
Renal cancer
Giant cell astrocytoma
Carney complex (160980, 605244) Myxoid subcutaneous tumors Dominant PRKARIA
Primary nodular adrenocortical
hyperplasia
Testicular Sertoli cell tumor
Atrial myxoma
Pituitary adenoma
Mammary fibroadenoma
Thyroid carcinoma
Schwannoma
Muir-Torre syndrome (158320) Sebaceous carcinoma Dominant MLH1
Sebaceous epitheliomas MSH2
Sebaceous adenomas
Keratoacanthomas
Colon cancer
Laryngeal carcinoma
Malignant gastrointestinal tract
tumors
Malignant genitourinary tract
tumors
Xeroderma pigmentosum (278700, Skin cancer Recessive XPA,B,C,D,E,F,G
278720, 278760, 74740, 278780, Melanoma POLH
278750, 133510) Leukemia
Rothmund Thomson syndrome (268400) Basal-cell carcinoma Recessive RECOL4
Squamous-cell carcinoma
Osteogenic sarcoma
Genitourinary Cancer Predisposition Syndromes
Simpson-Golabi-Behmel syndrome (312870) Embryonal tumors X-linked GPC3
Wilms’ tumor recessive
Von-Hippel-Lindau syndrome (193300) Hemangioblastomas of retina and Dominant VHL
central nervous system
Renal cell cancer
Pheochromocytomas
Beckwith-Wiedemann syndrome (130650) Wilms’ tumor Dominant CDKN1C
Hepatoblastoma NSD1
Adrenal carcinoma
Gonadoblastoma
Wilms’ tumor syndrome (194070) Wilms’ tumor Dominant WT1
WAGR: Wilms’ tumor, aniridia, genitourinary Wilms’ tumor Dominant WT1
abnormalities, mental retardation (194072) Gonadoblastoma
Birt-Hogg-Dube syndrome (135150) Renal tumors Dominant FLCN
Papillary renal cancer syndrome (605074) Papillary renal cancer Dominant MET, PRCC
Hereditary leiomyomatosis and renal cell Type 2 papillary renal cancer Dominant FH
cancer (150800)
Constitutional t(3;8) translocation (603046) Renal cell cancer Dominant TRC8
Hereditary bladder cancer (109800) Bladder cancer Sporadic Unknown
Unknown
Hereditary testicular cancer (273300) Testicular cancer Possibly Unknown
X-linked Unknown
Possibly
recessive
(continued)
C l inic a l C a nc e r G e no m ic s   •   5 4 5
Table 36.1 CONTINUED

SYNDROME (OMIM ENTRY ) COMPONENT TUMORS MODE OF GENES

INHERITANCE
Hematological Malignancy Predisposition Syndromes
Bloom syndrome (210900) Leukemia Recessive BLM
Carcinoma of the tongue
Squamous cancers
Wilms’ tumor
Colon cancer
Fanconi anemia (227650) Leukemia Recessive FANCA,B,C
Squamous cancers FANCD1,D2,
Skin carcinoma FANCE,F,G
Hepatoma FANCL
Shwachman-Diamond syndrome (260400) Myelodysplasia Recessive SBDS
Acute myelogenous leukemia
Nijmegen breakage syndrome (251260) Lymphoma Recessive NBS1
Glioma
Medulloblastoma
Rhabdomyosarcoma
Autoimmune lymphoproliferative syndrome/ Hodgkin’s and non-Hodgkin’s Dominant FAS
Canale-Smith syndrome (601859) lymphoma FASLG
CASP10
Immunodeficiency Syndromes
Wiskott-Aldrich (301000) Hematopoietic malignancies X-linked WAS
recessive
Common variable immune deficiency Lymphomas Recessive Unknown
(240500) Dominant Unknown
Severe combined immune deficiency B-cell lymphoma X-linked IL2RG
(102700, 300400, 312863, 601457, 600802, recessive ADA
602450) Recessive JAK3
RAG1
RAG2
IL7R
CD45
Artemis
X-linked lymphoproliferative syndrome Lymphoma X-linked SH2D1A
(308240) recessive
Central Nervous System/Vascular Cancer Predisposition Syndromes
Retinoblastoma (180200) Retinoblastoma Dominant RB1
Osteosarcoma
Soft-tissue sarcoma
Melanoma
Glioblastoma
Rhabdoid predisposition syndrome (601607) Rhabdoid tumors Dominant SNF5/N11
Medulloblastoma
Choroid plexus tumors
Primitive neuroectodermal tumors
Sarcoma/Bone Cancer Predisposition Syndromes
Li-Fraumeni syndrome (151623) See above
Retinoblastoma (180200) See above
Multiple exostoses (133701) Chondrosarcoma Dominant EXT1
EXT2
Carney complex (160980, 605244) See above Dominant PRKAR1A
Werner syndrome (277700) See above Recessive WRN
OMIM: Online Mendelian Inheritance in Man database.

5 4 6   •   G e no m ic s in C l inic a l P r actic e
 BRCA1, 20%

Unknown, 53–54%
Sporadic Familial
Hereditary
BRCA2, 20%

~5% Highly Penetrant Genes:

PTEN, TP53, STK11, CDH1,
NF1, SDHB/D, KLLN

Relative contribution of predisposition genes to heritable breast cancer. The majority of cases of breast cancer are sporadic, with BRCA1
Figure 36.3 
and BRCA2 germline mutations accounting for the majority of hereditary breast cancers. Other highly penetrant genes contributing to heritable
breast cancer syndromes together probably make up 5% of all hereditary and familial cases.

Hereditary Breast and Ovarian Cancer Syndrome
Hereditary breast–ovarian cancer (HBOC, MIM 113705)
syndrome is the most common form of heritable breast can-
cer and is caused by germline mutations in BRCA1 on 17q11
and BRCA2 on 13q12-q13.7,8 The prevalence of BRCA1/2
mutations in the general population is estimated to be
between 1 in 500 and 1 in 1000,9,10 although this is much
higher in certain founder populations, such as the Ashkenazi
Jewish and Icelandic populations (Table 36.2).11–27 In a
population-based study, BRCA1/2 mutations were found in
6% of women diagnosed with breast cancer at an age of less
than 36 years and in 4% of women diagnosed at ages from 36
to 45.28 Personal and family history characteristics suggestive
of HBOC are listed in Figure 36.4.
The proportion of HBOC attributable to BRCA1
or BRCA2 (BRCA1/2) mutations is poorly defined, and
estimates depend upon the population studied, the num-
ber of breast and ovarian cancer cases in the family, and
the mutation-detection technique utilized. In families
with two or more cases of breast cancer (<50 years) and at
least one case of ovarian cancer, more than 90% will carry
a deleterious germline BRCA1/2 mutation. In families
with site-specific breast cancer, these figures are lower and

Table 36.2  FOUNDER MUTATIONS IN BRCA1 AND BRCA2

POPULATION MUTATIONS IN BRCA1 MUTATIONS IN BRCA2 REFERENCES

African-American M1775R, 1832del5, 5296del4 Gao et al. (1997)
Ashkenazi-Jewish 185delAG, 5382insC 6174delT Neuhausen et al. (1996), Struewing et al. (1995), and
Tonin et al.(1996)
Dutch 2904delAA, del510, del3835 5573insA Peelen et al. (1997) and Petrij-Bosch et al. (1997)
French-Canadian C4446T 8765delAG Simard et al. (1994) and Tonin et al. (1999)
English 4184del4 2157delG Evans et al. (2004)
Icelandic 999del5 Thorlacius et al. (1996)
Norwegian 1675delA, 1135insA Andersen et al. (1996)
Polish 5382insC, C61G, 4153delA Gorski et al. (2000)
Russian 5382insC, 4153delA Gayther et al. (1997)
Scottish 2800delAA Liede et al. (2000)
Swedish Q563X, 3166ins5, 4486delG Hakansson et al. (1997) and Johannsson et al. (1996)
1201del1,2594delC

C l inic a l C a nc e r G e no m ic s   •   5 4 7
 Personal History
• Breast Cancer Diagnosed at less than 40 years of age
• Bilateral breast cancer, especially when first cancer is
diagnosed at less than 50 years of age
• History of breast and ovarian cancer

Family History
• Two or more family members diagnosed with breast cancer
at 50 years of age or younger
• Both breast and ovarian cancer in the family
• One or more family members diagnosed with breast cancer
at 50 years of age or younger and of Ashkenazi Jewish
ancestry
• One or more family members diagnosed with ovarian cancer
and of Ashkenazi Jewish ancestry
• Male breast cancer

Figure 36.4  Personal and family history characteristics suggestive of inherited breast and ovarian cancer syndrome.

range from 30% to 81%.29 In addition, a significant pro-
portion of young breast cancer cases unselected for family
history will carry a BRCA1/2 mutation. This proportion
is as high as 20% in women of Ashkenazi Jewish ances-
try diagnosed before the age of 40.14,30 The cancer risks in
BRCA1-mutation-positive individuals are confined largely
to the breast and ovaries31 (Figure 36.5), while pancreatic
and prostate cancer, melanoma, and other cancer risks are
increased in BRCA-mutation-positive individuals.32 Male
BRCA1/2-mutation-positive individuals have an increased
risk of developing prostate cancer (16% lifetime risk by age
70), and male BRCA2-mutation-positive individuals are at
risk for breast cancer (5–10% lifetime risk).12
The mutations in BRCA1/2 causing HBOC are inher-
ited in an autosomal dominant fashion and are highly pen-
etrant. Mutations in BRCA1 typically confer a lifetime risk
between 50% and 85% of developing female breast cancer
and a 15% to 40% lifetime risk of ovarian cancer.33 BRCA2
germline mutations can lead to a lifetime breast cancer risk
between 6% and 7.5% for men, 50% and 85% for women,
and approximately 14% and 27% for ovarian cancer.32,34,35
BRCA1 encodes a 1863-amino-acid polypeptide.7
The N-terminus contains a zinc-finger domain that inter-
acts both directly and indirectly with DNA.7 Exon 11 of
BRCA1 encodes over 60% of the protein, contains two
nuclear localization signals, and interacts with RAD51,36
p53,37 RB1,38 and MYC.39 The C-terminus contains a tran-
scription activation domain and interacts with RNA poly-
merase II40 and BRCA2,41 as well as other proteins. BRCA1
has been shown to play a role in a multitude of cellular pro-
cesses, including, but not limited to, DNA transcription,
cell cycle regulation, DNA damage repair, and apoptosis.42
It acts as a tumor suppressor, which is supported by the fact
that many BRCA1-associated tumors show loss of heterozy-
gosity (LOH) of the wild-type allele.43,44 The BRCA2 pro-
tein has also been implicated in DNA repair and has been
shown to regulate the activity of RAD51, which is required
for homologous recombination leading to double-stranded
DNA repair.45 Interestingly, biallelic inactivating mutations
in BRCA2 were identified in patients with Fanconi anemia
group D1 (FANC-D1)46, linking two seemingly unrelated
syndromes to a common DNA damage response pathway.
Evidence for genotype–phenotype correlations exists.35,47,48
Mutations in and around exon 11 of BRCA2, which is often
referred to as the “ovarian cancer cluster region” (OCCR),
confer a higher ovarian cancer risk and lower breast cancer
risk than mutations in other areas of the gene.24,49
Identifying Patients for Genetics Risk Assessment
Individuals with a greater than 5–10% probability of having
a BRCA1/2 mutation are considered candidates for genetics

5 4 8   •   G e no m ic s in C l inic a l P r actic e
 Brain
General population <1%
HNPCC 1–3%
Melanoma
General population 2%
Thyroid PHTS 6%
General population 1%
PHTS 35%

Breast Stomach
General population 12% General population <1%
HBOC-BRCA1 50–85% HNPCC 11–19%
HBOC-BRCA2 50–85%
PHTS ~85%
Renal Cell
General population 1.6%
Hepatobiliary Tract PHTS 34%
General population <1%
HNPCC 2–7% Small Bowel
General population <1%
Pancreatic HNPCC 1–4%
General population <1%
HBOC-BRCA1 2.3% Ovarian
HBOC-BRCA2 2.3% General population 1.6%
HNPCC 4% HBOC-BRCA1 15–40%
Colon HBOC-BRCA2 14–25%
General population 5.5% HNPCC 3–13%
HNPCC 60–80%
Endometrial
PHTS 9%
General population 2.6%
Urinary Tract HNPCC 20–60%
General population <1% PHTS 28%
HNPCC 4–5%

Figure 36.5 
Lifetime risks for cancer in patients with HBOC, HNPCC, and PHTS. Lifetime cancer risks for each hereditary cancer syndrome are
shown together with organ-specific lifetime risk for cancer in the general population.

referral for risk assessment and consideration of gene testing
in the context of genetic counseling. Multiple models are
available to aid in determining good candidates for genetic
testing (Table 36.3).50 As mentioned above, most germline
BRCA1/2 mutations are caused by point mutations or small
deletions or insertions. However, large genomic rearrange-
ments appear to be the source of a significant minority of
BRCA1/2 mutations,51 and these should be considered if
testing by full PCR-based Sanger sequencing is negative on
an individual basis.
Cancer Surveillance Recommendations
Enhanced surveillance is recommended to promote early
detection of breast cancer in women with a known or
suspected genetic mutation. Table 36.4 summarizes the
recommendations of a National Institutes of Health (NIH)
consensus panel and other recommendations for screening
women who test positive for a BRCA1 or BRCA2 mutation.
Ovarian cancer screening measures available (transvaginal
ultrasound examination and serum CA-125 concentration)
have limited sensitivity and specificity and have not been
shown to reduce mortality.52 Thus, the standard of care is
prophylactic bilateral salpingo-oophorectomy, with or
without total hysterectomy.
Risk-Reduction Strategies
Several risk-reduction strategies have been proposed to
help patients lower their risk of developing breast or ovar-
ian cancer. In the Breast Cancer Prevention Trial, the use of
tamoxifen for primary prevention of breast cancer reduced

C l inic a l C a nc e r G e no m ic s   •   5 4 9
Table 36.3  PROBABILITY MODELS TO ESTIMATE THE LIKELIHOOD OF A BRCA MUTATION

PROBABILITY MODEL GENE PREDICTORS DISADVANTAGES

Couch BRCA1 Ovarian cancer Based on smaller sample size
Ashkenazi Jewish descent Not useful for families with site-specific ovarian
Average age of breast cancer diagnosed cancer
<55 years old Need extrapolation for unaffected women
Shattuck-Eldens BRCA1 Younger age at diagnosis Not for use in women diagnosed with breast can-
Bilateral breast cancer at diagnosis cer before age 30
Ovarian cancer Need extrapolation for unaffected women
Relative with breast and/or ovarian cancer
Ashkenazi Jewish descent
Fran BRCA1 and BRCA2 Ovarian cancer Not for use in women diagnosed with breast can-
Bilateral breast cancer cer at 50 years or older
Age <40 years at breast cancer diagnosis Need for extrapolation for unaffected women
BRCAPRO BRCA1 and BRCA2 Cancer penetrance in mutation carriers Need computer software
Cancer status Time-consuming
Age of first- and second-degree relatives Varying probability depending on affected rela-
affected tive chosen for analysis

the risk of invasive breast cancer by 49%.53 The women in
this trial were not known mutation carriers but rather were
determined to be at increased risk using the Gail model, a
model devised to assess the five-year lifetime risk for breast
cancer by taking into account a woman’s personal and fam-
ily histories (Table 36.5 compares the two commonly used
breast cancer risk-assessment models). Tamoxifen has sub-
sequently been shown to reduce breast cancer incidence in
BRCA1/2 mutation carriers.43,54–56 Before starting women
on tamoxifen, its risks, including the increased risk of
thrombosis, and its benefits should be discussed.
Prophylactic mastectomy and oophorectomy can pre-
vent cancers in mutation carriers, but surgical menopause
may have other deleterious effects, and mastectomy in
particular is not acceptable to some women. A  retrospec-
tive study on a cohort of 639 women with a strong fam-
ily history of breast cancer demonstrated that prophylactic
mastectomy reduced the risk of breast cancer by at least
90%51 and increased life expectancy by three to five years.57
Similar results were noted in a prospective study of over
400 women who were mutation positive.58 Prophylactic
removal of the ovaries is presented as an option to women
who are BRCA1/2-mutation-positive, supported by mul-
tiple prospective studies confirming that such surger-
ies not only decrease the incidence of subsequent breast
and ovarian cancer, but also find occult early-ovarian
neoplasms.59,60 Individuals at high a priori risk of being
BRCA1/2-mutation-positive (Figure 36.4) should strongly
consider genetic testing for presence of BRCA1/2 mutation
prior to surgery for their primary cancer diagnosis.
C OWD E N S Y N D RO M E
Cowden syndrome (CS; [Mendelian Inheritance in Man]
MIM 158350) is an autosomal dominant disorder and part
of the PTEN hamartoma tumor syndrome (PHTS), which
also includes subsets of Bannayan-Riley-Ruvalcaba syn-
drome (BRRS), Proteus syndrome (PS), and Proteus-like
syndrome.61,62 PHTS is molecularly defined as having a
germline PTEN mutation irrespective of clinical syndrome.

Table 36.4 SCREENING RECOMMENDATIONS FOR BRCA1 OR BRCA2 MUTATION CARRIERS

SCREENING TEST STARTING AGE (YEARS) INTERVAL

Breast cancer Breast self-exam 18–21 Monthly
Clinical breast exam 25–35 6–12 mthly
Mammogram 25–35 6–12 mthly
Ultrasound (US) breast Possible adjunct to mammogram
MRI breast Possible adjunct to mammogram
Ovarian cancer Clinical pelvic exam Uncertain 6 mthly
Transvaginal US 25–35 6–12 mthly
CA-125 25–35 6–12 mthly

5 5 0   •   G e no m ic s in C l inic a l P r actic e
Table 36.5  PROS AND CONS FOR GAIL AND CLAUS MODELS IN BREAST CANCER RISK ASSESSMENT

RISK ASSESSMENT FACTORS IN CONSIDERATION PROS CONS

MODEL
Gail model Age Calculates factors other than family Does not consider: second-degree rela-
Age at first live birth history tives, age of affected relatives, family his-
Age at menarche Calculates individual risk tory of ovarian cancer, ethnicity
Number of first-degree relatives Calculates risk over Overestimates risk in those with benign
with breast cancer specified time interval breast biopsies
Number of breast biopsies
Claus model Number of first- and second-degree Consideration of first and Risk based only on family history
relatives with breast cancer second-degree relatives
Age of diagnosis for affected Risk based on specified relationship
relatives to affected relative
Calculation of lifetime risk or risk
over specified 10-year interval

CS is a multiple hamartoma syndrome with a high risk of
benign and malignant tumors of the thyroid, breast, endo-
metrium, and other organs. BRRS is a congenital disorder
characterized by macrocephaly, intestinal polyposis, lipo-
mas, and pigmented macules of the glans penis. PS is a
complex, highly variable disorder involving congenital mal-
formations and overgrowth of multiple tissues. Proteus-like
syndrome is undefined but refers to individuals with signifi-
cant clinical features of PS who do not meet the diagnostic
criteria for PS.
The estimated incidence of CS is approximately one in
200,000,63,64 but as many of the features of CS are common
in the general population (e.g., fibrocystic breast disease,
uterine fibroids), the incidence may be even higher. The
most commonly reported manifestations of CS include
macrocephaly, mucocutanous lesions, fibrocystic breast dis-
ease, thyroid abnormalities, multiple uterine leiomyomata,
and gastrointestinal polyps (Table 36.6).65
CS is a highly penetrant genetic disorder. More than
90% of individuals with PTEN mutations are believed to
manifest some feature of the syndrome (although rarely
cancer) by age 20, and by age 30 nearly 100% of muta-
tion carriers are believed to have developed at least some
of the mucocutanous signs. Affected individuals also
have an increased risk for several malignancies, including
female breast cancer (85% lifetime risk), thyroid cancer
with an over-representation of follicular thyroid cancer
and endometrial cancer (Figure 36.6).66–68 Consensus diag-
nostic criteria for CS were first developed in 1996 by the
International Cowden Consortium (ICC), and they form
the basis for the National Comprehensive Cancer Network
Guidelines (Table 36.7).65 Approximately 25% of individ-
uals who meet the strict diagnostic criteria for CS have a
pathogenic PTEN mutation, including large deletions.66
The CS susceptibility locus was mapped to 10q22-q23.63
Subsequently, germline mutations in PTEN, localized to
10q23.3, were found in CS.69 PTEN spans nine exons,
with a transcript length of 3,417 bps, encoding for a 403
amino acid protein (Figure 36.6). The C-terminal domain
contains important subdomains that are common to other
signal-transducing molecules. First, PTEN contains a C2
domain, which is associated with phospholipid-binding

Table 36.6  CANCER RISKS AND SCREENING RECOMMENDATIONS FOR PTEN-HAMARTOMA TUMOR

SYNDROME (PHTS)

CANCER GENERAL POPULATION LIFETIME RISK AGE AT SCREENING RECOMMENDATIONS

RISK (%) WITH PHTS (%) PRESENTATION
Breast 12 ~85 40s Starting at age 30: annual mammogram, con-
sider MRI for patients with dense breast
Thyroid 1 35 30s–40s Baseline ultrasound at diagnosis, annual ultra-
sound and clinical examination
Endometrial 2.6 28 40s–50s Starting at age 30: annual endometrial biopsy or
transvaginal ultrasound
Renal cell 1.6 34 50s Starting at age 40: renal imaging every 2 years
Colon 5 9 40s Starting at age 40: colonoscopy every 2 years
Melanoma 2 6 40s Annual dermatological examination

C l inic a l C a nc e r G e no m ic s   •   5 5 1
 15 185 351 403
Domain Phosphatase domain C2 domain PDZ

PEST
Figure 36.6 
Functional domains of PTEN. The domain structure of phosphatase and tensin homologue mutated on chromosome ten (PTEN). PTEN
is a 403-amino acid protein that comprises a phosphatase domain, a C2 domain, and a PDZ-binding domain. PDZ domains are significant regions
for protein–protein interactions that play a vital role in cellular signal transduction. The N-terminal domain contains the phosphatase domain (the
enzymatic activity of PTEN), and it is therefore not surprising that the majority of PTEN mutations occur within this domain.

regions, including phospholipid membranes. Additionally,
the C terminus features a PDZ-binding motif, which inter-
acts strongly with the phosphatase domain. PDZ domains
are significant regions for protein–protein interactions that
play a vital role in cellular signal transduction.70 Removal
of the PDZ domain reduces the ability of PTEN to inhibit
one of its substrates, AKT. The C terminus also contains
Table 36.7  INTERNATIONAL COWDEN SYNDROME
CONSORTIUM OPERATIONAL CRITERIA FOR THE
DIAGNOSIS OF COWDEN SYNDROME (VER. 2000)
Pathognomonic criteria (mucocutaneous lesions):
Trichilemmomas, facial
Acral keratoses
Papillomatous papules
Mucosal lesions
Major criteria:
Breast carcinoma
Thyroid carcinoma (non-medullary), especially follicular thyroid
carcinoma
Macrocephaly (megalencephaly, e.g., ≥97th percentile)
Lhermitte–Duclos disease (LDD)
Endometrial carcinoma
Minor criteria:
Other thyroid lesions (e.g., adenoma or multinodular goiter)
Mental retardation (e.g., IQ ≤75)
Gastrointestinal hamartomas
Fibrocystic disease of the breast
Lipomas
Fibromas
GU tumors (e.g., renal cell carcinoma or uterine fibroids) or
malformation
Operational diagnosis in an individual:
1. Mucocutaneous lesions alone if (a) there are six or more facial
papules, of which three or more must be trichilemmoma;
(b) cutaneous facial papules and oral mucosal papillomatosis;
(c) oral mucosal papillomatosis and acral keratosis; (d) palmo-
plantar keratosis, six or more
2. Two major criteria, of which one must include macrocephaly or
LDD
3. One major and three minor criteria
4. Four minor criteria
Operational diagnosis in a family where one individual is diagnos-
tic for Cowden:
1. One or more pathognomonic criterion/ia
2. Any one major criterion with or without minor criteria
3. Two minor criteria
PEST sequences, in addition to several phosphorylation
sites located in the last 50 amino acids, which both appear
to play important roles in PTEN protein stability. The
N-terminal domain contains the phosphatase domain (the
enzymatic activity of PTEN) and it is, therefore, not sur-
prising that the majority of PTEN mutations occur within
this domain.71,72
Recent research efforts to identify other susceptibility
genes in CS have resulted in six other germline susceptibil-
ity genes. Approximately 10% of individuals with classic
CS or CS-like (CSL) phenotypes carry germline variants
in the genes encoding three of the four subunits of suc-
cinate dehydrogenase (SDH) or mitochondrial complex
II.73 Single-exon KLLN, on 10q23, encodes KILLIN, and
shares a bidirectional promoter with PTEN. Up to 30% of
individuals with CS/CSL phenotypes, without germline
PTEN or SDHx mutations, were recently found to carry
germline KLLN promoter hypermethylation.74 Another 9%
of unrelated CS individuals without germline PTEN muta-
tions were found to have germline PIK3CA mutations, and
2% harbored germline AKT1 mutations.75 Functionally,
all mutation-positive protein lysates showed a significant
increase in P-Thr308-AKT1 levels, further supporting their
potential role as novel CS-susceptibility genes.75
Identifying Patients for Genetics Risk Assessment
In part because of phenotypical heterogeneity, and in part
because of the high frequency of de novo PTEN germline
mutations,76 a family history of associated cancers may not
be apparent. Additionally, many of the benign features of
CS are common in the general population, making the
diagnosis of CS a challenge for most clinicians. To aid clini-
cal diagnosis, a clinical predictor (Cleveland Clinic PTEN
Risk Calculator) was developed based on clinical features
derived from a prospective study of over 3000 patients
suspected of having CS.66 The questionnaire-based clini-
cal decision tool is available online (http://www.lerner.ccf.​
org/gmi/ccscore/) to assist clinicians at the point of patient
care. Based on a patient’s presentation, a score will be
derived that corresponds to an estimated risk for germline

5 5 2   •   G e no m ic s in C l inic a l P r actic e
PTEN mutation (Figure 36.7) to guide clinicians for refer-
ral to genetics professionals.
Surveillance and Management of Cowden Syndrome
The mucocutaneous manifestations of Cowden syndrome
are rarely life-threatening. If the patient is asymptomatic,
observation alone is prudent. When symptomatic, topical
agents (e.g., 5-fluorouracil), curettage, cryosurgery, or laser
ablation may provide only temporary relief.77 Treatment
for the benign and malignant manifestations of PHTS is
the same as for their sporadic counterparts. Some women
at increased risk for breast cancer consider prophylactic
mastectomy, especially if breast tissue is dense or if repeated
breast biopsies have been necessary. The recommendation
of prophylactic mastectomy is a generalization for women
at increased risk for breast cancer from a variety of causes,
not just from PHTS, and is best managed by breast sur-
geons with a specialty interest in high-risk breast cancer
patients.
The most serious consequences of PHTS relate to the
increased risk of cancers, including breast, thyroid, endome-
trial, and (to a lesser extent) renal cancers. In this regard, the
most important aspect of management of any individual
with a PTEN mutation is increased cancer surveillance to
detect any tumors at the earliest, most treatable stages.
Based on new information on lifetime risk estimates
of CS component cancers,68 expert opinion currently rec-
ommends an annual comprehensive physical examination
with special attention paid to skin and thyroid (all ages),
and breast (adults) as well as annual formal dermatologi-
cal examination, starting at 18 years. Screening for various
component cancers may start at the ages listed in Table 36.6,
or should start 5–10 years before the youngest diagnosis of
that particular cancer type in the family, whichever is earlier.
H E R E D I TA RY C O L O R EC TA L C A N C E R
Most (70–80%) colorectal cancers (CRC) are sporadic.
Around 20% to 30% of cases have a familial component;
that is, there are one or more affected first- or second-degree
relatives (or both) and a potentially definable genetic
basis.78,79 In the United States, around 10% to 15% of adults
have a history of CRC in a first-degree relative.80 Molecular
genetics has made a significant impact by identifying germ-
line and somatic mutations associated with the develop-
ment of CRC.81 Single-gene germline mutations conferring
a hereditary susceptibility (high risk) to CRC account for

Case 1 Case 2 Case 3

Breast Cancer Macrocephaly (6) Single
at Age 55(1) Breast Cancer hamartomatous
Thyroid Cancer at Age 38(4) polyp (10)
at Age 44(4) Hashimoto’s
thyroidits (4)
Skin lipomas (1)

Total Score
0 5 10 15 20 25 30 35 40

Risk for
PTEN mutations 0.005 0.01 0.02 0.03 0.05 0.1 0.15 0.30 0.6 0.8 0.95 0.99

Cleveland Clinic (CC) PTEN Risk Calculator: three illustrative cases. The CC score (Total Score) is first derived by a sum of the weights
Figure 36.7 
of positive clinical features. To illustrate, three hypothetical cases are presented, each corresponding to a CC score of 5 points, 10 points, and 15
points, respectively. Case 1 may present with breast cancer at age 55 (1 point), with background of thyroid cancer at age 44 (4 points), with a
final score of 5 and corresponding prior probability <1% of having a PTEN mutation. Case 2 may present with breast cancer at age 38 (4 points)
and concurrent macrocephaly (6 points), with a final score of 10 and corresponding prior probability of 3%. Case 3 may present with a single
hamartomatous gastrointestinal polyp (10 points) found on endoscopy, Hashimoto’s thyroiditis (4 points), and lipomas (1 point), for a final score of
15 and corresponding prior probability of 10%. Adapted from Tan et al, American Journal of Human Genetics 2011.66

C l inic a l C a nc e r G e no m ic s   •   5 5 3
round 6% to 7% of all CRCs, the most common of which
is hereditary non-polyposis colorectal cancer (Table 36.1).79
Hereditary Non-Polyposis Colorectal
Cancer Syndrome
Hereditary non-polyposis colorectal cancer (HNPCC,
MIM 114500) was described as a clinical entity, character-
ized by a strong family history of early-onset colon cancer
and associated cancers without numerous polyps, long
before causative genes were identified.82 HNPCC is the
most common form of hereditary colorectal cancer, affect-
ing as many as one in 400 individuals, and it is caused by a
germline mutation in any one of five DNA mismatch-repair
(MMR) genes (MLH1, MSH2, MSH6, PMS1, and
PMS2). To honor the physician who described the first
families with HNPCC, Henry Lynch, individuals/families
with HNPCC who are found to have a germline MMR
gene mutation are designated as having Lynch syndrome.83
Molecular screening of all patients with colorectal can-
cer has shown that about 3% of these cancers are attribut-
able to germline MMR gene alterations.84–87 Penetrance in
HNPCC is typically high, and the phenotype is quite vari-
able within families. Males with HNPCC have virtually a
100% chance of developing colorectal cancer by age 70.88
Women appear to have a higher lifetime risk of endometrial
cancer (42–60%) than colorectal cancer (54%).88–91 The
lifetime risk for stomach cancer is 13–19%, although this
may be much higher in families of Asian descent.92 Women
with HNPCC have a 9–12% lifetime risk of developing
ovarian cancer.88 Other cancers seen in HNPCC include
cancers of the small intestine, pancreas, brain, hepatobili-
ary tract, and urinary tract, for which the lifetime risk is in
the range of 2–4%88 (Figure 36.8). Sebaceous gland tumors,
multiple keratoacanthomas, basal cell carcinomas, and pos-
sibly breast cancer may be more common in the Muir-Torre
variant (MIM 158320)  of HNPCC. The Turcot variant
(MIM 276300) should be considered when the family his-
tory includes glioblastoma multiforme.
There are differences in phenotype, depending on the
MMR gene involved. Endometrial cancer may be more
common in families with germline MSH6 alterations as
compared to the other mismatch repair genes.79,93,94 In the
Muir-Torre variant of HNPCC, the vast majority of germ-
line mutations identified have been in MSH2, with a few
families harboring MLH1 alterations. One-third of patients
with Turcot syndrome have mutations in MSH2. In certain
populations, founder mutations in MLH1 and MSH2 have
been identified in Lynch syndrome families. In the Finnish
population, for example, two mutations in MLH1 account
5 5 4   •   G e no m ic s in C l inic a l P r actic e
for 63% of all disease-causing mutations in families with
HNPCC.95 Other founder mutations include those seen
in the Swiss, the Chinese, and Ashkenazi Jews.96–98 Notably,
genomic rearrangements accounted for 24% of all muta-
tions in a series of 59 clinically selected North American
HNPCC families.99 Testing for large genomic rearrange-
ments and germline mutations that affect promoter regions
or deep intronic regulatory regions may be missed by the
use of conventional methods alone but may be warranted,
especially for families for whom tumor immunohistochem-
istry testing suggests a germline mutation but PCR-based
Sanger sequencing of germline DNA is negative.100–102
Identifying Patients for Genetics Risk Assessment
The International Collaborative Group on HNPCC estab-
lished diagnostic criteria, known as “the Amsterdam cri-
teria,” in 1991 (Table 36.8).103 These criteria were further
modified in 1999 (Amsterdam II Criteria) in an attempt
to incorporate extracolonic cancers.104 However, stud-
ies have shown that only 40–80% of families that meet
the original criteria, and no more than 50% meeting the
modified criteria, have germline mutations in the associated
mismatch-repair genes.105 For this reason, a third set of crite-
ria, the Bethesda criteria, were developed. When compared
to the Amsterdam I and II criteria, the Bethesda criteria are
the most sensitive, especially in identifying HNPCC fami-
lies with pathogenic mutations.106 The modified Bethesda
criteria will identify people without a strong family history
of colorectal cancer who are considered at high-risk.107
Microsatellite Instability (MSI) Testing
DNA replication errors characterize tumors with loss-of-
function mutations in MMR genes. These can be detected
as microsatellite instability (MSI), which is the finding that,
in the same individual, the number of repeats in a given
repeating sequence of DNA varies from cell to cell instead
of being constant. More than 90% of colorectal cancers in
people with germline DNA MMR gene mutations have
high MSI, whereas less than 15% of sporadic colorectal can-
cers do (Figure 36.8).
Although the sensitivity of the Bethesda Guidelines is
better than 72%, these guidelines are burdensome to recall
and have been shown to be poorly implemented in clinical
practice. In 2009, the Evaluation of Genomic Applications
in Practice and Prevention (EGAPP) Working Group rec-
ommended that all newly diagnosed patients with colorec-
tal cancer be screened for HNPCC through MSI testing or
immunohistochemistry testing of tumor samples.108 While
 (A)
Normal Dysplastic
crypt
Inactivation of Mutation Adenomatous CIN 18q loss Inactivation Colon Cancer
APC (5q loss) of K-Ras lesion defect DCC, DPC4 of p53
(17p loss)

(B) Familial MSI-H

pathway (HNPCC)
Germline MMR
mutational inactivation
(MLHL, MSH2, MSH6,
PMS2, PMS1) APC Mutation
β-catenin of K-Ras Adenomatous Colon Cancer
Axin2 lesion

Mutational inactivation of genes with

Normal microsatellites: TGFβIIR, BAX

APC Mutation
β-catenin of B-Raf
Somatic MMR Axin2 (>K-Ras)
epigenetic inactivation
(MLH1)
Serrated Colon Cancer
adenomatous lesion
Sporadic MSI-H
pathways (CIMP) Inactivation of tumor suppresor genes
by promoter hyprmethylation

Figure 36.8 Genetic model of sporadic and MMR-related colorectal cancers. A: Sporadic colorectal cancers (CRC) are believed to arise from
adenomatous polyps, resulting from accumulation of mutations in oncogenes and tumor-suppressor genes. Mutations in adenomatous polyposis
coli (APC), β-catenin, or other components of this pathway mediate the transition of single pre-neoplastic cells to aberrant crypt foci and then
to adenoma and colorectal carcinoma. B: In about 20% of colorectal cancers, mismatch repair (MMR) function is inactivated, either by somatic
mutations or by epigenetic inactivation leading to microsatellite instability (MSI-H). Mutational inactivation of MMR genes is most commonly
observed as the second hit in patients who already carry germline mutations in MMR genes and fall under the hereditary nonpolyposis colorectal
cancer syndrome. Epigenetic inactivation of MMR genes most commonly affects hypermethylation of the MLH1 promoter. Adapted from Fearon and
Bommer, Devita, Hellman and Rosenberg’s Cancer: Principles and Practice of Oncology, 8 ed.179
th

the adoption of such universal institution-wide practice has
been problematic and varied across institution types,109 a
recent study has shown that successful implementation of
universal screening can be achieved with a high uptake of
diagnostic genetic services. To achieve the greatest success,
the program at a minimum must have representation from
the fields of colorectal surgery, gastroenterology, gynecol-
ogy, pathology, and genetics.110
Risk Reduction for HNPCC-associated Cancers
Colorectal cancer-risk reduction in HNPCC is mainly
through colonoscopy to identify and remove polyps.
Surveillance for various types of cancer in individuals with
HNPCC is based on expert recommendations and on case
series that have demonstrated a reduction in CRC incidence
with periodic colonoscopic surveillance111–113 (Table 36.9).
H E R E D ITA RY T H Y RO I D C A N C E R
While thyroid cancer only accounts for approximately 1%
of all newly diagnosed cancer cases, it is the most common
malignancy of endocrine organs. While medullary thyroid
carcinoma (MTC) represents only 10% of all thyroid can-
cers, 10 to 25% of patients with MTC are due to mutation
in a high penetrance predisposition gene.72,114 Molecular
studies have led to increased appreciation of the heteroge-
neity of thyroid cancers, with hereditary predisposition,
somatic mutation, and epigenetic modulation all contribut-
ing to tumor behavior.115 For example, the discovery of the
RET (REarranged during Transfection) proto-oncogene as
the causative gene for multiple endocrine neoplasia type 2
(MEN 2) enabled gene-based molecular diagnosis, predic-
tive testing, and genotype-specific management for affected
individuals.
MU LT I P L E E N D O C R I N E N EO P L A S I A
2 ( M E N 2)
MEN 2 is an autosomal dominant cancer syndrome classified
into three clinical subtypes: MEN 2A (MIM 171400), famil-
ial medullary thyroid carcinoma (FMTC; MIM 155240), and
MEN 2B (MIM 162300) (Table 36.10). All three subtypes
involve high risk of developing MTC; MEN 2A and MEN

C l inic a l C a nc e r G e no m ic s   •   5 5 5
Table 36.8  HEREDITARY NONPOLYPOSIS COLORECTAL CANCER (HNPCC) CRITERIA

Original Amsterdam criteria (Amsterdam criteria I)There should be at least three relatives with colon cancer and:
1. One should be a first-degree relative of the other two.
2. At least two successive generations should be affected.
3. At least one should be diagnosed before age 50.
4. Familial adenomatous polyposis should be excluded.
5. Tumors should be verified by pathological examination.
Revised Amsterdam criteria (Amsterdam criteria II)There should be at least three relatives with an HNPCC-associated cancer (cancer of the
colorectum, endometrium, small bowel, ureter, or renal pelvis) and:
1. One should be a first-degree relative to the other two.
2. At least two successive generations should be affected.
3. At least one should be diagnosed before age 50.
4. Familial adenomatous polyposis should be excluded.
5. Tumors should be verified by pathological examination.
Bethesda guidelines (ver. 2004)Tumors from individuals should be tested for microsatellite instability (MSI) in the following situations:
1. Colorectal cancer diagnosed in a patient who is less than 50 years of age
2. Presence of synchronous, metachronous colorectal, or other HNPCC-associated tumors, regardless of age
3. Colorectal cancer with the MSI-H histology diagnosed in a patient who is less than 60 years of age
4. Colorectal cancer diagnosed in one or more first-degree relatives with an HNPCC-related tumor, with one of the cancers being diagnosed
under age 50 years
5. Colorectal cancer diagnosed in two or more first- or second-degree relatives with HNPCC-related tumors, regardless of age

2B have an increased risk for pheochromocytoma (PC).116
MEN 2A also has an increased risk for primary hyperpara-
thyroidism (HPT). Additional features in MEN 2B include
mucosal neuromas of the lips and tongue, distinctive facies
with enlarged lips, ganglioneuromatosis of the gastrointestinal
tract, and an asthenic “Marfanoid” body habitus.117
All three subtypes are caused by germline RET muta-
tions.118–121 RET was the first proto-oncogene to be impli-
cated in an inherited cancer-susceptibility syndrome. The
frequency range of germline RET mutations reported for
apparently sporadic MTC cases varies widely, from approxi-
mately 5% based on direct testing to about 25% using fam-
ily history–based assessments.122–127
RET comprises 21 exons and encodes a transmembrane
receptor tyrosine kinase expressed in tissues and tumors
derived mainly from the neural crest.128–133 The structure
of the RET receptor is like that of most receptor tyro-
sine kinases, with a large extracellular domain, compris-
ing a cysteine-rich domain and a cadherin-like domain,
a transmembrane domain, and an intracellular tyrosine
kinase domain134,135 (Figure 36.9). However, the RET recep-
tor is unique in that it requires a multicomponent complex
to trigger activation. RET has to bind one of four GFRα
family of GPI-linked co-receptors before binding one of
four members of the glial cell line–derived neurotrophic
factor family of ligands (GDNF; neurturin, persephin, and
artemin) in a heterohexameric complex.136–138 RET signals
down a number of well-characterized signal transduction
pathways, including the MAP kinase–RAS–RAF and
phosphoinositide 3-kinase pathways.139,140 MEN 2A- and
FMTC-related mutations affecting cysteine codons result
in disulphide bond–mediated inter-receptor binding and
thus constitutive activation.141,142
Identifying Patients for Genetics Risk Assessment
The diagnostic criteria for MEN 2A are the presence
of MTC, PC, and/or HPT, and a germline RET gene

Table 36.9 SCREENING RECOMMENDATIONS FOR PATIENTS WITH CONFIRMED OR HIGHLY SUSPECTED

HNPCC

CANCER RISK PROCEDURE AGE TO START FREQUENCY

(YEARS) (YEARS)
Colon cancer Colonoscopy 20–25 1–3
Endometrial cancer and ovarian Gynecological examination ( endometrial sampling), 30–35 1–2
cancer transvaginal ultrasound, CA-125
Urinary tract cancer Ultrasonography, urine cytology 30–35 1–2
Stomach cancer Gastroscopy 30–35 1–2
Skin cancer Clinical exam 30–35 1

5 5 6   •   G e no m ic s in C l inic a l P r actic e
 Table 36.10  CLINICAL FEATURES OF MEN 2 SUBTYPES

SUBTYPE MEDULLARY THYROID CANCER PHEOCHROMOCYTOMA PARATHYROID DISEASE

MEN 2A 95% 50% 20–30%
FMTC 100% 0% 0%
MEN 2B 100% 50% Uncommon

Codon
Mutation MEN 2A FMTC MEN 2B

609

611

Cysteine-rich 618
Doamain
620

630

634

Transmembrane

768
Tyrosine Kinase
Doamain 790

804

*(if in combination
with other variants)

883
Tyrosine Kinase
891
Doamain
918

-Intermediate Risk -High Risk -Highest Risk

Germline missense mutations in RET associated with MEN 2A, FMTC, and MEN 2B. Schematic diagram of the RET receptor and
Figure 36.9 
distribution of mutated codons associated with different risk levels for aggressive MTC in MEN 2 syndromes. The most common MEN 2–
associated mutations are reported. For codon 804, there is observed phenotypical heterogeneity. There is an increased risk for aggressive MTC as
well as pheochromocytomas in patients harboring a codon 804 mutation with a second RET mutation or variant, and such individuals look MEN
2B–like. While this schematic illustrates the most common germline RET mutations causing MEN 2, note that somatic M918T mutations are very
common in sporadic MTC tissue.

C l inic a l C a nc e r G e no m ic s   •   5 5 7
mutation. The mutation frequency is greater than 98%, thus
few families present a diagnostic dilemma when fewer than
three clinical features are present. In cases in which only
one or two clinical features are present, the diagnosis can be
made either when a first-degree relative shows MEN 2A fea-
tures or when a RET mutation is identified. At a minimum,
the diagnosis can be made when two clinical features are
present, even when an autosomal dominant pattern is not
evident and a RET mutation has not been demonstrated.143
FMTC is, by definition, phenotypically limited to MTC.144
The importance of identifying all MEN 2–associated
MTC cases for optimal patient and family management
has led to a recommendation to offer pre- and post-test
genetic counseling and genetic testing for RET mutations
to all patients presenting with MTC, irrespective of other
features and family history.143
Genetic Testing and Genotype–Phenotype
Correlations in MEN 2
Germline RET mutations have been identified in approxi-
mately 95% of all MEN 2 cases, with 98% of MEN 2A pro-
bands found to have a mutation, 97% in MEN 2B, and 85%
in FMTC.145–147 The characteristic mutational spectrum
found in MEN 2A includes missense mutations in one of
cysteine codons 609, 611, 618, 620 (exon 10), or 634 (exon
11).145,148 Genotype–phenotype analyses reveal that codon
634 mutations are associated with the presence of PC and
hyperparathyroidism (HPT).145 In particular, the C634R
mutation is probably associated with the development of
HPT.145,149,150 FMTC-associated mutations occur at the
same cysteine codons as those in MEN 2A, although muta-
tions at codons 609–620 are more proportionately frequent
in FMTC than in MEN 2A. Germline M918T and A883F
mutations occur in 95% and ~2%, respectively, of MEN 2B
patients.145–147 These two mutations are unique to MEN 2B
and have never been observed in MEN 2A or FMTC.
Genetic testing for RET mutations is therefore recom-
mended to proceed in stepwise fashion, based on the high
rate of exon 10 or 11 mutation in MEN 2A and character-
istically affected codons in MEN 2B.143,148,151 The various
genotypes also have an impact on penetrance. For example,
it has now become obvious that RET codons’ 918, 883, and
634 mutations have the highest penetrance, predisposing to
MEN 2B and MEN 2A with MTC, PC, and HPT involve-
ment, respectively.152,153 In contrast, germline V804M
mutations and perhaps cysteine codon 609/611 mutations
have the lowest penetrance and the older ages of onset.145
A recent study showed that codon-associated penetrance by
age 50 ranged from 60% (codon 611) to 86% (620). More
5 5 8   •   G e no m ic s in C l inic a l P r actic e
advanced stage and increasing risk of metastases correlated
with mutation in codon position (609→620) the closer one
approached the juxtamembrane domain.154
Surveillance in MEN 2
MEN 2 showcases how genetic testing facilitates presymp-
tomatic diagnosis and prevention, which are tailored to the
specific genotype.155 Codon-specific risks and age of onset
have been described for PC.156,157 It has been suggested that
annual screening for PC may be warranted, beginning at
age 10 years, for carriers of mutations in codons 918, 634,
and 630. Carriers with mutations at other codons could ini-
tiate screening at age 20 years.157
Management in MEN 2 Is Genotype-Dependent
Total thyroidectomy with central cervical lymph-node dissec-
tion is the recommended surgical procedure for MEN2 muta-
tion carriers.143,158 Earlier age at surgery results in less chance of
recurrence, but there have been differing opinions about the
optimal timing of surgery. Genotype–phenotype correlations
in MEN 2 are strong, both in terms of being mutation-specific
and as highly predictive of age of MTC-onset, which makes
genotyping an indispensable tool in planning the age of pro-
phylactic surgery on an individual basis143,155,158 (Table 36.11).
An approach put forth by the National Comprehensive
Cancer Network143,158 is to stratify risk by genotype into the
categories, and to time surgery, as described in Table 36.11.72,143
Somatic RET mutations are also found in at least 50% of spo-
radic MTCs. Importantly, this link between MTC and RET
has recently led to the USFDA’s approval of vandetanib in
2011, an orally bioavailable inhibitor of RET, VEGFR-2,
and EGFR, for use in advanced MTC, even cases arising in
the sporadic setting, following dramatic improvements in
progression-free survival over placebo.159
S O M AT I C C A N C E R G E N O M I C S : B EYO N D
I N H E R I T E D C A N C E R SY N D RO M E S
Early successes in targeted cancer therapy have helped
enlighten our understanding. The best known of these is
extending imatinib’s use in chronic myeloid leukemia to
gastrointestinal stromal tumors (GIST), the majority of
which sport a receptor tyrosine kinase (called c-Kit) that is
activated by mutation,160,161 heralding an era of increasingly
extensive systematic sequencing of cancer genomes over the
past few years. This sequencing has identified several new
mutated somatic cancer genes encoding for proteins that
serve as targets for new therapeutics, particularly those that
Table 36.11  GENOTYPE-GUIDED MANAGEMENT RECOMMENDATIONS FOR MEDULLARY THYROID CANCER

ATA RISK MUTATIONS AGE OF AGE TO BEGIN SCREENING

LEVEL PROPHYLACTIC FOR FOR PARATHYROID
SURGERY PHEOCHROMOCYTOMA DISEASE
Level D (high- p.Ala883Phe p.Met918Thr As soon as possible in 8 yrs NA
est risk) p.Val804Met+p.Glu805Lys first year of life
p.Val804Met+p.Tyr806Cys
p.Val804Met+p.Ser904Cys
Level C p.Cys634Arg/Gly/Phe/Ser/Trp/Tyr <5 yrs 8 yrs 8 yrs
Level B p.Cys609Phe/Arg/Gly/Ser/Tyr Consider <5 yrs Codon 630 mutation: 8 yrs Codon 630 muta-
p.Cys611Arg/Gly/Phe/Ser/Trp/Tyr All others: 20 yrs tion: 8 yrs. All
p.Cys618Arg/Gly/Phe/Ser/Tyr ­others: 20 yrs
p.Cys620Arg/Gly/Phe/Ser/Trp/Tyr
p.Cys630Arg/Phe/Ser/Tyr
p.Asp631Tyr p.633/9 bp dup
p.634/12 bp dup p.Val804Met+p.Val778Ile4
Level A p.Arg321Gly p.531/9 bp dup May delay beyond age 20 yrs 20 yrs
p.532 dup p.Cys515Ser p.Gly533Cys 5 yrs if normal annual
p.Arg600Gln p.Lys603Glu basal and/or stimu-
p.Tyr606Cys p.635/insert ELCR; lated serum calcitonin;
p.Thr636Pro p.Lys666Glu p.Glu768Asp normal annual neck
p.Asn777Ser p.Leu790Phe p.Val804Leu/Met ultrasound; family his-
p.Gly819Lys p.Arg833Cys p.Arg844Gln tory of less aggressive
p.Arg866Trp p.Ser891Ala p.Arg912Pro MTC
Adapted from the American Thyroid Association Guidelines Task Force (Kloos et al., Thyroid; 2009143).

encode enzymes such as kinases and in which the muta-
tions result in constitutive activation. Examples include
BRAF, PIK3CA, AKT1, and IDH1,162–166 and searches
for inhibitors of the proteins encoded by many of these
mutated genes are in progress. A notable example of a pro-
tein activated through somatic mutation in cancer is BRAF,
a serine–threonine kinase. Activating somatic mutations in
BRAF were identified by a systematic search of candidate
cancer genes in a panel of diverse cancers; BRAF was then
shown to be mutated in about two-thirds of malignant
melanomas.165 Because of the limited treatment options for
metastatic malignant melanoma, the discovery of BRAF
mutations triggered multiple drug-discovery programs in
the academic and pharmaceutical sectors. Vemurafenib,
a potent inhibitor of mutated BRAF,167 has marked anti-
tumor effects against melanoma cell lines with the BRAF
V600E mutation, but not against cells with wild-type
BRAF.167,168 Vemurafenib produced improved rates of over-
all and progression-free survival in patients with previously
untreated melanoma with the BRAF V600E mutation169
and is currently the first-line recommended therapy for
melanomas with the BRAF V600E mutation.
The use of gene-expression signatures has further
empowered the identification of prognostic subclasses.
For example, gene-expression profiles have been used to
define the risk of relapse in patients with early-stage breast
cancer. Three profiles have shown prognostic ability:  the
70-gene profile MammaPrint, approved by the Food and
Drug Administration (FDA); the 21-gene recurrence
score Oncotype DX; and the 76-gene outcome Rotterdam
signature.170–172 These profiles, which are obtained with
the use of tumor-derived mRNA assayed on either DNA
microarrays or by quantitative reverse-transcriptase
polymerase-chain-reaction (RT-PCR) assay, are being used
in clinical practice to predict prognosis and hence to influ-
ence therapeutic intervention (Figure 36.10). Reflecting
this, the use of the Oncotype DX assay has been included
in the guidelines of the American Society of Clinical
Oncology for the evaluation of patients with node-negative,
ER-positive disease to identify those who may obtain the
most benefit from adjuvant tamoxifen and those who may
not require chemotherapy. This has led to a significant
change in how medical oncologists manage breast cancer
that is at low clinical risk for recurrence. A  recent multi-
center study of physician practice showed that the treat-
ment recommendation of medical oncologists has changed
for almost a third of patients in this subgroup on the basis
of this assay.173
C O N C LU S I O N S
The ultimate goal in oncology is to markedly decrease
death from cancer. The past 50 years of research have dem-
onstrated that the architecture of inherited genetic suscep-
tibility to cancers is defined by a collage of predisposition

C l inic a l C a nc e r G e no m ic s   •   5 5 9

Breast Adenocarcinoma
Figure 36.10 
Use of a gene-expression profile to define the risk of relapse in patients with early-stage breast cancer and to aid choice of adjuvant
treatment. Gene-expression signatures can be incorporated with clinical decision-making. Patients with a gene-expression profile predicting a high
risk of relapse can be offered the more aggressive option of adjuvant hormonal and chemotherapy for early-stage hormone-responsive breast cancers;
in contrast, patients with low-risk gene-expression profiles may do as well with hormonal therapy alone.
alleles with different levels of risk and prevalence in the
population. Improved detection of patients harboring del-
eterious mutations has allowed tailored, gene-enabled man-
agement both in terms of novel targeted therapy, as well as
by a targeted-surveillance approach for affected individuals.
However, technology continues to move apace, dramati-
cally reducing costs and making genome sequencing fairly
routine. The optimal management of low to moderately
penetrant genes conferring modestly increased relative risks
remains incompletely defined. The next frontier would be to
determine what the modifiers of specific risks for Mendelian
high-penetrance alleles are and to better understand how
low-penetrance susceptibility alleles “cross-talk” with
one another, to facilitate greater precision in risk predic-
tion. Additionally, recent efforts such as the Collaborative
Oncological Gene-Environment Study (COGS) shed light
on new genetic susceptibility loci for breast, ovarian, and
prostate cancers.174–178 These newly identified susceptibility
loci explain an increasing proportion of the familial risk of
these cancers. Understanding the biological basis of how
these germline susceptibility genes interact with somatically
acquired alterations will be key to ongoing efforts for novel
cancer-prevention and therapeutic strategies.
R E F E R E N C ES
1. Broca P. Traits des tumeurs: Tome premier, Des tumeurs en general.
1866; http://www.archive.org/stream/traitdes​tumeurs02brocgoog
- page/n12/mode/1up. Accessed Mar 2013.
2. Warhin A. Heredity with reference to carcinoma as shown by
the study of cases examined in the pathological laboratory of the
University of Michigan. Arch Intern Med. 1913;12:546–555.
3. U.S. Department of Health and Human Services. Healthy People
2020: http://www.healthypeople.gov/2020/default.​aspx. Accessed
Mar 8, 2013.
5 6 0   •   G e no m ic s in C l inic a l P r actic e
High Molecular Risk
Hormone Therapy
+
Chemotherapy
Gene Expression Hormone Therapy
Profiling Alone
Low Molecular Risk
4. Fong PC, Boss DS, Yap TA, et al. Inhibition of poly(ADP-ribose)
polymerase in tumors from BRCA mutation carriers. N Engl J Med.
Jul 9, 2009;361(2):123–134.
5. Byrski T, Gronwald J, Huzarski T, et  al. Pathologic complete
response rates in young women with BRCA1-positive breast
cancers after neoadjuvant chemotherapy. J Clin Oncol. Jan 20,
2010;28(3):375–379.
6. Offit K. Genomic profiles for disease risk: predictive or premature?
JAMA. Mar 19, 2008;299(11):1353–1355.
7. Miki Y, Swensen J, Shattuck-Eidens D, et al. A strong candidate for
the breast and ovarian cancer susceptibility gene BRCA1. Science.
Oct 7, 1994;266(5182):66–71.
8. Wooster R, Bignell G, Lancaster J, et  al. Identification of the
breast cancer susceptibility gene BRCA2. Nature. Dec 21–28,
1995;378(6559):789–792.
9. Claus EB, Risch N, Thompson WD. Genetic analysis of breast can-
cer in the cancer and steroid hormone study. Am J Hum Genet. Feb
1991;48(2):232–242.
10. Ford D, Easton DF. The genetics of breast and ovarian cancer. Br J
Cancer. Oct 1995;72(4):805–812.
11. Roa JC, Baeza R, Romeo E, Andrade L, Foradori A, Gonzalez S.
[Breast cancer: comparison between biochemical and immunohis-
tochemical determination of estrogen receptors]. Rev Med Chil.
Mar 1996;124(3):307–312.
12. Struewing JP, Hartge P, Wacholder S, et  al. The risk of can-
cer associated with specific mutations of BRCA1 and BRCA2
among Ashkenazi Jews. N Engl J Med. May 15, 1997;336(20):
1401–1408.
13. Gao Q, Neuhausen S, Cummings S, Luce M, Olopade OI. Recurrent
germ-line BRCA1 mutations in extended African American
families with early-onset breast cancer. Am J Hum Genet. May
1997;60(5):1233–1236.
14. Neuhausen S, Gilewski T, Norton L, et  al. Recurrent BRCA2
6174delT mutations in Ashkenazi Jewish women affected by breast
cancer. Nat Genet. May 1996;13(1):126–128.
15. Tonin P, Weber B, Offit K, et al. Frequency of recurrent BRCA1 and
BRCA2 mutations in Ashkenazi Jewish breast cancer families. Nat
Med. Nov 1996;2(11):1179–1183.
16. Peelen T, van Vliet M, Petrij-Bosch A, et  al. A high proportion
of novel mutations in BRCA1 with strong founder effects among
Dutch and Belgian hereditary breast and ovarian cancer families.
Am J Hum Genet. May 1997;60(5):1041–1049.
17. Petrij-Bosch A, Peelen T, van Vliet M, et al. BRCA1 genomic dele-
tions are major founder mutations in Dutch breast cancer patients.
Nat Genet. Nov 1997;17(3):341–345.
 18. Simard J, Tonin P, Durocher F, et al. Common origins of BRCA1
mutations in Canadian breast and ovarian cancer families. Nat
Genet. Dec 1994;8(4):392–398.
19. Tonin PN, Mes-Masson AM, Narod SA, Ghadirian P, Provencher
D. Founder BRCA1 and BRCA2 mutations in French Canadian
ovarian cancer cases unselected for family history. Clin Genet. May
1999;55(5):318–324.
20. Evans DG, Neuhausen SL, Bulman M, Young K, Gokhale D,
Lalloo F. Haplotype and cancer risk analysis of two common
mutations, BRCA1 4184del4 and BRCA2 2157delG, in high
risk northwest England breast/ovarian families. J Med Genet. Feb
2004;41(2):e21.
21. Thorlacius S, Olafsdottir G, Tryggvadottir L, et al. A single BRCA2
mutation in male and female breast cancer families from Iceland with
varied cancer phenotypes. Nat Genet. May 1996;13(1):117–119.
22. Andersen TI, Borresen AL, Moller P. A common BRCA1 mutation
in Norwegian breast and ovarian cancer families? Am J Hum Genet.
Aug 1996;59(2):486–487.
23. Gorski B, Byrski T, Huzarski T, et  al. Founder mutations in the
BRCA1 gene in Polish families with breast-ovarian cancer. Am J
Hum Genet. Jun 2000;66(6):1963–1968.
24. Gayther SA, Mangion J, Russell P, et al. Variation of risks of breast
and ovarian cancer associated with different germline mutations of
the BRCA2 gene. Nat Genet. Jan 1997;15(1):103–105.
25. Liede A, Cohen B, Black DM, et al. Evidence of a founder BRCA1
mutation in Scotland. Br J Cancer. Feb 2000;82(3):705–711.
26. Hakansson S, Johannsson O, Johansson U, et al. Moderate frequency
of BRCA1 and BRCA2 germ-line mutations in Scandinavian famil-
ial breast cancer. Am J Hum Genet. May 1997;60(5):1068–1078.
27. Johannsson O, Ostermeyer EA, Hakansson S, et  al. Founding
BRCA1 mutations in hereditary breast and ovarian cancer in south-
ern Sweden. Am J Hum Genet. Mar 1996;58(3):441–450.
28. Peto J, Collins N, Barfoot R, et  al. Prevalence of BRCA1 and
BRCA2 gene mutations in patients with early-onset breast cancer. J
Natl Cancer Inst. Jun 2, 1999;91(11):943–949.
29. Ford D, Easton DF, Stratton M, et  al. Genetic heterogeneity and
penetrance analysis of the BRCA1 and BRCA2 genes in breast can-
cer families. The Breast Cancer Linkage Consortium. Am J Hum
Genet. Mar 1998;62(3):676–689.
30. Offit K. Breast cancer and BRCA1 mutations. N Engl J Med. May 2,
1996;334(18):1197–1198; author reply, 1199–1200.
31. Brose MS, Rebbeck TR, Calzone KA, Stopfer JE, Nathanson KL,
Weber BL. Cancer risk estimates for BRCA1 mutation carriers
identified in a risk evaluation program. J Natl Cancer Inst. Sep 18,
2002;94(18):1365–1372.
32. Cancer risks in BRCA2 mutation carriers. The Breast Cancer Linkage
Consortium. J Natl Cancer Inst. Aug 4, 1999;91(15):1310–1316.
33. Armstrong K, Eisen A, Weber B. Assessing the risk of breast cancer.
N Engl J Med. Feb 24, 2000;342(8):564–571.
34. Barnes-Kedar IM, Plon SE. Counseling the at risk patient in
55. Duffy SW, Nixon RM. Estimates of the likely prophylactic effect of
the BRCA1 and BRCA2 Era. Obstet Gynecol Clin North Am. Jun
2002;29(2):341–366, vii.
35. Easton DF, Steele L, Fields P, et al. Cancer risks in two large breast
cancer families linked to BRCA2 on chromosome 13q12–13. Am J
Hum Genet. Jul 1997;61(1):120–128.
36. Scully R, Chen J, Plug A, et al. Association of BRCA1 with Rad51 in
mitotic and meiotic cells. Cell. Jan 24, 1997;88(2):265–275.
37. Zhang H, Somasundaram K, Peng Y, et al. BRCA1 physically associ-
ates with p53 and stimulates its transcriptional activity. Oncogene.
Apr 2, 1998;16(13):1713–1721.
38. Aprelikova ON, Fang BS, Meissner EG, et  al. BRCA1-associated
growth arrest is RB-dependent. Proc Natl Acad Sci U S A. Oct 12,
1999;96(21):11866–11871.
39. Wang Q, Zhang H, Kajino K, Greene MI. BRCA1 binds c-Myc
and inhibits its transcriptional and transforming activity in cells.
Oncogene. Oct 15, 1998;17(15):1939–1948.
C l inic a l C a nc e r G e no m ic s   •   5 6 1
40. Scully R, Chen J, Ochs RL, et al. Dynamic changes of BRCA1 sub-
nuclear location and phosphorylation state are initiated by DNA
damage. Cell. Aug 8, 1997;90(3):425–435.
41. Chen J, Silver DP, Walpita D, et al. Stable interaction between the
products of the BRCA1 and BRCA2 tumor suppressor genes in
mitotic and meiotic cells. Mol Cell. Sep 1998;2(3):317–328.
42. Somasundaram K. Breast cancer gene 1 (BRCA1): role in cell cycle
regulation and DNA repair—perhaps through transcription. J Cell
Biochem. Apr 15, 2003;88(6):1084–1091.
43. Smith SA, Easton DF, Evans DG, Ponder BA. Allele losses in the
region 17q12–21 in familial breast and ovarian cancer involve the
wild-type chromosome. Nat Genet. Oct 1992;2(2):128–131.
44. Cornelis RS, Neuhausen SL, Johansson O, et al. High allele loss rates
at 17q12-q21 in breast and ovarian tumors from BRCAl-linked
families. The Breast Cancer Linkage Consortium. Gene Chromosome
Canc. Jul 1995;13(3):203–210.
45. Jasin M. Homologous repair of DNA damage and tumorigenesis: the
BRCA connection. Oncogene. Dec 16, 2002;21(58):8981–8993.
46. Howlett NG, Taniguchi T, Olson S, et  al. Biallelic inac-
tivation of BRCA2 in Fanconi anemia. Science. Jul 26,
2002;297(5581):606–609.
47. Thompson D, Easton D. Variation in BRCA1 cancer risks by
mutation position. Cancer Epidemiol Biomarkers Prev. Apr
2002;11(4):329–336.
48. Gayther SA, Warren W, Mazoyer S, et  al. Germline mutations of
the BRCA1 gene in breast and ovarian cancer families provide
evidence for a genotype–phenotype correlation. Nat Genet. Dec
1995;11(4):428–433.
49. Thompson D, Easton D. Variation in cancer risks, by mutation
position, in BRCA2 mutation carriers. Am J Hum Genet. Feb
2001;68(2):410–419.
50. Panchal SM, Ennis M, Canon S, Bordeleau LJ. Selecting a BRCA
risk assessment model for use in a familial cancer clinic. BMC Med
Genet. 2008;9:116.
51. Palma MD, Domchek SM, Stopfer J, et  al. The relative contribu-
tion of point mutations and genomic rearrangements in BRCA1
and BRCA2 in high-risk breast cancer families. Cancer Res. Sep 1,
2008;68(17):7006–7014.
52. Clarke-Pearson DL. Clinical practice. Screening for ovarian cancer.
N Engl J Med. Jul 9, 2009;361(2):170–177.
53. Fisher B, Costantino JP, Wickerham DL, et al. Tamoxifen for pre-
vention of breast cancer: report of the National Surgical Adjuvant
Breast and Bowel Project P-1 Study. J Natl Cancer Inst. Sep 16,
1998;90(18):1371–1388.
54. King MC, Wieand S, Hale K, et  al. Tamoxifen and breast cancer
incidence among women with inherited mutations in BRCA1 and
BRCA2:  National Surgical Adjuvant Breast and Bowel Project
(NSABP-P1) Breast Cancer Prevention Trial. JAMA. Nov 14,
2001;286(18):2251–2256.
tamoxifen in women with high risk BRCA1 and BRCA2 mutations.
Br J Cancer. Jan 21, 2002;86(2):218–221.
56. Narod SA, Brunet JS, Ghadirian P, et al. Tamoxifen and risk of con-
tralateral breast cancer in BRCA1 and BRCA2 mutation carriers: a
case-control study. Hereditary Breast Cancer Clinical Study Group.
Lancet. Dec 2, 2000;356(9245):1876–1881.
57. Schrag D, Kuntz KM, Garber JE, Weeks JC. Decision analysis—
effects of prophylactic mastectomy and oophorectomy on life expec-
tancy among women with BRCA1 or BRCA2 mutations. N Engl J
Med. May 15, 1997;336(20):1465–1471.
58. Rebbeck TR, Friebel T, Lynch HT, et  al. Bilateral prophylactic
mastectomy reduces breast cancer risk in BRCA1 and BRCA2
mutation carriers: the PROSE Study Group. J Clin Oncol. Mar 15,
2004;22(6):1055–1062.
59. Rebbeck TR. Prophylactic oophorectomy in BRCA1 and BRCA2
mutation carriers. Eur J Cancer. Nov 2002;38 Suppl 6:S15–17.
60. Kauff ND, Satagopan JM, Robson ME, et  al. Risk-reducing
salpingo-oophorectomy in women with a BRCA1 or BRCA2 muta-
tion. N Engl J Med. May 23, 2002;346(21):1609–1615.
61. Zhou XP, Woodford-Richens K, Lehtonen R, et al. Germline muta-
tions in BMPR1A/ALK3 cause a subset of cases of juvenile pol-
yposis syndrome and of Cowden and Bannayan-Riley-Ruvalcaba
syndromes. Am J Hum Genet. Oct 2001;69(4):704–711.
62. Waite KA, Eng C. Protean PTEN: form and function. Am J Hum
Genet. Apr 2002;70(4):829–844.
63. Nelen MR, Padberg GW, Peeters EA, et al. Localization of the gene
for Cowden disease to chromosome 10q22–23. Nat Genet. May
1996;13(1):114–116.
64. Nelen MR, Kremer H, Konings IB, et al. Novel PTEN mutations
in patients with Cowden disease: absence of clear genotype–pheno-
type correlations. Eur J Hum Genet. Apr 1999;7(3):267–273.
65. Eng C. Will the real Cowden syndrome please stand up:  revised
diagnostic criteria. J Med Genet. Nov 2000;37(11):828–830.
66. Tan MH, Mester J, Peterson C, et al. A clinical scoring system for
selection of patients for PTEN mutation testing is proposed on the
basis of a prospective study of 3042 probands. Am J Hum Genet. Jan
7, 2011;88(1):42–56.
67. Ngeow J, Mester J, Rybicki LA, Ni Y, Milas M, Eng C. Incidence
and clinical characteristics of thyroid cancer in prospective series of
individuals with Cowden and Cowden-like syndrome characterized
by germline PTEN, SDH, or KLLN alterations. J Clin Endocrinol
Metab. Dec 2011;96(12):E2063–2071.
68. Tan MH, Mester JL, Ngeow J, Rybicki LA, Orloff MS, Eng C.
Lifetime cancer risks in individuals with germline PTEN mutations.
Clin Cancer Res. Jan 15, 2012;18(2):400–407.
69. Liaw D, Marsh DJ, Li J, et  al. Germline mutations of the PTEN
gene in Cowden disease, an inherited breast and thyroid cancer syn-
drome. Nat Genet. May 1997;16(1):64–67.
70. Fanning AS, Anderson JM. PDZ domains:  fundamental building
blocks in the organization of protein complexes at the plasma mem-
brane. J Clin Invest. Mar 1999;103(6):767–772.
71. Eng C. PTEN:  one gene, many syndromes. Hum Mutat. Sep
2003;22(3):183–198.
72. Zbuk KM, Eng C. Cancer phenomics: RET and PTEN as illustra-
tive models. Nat Rev Cancer. Jan 2007;7(1):35–45.
73. Ni Y, Zbuk KM, Sadler T, et al. Germline mutations and variants
in the succinate dehydrogenase genes in Cowden and Cowden-like
syndromes. Am J Hum Genet. Aug 2008;83(2):261–268.
74. Bennett KL, Mester J, Eng C. Germline epigenetic regulation of
KILLIN in Cowden and Cowden-like syndrome. JAMA. Dec 22,
2010;304(24):2724–2731.
75. Orloff MS, He X, Peterson C, et al. Germline PIK3CA and AKT1
mutations in Cowden and Cowden-like syndromes. Am J Hum
Genet. Jan 10, 2013;92(1):76–80.
76. Mester J, Eng C. Estimate of de novo mutation frequency in pro-
bands with PTEN hamartoma tumor syndrome. Genet Med. Sep
2012;14(9):819–822.
77. Hildenbrand C, Burgdorf WH, Lautenschlager S. Cowden syndrome-
diagnostic skin signs. Dermatology. 2001;202(4):362–366.
78. Kastrinos F, Syngal S. Recently identified colon cancer pre-
dispositions:  MYH and MSH6 mutations. Semin Oncol. Oct
2007;34(5):418–424.
79. Lynch HT, de la Chapelle A. Hereditary colorectal cancer. N Engl J
Med. Mar 6, 2003;348(10):919–932.
80. Fuchs CS, Giovannucci EL, Colditz GA, Hunter DJ, Speizer
FE, Willett WC. A prospective study of family history and the
risk of colorectal cancer. N Engl J Med. Dec 22, 1994;331(25):
1669–1674.
81. Markowitz SD, Bertagnolli MM. Molecular origins of can-
cer:  molecular basis of colorectal cancer. N Engl J Med. Dec 17,
2009;361(25):2449–2460.
82. Warthin AS. Heredity of carcinoma in man. Ann Intern Med.
1931;4:681–696.
5 6 2   •   G e no m ic s in C l inic a l P r actic e
83. Lynch HT, Shaw MW, Magnuson CW, Larsen AL, Krush AJ.
Hereditary factors in cancer. Study of two large Midwestern kin-
dreds. Arch Intern Med. Feb 1966;117(2):206–212.
84. Salovaara R, Loukola A, Kristo P, et al. Population-based molecu-
lar detection of hereditary nonpolyposis colorectal cancer. J Clin
Oncol. Jun 2000;18(11):2193–2200.
85. Aaltonen LA, Salovaara R, Kristo P, et  al. Incidence of heredi-
tary nonpolyposis colorectal cancer and the feasibility of
molecular screening for the disease. N Engl J Med. May 21,
1998;338(21):1481–1487.
86. Hampel H, Frankel WL, Martin E, et al. Feasibility of screening
for Lynch syndrome among patients with colorectal cancer. J Clin
Oncol. Dec 10, 2008;26(35):5783–5788.
87. Hampel H, Frankel WL, Martin E, et al. Screening for the Lynch
syndrome (hereditary nonpolyposis colorectal cancer). N Engl J
Med. May 5, 2005;352(18):1851–1860.
88. Aarnio M, Sankila R, Pukkala E, et  al. Cancer risk in mutation
carriers of DNA-mismatch-repair genes. Int J Cancer. Apr 12,
1999;81(2):214–218.
89. Vasen HF, Wijnen JT, Menko FH, et  al. Cancer risk in families
with hereditary nonpolyposis colorectal cancer diagnosed by
mutation analysis. Gastroenterology. Apr 1996;110(4):1020–1027.
90. Vasen HF, Sanders EA, Taal BG, et al. The risk of brain tumours
in hereditary non-polyposis colorectal cancer (HNPCC). Int J
Cancer. Feb 8, 1996;65(4):422–425.
91. Millar AL, Pal T, Madlensky L, et  al. Mismatch repair gene
defects contribute to the genetic basis of double primary can-
cers of the colorectum and endometrium. Hum Mol Genet. May
1999;8(5):823–829.
92. Park YJ, Shin KH, Park JG. Risk of gastric cancer in hereditary
nonpolyposis colorectal cancer in Korea. Clin Cancer Res. Aug
2000;6(8):2994–2998.
93. Wagner A, Hendriks Y, Meijers-Heijboer EJ, et  al. Atypical
HNPCC owing to MSH6 germline mutations: analysis of a large
Dutch pedigree. J Med Genet. May 2001;38(5):318–322.
94. Wijnen J, de Leeuw W, Vasen H, et al. Familial endometrial cancer
in female carriers of MSH6 germline mutations. Nat Genet. Oct
1999;23(2):142–144.
95. Nystrom-Lahti M, Kristo P, Nicolaides NC, et al. Founding muta-
tions and Alu-mediated recombination in hereditary colon cancer.
Nat Med. Nov 1995;1(11):1203–1206.
96. Hutter P, Couturier A, Scott RJ, et al. Complex genetic predisposi-
tion to cancer in an extended HNPCC family with an ancestral
hMLH1 mutation. J Med Genet. Aug 1996;33(8):636–640.
97. Chan TL, Yuen ST, Ho JW, et al. A novel germline 1.8-kb deletion of
hMLH1 mimicking alternative splicing: a founder mutation in the
Chinese population. Oncogene. May 24, 2001;20(23):2976–2981.
98. Foulkes WD, Thiffault I, Gruber SB, et al. The founder mutation
MSH2*1906G—>C is an important cause of hereditary nonpol-
yposis colorectal cancer in the Ashkenazi Jewish population. Am J
Hum Genet. Dec 2002;71(6):1395–1412.
99. Wagner A, Barrows A, Wijnen JT, et al. Molecular analysis of hered-
itary nonpolyposis colorectal cancer in the United States:  high
mutation detection rate among clinically selected families and
characterization of an American founder genomic deletion of the
MSH2 gene. Am J Hum Genet. May 2003;72(5):1088–1100.
100. Wijnen JT, Vasen HF, Khan PM, et al. Clinical findings with impli-
cations for genetic testing in families with clustering of colorectal
cancer. N Engl J Med. Aug 20, 1998;339(8):511–518.
101. Gille JJ, Hogervorst FB, Pals G, et al. Genomic deletions of MSH2
and MLH1 in colorectal cancer families detected by a novel muta-
tion detection approach. Br J Cancer. Oct 7, 2002;87(8):892–897.
102. Nakagawa H, Yan H, Lockman J, et al. Allele separation facilitates
interpretation of potential splicing alterations and genomic rear-
rangements. Cancer Res. Aug 15, 2002;62(16):4579–4582.
103. Vasen HF, Mecklin JP, Khan PM, Lynch HT. The International
Collaborative Group on Hereditary Non-Polyposis Colorectal
 Cancer (ICG-HNPCC). Dis Colon Rectum. May 1991;34(5):
424–425.
104. Vasen HF, Watson P, Mecklin JP, Lynch HT. New clinical criteria
for hereditary nonpolyposis colorectal cancer (HNPCC, Lynch
syndrome) proposed by the International Collaborative Group on
HNPCC. Gastroenterology. Jun 1999;116(6):1453–1456.
105. Bisgaard ML, Jager AC, Myrhoj T, Bernstein I, Nielsen FC.
Hereditary non-polyposis colorectal cancer (HNPCC):  pheno-
type–genotype correlation between patients with and without
identified mutation. Hum Mutat. Jul 2002;20(1):20–27.
106. Syngal S, Fox EA, Li C, et al. Interpretation of genetic test results for
hereditary nonpolyposis colorectal cancer: implications for clinical
predisposition testing. JAMA. Jul 21, 1999;282(3):247–253.
107. Umar A, Boland CR, Terdiman JP, et  al. Revised Bethesda
Guidelines for hereditary nonpolyposis colorectal cancer (Lynch
syndrome) and microsatellite instability. J Natl Cancer Inst. Feb 18,
2004;96(4):261–268.
108. Evaluation of Genomic Applications in Practice and Prevention
(EGAPP) Working Group: Recommendations from the EGAPP
Working Group:  genetic testing strategies in newly diagnosed
individuals with colorectal cancer aimed at reducing morbidity
and mortality from Lynch syndrome in relatives. Genet Med. Jan
2009;11(1):35–41.
109. Beamer LC, Grant ML, Espenschied CR, et  al. Reflex immu-
nohistochemistry and microsatellite instability testing of
colorectal tumors for Lynch syndrome among US cancer pro-
grams and follow-up of abnormal results. J Clin Oncol. Apr 1,
2012;30(10):1058–1063.
110. Heald B, Plesec T, Liu X, et al. Implementation of universal micro-
satellite instability and immunohistochemistry screening for diag-
nosing Lynch syndrome in a large academic medical center. J Clin
Oncol. Feb 11, 2013.
111. Burke W, Petersen G, Lynch P, et al. Recommendations for follow-
up care of individuals with an inherited predisposition to cancer.
I. Hereditary nonpolyposis colon cancer. Cancer Genetics Studies
Consortium. JAMA. Mar 19, 1997;277(11):915–919.
112. Smith RA, Cokkinides V, Eyre HJ. American Cancer Society
guidelines for the early detection of cancer, 2003. CA Cancer J
Clin. Jan–Feb 2003;53(1):27–43.
113. Jarvinen HJ, Mecklin JP, Sistonen P. Screening reduces colorectal
cancer rate in families with hereditary nonpolyposis colorectal can-
cer. Gastroenterology. May 1995;108(5):1405–1411.
114. Richards ML. Familial syndromes associated with thyroid cancer in
the era of personalized medicine. Thyroid. Jul 2010;20(7):707–713.
115. Xing M. Molecular pathogenesis and mechanisms of thyroid can-
cer. Nat Rev Cancer. Mar 2013;13(3):184–199.
116. Gimm O. Multiple endocrine neoplasia type 2:  clinical aspects.
Front Horm Res. 2001;28:103–130.
117. Gorlin RJ, Sedano HO, Vickers RA, Cervenka J. Multiple mucosal
neuromas, pheochromocytoma and medullary carcinoma of the
thyroid—a syndrome. Cancer. Aug 1968;22(2):293–299 passim.
118. Mathew CG, Chin KS, Easton DF, et al. A linked genetic marker
for multiple endocrine neoplasia type 2A on chromosome 10.
Nature. Aug 6–12, 1987;328(6130):527–528.
119. Simpson NE, Kidd KK. Where is the locus for multiple endocrine
neoplasia type 2A? Henry Ford Hosp Med J. 1987;35(2–3):168–171.
120. Gardner E, Papi L, Easton DF, et al. Genetic linkage studies map
the multiple endocrine neoplasia type 2 loci to a small interval on
chromosome 10q11.2. Hum Mol Genet. Mar 1993;2(3):241–246.
121. Mulligan LM, Kwok JB, Healey CS, et al. Germ-line mutations of
the RET proto-oncogene in multiple endocrine neoplasia type 2A.
Nature. Jun 3, 1993;363(6428):458–460.
122. Eng C, Mulligan LM, Smith DP, et  al. Mutation of the RET
protooncogene in sporadic medullary thyroid carcinoma. Gene
Chromosome Canc. Mar 1995;12(3):209–212.
123. Wohllk N, Cote GJ, Evans DB, Goepfert H, Ordonez NG,
Gagel RF. Application of genetic screening information to the
C l inic a l C a nc e r G e no m ic s   •   5 6 3
management of medullary thyroid carcinoma and multiple endo-
crine neoplasia type 2. Endocrinol Metab Clin North Am. Mar
1996;25(1):1–25.
124. Schuffenecker I, Ginet N, Goldgar D, et al. Prevalence and parental
origin of de novo RET mutations in multiple endocrine neoplasia type
2A and familial medullary thyroid carcinoma. Le Groupe d’Etude des
Tumeurs a Calcitonine. Am J Hum Genet. Jan 1997;60(1):233–237.
125. Wiench M, Wygoda Z, Gubala E, et al. Estimation of risk of inher-
ited medullary thyroid carcinoma in apparent sporadic patients. J
Clin Oncol. Mar 1, 2001;19(5):1374–1380.
126. Decker RA, Peacock ML, Borst MJ, Sweet JD, Thompson NW.
Progress in genetic screening of multiple endocrine neoplasia type
2A: is calcitonin testing obsolete? Surgery. Aug 1995;118(2):257–
263; discussion 263–254.
127. Brandi ML, Gagel RF, Angeli A, et al. Guidelines for diagnosis and
therapy of MEN type 1 and type 2. J Clin Endocrinol Metab. Dec
2001;86(12):5658–5671.
128. Takahashi M, Ritz J, Cooper GM. Activation of a novel human
transforming gene, RET, by DNA rearrangement. Cell. Sep
1985;42(2):581–588.
129. Takahashi M, Cooper GM. RET transforming gene encodes a
fusion protein homologous to tyrosine kinases. Mol Cell Biol. Apr
1987;7(4):1378–1385.
130. Takahashi M. Structure and expression of the RET transforming
gene. IARC Sci Publ. 1988(92):189–197.
131. Pachnis V, Mankoo B, Costantini F. Expression of the c-RET
proto-oncogene during mouse embryogenesis. Development. Dec
1993;119(4):1005–1017.
132. Nakamura T, Ishizaka Y, Nagao M, Hara M, Ishikawa T.
Expression of the RET proto-oncogene product in human nor-
mal and neoplastic tissues of neural crest origin. J Pathol. Mar
1994;172(3):255–260.
133. Tsuzuki T, Takahashi M, Asai N, Iwashita T, Matsuyama M, Asai
J. Spatial and temporal expression of the RET proto-oncogene
product in embryonic, infant and adult rat tissues. Oncogene. Jan 5,
1995;10(1):191–198.
134. Pasini B, Hofstra RM, Yin L, et al. The physical map of the human
RET proto-oncogene. Oncogene. Nov 2, 1995;11(9):1737–1743.
135. Eng C. RET proto-oncogene in the development of human cancer.
J Clin Oncol. Jan 1999;17(1):380–393.
136. Durbec P, Marcos-Gutierrez CV, Kilkenny C, et al. GDNF signal-
ling through the RET receptor tyrosine kinase. Nature. Jun 27,
1996;381(6585):789–793.
137. Jing S, Wen D, Yu Y, et al. GDNF-induced activation of the RET
protein tyrosine kinase is mediated by GDNFR-alpha, a novel
receptor for GDNF. Cell. Jun 28, 1996;85(7):1113–1124.
138. Kotzbauer PT, Lampe PA, Heuckeroth RO, et  al. Neurturin, a
relative of glial-cell-line-derived neurotrophic factor. Nature. Dec
5, 1996;384(6608):467–470.
139. van Weering DH, Bos JL. Signal transduction by the receptor tyro-
sine kinase RET. Recent Results Cancer Res. 1998;154:271–281.
140. Besset V, Scott RP, Ibanez CF. Signaling complexes and protein–
protein interactions involved in the activation of the Ras and phos-
phatidylinositol 3-kinase pathways by the c-RET receptor tyrosine
kinase. J Biol Chem. Dec 15, 2000;275(50):39159–39166.
141. Borrello MG, Pelicci G, Arighi E, et al. The oncogenic versions of
the RET and TRK tyrosine kinases bind Shc and Grb2 adaptor
proteins. Oncogene. Jun 1994;9(6):1661–1668.
142. Santoro M, Carlomagno F, Romano A, et al. Activation of RET as
a dominant transforming gene by germline mutations of MEN2A
and MEN2B. Science. Jan 20, 1995;267(5196):381–383.
143. Kloos RT, Eng C, Evans DB, et al. Medullary thyroid cancer: man-
agement guidelines of the American Thyroid Association. Thyroid.
Jun 2009;19(6):565–612.
144. Farndon JR, Leight GS, Dilley WG, et al. Familial medullary thy-
roid carcinoma without associated endocrinopathies:  a distinct
clinical entity. Br J Surg. Apr 1986;73(4):278–281.
 145. Eng C, Clayton D, Schuffenecker I, et al. The relationship between
specific RET proto-oncogene mutations and disease phenotype in
multiple endocrine neoplasia type 2. International RET mutation
consortium analysis. JAMA. Nov 20, 1996;276(19):1575–1579.
146. Gimm O, Marsh DJ, Andrew SD, et  al. Germline dinucleotide
mutation in codon 883 of the RET proto-oncogene in multiple
endocrine neoplasia type 2B without codon 918 mutation. J Clin
Endocrinol Metab. Nov 1997;82(11):3902–3904.
147. Smith DP, Houghton C, Ponder BA. Germline mutation of RET
codon 883 in two cases of de novo MEN 2B. Oncogene. Sep 4,
1997;15(10):1213–1217.
148. Mulligan LM, Eng C, Healey CS, et al. Specific mutations of the
RET proto-oncogene are related to disease phenotype in MEN 2A
and FMTC. Nat Genet. Jan 1994;6(1):70–74.
149. Mulligan LM, Eng C, Healey CS, et al. A de novo mutation of the
RET proto-oncogene in a patient with MEN 2A. Hum Mol Genet.
Jun 1994;3(6):1007–1008.
150. Schuffenecker I, Virally-Monod M, Brohet R, et al. Risk and pen-
etrance of primary hyperparathyroidism in multiple endocrine
neoplasia type 2A families with mutations at codon 634 of the
RET proto-oncogene. Groupe D’etude des Tumeurs a Calcitonine.
J Clin Endocrinol Metab. Feb 1998;83(2):487–491.
151. Mulligan LM, Marsh DJ, Robinson BG, et al. Genotype–pheno-
type correlation in multiple endocrine neoplasia type 2: report of
the International RET Mutation Consortium. J Intern Med. Oct
1995;238(4):343–346.
152. Eng C, Mulligan LM, Smith DP, et al. Low frequency of germline
mutations in the RET proto-oncogene in patients with apparently
sporadic medullary thyroid carcinoma. Clin Endocrinol (Oxf ). Jul
1995;43(1):123–127.
153. Eng C. Multiple endocrine neoplasia type 2 and the practice
of molecular medicine. Rev Endocr Metab Disord. Nov 2000;
1(4):283–290.
154. Frank-Raue K, Rybicki LA, Erlic Z, et al. Risk profiles and pene-
trance estimations in multiple endocrine neoplasia type 2A caused
by germline RET mutations located in exon 10. Hum Mutat. Jan
2011;32(1):51–58.
155. Machens A, Dralle H. Genotype–phenotype based surgical con-
cept of hereditary medullary thyroid carcinoma. World J Surg. May
2007;31(5):957–968.
156. Yip L, Cote GJ, Shapiro SE, et  al. Multiple endocrine neoplasia
type 2: evaluation of the genotype–phenotype relationship. Arch
Surg. Apr 2003;138(4):409–416; discussion 416.
157. Machens A, Brauckhoff M, Holzhausen HJ, Thanh PN, Lehnert
H, Dralle H. Codon-specific development of pheochromocytoma
in multiple endocrine neoplasia type 2. J Clin Endocrinol Metab.
Jul 2005;90(7):3999–4003.
158. Ogilvie JB, Kebebew E. Indication and timing of thyroid surgery
for patients with hereditary medullary thyroid cancer syndromes.
J Natl Compr Canc Netw. Feb 2006;4(2):139–147.
159. Wells SA, Jr., Gosnell JE, Gagel RF, et al. Vandetanib for the treat-
ment of patients with locally advanced or metastatic hereditary med-
ullary thyroid cancer. J Clin Oncol. Feb 10, 2010;28(5):767–772.
160. Hirota S, Isozaki K, Moriyama Y, et al. Gain-of-function mutations
of c-kit in human gastrointestinal stromal tumors. Science. Jan 23,
1998;279(5350):577–580.
161. Demetri GD, von Mehren M, Blanke CD, et al. Efficacy and safety
of imatinib mesylate in advanced gastrointestinal stromal tumors.
N Engl J Med. Aug 15, 2002;347(7):472–480.
5 6 4   •   G e no m ic s in C l inic a l P r actic e
162. Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung c­ ancer:
correlation with clinical response to gefitinib therapy. Science.
Jun 4, 2004;304(5676):1497–1500.
163. Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatin-
paclitaxel in pulmonary adenocarcinoma. N Engl J Med. Sep 3,
2009;361(10):947–957.
164. Samuels Y, Wang Z, Bardelli A, et  al. High frequency of muta-
tions of the PIK3CA gene in human cancers. Science. Apr 23,
2004;304(5670):554.
165. Davies H, Bignell GR, Cox C, et al. Mutations of the BRAF gene
in human cancer. Nature. Jun 27, 2002;417(6892):949–954.
166. Stephens P, Hunter C, Bignell G, et  al. Lung cancer:  intra-
genic ERBB2 kinase mutations in tumours. Nature. Sep 30,
2004;431(7008):525–526.
167. Bollag G, Hirth P, Tsai J, et al. Clinical efficacy of a RAF inhibitor
needs broad target blockade in BRAF-mutant melanoma. Nature.
Sep 30, 2010;467(7315):596–599.
168. Tsai J, Lee JT, Wang W, et al. Discovery of a selective inhibitor of
oncogenic B-Raf kinase with potent antimelanoma activity. Proc
Natl Acad Sci U S A. Feb 26, 2008;105(8):3041–3046.
169. Chapman PB, Hauschild A, Robert C, et  al. Improved survival
with vemurafenib in melanoma with BRAF V600E mutation.
N Engl J Med. Jun 30, 2011;364(26):2507–2516.
170. van ‘t Veer LJ, Dai H, van de Vijver MJ, et al. Gene expression pro-
filing predicts clinical outcome of breast cancer. Nature. Jan 31,
2002;415(6871):530–536.
171. Paik S, Shak S, Tang G, et al. A multigene assay to predict recur-
rence of tamoxifen-treated, node-negative breast cancer. N Engl J
Med. Dec 30, 2004;351(27):2817–2826.
172. Wang Y, Klijn JG, Zhang Y, et al. Gene-expression profiles to pre-
dict distant metastasis of lymph-node-negative primary breast can-
cer. Lancet. Feb 19–25, 2005;365(9460):671–679.
173. Lo SS, Mumby PB, Norton J, et al. Prospective multicenter study of
the impact of the 21-gene recurrence score assay on medical oncol-
ogist and patient adjuvant breast cancer treatment selection. J Clin
Oncol. Apr 1, 2010;28(10):1671–1676.
174. Burton H, Chowdhury S, Dent T, Hall A, Pashayan N, Pharoah
P. Public health implications from COGS and potential for risk
stratification and screening. Nat Genet. Mar 27, 2013;45(4):
349–351.
175. Sakoda LC, Jorgenson E, Witte JS. Turning of COGS moves for-
ward findings for hormonally mediated cancers. Nat Genet. Mar
27, 2013;45(4):345–348.
176. Michailidou K, Hall P, Gonzalez-Neira A, et al. Large-scale geno-
typing identifies 41 new loci associated with breast cancer risk. Nat
Genet. Mar 27, 2013;45(4):353–361.
177. Bojesen SE, Pooley KA, Johnatty SE, et  al. Multiple indepen-
dent variants at the TERT locus are associated with telomere
length and risks of breast and ovarian cancer. Nat Genet. Mar 27,
2013;45(4):371–384.
178. Eeles RA, Olama AA, Benlloch S, et al. Identification of 23 new
prostate cancer susceptibility loci using the iCOGS custom geno-
typing array. Nat Genet. Mar 27, 2013;45(4):385–391.
179. Feardon ER, Bommer GT. Molecular biology of colorectal cancer.
In: DeVita VT, Lawrence TS, Rosenberg SA, eds. Devita, Hellman
and Rosenberg’s Cancer:  Principles and Practice of Oncology.
Philadelphia, PA: Lippincott Williams & Wilkins; 2008.
 37.
GENOMICS AND INFECTIOUS
DISEASES: SUSCEPTIBILIT Y, RESISTANCE, RESPONSE,
AND ANTIMICROBIAL THERAPY
Michaela Fakiola, Wei Lu, Sarra E. Jamieson, and Christopher S. Peacock
INTRODUCTION
Mankind has been afflicted by infectious disease through-
out history, and the global burden of infectious disease is
still extremely high. World Health Organization (WHO)
figures indicate that infectious diseases are the second lead-
ing cause of death after cardiovascular diseases1 (see Figure
37.1). Of the 56.8  million deaths worldwide in 2008, an
estimated 12.3  million (around 22%) were due to infec-
tious or parasitic disease, with a large proportion (43%)
of this burden falling on children (0–14 yrs. of age). Of
those 12.3  million deaths, large proportions were due to
common infectious diseases such as lower respiratory tract
infections (3.46 million deaths in 2008) and diarrheal dis-
eases (2.46  million deaths). As well as mortality figures,
the overall burden placed on a population by a particu-
lar disease can also be measured in disability-adjusted life
years (DALYs).2 This measure allows not only the number
of years of healthy life lost due to early mortality, but also
the number of years of healthy life lost due to disability or
poor health to be combined and used to quantify overall
disease burden. Using this measure, the global burden of
infectious (including respiratory infections) and parasitic
diseases was 806 million DALYs in 2004, compared to car-
diovascular diseases with a burden of around 288 million
DALYs in 2004.1
Whilst there have been many advances in the last few
decades in the development of antimicrobials, vaccines, and
novel therapeutics that have undoubtedly represented huge
advances in mankind’s fight against infection, it is clear that
infectious disease still represents a significant concern in the
modern world. Current research is now taking full advantage
of the significant technological and bioinformatic advances
that have occurred over the last five to ten years to better
understand human susceptibility to infection, the immune
565
response, and the rapidly evolving evasion mechanisms of
the pathogen, as well as the complete human microbiota,
in order to develop and improve diagnostic protocols, treat-
ment regimens, and prevention strategies. This chapter will
highlight some of this ongoing research.
H U M A N G E N ET I C S U S C E P T I B I L I T Y TO
INFECTIOUS DISEASE
Human susceptibility to infectious disease is multifactorial,
being the result of interactions between environmental fac-
tors, host-related factors, and the infecting pathogen itself.
Whilst infection with the pathogen is required for disease,
it is not the sole requirement, as not all exposed individu-
als will go on to develop clinical disease. Accordingly,
there is now a large body of evidence demonstrating that
host genetics plays one important role in the susceptibil-
ity, or resistance, of individuals to infectious disease. This
includes epidemiological evidence from ethnic differences
in infectious disease susceptibility,3,4 coupled with familial
studies of rare families who exhibit Mendelian susceptibil-
ity to weakly pathogenic infectious agents (reviewed by5,6)
familial aggregation studies, complex segregation analyses,
as well as twin and adoptee studies—all of which confirm
the role played by the host’s genetics in infectious disease
susceptibility.7
Whilst epidemiological and familial studies provide con-
vincing evidence that genetic variation in human populations
contributes to infectious disease susceptibility, they provide
no information about the genes involved. The investigation
of human genetic loci of susceptibility, disease progression,
and response to treatment has, for many years, utilized strat-
egies that broadly fall into the categories of genome-wide
linkage scans (GWLSs) or hypothesis-driven candidate-gene

22%
30%
Cardiovascular diseases Infectious diseases
Respiratory diseases Injuries (all)
Digestive diseases Neuropsychiatric conditions
Figure 37.1  Leading causes of worldwide deaths (WHO estimates, 2008). Total deaths—56.8 million.
studies; both approaches have proven successful to a degree.
For example, GWLSs have been used to identify regions of
the genome carrying disease susceptibility loci (DSL) for
schistosomiasis,8 tuberculosis,9,10 leprosy,10,11 and other infec-
tious diseases.7 However, as most linkage peaks cover a rela-
tively broad region (10–20 megabases [Mb]) of the genome
encompassing numerous genes, only one such DSL has sub-
sequently been fine-mapped sufficiently to identify the puta-
tive causative gene, that being the PARK/PACRG gene region
in leprosy.12 The candidate-gene approach has potentially
been more successful, with genes identified as putative can-
didates on the basis of homology with previously identified
murine-susceptibility genes or on the basis of plausible bio-
logical function.7 However, the genetic risk factors involved in
that susceptibility to infection are expected to have relatively
small effect sizes (i.e., an odds ratio <1.5), requiring large
sample sizes to detect them with credible statistical support.
The majority of reported candidate-gene studies in infectious
disease have generally been underpowered (usually based on
fewer than 500 cases), mainly due to limitations of budgets
and problems of case recruitment, and remain unreplicated in
independent studies.
G E N O M E -W I D E A S S O C I AT I O N S T U D I E S
Recently there has been a shift from these approaches to
genome-wide association studies (GWASs). The GWAS has
5 6 6   •   G enomics in C linical P ractice
14%
7%
9%
6%
4%
2%
6%
Malignant/other neoplasms
Maternal/Congenital/Perinatal
Other
a distinct advantage in that it enables the hypothesis-free
investigation of the genetic component influencing dis-
ease pathogenesis. GWAS is a powerful tool that com-
pares the frequency of several hundred thousand to more
than a million single-nucleotide polymorphisms (SNPs)
across the whole genome in hundreds or thousands of unre-
lated cases versus ethnically matched controls. Although
the population-based design is most commonly used,
family-based designs, such as parent/offspring trios, have
also been employed. In both instances, novel and previously
unsuspected genetic risk factors can be identified, provid-
ing a broader view of the gene network and pathways impli-
cated in host responses.
GWASs have become feasible due to rapid advances in
high-throughput array technologies that enable the geno-
typing of a dense map of markers across the genome. The
successful design of cost-effective genotyping platforms by
Affymetrix and Illumina has been facilitated by the comple-
tion of the Human Genome Project, the International SNP
Map,13 HapMap,14 and Human Genome Diversity Project,15
all of which provide a rich resource of known SNPs within
the human genome, the frequency at which they are
observed, and the patterns of linkage disequilibrium (LD;
the non-random association of alleles) between them.
Furthermore, the 1000 Genome Project16,17 (1KGP;http://
www.1000genomes.org/), which aims to provide next-gen-
eration sequencing data on 2,500 people from 25 popula-
tions, has further enriched the public catalogue of known
human variants, in particular for rare variants. This infor-
mation is of particular importance in designing SNP arrays
that not only have adequate genome coverage, but that also
utilize knowledge of population-specific LD patterns, thus
allowing the use of imputation. This is a method to predict
the genotypes of SNPs that are not directly genotyped on
an array,18 thus further increasing the density of SNPs that
can be analyzed.
Within the past 10 years, the catalogue of GWAS pub-
lications, listed in a number of databases,19–21 including
the National Human Genome Research Institute (http://
www.genome.gov/gwastudies/), has increased consider-
ably. Considered a landmark GWAS in 2007, the Wellcome
Trust Case-Control Consortium (WTCCC;http://www.
wtccc.org.uk/),22 which utilized 2,000 well-characterized
cases for each of seven major common diseases and a shared
group of 3,000 controls, identified several loci with reliable
significant associations to one or more of the common dis-
eases studied. Whilst this study clearly illustrated the power
of the GWAS approach, it also highlighted several statistical
considerations, including the requirement for multiple test-
ing correction, in order to obtain genome-wide significance
levels (P <5 × 10–7) and for replication of the observed asso-
ciations in a second independent cohort.22
Following on from the successful application of GWAS
in common non-communicable diseases, this approach has
subsequently been employed to further our understanding
of human genetic risk factors in infectious disease. However,
only a fraction of the published GWASs have been under-
taken on diseases with a known underlying infectious agent,
such as HIV infection, viral hepatitis, leprosy, tuberculosis,
and malaria. These have been conducted on a wide spec-
trum of clinical phenotypes following infection in order
to decipher the genetic determinants of pathogenesis that
could be used as prediction tools in diagnostics or potential
targets for new therapeutic and vaccine strategies. Some of
the key studies are outlined here.
H U M A N I M MU N O D E F I C I E N C Y
VI RU S ( H I V )
In terms of understanding the genetic component of
communicable-disease susceptibility, the GWAS has most
frequently been applied to understanding genetic risk fac-
tors in human immunodeficiency viral (HIV) infection.
This is a reflection of HIV’s global importance, coupled
with the growing demand to better understand its hetero-
geneity and develop more effective antiretroviral drugs. In
previous candidate-gene studies, HIV-1 infection, disease
G enomics and I nfectious D iseases   •   5 6 7
progression, and response to treatment have been strongly
associated with alleles of the human leukocyte antigen
B (HLA-B) and a 32-base pair deletion in chemokine
receptor 5 (CCR5-∆32), with less consistent associations
reported for additional susceptibility genes.23 In agreement,
GWASs have been consistent in the identification of the
major histocompatibility complex (MHC) as the strongest
genetic determinant for various viral phenotypes, of which
viral load at set point and progression to AIDS (acquired
immune deficiency syndrome) have been the most com-
monly investigated.
Fellay and colleagues24 performed the first GWAS of
HIV-1 infection based on a European multi-center consor-
tium, the Euro-CHAVI (Center for HIV/AIDS Vaccine
Immunology) cohort. The strongest associations with
plasma HIV-RNA levels were observed at HLA complex
P5 (HCP5; rs2395029) and HLA-C (rs9264942). A third
genomic loci for disease progression was reported near
zinc ribbon domain–containing 1(ZNRD1), an RNA
polymerase I  subunit. These three MHC variants, which
explained 14.1% of the variation observed in circulating
viral levels during the asymptomatic phase of infection,
have subsequently been replicated in multiple indepen-
dent studies.25–27 A  quantitative trait locus (QTL) within
MHC class I that regulates the CD4:CD8 T-cell ratio in
the general population was also associated with natural con-
trol of HIV-1 viremia.28 In a much larger GWAS of 2,554
infected individuals, Fellay and colleagues29 confirmed the
central role of HLA-B and HLA-C and showed additional
independent signals at MHC. However, apart from the
chemokine receptor cluster, they failed to support other
non-MHC associations previously reported by a plethora
of candidate gene studies.
HCP5 is a human endogenous retroviral element with
sequence homology to HIV-1 pol,30 and it has been sug-
gested that it could potentially modulate HIV replication
via an antisense RNA-interfering mechanism.24 The asso-
ciated HCP5 rs2395029 variant is also known to act as a
proxy in Caucasians for the classical HLA-B*5701 allele,
which confers protection against HIV disease progression
and high viral load.31,32 The role of the HLA-B*57 group
was shown in GWASs of African American populations,33,34
with the classical HLA-B*5703 allele explaining 10% of the
variation in viral load.33 A  more detailed exploration has
revealed that specific amino acid positions (67, 70, and
97) in the peptide binding groove of the HLA-B proteins
are associated with disease outcome.34 These amino acids
are involved in the conformation of viral peptide presen-
tation following infection and thus could explain, at least
partly, the major genetic effect of HLA-B on HIV-control.
In support of this, HLA-B alleles produced the strongest
associations with T-cell responses to the MRKAd5 HIV
gag/pol/nef vaccine, highlighting their potential impor-
tance for vaccine design.35
A GWAS of large structure variations has also
reported epistatic interactions between the killer cell
immunoglobulin-like receptors (KIR) and HLA class  I,
where the copy number of the inhibitory KIR3DL1 and
activating KIR3DS1 receptors is associated with viral load
at set point in the presence of appropriate HLA-Bw4 mol-
ecules.36 The genetic associations were supported by func-
tional data showing that the expansion of natural killer
(NK) cells and the subsequent effective inhibition of HIV-1
replication, are determined by the number of KIR3DL1
and KIR3DS1 copies.
In addition to HLA-B, the ability of HLA-C to pres-
ent viral epitopes to cytotoxic CD8+ T cells, in conjunction
with the fact that it is unaffected by Nef-mediated HLA
down-regulation,37–39 suggests a role for HLA-C-driven
viral control, providing an alternative target in developing
therapeutic or vaccine strategies. The SNP rs9264942 has
been shown to be associated not only with plasma HIV-
RNA levels, but also with levels of HLA-C mRNA tran-
scripts and cell surface expression. However, it has been
proposed that rs9264942 is not itself the causal variant but
is instead tagging the functional true variant, suggested to
be rs67384697 in the 3′ untranslated region of HLA-C that
regulates binding of the microRNA hsa-miR-148 to its tar-
get site enabling regulation of HLA-C levels.40
Three HIV GWASs25,27,41 have also been conducted
on French populations recruited under the umbrella of
the National Agency for AIDS Research (ANRS). The
first of these studies highlighted the importance of the
MHC class III region, in addition to class I, in both plasma
HIV-RNA levels (a marker of HIV-replication) and cellu-
lar HIV-DNA levels.25 The latter viral phenotype, which
reflects the HIV reservoir,42 was further associated to novel
genomic loci on chromosomes 17 and 8.  The protective
alleles of the identified class I and III polymorphisms were
found to be enriched in an independent cohort of long-term
HIV controllers.43 The two subsequent GWA studies used
the extreme non-progressor phenotype in the French GRIV
(Genomics of Resistance to Immunodeficiency Virus)
cohort27 and a unique cohort of rapid disease progressors.41
Results confirmed the previous associations at HCP5 and
ZNRD1 with disease non-progression and suggested novel
genetic loci, implicating genes with a known function in the
development of HIV infection.
Novel associations at additional genes outside of the
MHC have also been reported for progression, long-term
5 6 8   •   G enomics in C linical P ractice
non-progression to AIDS, or alternative HIV-1–related phe-
notypes, but discussion of them all is outside the scope of this
chapter. One study of interest has highlighted the potential
for sex-linked genetic variation to influence the rate of dis-
ease.44 Subsequent confirmation in HIV-1–infected patients
showed that an associated SNP at Xq21.1 explained 21–25%
of the variance in female log-progression phenotype.44
Another has investigated response to nevirapine treatment,
a generic combined tablet that often induces a rash, with
associated variants identified in the Coiled-coil alpha-helical
rod protein 1 (CCHCR1) gene.45 Whilst of interest, some of
these additional GWAS associations do not withstand strict
multiple testing correction for genome-wide significance.
Nevertheless, their validation in future studies is of impor-
tance as they have the potential to provide new knowledge
on disease pathogenesis and response to treatment.
H E PAT I T I S B
Chronic hepatitis B is one of the most common infectious
liver diseases, with more than 400 million people worldwide
being chronically infected with hepatitis B virus (HBV).46
Chronic infection is related to a broad spectrum of clini-
cal sequelae, ranging from asymptomatic infection to liver
cirrhosis and hepatocellular carcinoma (HCC),47 and
accounting for nearly 60% of liver cancers.46 The prevalence
of infection shows regional diversity,48 and epidemiological
data have indicated the importance of host genetics in the
outcome of HBV infection.49
All GWASs of HBV infection to date have been per-
formed on East Asian populations, which contribute the
largest number of chronic carriers worldwide.48 The first
GWAS in Japanese and Thai cohorts reported associa-
tions with SNPs at the previously identified MHC class II
HLA-DPA1 (odds ratio [OR] = 0.56; 95% confidence
interval [CI] = 0.51–0.61) and HLA-DPB1 (OR = 0.57;
95% CI = 0.52–0.62) region.50 Direct sequencing of
exon 2 encoding the DPs antigen-binding site further
identified the classical allele haplotypes associated with
protection (DPA1*0103–DPB1*0402 and DPA1*0103–
DPB1*0401) and susceptibility (DPA1*0202–DPB1*0501
and DPA1*0202–DPB1*0301) to chronic hepatitis B.
Utilizing a larger cohort, the authors also revealed an inde-
pendent effect of the HLA-DQ locus, and defined the
protective (DQA1*0102–DQB1*0604 and DQA1*0101–
DQB1*0501) and risk (DQA1*0102–DQB1*0303 and
DQA1*0301-DQB1*0601) haplotypes.51
Post-vaccination antibody titers in recipients of two-
dose hepatitis B vaccination have also been investigated in a
two-stage GWAS by Png and colleagues.52 Three indepen-
dent signals with a cumulative effect were identified within
the HLA region (HLA-DRB1, HLA-DPB1, and a gene-rich
HLA class III interval) in agreement with the findings of
the first GWAS on chronic HBV infection.50 In light of this
finding, the question has been raised as to whether the same
genetic risk factors that confer HBV susceptibility also con-
tribute to low post-vaccine antibody titer, suggesting that
vaccination is likely to be ineffective in the individuals most
at risk. A more recent Japanese/Korean GWAS combined
with subsequent meta-analysis of six independent cohorts
has confirmed that the role of HLA-DP variants in both
protection against chronic infection and viral clearance is
widely consistent across East Asian populations.53
H E PAT I T I S C
Reduced efficacy, poor tolerance, and population differ-
ences in the treatment of genotype 1 chronic hepatitis C,
a leading cause of cirrhosis, hepatocellular carcinoma, and
liver transplantation in the United States,54–56 have led to a
number GWAS studies focusing on the genetic determi-
nants of response to treatment with pegylated interferon-a
(PEG-IFN-α) and ribavirin (RBV). The first GWAS of
hepatitis C virus (HCV) infection in the individualized
dosing efficacy versus flat dosing to assess optimal pegylated
interferon therapy (IDEAL) study57 cohort identified the
rs12979860 polymorphism upstream of the interleukin 28B
(IL28B) as the major genetic factor that influences both
the rate of sustained virological response (SVR) as well as
the baseline viral load.58 The frequency of the rs12979860
C allele associated with better response, which was para-
doxically highly correlated with SVR rates across diverse
ethnic groups, could explain approximately half of the
higher response rate in populations of European ancestry
compared to African-Americans. The study also supports
previous reports that individual genotype is a better predic-
tor of response to treatment than commonly used ethnic
labels.59 The lower frequency of the C allele in the chronically
infected patients compared to ethnically matched controls
has sparked further investigations, which strongly associated
the homozygote rs12979860 C/C genotype with natural
clearance of the virus in both European and African individu-
als.60 The effect of IL28B polymorphisms has been indepen-
dently replicated in GWAS studies of Japanese,61 Australian,62
and European63 populations, with supportive functional
data for the lower expression of IL28 in carriers of the risk
allele61 or of the non-responder haplotype.62 Interestingly, the
effect sizes reported across studies varied from OR  =  1.98
G enomics and I nfectious D iseases   •   5 6 9
(95% CI  =  1.57–2.52) in the Australian population,62 to
OR = 27.2 (95% CI = 13.9–53.4) in the Japanese cohort,61
which could be a consequence of the different allele frequen-
cies of the associated variants in those populations. Moreover,
in both HCV genotype-1 and genotype-3–infected patients
carrying the unfavorable IL28B genotypes, the difference in
SVR rates between those with high versus low ferritin levels,
where serum ferritin levels are reported to predict treatment
failure (P <0.0001),64 was more than 30%. Finally, in patients
with heterozygous IL28B genotype, low-density lipoprotein
cholesterol (LDL-C) levels in combination with reduced
HCV RNA burden could predict response to PEG-IFN/
RBV treatment.65
More recently, a number of GWASs have reported addi-
tional genetic risk factors associated with certain side effects
in response to PEG-IFNα/RBV treatment. Variants at ITPA/
DDRGK1 gene region were associated with IFN-induced
thrombocytopenia66,67 but also with protection against
RBV-induced hemolytic anemia.66 Fibrosis progression in
HCV-infected patients was associated with variants in the
ring finger protein 7 (RNF7) gene, an antioxidant confer-
ring protection against apoptosis, and the functionally
related C-mer proto-oncogene tyrosine kinase (MERTK)
and tubby like protein 1 (TULP1) genes, which are involved
in macrophage-mediated phagocytosis of apoptotic cells.68
These GWAS findings present an opportunity for the
development of clinically relevant diagnostic tests based on
host genotype in order to identify the 50–60% of patients
likely to be refractory to standard treatment regimens.61,62
Given the high cost and adverse reactions of PEG-IFN-α/
RBV treatment, such a tool could considerably reduce the
annual cost of unsuccessful clinical management whilst
avoiding unnecessary patient distress. The GWAS associa-
tions also highlight the role of the Type III interferon (IFN)
signaling pathway as a potential target for designing new
therapeutic approaches for HCV infection. Type III IFNs
(IL28B, IL28A, and IL29) are upregulated by viral infec-
tion and IFNα.69,70 They subsequently promote toll-like
receptor-mediated antiviral protection71–73 by inducing
the expression of IFN-stimulated genes but in a slower and
more sustained manner than IFNα.71,74 Hence, it has been
suggested that they can serve as an alternative therapy to
Type I IFNs.72 Indeed, the use of IL29 in HCV treatment is
currently being investigated in Phase 2 clinical trials.62
M YC O B AC T E R I A L I N F E C T I O N S
Tuberculosis (TB) and leprosy present with diverse clinical
phenotypes due to infection with the evolutionary related
intracellular mycobacteria, Mycobacterium tuberculosis
and M. leprae. Epidemiological data have long indicated a
genetic component to susceptibility to both diseases, but
the evidence for reported genetic variants is currently more
robust for leprosy than for tuberculosis.75,76 Consistent with
this, GWASs undertaken to delineate genetic determinants
of TB susceptibility have been more challenging than those
for leprosy.
In a GWAS of TB undertaken within West African
populations, comprising a cohort from Ghana and Gambia,
Thye and coworkers77 reported a novel susceptibility locus
in a gene desert on chromosome 18q11.2, with genome-
wide significance achieved following combined analysis of
the discovery and replication African cohorts (combined
P = 6.8×10−9; OR = 1.19, 95% CI 1.12–1.26). A nearby
gene encoding the GATA-binding factor 6 (GATA6) has
been proposed as the most plausible biological candidate.
Subsequent imputation analysis of the Ghanaian genome-
wide dataset using the 1KGP populations as a reference
revealed a second locus on chromosome 11p13 downstream
of the Wilms tumor 1 (WT1) gene,78 which was replicated
in three independent cohorts from Gambia, Russia, and
Indonesia (combined P = 2.57 × 10−11).
Efforts to identify genome-wide signals for TB in Asian
populations have been less successful. A two-stage GWAS
of pulmonary TB in Indonesians with further replication
in a Russian cohort has reported associations with suscep-
tibility loci close to immune-related genes, although none
achieved genome-wide significance.79 A  first-pass GWAS
in Japanese and Thai populations also failed to report con-
vincing association signals.80 In subsequent meta-analysis
focusing on age at onset of TB, the HSEP1-MAFB locus on
chromosome 20q12 was identified as a novel genetic factor
for TB susceptibility in the young, with a moderate effect
size (OR = 1.73, 95% CI 1.42–2.11).
On the other hand, the first GWAS of leprosy carried out
in the Han Chinese population revealed several novel sus-
ceptibility loci (CCDG122, C13orf31, NOD2, TNFSF15,
RIPK2, and LRRK2), in addition to confirming associa-
tions at the previously known HLA-DRB1 gene.81 Five of
the reported loci feature in a single network of genes that
regulates innate immunity, the NOD2 (nucleotide-binding
oligomerization domain containing 2)-mediated signaling
pathway. Intriguingly, this pathway also includes PARK2,
a previously known major risk factor for leprosy identified
via GWLS.12 The study also showed evidence of heteroge-
neity, with most of the observed associations being stronger
with the multibacillary form of leprosy than with the pauci-
bacillary form. Using imputation methods in combination
with an expanded discovery controls set and an extensive
5 7 0   •   G enomics in C linical P ractice
cohort of replication samples, Zhang and colleagues82 have
reported two additional loci, on 1p31.3 and 6q24.3, with
interleukin 23 receptor (IL23R) and Ras-related protein
Rab-32 (RAB32) genes being the most plausible biologi-
cal candidates. Furthermore, pairwise interaction analysis
across all nine loci has revealed an interaction between the
NOD2 and RIPK2 (receptor-interacting serine-threonine
kinase 2) genes.
Using a gene-centric 50K microarray, Wong and col-
leagues83 have identified, in addition to HLA-DRB1/
DQA1, a non-synonymous change at Toll-like receptor 1
(TLR1) as a major genetic risk factor for susceptibility to
M. leprae in an Indian population. This GWAS also repli-
cated the associations at CCDG122 and C13orf31 on chro-
mosome 13, but not those at NOD2, TNFSF15, or RIPK2
variants.84 Conversely, the population-specific effect seen at
TLR1 was not replicated in the Zhang82 study, suggesting
population heterogeneity in leprosy susceptibility, or differ-
ences in allele frequencies and the effectiveness of commer-
cially available arrays across different populations. Overall,
the GWASs reported to date for TB and leprosy susceptibil-
ity have not confirmed susceptibility loci previously iden-
tified by GWLS or candidate-gene studies. Furthermore,
an overlap of genetic susceptibility loci between these two
diseases (excluding HLA) was not evident, demonstrating
the different mechanisms of pathogenesis underlying these
mycobacterial infections.
MALARIA
Malaria, caused by the Plasmodium falciparum parasite,
was responsible for more than 800,000 deaths in 2008,1
with the majority of these deaths occurring in sub-Saha-
ran Africa.85 In endemic populations, the genetic compo-
nent of disease risk approaches 25%; however, only 2%
is attributed to the classical hemoglobin S (HbS; caus-
ing sickle-cell anemia) variant of the hemoglobin beta
(HBB) gene, the strongest genetic risk factor identified
to date.86 A GWAS of malaria susceptibility in Gambia
by the Malaria Genomic Epidemiology Network87 and
the WTCCC was the first to be conducted on an African
population.88 The discovery cohort used by Jallow and
colleagues88 comprised 958 cases and 1,382 controls geno-
typed on the Affymetrix GeneChip 500K Mapping Array
Set, while replication was sought in a cohort of over 3,400
children using the Sequenom iPlex platform. The stron-
gest association signal was seen for SNPs on chromosome
11p15, in close proximity to HBB. However, the study did
not provide support for other previously reported malaria
 Free ebooks ==> www.Ebook777.com
susceptibility/resistance loci, such as the glucose-6-phos-
phate dehydrogenase (G6PD), the ABO blood group, and
HBA1-HBA2 genes. This may be partly explained by the
limited coverage of the Affymetrix 500K SNP array of the
variation seen in African populations, as well as the low
minor allele frequency of known functional variants in the
Gambian population.
In a second GWAS of severe falciparum malaria, 2,153
individuals from Ghana were genotyped on the more
extensive Affymetrix Genome-Wide Human SNP Array
6.0.89 Genome-wide imputation of genotypes at SNPs not
represented on the array based on a reference panel of 174
African samples from the 1KGP provided additional gen-
otypes for more than five million SNPs. Upon combined
analysis with more than 3,500 replication samples, four loci
achieved genome-wide significance (P < 5 × 10–7), includ-
ing the known causal variants at HBB and ABO. Two fur-
ther novel association signals were observed for variants
within the ATPase, Ca++ transporting, plasma membrane
4 (ATP2B4) gene and upstream of the MARVEL domain
containing 3 (MARVELD3; encoding a tight-junction
protein) gene, highlighting the effectiveness of the GWAS
approach combined with imputation for revealing novel
genetic risk factors for an infectious disease, even in African
populations.
OT H E R I N F E C T I O U S D I S E A S E S
In addition to the examples given above, the GWAS
approach has been employed to elucidate the host genetic
component of several other parasitic, bacterial, and viral
infections and has highlighted genes of interest in disease
pathogenesis. These include GWASs of visceral leishmania-
sis,90 meningococcal disease caused by the bacteria Neisseria
meningitides,91 seropositivity against the L1 capsid protein
of human papilloma virus (HPV),92 antibody and cytokine
responses to smallpox vaccination,93–95 and dengue-related
hypovolemic shock.96 The GWAS approach has also been
applied to further our understanding of recent pandem-
ics, such as the 2009 H1N1 influenza A virus (A[H1N1]
pdm09) pandemic. Results of the first small GWAS focused
on individuals who developed severe pneumonia following
A[H1N1]pdm09 infection and identified a role for genes
in immune complex processing (FCGR2A) and comple-
ment activation (C1QBP).97 A subsequent GWAS further
highlighted a role for the CD55 complement regulatory
protein in the severity of infection, a result that was sup-
ported with functional data showing allele-specific effects
on CD55 expression.98
G enomics and I nfectious D iseases   •   5 7 1
www.Ebook777.com
C AU S AT I VE G E N ET I C L I N K S
B ET W E E N I N F E C T I O U S A N D
N O N - C O M MU N I C A B L E D I S E A S E S
As well as looking at diseases that have an obvious infec-
tious etiology, the GWAS approach has also been applied
to non-communicable diseases with a known or suspected
infectious etiology, including oncovirus-mediated carcino-
mas, Kawasaki disease, and diabetes. Results are helping
delineate mechanisms of carcinogenesis associated with
infection and, for some conditions, providing novel or addi-
tional support for an underlying infectious etiology.
Of the oncoviral induced carcinomas, HBV-associated
HCC has been most thoroughly investigated via GWAS
to date. In the first GWAS to look at genetic risk factors
underlying development of HCC following chronic HBV
infection, Zhang and coworkers99 identified a susceptibil-
ity locus at 1p36.22 (UBE4B-KIF1B-PGD) in a cohort
of Chinese ancestry. The common genetic variant explains
3% of the HCC familial relative risk and was associated
with a population-attributable fraction of 24.1%. Further
studies in Asian HBV carriers with liver cirrhosis or HCC
highlighted variants in the “antigen presentation and pro-
cessing” pathway,100 as well as a variant at an excitatory neu-
rotransmitter receptor subunit (GRIN2A gene).101 More
recently, the GRIK1 gene and glutamate signaling were also
implicated in HBV-related HCC.102 In addition, a small
GWAS identified an expression sequence tag (EST) lying
in the peak region of association that showed decreased
expression levels in HCC tumours.103 The EST was further
shown to have characteristics of a long, non-coding RNA, a
class of non-protein coding transcripts that have numerous
regulatory functions, and could represent a novel diagnos-
tic and prognostic biomarker. Susceptibility to nasopharyn-
geal carcinoma following Epstein-Barr virus infection has
also been shown to be genetically regulated in a GWAS of
Taiwanese samples. Results here implicated the HLA-A and
GABBR1 (gamma aminobutyric acid b receptor 1) genes in
its development.104
Kawasaki disease (KD) is a pediatric systemic vasculitis
of unknown etiology with evidence for genetic regulation
of disease susceptibility.105 Data suggest that an infectious
agent(s) is the underlying cause of disease, but no specific
pathogen has been identified as yet.106 The first GWAS of
KD, carried out in a European cohort, identified associa-
tion at several genes (LNX1, CAMK2D, ZFHX3, CSMD1,
and TCP1) that were shown to cluster in pathways impor-
tant in inflammation, apoptosis, and cardiovascular pathol-
ogy.107 Subsequent functional analyses confirmed decreased
pre-treatment transcriptional expression of these genes,
confirming their likely role in KD. Novel associations in
immune-related genes were also revealed in GWASs of
KD in Chinese108 and Koreans.109 More recently, a third
KD GWAS highlighted a novel association with a func-
tional variant in the IgG receptor gene FCGR2A that could
potentially advance knowledge of disease pathogenesis
and response to treatment.110 An infectious trigger has also
been proposed for complex diseases, such as Type I diabe-
tes (T1D).111 Results of a large GWAS112 in Caucasians fol-
lowed by large scale resequencing113 to identify genetic risk
factors underlying T1D identified several rare variants in
the interferon induced with helicase C domain 1 (IFIH1)
gene that independently decreased T1D risk. This is of
interest, as the IFIH1 gene encodes the cytoplasmic RNA
sensor MDA5, which senses and initiates antiviral activity
against RNA viruses. These results lend support to the the-
ory that T1D could be triggered by a viral infection, leading
to the possibility of prevention via vaccination.
Intriguingly, a number of GWAS variants identified for
infectious diseases coincide with association signals reported
for common noncommunicable diseases, such as psoriasis and
inflammatory bowel disease. This might indicate a balance in
evolutionary genetics in which genetic variants that confer pro-
tection against pathogens also predispose negatively to exac-
erbated inflammatory responses.114 For example, it has been
suggested that people who have psoriasis are more likely to
carry variants that confer protection against HIV-infection.115
Variants associated with HIV infection—including the amino
acid residues at positions 67, 70, and 97 within HLA-B,
the causative rs67384697 variant at HLA-C, the HCP5
rs2395029 variant, and the non-synonymous rs1576 SNP
at PSORS1C3—were found to be major genetic risk factors
of susceptibility to the autoimmune diseases of psoriasis and
psoriatic arthritis.116–118 There have also been cases of remark-
able overlap between the genetic loci underlying communi-
cable and noncommunicable disease. For example, results of
GWASs for leprosy compared to those for Crohn’s disease
(CD) and ulcerative colitis, two common inflammatory
bowel diseases, implicate a common immunological pathway
in the development of all these diseases.22,119–122 This is of inter-
est, as leprosy and CD are both granulomatous diseases, and
the observation of shared genetic risk factors provides support
to theories proposing a potential causative role for mycobacte-
rial infection in the development of CD.123
GWA S — C H A L L E N G E S A N D S U C C E S S E S
Genome-wide association studies for diseases of a poly-
genic nature encompass specific challenges, some of which
5 7 2   •   G enomics in C linical P ractice
are discussed here. Contrary to monogenic disorders, in
polygenic conditions, genetic polymorphisms may not
be instrumental to disease pathogenesis but instead may
modulate host responses to a pathogen. As such, the effect
of each associated variant on the phenotype is likely to be
weak or modest. GWAS findings for infectious diseases
have certainly expanded our knowledge of the genes and
pathways involved but have not always met the initial high
expectations for major advances in the field. Although
there is compelling epidemiological evidence for the role
of genetics in infectious disease susceptibility, power limita-
tions due to small sample sizes is one of the major problems
of genetic study designs.124 When small genetic effects are
expected (i.e., OR ≤1.5), sample size is the crucial factor
determining the power of the study to detect the effect. As
shown by Burgner and colleagues,124 in order for a study to
have sufficient power to detect association at common vari-
ants (i.e., minor allele frequency [MAF] >0.1) with mod-
est effect size (i.e., OR  =  1.5), at least 1,500 case-control
pairs or case-parent trios are required. For example, a very
large cohort of patients and controls (>11,000 individuals)
had to be employed in the Thye and colleagues77 study of
TB to allow identification of common variants with small
effect (OR  =  1.19). In contrast, a much smaller GWAS
(326 individuals) of treatment response to HCV-infection
successfully detected association at genome-wide signifi-
cance due to the much larger effect of the locus detected
(OR  =  18.5–27.2).61 Finally, a GWAS of HIV-infection
focusing on a group of rapid progressors has highlighted the
use of extreme phenotypes as a powerful tool in identifying
genetic associations, even with a small sample size.41
Many of the reported GWASs for infectious diseases
have used cohorts of smaller size (from few hundreds to
a couple of thousand individuals) compared to noncom-
municable diseases (some of which have used tens of
thousands), partly reflecting difficulties in recruiting large
sample sizes for infectious diseases in endemic countries.125
This has resulted in association signals that approach but do
not always reach genome-wide significance (P = 5 × 10–7).22
This strict statistical cutoff aims to filter the plethora of
observed associations generated by the GWA approach in
order to minimize the reporting of false discoveries. A direct
consequence is that truly positive associations with lower P
values can be missed. Many GWAS have compensated for
the lack of genome-wide significance by employing a com-
bined analysis of the discovery and replication cohorts.
However, further evidence of association with interest-
ing candidates should be sought in independent studies in
order to increase confidence in the findings. Overall, the
ORs reported in GWASs for infectious diseases range from
small to modest; the only current exception being the rather
high ORs reported for IL28B61 and DDRGKI/ITPA66 in
Japanese patients with chronic hepatitis C. A consequence
of small–modest ORs is the generally small proportion of
the total genetic variance that can be attributed to these
genetic risk factors, often not exceeding 20%, leaving a
large proportion of the known heritability unaccounted
for or missing. Meta-analysis methods that combine find-
ings from multiple, often underpowered studies will be
valuable in identifying new susceptibility loci and provid-
ing more accurate estimates of effect size.126 But future
studies with larger sample sizes should focus on gene–gene
and gene–environment interactions, and structural or rare
variants to potentially increase information on this missing
heritability.127
The case-control design, commonly used in GWAS stud-
ies, can also present a challenge in the presence of (poten-
tially unknown) genetic substructure within the population
under investigation. Such substructure results from ancestry
differences as well as cryptic relatedness amongst individu-
als, and it can lead to false positive-association signals.128
In this respect, the use of a family-based design in GWASs
presents a distinct advantage due to its robust population
substructure.129 Contrary to the candidate-gene study
approach, the existence of population substructure can be
formally addressed in the GWAS by employing principal
component128 or linear mixed model130 analyses. Ultimately,
these analyses aim to reduce the genomic-inflation factor
(λ), an indicator of the degree of false positive associations in
the data. Whilst most GWA studies used the genome-wide
genetic data available to identify population substructure,
Jallow and colleagues88 showed that self-reported ethnicity
could also account for the observed population substruc-
ture in their study of malaria in West Africans. Although
this might be a useful consideration in case-control studies
that lack genome-wide data, further evidence is required to
demonstrate its applicability across diverse populations.
GWA studies in African populations are also particu-
larly challenging due to the greater genetic diversity and
shorter stretches of LD than seen in Caucasian popula-
tions.131 This issue is clearly illustrated in the malaria GWAS
carried out in Africa,88 where the P values for SNPs at the
known HbS locus present on the array used were rather
weak (best P = 3.9 × 10–7). However, when the actual causal
variant rs334, which is not represented or well tagged on
the array used, was directly genotyped, a much stronger
association was detected (P = 1.3 × 10–28). A similar issue
was seen for the known functional variants at the ABO and
GP6D loci. In this instance, the use of imputation meth-
ods to predict genotypes at known functional variants using
G enomics and I nfectious D iseases   •   5 7 3
the HapMap Yoruba population as reference panel did not
allow the association with rs334 to be detected (P = 0.06).
This issue highlights the need for optimal design of geno-
typing platforms for non-Caucasian populations88 and
demonstrates the methodological challenges of imputation
in African populations. For imputation methods to be use-
ful, an appropriate reference panel that is representative of
the imputed population is essential.87 The current 1KGP is
now providing robust reference panels for accurate imputa-
tion in African populations and a recent GWAS of malaria
has underlined the value of these data.89
In many instances, the GWAS approach has failed to
provide support for findings that emerged from earlier
GWLSs and candidate gene studies. This might indicate the
occurrence of false-positive associations in previous reports.
However, it is worth noting that the well-established CCR5-
∆32 association with HIV-infection was often not identified
with the required genome-wide significance in the reported
GWASs of HIV. Hence, the lack of replication across studies
may be a consequence of other parameters and study limita-
tions, which can include differences in the clinical phenotype
definition or study design, limited sample sizes, overesti-
mated effect size of the original association (referred to as the
“winner’s curse”), or allelic heterogeneity (different causative
alleles within a locus). Additional explanations may lie in
allele frequency and haplotypical structure differences across
diverse populations. As a result, the markers tagging the true
causative variants might differ between two populations, thus
deviations in SNP coverage and low tagging efficiency of
SNP arrays may hamper the replication of association signals.
In spite of the various study limitations, GWASs of infec-
tious diseases have offered new insights into genetic risk
factors underlying human susceptibility. The strong associa-
tions with evidence of replication in multiple cohorts could
be utilized in routine diagnostics to facilitate predictions on
the disease course and the effectiveness of alternative treat-
ments, or such risk factors can be considered as new targets
for therapy and vaccine development. In summary, GWAS
findings to date have
1. confirmed previous associations (e.g., HLA loci for
HIV and HBV infections), increasing confidence in
their importance;
2. revealed novel genetic loci and pathways with a
plausible role in disease pathogenesis (e.g., Type III
interferon signaling pathway for HCV, NOD-signaling
pathway for leprosy); and
3. identified associations in previously unexplored regions
(e.g., novel loci for HIV-infection and tuberculosis),
 which have the potential to provide new knowledge of
the mechanisms of disease.
Many of the reported loci are related to the adaptive (e.g.,
HIV and HBV infections) or innate (e.g., hepatitis C, lep-
rosy, meningococcal disease) immune system.
While in some instances the causative variant has been
proposed (e.g., HIV infection), the successful identifica-
tion of other causal variants is likely to require a com-
bination of methods.132 These include imputation and
meta-analysis approaches, pathway analyses, sequencing
strategies, and gene-expression studies, as well as solid
functional evidence.133 Evolutionary approaches in search
of natural-selection signatures can further delineate genes/
polymorphisms with a role in host defenses and enhance
our knowledge of how pathogen-driven selective pressure
impacts allelic variability.134,135 Plausible theories regard-
ing the genetic architecture of susceptibility to each infec-
tious disease could guide future strategies to determine
additional genetic risk factors.125 Further identification of
common variants associated with disease, for example, will
be facilitated by improved SNP coverage on commercially
available microarrays. However, next-generation sequenc-
ing, either of the exome (the coding portion of the genome)
or of the whole genome, will be required to identify the rare
variants involved in disease susceptibility. In addition, the
plethora of reported associations within immune-related
genes emphasizes the need for a systematic investigation of
the MHC region as well as a more targeted approach to dis-
sect information from a large number of genes implicated in
innate and acquired host responses.136
N E X T- G E N E R AT I O N
SEQUENCING
Along with the advances made in the availability of dense
genotyping arrays that allow genome-wide assessment
of common variations, there have also been significant
advances made in sequencing technologies. Sanger sequenc-
ing, the technology of choice for over two decades, has
numerous limitations, mainly around throughput poten-
tial, scalability, and cost. Nonetheless, Sanger sequencing
has proven invaluable in expanding our knowledge of the
genetic determinants of infectious disease susceptibility,
allowing the re-sequencing of candidate genes to identify
putative causal variants. However, over the last few years,
the next generation of sequencing technology (NGS),
which relies on several stages, including template prepa-
ration, sequencing, imaging, and data analysis, has been
5 7 4   •   G enomics in C linical P ractice
rapidly developed. These NGS methods allow the complete
exome or entire genome to be sequenced rapidly and accu-
rately and at a depth of coverage that allows novel variants
to be identified. These features, combined with falling costs,
mean it is now feasible to fully sequence at least the exome
in sample sizes large enough to identify host genetic factors
that play a role in disease susceptibility.
Exome sequencing has proven especially successful for
monogenic disorders, where we expect the mutation to
lie within the coding region of the genome. This includes
a number of single-gene defects predisposing to infectious
disease. For example, exome sequencing allowed the iden-
tification of a causative mutation underlying childhood
susceptibility to Kaposi sarcoma following infection with
human herpes virus-8 (HHV-8).137 Exome sequencing of
one patient identified a homozygous splice site mutation
in the stromal interaction molecule 1 (STIM1) gene that
resulted in the abolition of mRNA splicing and protein
production leading to T-cell deficiency. Exome sequencing
has also been used to identify germline mutations in the sig-
nal transducer and activator of the transcription 1 (STAT1)
gene that underlie chronic mucocutaneous candidiasis, a
persistent and recurrent Candida albicans infection of the
nails, skin, or oral mucosae.138 Finally, exome sequencing
allowed the recent identification of the ubiquitin-like mod-
ifier (ISG15) gene as underlying Mendelian susceptibility
to mycobacterial disease (MSMD), a rare monogenic dis-
order characterized by severe clinical disease upon infection
with weakly virulent mycobacterial species.139 Two muta-
tions identified in two unrelated families were not found in
1,256 control subjects. Functional studies suggest the muta-
tions are loss-of-function mutations resulting in a lack of
ISG15 gene product and a subsequent impairment of IFNγ
production.
Given that the exome represents only around 1%
of the entire genome, it is perhaps not surprising that
exome sequencing is currently more commonly applied
to understanding human genetic risk to infection than is
whole-genome sequencing. However, the applicability of
exome sequencing to polygenic disorders, in which variants
influencing disease risk may not be confined to the cod-
ing regions, is less well established. One potential strategy
when using exome sequencing to look at a polygenic trait
is to select the individuals with phenotypes that fall at the
extreme ends of the phenotype distribution, under the
assumption that risk and protective alleles will be enriched
in these individuals, which would allow success with a more
modest sample size. This approach has been successfully
applied to look at susceptibility to Pseudomonas aeruginosa
infection in individuals with cystic fibrosis.140 Sequencing
the exomes of individuals with either very early or very
late-onset P.  aeruginosa infection revealed that variants in
dynactin 4 (DCTN4) were associated with time to onset of
chronic P. aeruginosa infection (P = 2.2 × 10−6). Given that
chronic P. aeruginosa infection in cystic fibrosis sufferers is
associated with reduced lung function and survival, these
results could help to determine which individuals require
more aggressive screening and therapy.
T R A N S C R I P TO M E A N A LYS I S
Transcriptomics is the global study of gene expression at
the RNA level. As well as looking at mRNA, transcrip-
tomics can also include the analysis of other RNA species
such rRNA, tRNA, and numerous non-coding RNA spe-
cies within a specific cell type or within a tissue. In paral-
lel to the quest for genetic determinants of disease, there
has also been significant focus on understanding the host
transcriptome: especially how it is modified in response to
environmental stimuli, pathogens, or autoimmune reac-
tions. It has been possible to assess genome-wide transcrip-
tional changes using hybridization-based microarrays for
some years, and this technology has provided new insights
into the molecular mechanisms of pathogenesis. Such
studies utilize biologically relevant tissue samples, such as
whole blood, spleen, liver, bone marrow, or specific cell
populations (such as peripheral blood mononuclear cells)
to identify patient-based patterns of expression that may
serve as putative biomarkers of disease progression or treat-
ment success.141 Among the first investigations to identify
pathogen-induced signatures of differentially regulated
genes were those in HCV-infected patients with cirrhotic
livers142 and in patients with influenza A, Escherichia coli,
or Streptococcus pneumoniae infections.141,143 To date, gene
expression profiles have been demonstrated for a num-
ber of infectious diseases, including HIV-infection,144–147
HPV-induced carcinomas,148–152 HCV-related diseases,153,154
tuberculosis,155–161 malaria,162 and influenza.144,163–166
A systematic review of all these studies is beyond the
scope of this chapter, but one study is highlighted here as
an example of the clinical relevance of genome-wide host
transcriptome data. A genome-wide transcriptional study
conducted by Berry and colleagues155 in active pulmonary
TB patients identified a whole-blood signature of 393
genes that correlated with disease severity which dimin-
ished upon anti-tuberculosis treatment. Comparison to
transcriptome signatures obtained for other bacterial and
autoimmune diseases revealed a subset of 86 genes whose
differential expression was unique in active TB patients.
G enomics and I nfectious D iseases   •   5 7 5
Pathway analysis revealed an over-representation of both
Type I and II IFN-signaling pathways, with blood neutro-
phils, and to a lesser extent, monocytes, being the main
contributors. Importantly, in terms of clinical relevance,
10–25% of individuals with latent TB infection showed
a transcriptome profile similar to that of active TB cases,
providing a potential tool for identifying those at risk of
progressing to active disease. Follow-on studies have also
identified a transcriptional signature specific to patients
undergoing anti-tuberculosis treatment that may allow the
development of personalized treatment regimens and novel
biomarkers of successful treatment.167
The recent advances in NGS technologies that allow
the exome and whole genome to be sequenced are now
also being applied to allow the entire transcriptome to
be sequenced and quantified, a technique referred to as
RNA-Seq. RNA-Seq has some attractions over transcrip-
tome profiling via hybridisation-based microarray, in that
RNA-Seq allows the capture and profiling of the entire
transcriptome and so has revealed previously unknown
complexities, such as an abundance of long non-coding
RNAs, micro-RNAs, and alternatively spliced variants.168
RNA-Seq methods are now being applied in order to more
fully understand the host response to infection. For exam-
ple, RNA-Seq has been used to determine factors that influ-
ence individuals infected with Leishmania braziliensis to
progress from cutaneous leishmaniasis (CL), characterized
by a single lesion, to mucocutaneous leishmaniasis (ML),
a progressively disfiguring lesion of the nasopharynx.169
RNA-Seq was applied to transcripts obtained from lesion
biopsy samples and revealed differential expression of 867
genes, 434 of which were upregulated in CL patients and
433 of which were upregulated in ML patients. Of these,
the transcriptional changes at 13 genes were shown to
predict a higher probability of ML, suggesting they could
be developed as novel biomarkers of long-term clinical
outcome. Interestingly, this study also demonstrated that
some RNA-Seq reads mapped to the Leishmania genome,
although not at high enough coverage to allow meaning-
ful analysis. Nonetheless, this observation is of interest and
suggests that it may be possible to simultaneously assess the
transcriptome of the host and the parasite.
MICROBIAL GENOMICS
These advances in genomic technology are not just being
applied to the investigation of our own genome; they are also
being employed to scrutinize the genomes and transcrip-
tomes of the pathogens themselves in much greater depth
than has ever been possible before. The last five years have
seen increased application of next-generation sequencing in
the clinical setting, particularly in the field of metagenom-
ics, where many high-profile multicenter studies are now
being published. The cost of sequencing has plummeted,
bringing this methodology into the realm of other diag-
nostic tests. This, combined with the availability of smaller
benchtop sequencers and user-friendly analysis software,
has made this technology available to most research groups,
even those with smaller budgets and limited infrastructure.
Although there are still practical issues related to translating
this technology for use in clinical medicine, sequencing is
rapidly becoming a common tool in the molecular biology
laboratory and is now making its way into the diagnostic
arena.
N E X T- G E N E R AT I O N S E Q U E N C I N G I N
T H E C L I N I C A L E N VI R O N M E N T
Cell culture, serology, and biochemical assays have been
the methods of choice to identify and type human patho-
gens for decades, but these methods are now being com-
plemented or replaced with molecular methods based
on knowledge from microbial sequencing. This is due
to the fact that many phenotypical methods, including
culture and biochemical detection, are specific to the
pathogen under investigation. Furthermore, the cultur-
ing techniques that have been the mainstay of micro-
bial diagnostic practice for more than 50  years remain
time-consuming, with some microbes taking weeks or
months to grow, while others still cannot be cultured.
Since the advent of polymerase chain reaction (PCR) in
the 1980s, the ability to determine the DNA and RNA
sequences has revolutionized all areas of biomedical sci-
ence, including microbial diagnostics. There are many
advantages to genetics-based methods, not least the fact
that PCR amplification, hybridization, and sequencing
can (generally) be applied across the board. Molecular
tests derived from sequence information can also be used
in all areas of the clinical microbiological laboratory,
including pathogen detection and typing, determination
of drug resistance, surveillance of infections and out-
breaks, and analysis of whole microbial communities (or
microbiomes).
Development of these methods is flourishing in the
wake of a revolution in microbial sequencing. Although
the first complete viral genome was sequenced as far back
as 1977,170 the first fully sequenced bacterium was that
of Haemophilus influenzae, almost 20 years later.171 More
5 7 6   •   G enomics in C linical P ractice
recently, the advent of next-generation sequencing has
dramatically altered the landscape of pathogen detection
in four areas. First, the massive increase in data gener-
ated by these machines has not only provided increased
information on microbial genomes, it also decreased the
cost of sequencing (see Figure 37.2). This continuing
trend is now bringing the cost into the realm of the diag-
nostic laboratory. It is currently possible to completely
sequence a bacterial genome for less than $200 and a viral
genome for even less.172,173 Second, unlike using Sanger
sequencing of microbial genomes, NGS does not require
the time-consuming preparation of a cloning library.
Instead, the DNA is randomly sheared to form a DNA
library, distributed across millions of sequencing nodes,
after which it undergoes massively parallel amplification
and sequencing. This means it also possible to directly
sequence both whole cultured bacteria and mixed clini-
cal or environmental samples. Third, because the DNA
is randomly sheared prior to sequencing, there are no
significant biases induced by the secondary and tertiary
DNA structures that impact on cloning libraries for
Sanger sequencing. Although there are still sequence
types, such as AT- or GC-rich genomes or those with
long stretches of a single base, that effect sequencing,173
it is generally accepted that the DNA is sequenced evenly
across the DNA library. Finally, the ability to multiplex
DNA libraries through addition of unique barcodes has
added flexibility to the system, allowing smaller genomes
or amplicons to be run economically.
To date, the market has been populated by three
competing technologies: 454 pyrosequencing, Solexa/
Illumina HiSeq, and ABI SOLiD platforms. The impetus
behind the development of these technologies was primar-
ily driven by human genome sequencing programs. This
resulted in an emphasis on increasing sequencing volume
per run at the expense of flexibility in sample provision,
meaning the cost per run remained prohibitively expen-
sive for microbial genomes. However, in the last couple
of years this has changed, due to the ability to multiplex
many samples in a single run, making NGS cost-effective
enough to be applied to microbial genome. The recent
introduction of smaller benchtop sequencing machines,
including the MiSeq from Illumina, Ion Torrent from
ABI, and Roche 454 FLX Junior, has further expanded
the potential uses of NGS. These machines generate less
data without significantly increasing the cost per base; are
affordable, with a reduced initial investment and lower
running costs; and offer user-friendly sample prepara-
tion, sequencing, and analysis. In 2012, these benchtop
machines started to appear in larger microbiological
 $ 10,000.0 4000

Sanger 454 Illumina/ABSOLiD

Sequencing Sequencing Sequencing 3500
$ 1,000.0

3000

$ 100.0
2500

$ 10.0 2000

1500
$ 1.0

Complete Genomes 1000

Cost of Sequencing
$ 0.1 Per Mb $
500

$ 0.0 0
99
00

01

02

03

04

05

06

07

08

09

10

11

12
19
20

20

20

20

20

20

20

20

20

20

20

20

20
Figure 37.2 
Graphical representation of the rapid decrease in sequencing costs leading to an increased number of completed genome projects. The
availability of new technology and the advent of next-generation sequencing have dramatically reduced sequencing costs. Estimates of costs are
based on the generation of a megabase of raw sequence data sourced from the NHGRI (www.genome.gov/​sequencingcosts​), and not the multiple
times coverage of finished sequence required for genome assembly. The completed genome data are taken from the Genomes Online database
(http://www.genomesonline. org).

diagnostic laboratories when the U.S. Food and Drug
Administration (FDA) began testing the practicality of
these machines through placement of MiSeq instruments
in both state and federal laboratories.174 As these bench-
top sequencers become more robust and easier to use,
there is even the possibility of incorporating this technol-
ogy into rural or mobile laboratories, as has been done
with other molecular equipment.
In addition, there are also new contenders bringing
novel technologies to the NGS market in the form of
third-generation machines that have potential benefits
beyond that provided by next-generation machines. The
first of these to be commercially available, the PacBio from
Pacific Biosciences, has applied itself to the microbial
genomics market by mixing high-quality sequence data
obtained from short, accurate sequencing reads (<1000
base pairs) with the assembly information obtained from
longer (<10,000 bp) reads. The combination of these read
types leads to improved genome assembly. The develop-
ment of nanopore-sequencing technology also promises
the possibility of cheap, highly portable, easy-to-use
machines that will deliver results in real time directly
to a laptop or mobile device. Oxford Nanopore recently
unveiled two highly portable machines, one not much
larger than a USB flash drive (MinIon), and the other the
size of a DVD player (GridIon), which can sequence a
bacterial genome in a day. It is anticipated that availabil-
ity of this technology will lead to more widespread use
in clinical and rural environments (www.nano​poretech.
com/​).
B AC T E R I A L G E N O M E
SEQUENCING
While the majority of diagnostic services use PCR-based
methods in conjunction with cell culture, we are now
moving into a new era of genetic diagnostics carried
out directly on clinical samples. This move is leading
to faster results, something that can be critical in deter-
mining patient treatment or when dealing with an infec-
tious disease outbreak. The potential of whole-genome
sequencing adds another major advantage to the clinical
arena. Whereas phenotypical and molecular tests have
to be targeted to answer one issue, such as identifica-
tion, isolate-typing, antibiotic resistance, response to
vaccination, or pathogen surveillance, whole-genome
sequencing is an all-encompassing test that can simulta-
neously address all these requirements.175,176 Furthermore,
whole-genome sequencing requires no prior knowledge

G enomics and I nfectious D iseases   •   5 7 7

of the pathogen’s genetic sequence, making this tech-
nology more amenable to identifying the emergence
of new antibiotic resistance mechanisms or the emer-
gence of novel human pathogens. Given the high muta-
tion rate of microbial species and the frequency of
new zoonotic infections, this characteristic will prove
very useful. For example, in 2008 during an outbreak
of multidrug-resistant Acinetobacter baumannii in the
United Kingdom, the standard molecular methods used
to identify the bacteria (in this case, pulsed field gel elec-
trophoresis) could not distinguish between isolates from
six cases. The use of NGS identified three variants that
were subsequently used to identify the initial case.177
While generating a complete bacterial genome sequence
is feasible, there are numerous technical issues that currently
hinder the integration of whole-genome sequencing into
the clinical setting. Although sequencing a bacterial genome
can be done in one to two days, interpretation of the results
can take much longer, depending on the resources avail-
able. The volume and complexity of the data generated can
be overwhelming, and analysis of such data requires good
computing power, specialist analytical software, and exten-
sive bioinformatic support.
Another issue is the requirement of data for comparison
purposes. In isolation, the sequence of a pathogen does not
reveal the wealth of data once predicted; instead, its value
comes from comparative analysis to the genome sequences
of other, well-phenotyped isolates and strains from differ-
ent species. For 20+ of the commonest human pathogens,
the genome sequence of various clinical isolates is available;
however, there are hundreds of rare or neglected pathogens
that lack critical comparative data. The expansion of the
public bacterial genome databases has occurred rapidly, due
to several large sequencing initiatives. The first large micro-
bial genome program was launched by the U.S. Department
of Energy in 2004 (http://microbialgenomics.energy.gov/
mgp.​shtml​). This ten-year project has had a significant impact
on our understanding of pathogens and related infectious
diseases.178 New initiatives, such as the 10,000 Microbial
Genomes Project from the Beijing Genome Institute
(BGI) (http://ldl.genomics.org.cn/​page/M-research.jsp​),
the 100K Foodborne Pathogen Genome Project (http://
100kgenome.vetmed.ucda​vis.edu/​), and the 900 genomes
identified in the Human Microbiome Project (https://
commonfund.​nih.gov/Hmp/​) will dramatically increase
the depth and diversity of data available. From a data-
handling point of view, the continually expanding size of
the genomic databases will increase the requirement for
memory-intensive search facilities, although specific tar-
geted databases, such as those for particular genes like 16S
5 7 8   •   G enomics in C linical P ractice
rRNA locus, and common pathogen- or species-specific
databases, will facilitate the process.
Perhaps a more immediate issue in the clinical context
is the cost, both of infrastructure and per test. The smaller
benchtop machines cost more than $100,000 to buy, and
while they can sequence a genome in a day, the consum-
able costs are still over a $1,000 per run. Multiplexing the
samples to combine multiple bacterial genomes in one
run will reduce this financial burden. For example, the
most economical of the benchtop machines, the MiSeq
from Illumina, can produce more than 2 gigabases (Gb) of
sequence data at a cost of $500 per Gb.173 This is sufficient
data for 5–10 bacterial genomes and reduces the cost per
genome to a couple of hundred dollars. However, this is
a rapidly changing landscape, and some of the new tech-
nologies expected on the market in the near future could
be more applicable to the routine clinical laboratory. The
highly portable machines under development by Oxford
Nanopore are likely to cost less than $1,000, making these a
potentially cheap addition to the routine laboratory. What
makes them possibly even more applicable to the clini-
cal environment is their potential for sequencing DNA
directly from samples such as blood, plasma, cerebrospinal
fluid, or urine without the need to extract and purify the
DNA first (http://www.nanoporetech.com/). This brings
the possibility of loading a sample directly into a handheld
device and recording the data directly to a laptop computer
much closer to reality.
Extracting relevant data from the large volume of
sequence data that are generated from an entire genome can
be difficult, requiring significant knowledge and resources.
One option to overcome this is to crowd-source the data
analysis, allowing external users to contribute. Perhaps the
best example of this comes from the 2011 Escherichia coli
outbreak in Germany. This virulent E. coli isolate spread
rapidly, causing almost 4,000 cases and 54 deaths. It had
an unusual clinical phenotype, with a high proportion
of adult cases and high levels of hemolytic-uremic syn-
drome (HUS). Serotyping identified the 0104:H4 strain,
which had only previously been associated with sporadic
cases. Within three days, the genomes of 10 isolates had
been sequenced, but to speed up analysis, the data were
released to an open-source global consortium of bioinfor-
maticans who collaborated on the analysis. This approach
resulted in the development of a specific, rapid diagnostic
DNA test that became available just five days later.179 This
had a dramatic impact on identifying and tracking new
cases and helped limit the extent of the outbreak. This
work also found that the outbreak strain contained char-
acteristics of two pathotypes, including a Shiga toxin 2a
(Stx2a)–encoding bacteriophage, explaining the severity
of the clinical phenotype.180
To date, the power of whole-genome sequencing has
been mostly utilized in generating targeted tests for iden-
tifying virulent isolates and tracking outbreaks. Sequencing
of emerging drug-resistant or virulent pathogens is critical
to understanding the dynamics of microbial pathogenicity
and to generating tests to detect new cases in time to allow
effective treatment. The data provided by whole-genome
sequencing have been applied to help identify and design
diagnostic tests for the increasing number of drug-resistant
microbes. For example, the genotype MTBDR
DNA-hybridization assay for drug-resistant tuberculosis
incorporates detection for resistance to a complete range of
drugs, including isoniazid, rifampicin, ethambutol, amino-
glycides, fluoroquinilones, and capreomycin.181 While the
initial study showed a 100% concordance with sequenc-
ing results to detect the resistance genotypes, testing on a
panel of 206 known drug-resistant isolates from the Italian
drug-resistance surveillance system showed it detected
91% of rifampin resistance and 67% of isolates resistant for
isoniazid.182
This methodology has been taken further with the
generation of microarrays covering thousands of unique
sequences, representing pathogenicity loci and mobile ele-
ments known to contribute to the acquisition of antibiotic
resistance mined from public databases. The major advan-
tage of these systems is that the same array can be used to
detect pathogenic variants and monitor outbreaks for a
vast range of bacterial species. For example, the Active
Surveillance of Pathogens oligonucleotide array was ini-
tially designed to identify 4,986 variants from 151 bacte-
rial species.183 More targeted approaches have also been
developed for clinical use, such as arrays developed to
detect antimicrobial resistance loci in gram-negative bac-
teria184 and for detecting three bacterial species that can
cause life-threatening bacteremia.185 However, there are
limitations to the array-based methods that NGS routinely
applied in the clinical setting will eventually help to over-
come. These include their current inability to identify novel
variants or classes of genes, the lack of resolution for phylo-
genetic analysis, and the requirement for larger volumes of
genomic material than are required for NGS.186
NGS has also been successfully utilized in tracking ret-
rospectively, looking at outbreaks of TB,187 salmonella,188
anthrax,189 and klebsiella.190 More importantly, it has also
been used in real-time during outbreaks of E. Coli,180 chol-
era,191 listeria,192 and acinetobacter.177 Targeted PCR-based
methods traditionally used for typing bacteria, such as ribo-
typing of the 16s rRNA gene,193 multilocus sequence typing
G enomics and I nfectious D iseases   •   5 7 9
(MLST),194 or detection of variable-number tandem repeat
(VNTR) regions, will detect the causative agent involved
in an outbreak, but they lack the resolution to differenti-
ate between the clonal isolates involved in the spread of
the infection. So, while information on the probable viru-
lence of the agent maybe be discernible, mapping the index
case or tracking the genetic changes during the outbreak
is not feasible when just a few loci are typed. In the last
few years, the greater availability of rapid, whole bacterial
genome sequencing has led to instances where variants in
the genome have been used to map the progress of an out-
break.175,195 As a result, researchers have been able to locate
the source and follow the transmission route of the infec-
tion, identify pathogenicity factors and antibiotic resistance
variants, and, critically, develop rapid tests specific for the
isolates involved that can be used while the outbreak is still
in progress.
VI R A L G E N O M E S E Q U E N C I N G
There have been numerous novel or emerging viruses
that have caused major disease outbreaks within the last
decade. The increase in global travel, combined with
high transmissibility, especially of respiratory viruses,
has required rapid response for identification of the
pathogens. During the severe acute respiratory syndrome
(SARS) outbreak in 2002, the causative agent, a novel
coronavirus, was identified by a DNA-based microar-
ray chip containing conserved viral sequences from
every publically available sequence database specifically
designed to rapidly identify and characterize viruses.196
Within a couple of weeks, the viral genome had been
sequenced, providing one of the first instances where the
genome sequence of the pathogen preceded knowledge of
its biology.197 Since then, identification of novel viruses is
usually undertaken using NGS.
However, there are many issues specific to viruses that
have to be considered when using sequencing on clinical
samples (reviewed in198). First, human viral pathogens are
made up of both DNA and RNA viruses that have to be
treated differently prior to sequencing. In addition to dif-
ferences in extraction of the nucleic acids, RNA has to
undergo reverse transcription to generate cDNA prior to
preparation of the sequencing library. Second, one of the
major issues related to detection in clinical samples is the
very low abundance of viral genomic material present in
the sample. While detection of viral sequences can be done
on material extracted directly from the clinical sample, the
low abundance requires a deep level of sequencing, with
a vast majority of the data-representing host DNA dis-
carded prior to analysis. This is shown clearly in a study
that undertook deep sequencing directly on DNA and
RNA extracted from swabs of patients with acute lower
respiratory infections. Results showed that only 0.05%
of the sequence data belonged to viruses, and just over
1% belonged to bacterial species.199 As such, even when
the viral load is enriched, relative to the host or bacterial
material, it will generally be at too low a concentration for
sequencing to be informative.
In view of the limited viral material obtainable from
clinical samples, the preparation steps must also include
an amplification step prior to sequencing, either of a
specific amplicon or of the whole sample using random
primers. Unlike bacteria, there are no genes common
to all viruses, so methods that target specific genes for
amplification lack the ability to identify novel or highly
divergent viral genomes. Many of the emerging viral dis-
eases are RNA viruses that cross over from animal ori-
gins, such as HIV, corona virus, influenza, Ebola, dengue,
hepatitis C virus, and hantavirus. A  novel method to
overcome this limitation when sequencing RNA viruses
is the sequence-independent single primer amplification
(SISPA) method. This incorporates a reverse transcrip-
tase conversion of the RNA to cDNA with the addition
of a known tag to the double-stranded cDNA that is gen-
erated. The cDNA is then amplified by primers specific to
the tag prior to sequencing.200
Viruses have very high rates of mutation, which,
together with high turnover and very short generation
times, results in a high level of genetic diversity, even
within the same patient. This allows viruses to adapt rap-
idly to environmental conditions and leading to adapta-
tion to antiviral drugs. There is now increasing use of NGS
for epidemiological studies to monitor viral evolution in
human populations and to monitor the response to viral
therapies (reviewed in198). Sequencing, along with being
used to identify the emergence of new drug-insensitive
or -resistant types, such as those responsible for integrase
resistance in HIV,201 can also be used to screen patients for
mutations that are known to render a treatment ineffec-
tive. Use of Selzentry, a CCR5-antagonist to treat HIV-1
patients is only effective against R5 viruses that infect cells
via the CCR5 chemokine receptor. The use of this drug in
patients with HIV viruses bound to the CXCR4 receptor,
the so-called X4 types, is ineffective. The development of a
test that combines both Sanger sequencing and NGS has
been used to identify the tropism of the virus, allowing
the treatment to be given only to those patients who will
benefit.202
5 8 0   •   G enomics in C linical P ractice
S E Q U E N C I N G T H E B AC T E R I A L
T R A N S C R I P TO M E
Over the past few years, the development of NGS and the
associated bioinformatics tools has widely expanded our
view of microbial gene regulation, physiology, and patho-
genicity. Direct sequencing of transcribed RNA is steadily
replacing the use of hybridisation-based oligonucleotide
arrays as the method of choice for studying the bacterial
transcriptome. This is primarily due to the fact that NGS
methods do not require prior knowledge of the genome
for array design, which enables the detection of novel
transcripts. Transcriptome analysis via NGS also provides
increased sensitivity and dynamic range.
RNA sequencing has been used to identify new genes
in previously finished genomes, to refine gene models,
and to identify novel expression systems that alter gene
expression in response to different environmental condi-
tions.203 This new knowledge from transcriptomics analy-
sis has subsequently provided a novel strategy to develop
new drugs or vaccines to treat infectious diseases.204 For
example, knowledge from primary transcriptome studies of
the human pathogens Helicobacter pylori and Burkholderia
cenocepacia is helping identify reliable vaccine candidates
and novel therapeutic targets.205,206 Also the detection of 5′
untranslated region (5′-UTR) and transcription start sites
in pathogenicity islands suggests a role for these UTRs in
virulence regulation.207 Recently, comparative transcrip-
tomics has been applied to the study of pathogenic and
non-pathogenic Listeria species, identifying a novel form
of bacterial regulatory RNA (long antisense transcripts,
lasRNAs) and providing a new view of bacterial resistance
and pathogenicity.208 While transcriptome sequencing has
not been readily applied directly to the clinical diagnostic
environment, transcriptomic knowledge is laying the foun-
dation for novel drug and vaccine discovery and bringing
the potential for new diagnostics.
M ETAG E N O M I C S
Until very recently, diagnostic or clinical methods have been
restricted to the detection of a single pathogen that causes a
clearly definable clinical phenotype, which in most cases is
at the severe or life-threatening end of the spectrum. While
only a minority of well-studied microbes cause disease,
the importance of the commensal microbial communities
in the human body is becoming increasingly recognized.
The last few years have seen several large, ambitious proj-
ects designed to examine the impact of these microbial
communities on maintenance of health and development of
disease. These projects have been driven by the availability
of NGS technologies.
The human microbial community, or microbiome,
plays a vital role in the maintenance of health and in the
development of a range of diseases, being associated with
autoimmune, inflammatory, and metabolic diseases, as well
as with cancer (reviewed in209). This is not surprising, given
that bacteria make up an estimated 90% of the cells in the
human body, outstripping human cells by a ratio of 10:1 and
viruses by a ratio of 100:1. The human gut alone contains as
many as 1,000 bacterial species that have co-evolved with
humans. In fact, this relationship is so close that it has been
used to confirm human migration patterns with microbial
changes found in Helicobacter pylori following the evolu-
tionary migration pathway of humans predicted by other
methods.210
This co-evolution has led to a symbiotic relationship
with the bacteria of our microbiome that benefits human
health in two ways. First, normal colonization by commen-
sal bacteria can prevent the establishment or dominance of
pathogenic species. The dramatic increase in outbreaks of
the gram-positive bacteria Clostridium difficile, the causative
agent of severe antibiotic-associated diarrhea, is an excel-
lent example of this replacement by pathogenic microbes.
While present naturally at low levels in some people, C. dif-
ficile only causes disease when the gut flora are disrupted,
usually following use of antibiotics to treat other bacterial
infections. Under these circumstances, the proliferation of
certain strains of C. difficile, which are not only resistant to
many antibiotics but also release toxins, can lead to severe
diarrhea and colitis. This is a major problem in environ-
ments where susceptible individuals reside, such as in hos-
pitals. In the United Kingdom from 2006 to 2008, it led
to 2.2% of all hospital deaths and 1.4% of deaths across the
country as a whole.211 Given the detrimental effect of antibi-
otics, more imaginative treatments for C. difficile infection
are being considered, such as fecal transplantation from
a healthy donor to reconstitute the gut microbiome.212
A recent study further identified just six bacteria that, when
given as a cocktail, suppressed the C.  difficile and helped
restore the normal balance of gut flora.213 Sequencing
showed that these six species are genetically diverse and
represented all four main groups of bacteria found in the
human gastrointestinal tract, and it paves the way for the
development of a controlled probiotic treatment for this
disease. Secondly, the human microbiota contribute signifi-
cantly to host metabolism, to such an extent that humans
are not considered to be metabolically viable without the
contribution from their colonizing gut bacteria. A  study
G enomics and I nfectious D iseases   •   5 8 1
in mice found that most of the metabolites present in the
blood are derived from microbial activity in the gut.214 Data
from the Human Microbiome Project have identified a
large number of dietary compounds, drugs, and even host
enzymes that are altered by gut bacteria, describing this area
of investigation as pharmacomicrobiomics.215
It is estimated that the bacteria in the gut microbiome
have a collective genome of 3 million unique genes, 150-fold
more than the human genome.176,216 While the fetus has a
sterile gut environment, it is populated shortly after birth by
close relatives, in particular, the mother. The development
of the microbiome is thought to be a result of selection
based on co-adaption between host and bacteria. While
this is critical for health, there are also many implications
for disease development, especially when the microbial
environment is perturbed (reviewed in209). In view of this,
there is a clear necessity to understand the human micro-
biome; however, our ability to directly sequence microbes
from environmental (or clinical) samples has shown that
less than 1% were identifiable by previous culturing meth-
odologies.217 Isolating and sequencing the genomes of
clinically relevant viruses, bacteria, and protozoan para-
sites have provided important insights into the biology
and pathogenic mechanisms of these organisms. However,
the prerequisite to isolate and culture the samples prior to
sequencing limited our understanding of the interaction
between these microorganisms, the microbiome, and the
host in the clinical context. Therefore, our ability to study
the microbiome has been transformed by the availability of
NGS, and metagenomic sequencing has now been used to
profile microbial communities from samples as diverse as
seawater, soil samples, various human body sites, and even
the fly splatter on car windshields.218 As costs fall and the
practical issues related to sampling microbial communities
in situ improve, the application of these methods to human
health will get broader.
The field of metagenomics covers two quite differ-
ent sequencing approaches—amplicon sequencing and
whole-genome shotgun sequencing. Amplicon sequencing
uses genes that contain both conserved and variable regions
and that are present in as broad a spectrum of microbial
organisms as possible. The use of universal primers that
can amplify the same region of a gene in multiple species
has allowed researchers to profile microbial communities
with even modest amounts of sequence data. The locus
predominantly used for amplicon metagenomic studies is
the non-coding 16S rRNA gene that is present in all bac-
teria and Archaea. It is ideally suited for the purpose, not
just for its ubiquitous presence in bacteria, but also for the
presence of highly conserved regions large enough for the
use of universal primers and of variable regions that are suf-
ficiently diverse to provide the ability to “speciate” microbes
in a complex sample. In addition, the popularity and wide
use of amplicon sequencing has led to the construction of
comprehensive 16S rRNA databases that can be used to
identify bacteria in new studies.219
However, this methodology has a number of drawbacks
in accurately estimating the microbial diversity within a
sample. First, there is no gene that is common to all viruses;
thus limiting the use of amplicon sequencing for metage-
nomic profiling to bacteria and Archaea. Even within bacte-
ria, the amplicon approach has issues: the 16S rRNA gene
is not a single locus within the genome; different bacterial
species have between one and 15 copies.220 In addition, PCR
bias can lead to differences in amplification of certain bacte-
rial groups.221 Both of these issues have implications for the
validity of the bacterial environmental diversity determined
by amplicon sequencing. Perhaps the major weakness of this
method is the paucity of functional data from the microbi-
ome. Sequencing of just the 16S rRNA locus provides little
or no information about the bacterial genetic component
or its capacity to induce disease.
The alternative method is to undertake whole-genome
shotgun sequencing (WGS), the same methodology that
is used for sequencing single genomes. The benefit of this
methodology lies in the unbiased sequencing of the whole
microbial “pan-genome.” For example, more than 3 mil-
lion bacterial genes have been sequenced from the human
gut microbiome alone;216 this is significantly more than in
the human genome. However, there are significant issues
related to both the vast quantity and the complexity of
data generated and the bioinformatics infrastructure
required to analyze it, meaning that the resources needed
for NGS currently prohibit its application in the routine
diagnostic setting. In addition, while sequencing of the
16S rRNA locus will only amplify bacterial genetic mate-
rial, WGS sequencing will generate data on any DNA
present in the sample, including host DNA. Given the
size difference between human and bacterial genomes,
this human genome “contamination” can become a signif-
icant issue in clinical samples, resulting in a vast majority
of sequence data having to be removed prior to analysis.
So far, the most studied human microbiome is that of
the human intestine. The microbiome of the gastrointes-
tinal tract has been described as an “organ” equivalent in
size to the liver,222 and as the most important connection
humans have with their external environment. The pres-
ence of microbes in the gut can physically alter the intes-
tinal structure, with differences in villus architecture and
lymphoid tissue seen in germ-free mice.223,224 The Human
5 8 2   •   G enomics in C linical P ractice
Microbiome Project (HMP), a NIH-driven project to
comprehensively characterize the human microbiota, is
using NGS methods to undertake four sub-projects:
1. Determination of microbiome diversity using 16S
rRNA sequencing;
2. Demonstration projects identifying links between the
human microbiome and health and disease;
3. Shotgun sequencing to identify microbial genes and
pathways in the microbiome; and
4. Sequencing of ~3000 individual microbes isolated from
the human body to generate a reference collection for
future comparisons.
The projects focused on human health and disease
include analyses of the gut microbiome in relation to
Crohn’s disease, ulcerative colitis, and obesity; analyses of
the cutaneous microbiome in relation to psoriasis, eczema,
and acne; analysis of the neonatal microbiome in relation
to necrotizing enterocolitis; and analysis of the viral biome
in febrile illness (http://​www.ncbi.nlm.nih.gov/biopro-
ject/46305​).
As well as finding associations with disease, these proj-
ects are also highlighting other characteristics of interest
that have implications for trying to control the microbi-
ome to promote health. First, analysis of the gut microbi-
ome has revealed a high level of gene transfer among the
microbes that colonize the gut.225 For example, a peptide/
nickel transport complex that was shown to be associated
with differing levels of host obesity appears to be trans-
ferred between bacteria in a habitat-driven process.226
Second, one of the most notable associations between
a specific species of the human microbiome and disease
was the determination that H.  pylori contributes to the
development of peptic ulcers.227 As well as predisposing
towards peptic ulcers, H.  pylori is also associated with
gastric adenocarcinoma,228 whilst showing a protective
effect in reflux esophagitis and early-onset asthma.229,230
Metagenomic sequencing of the stomach microbiome
has revealed that when H.  pylori is present, it totally
dominates, representing more than 90% of the bacterial
component. When it is absent, there is a high diversity of
microbial organisms that is absent from people infected
with H. pylori.231
Compared to studies on the gut microbiome, there has
been less application of metagenomic sequencing to other
sites of the human body. This is partly due to technical
difficulties in the application of this technique, including
access to samples from other body sites as well as the
contamination with host DNA. Open wounds, such as
those caused by trauma, surgery, or chronic conditions
like diabetic ulcers and venous insufficiency, are another
area that has been studied using NGS.232–234 In chronic
infected tissue, microbes that colonize the damaged tis-
sue elicit an inflammatory immune response, leading
to exacerbation of wound severity and delay in healing.
These complex microbial communities are often observed
in the form of biofilms. A  study in a diabetic mouse
model demonstrated that changes in the wound micro-
biome elicited a change in the expression of host genes
involved in the immune response, with some bacterial
profiles causing impaired healing.235 Although the ben-
eficial aspects of the microbiome are not well elucidated,
in chronic or acute infections there is evidence that some
benign bacteria could promote tissue repair (reviewed
in236). For example, the presence of the benign strain of
Lactobacillus reuteri in surgical wounds prevented abscess
formation by Staphylococcus aureus, suggesting a poten-
tial use of one bacterium to reduce the pathogenicity of
another.237 Related to this, a meta-analysis of the use of
probiotic or synbiotic (probiotic plus prebiotic) use prior
to elective surgery showed a reduction in sepsis post-
operatively after both probiotic and synbiotic use, and
reduced post-operative antibiotic usage after synbiotic
supplementation.238 This illustrates the influence of the
gut microbiome on overall health and points to the possi-
bility of simple nutritional strategies that may be effective
in reducing the incidence of postoperative infection.
THE NEAR FUTURE
The last several years have seen incredible advances in
genomic technologies that have significantly advanced our
understanding of human susceptibility to infectious disease.
As development of these genomic technologies continues,
they promise to offer an even greater understanding of
human variation, not only at the genetic and transcriptomic
levels, but also at the epigenetic, proteomic, and metabolo-
mic levels, giving us a more integrated understanding of the
host’s response to infection. Such endeavors have success-
fully been undertaken by the ENCODE (ENCyclopedia
Of DNA Elements) Project Consortium, which was estab-
lished by NHGRI in 2003 with the aim of creating an
expansive and integrated resource of all functional DNA
elements in the human genome.239
One rapidly growing area of interest is the human epig-
enome, which has become a focus of attention in many
G enomics and I nfectious D iseases   •   5 8 3
studies of human disease. The epigenome describes a series
of chemical modifications (including DNA methyla-
tion and histone modifications) that influence when and
where a gene is transcribed, leading to a change in pheno-
type without a change in DNA sequence. It has become
apparent that the epigenome is highly dynamic and that
it plays a vital role in the response of the host to environ-
mental stimuli. The development of NGS methods and
high-throughput arrays now allow the variation in the
human epigenome to be assessed under differing condi-
tions, allowing the role of the host epigenome in health
and disease to be assessed. For example, NGS methods
combined with antibody-enrichment steps can be used
to specifically target numerous histone modifications and
DNA methylation, so-called CHiP-Seq and MeDIP-Seq.
Several high-throughput arrays have also been launched
in the last few years that allow a more targeted analysis of
genome-wide DNA methylation levels to be assessed and
pave the way for epigenome-wide association studies, or
EWASs. Studies are now underway to advance our under-
standing of the role played by the human epigenome in our
susceptibility to infection and in our response to infection.
For many of these advanced genomic technologies, one
limitation that restricts their transfer to the clinical setting
is the requirement for relatively large volumes of starting
material (DNA or RNA). This requirement can be prohibi-
tively difficult to meet. One advance that will help to over-
come this limitation is the development of NGS methods
that require greatly reduced volumes of starting material,
even down to using a single cell for sequencing. This area
of development has been challenging, as achieving adequate
genome coverage necessitates a pre-amplification step that
can introduce sequence bias and may ultimately reduce the
test’s accuracy. The recent development of a novel ampli-
fication method—multiple annealing and looping-based
amplification cycles (MALBAC)—which appears to offer
high amplification uniformity across 93% of the genome,
will help address this issue.240 This area will be of particular
relevance to analysis of the transcriptome and epigenome,
which will vary on a cell-to-cell basis.
To date, human and pathogen transcriptomes are being
analyzed in isolation from each other, although there are
examples of the hybridisation-based array approach being
applied to both host and pathogen. However, a full under-
standing of the dynamic interactions that occur during
infection will require the simultaneous analysis of both
the host and the pathogen transcriptome. This will allow
the adaptations of both the host and the pathogen to be
fully characterized. NGS technology could help advance
this field, in a process that has been referred to as “dual
RNA-Seq.”241 Again, there are current limitations that
restrict this field that ongoing developments will, hopefully,
start to address. These include differences in RNA content
between the host and pathogen (both in amount and in
RNA species); limitations in starting material, especially of
the pathogen; differences in post-transcriptional modifica-
tions; and knowledge of the sequencing depth required to
provide meaningful data. However, developments are being
made that will address at least some of these issues; for
example, a novel method that allows the unbiased enrich-
ment of pathogen material from clinical samples.242
The development of third-generation sequencing
machines that are cheaper, smaller, easy to use, and capable
of using small volumes of clinical samples directly means it
is likely that sequencing will soon become a diagnostic tool
available to the pathology laboratory, used to identify causal
elements in the genome of both host and pathogen. However,
software development does not currently keep pace with
improvements in sequencing; therefore, the methodologies
used to analyze the data will be a probable limiting factor. In
2001, the $1,000 human genome cost was first raised at a pub-
lic meeting as something that would signify that we are now
able to use genome sequencing for personalized medicine
(“Beyond the Beginning: The Future of Genomics” meeting,
webcast available athttp://www.genome.gov/10001294).
The $1,000 exome is already commercially available
(www.23andme.com/exome), and it is highly likely that we
will soon see the advent of the $1,000 genome or probably
even less.243 It will be interesting to see if this milestone is the
beginning of a new era in clinical medicine.
C O N C LU S I O N S
The genomic era continues to revolutionize the way infec-
tious disease research is being undertaken, both on the
human host and on the infecting pathogen. The gen-
eration of new and exciting discoveries progress at a phe-
nomenal rate, with the real possibility of translational
outcomes in the clinical setting. The ongoing develop-
ment of genome-wide technologies will continue to sup-
port large-scale, high-throughput projects that allow both
human and pathogen genomes and transcriptomes to be
considered in their entirety, and will eventually allow them
to be considered together.
With respect to human susceptibility to infectious dis-
ease, the ongoing development and reducing costs of NGS
sequencing technologies will shortly allow the human
genome to be fully sequenced in large enough sample sizes
within the scope of a research project. In conjunction with
5 8 4   •   G enomics in C linical P ractice
large-scale projects such as the 1KGP, this will allow the role
of rare genetic variation, as well as common variation, in sus-
ceptibility to infectious disease to be determined. Reducing
costs, of both NGS and high-throughput SNP arrays, will
allow much larger cohort sizes (>10,000) to be investigated
and will eventually allow researchers to start untangling the
complex genetic interactions that exist between genes (gene
× gene interactions) and the environment (gene × environ-
ment interactions). Not only will this provide us valuable
information about the pathways involved in susceptibility
to infectious disease, it will hopefully lead us to the devel-
opment of novel therapies, diagnostics, and more targeted
application of appropriate drug therapies.
A decade ago the availability of a representative com-
plete genome sequence for microbial pathogens was seen as
the breakthrough required for completely interpreting the
pathogen genome, resulting in a new era of novel therapies
against infectious diseases. It has become apparent that the
rapid evolutionary rate of microbes, their ability to exchange
genomic material within communities, and the increasing
emergence of novel pathogens means that we must continue
to sequence isolates from all pathogens, not just monitor
outbreaks. Our ability to sequence the microbiome has
also improved our understanding of just how important
our relationship is with the microbial environment. This
knowledge is leading to an increasing emphasis on the ben-
efits of promoting a healthy microbiome to prevent chronic
disease. While the advent of next-generation sequencing
heralded a new era of non–hypothesis driven science, the
third generation of sequencing machines is likely to herald
a new era, one that may make this technology indispensable
in the routine clinical environment.
REFERENCES
1. World Health Organisation. Global health observatory. (WHO,
Geneva, 2008).
2. Murray, C.J., & Acharya, A.K. Understanding DALYs
(disability-adjusted life years). J Health Econ 16, 703–730 (1997).
3. Kwiatkowski, D.P. How malaria has affected the human genome and
what human genetics can teach us about malaria. Am J Hum Genet
77, 171–192 (2005).
4. Barreiro, L.B., et  al. Promoter variation in the DC-SIGN-encoding
gene CD209 is associated with tuberculosis. PLoS Med 3, e20 (2006).
5. Casanova, J.L., & Abel, L. Inborn errors of immunity to infec-
tion: the rule rather than the exception. J Exp Med 202, 197–201
(2005).
6. Casanova, J.L., et  al. From idiopathic infectious diseases to novel
primary immunodeficiencies. J Allergy Clin Immunol 116, 426–430
(2005).
7. Jamieson, S.E., & Peacock, C.S. Genomics and infectious dis-
eases:  susceptibility, resistance, response and anti-microbial ther-
apy. In Genomics and Clinical Medicine (ed. Kumar, D.) (Oxford
University Press, 2006). New York.
 8. Marquet, S., et  al. Genetic localization of a locus controlling the
intensity of infection by Schistosoma mansoni on chromosome
5q31-q33. Nat Genet 14, 181–184 (1996).
9. Bellamy, R., et al. Genetic susceptibility to tuberculosis in Africans: a
genome-wide scan. Proc Natl Acad Sci U S A 97, 8005–8009 (2000).
10. Miller, E.N., et al. Genome-wide scans for leprosy and tuberculosis
susceptibility genes in Brazilians. Genes Immun 5, 63–67 (2004).
11. Siddiqui, M.R., et  al. A major susceptibility locus for leprosy in
India maps to chromosome 10p13. Nat Genet 27, 439–441 (2001).
12. Mira, M.T., et al. Susceptibility to leprosy is associated with PARK2
and PACRG. Nature 427, 636–640 (2004).
13. Sachidanandam, R., et al. A map of human genome sequence varia-
tion containing 1.42  million single nucleotide polymorphisms.
Nature 409, 928–933 (2001).
14. International HapMap Consortium. A haplotype map of the human
genome. Nature 437, 1299–1320 (2005).
15. Cann, H.M., et al. A human genome diversity cell line panel. Science
296, 261–262 (2002).
16. Abecasis, G.R., et al. An integrated map of genetic variation from
1,092 human genomes. Nature 491, 56–65 (2012).
17. 1000 Genomes Project Consortium. A map of human genome vari-
ation from population-scale sequencing. Nature 467, 1061–1073
(2010).
18. Marchini, J., Howie, B., Myers, S., McVean, G., & Donnelly, P. A
new multipoint method for genome-wide association studies by
imputation of genotypes. Nat Genet 39, 906–913 (2007).
19. Hindorff, L.A., et al. Potential etiologic and functional implications
of genome-wide association loci for human diseases and traits. Proc
Natl Acad Sci U S A 106, 9362–9367 (2009).
20. Johnson, A.D., & O’Donnell, C.J. An open access database of
genome-wide association results. BMC Med Genet 10, 6 (2009).
21. Li, M.J., et  al. GWASdb:  a database for human genetic variants
identified by genome-wide association studies. Nucleic Acids Res 40,
D1047–1054 (2012).
22. Wellcome Trust Case Control Consortium. Genome-wide asso-
ciation study of 14,000 cases of seven common diseases and 3,000
shared controls. Nature 447, 661–678 (2007).
23. An, P., & Winkler, C.A. Host genes associated with HIV/
AIDS:  advances in gene discovery. Trends Genet 26, 119–131
(2010).
24. Fellay, J., et al. A whole-genome association study of major determi-
nants for host control of HIV-1. Science 317, 944–947 (2007).
25. Dalmasso, C., et al. Distinct genetic loci control plasma HIV-RNA
and cellular HIV-DNA levels in HIV-1 infection:  the ANRS
Genome Wide Association 01 study. PLoS One 3, e3907 (2008).
26. Le Clerc, S., et al. Screening low-frequency SNPS from genome-wide
association study reveals a new risk allele for progression to AIDS. J
Acquir Immune Defic Syndr 56, 279–284 (2011).
27. Limou, S., et  al. Genomewide association study of an
AIDS-nonprogression cohort emphasizes the role played by HLA
genes (ANRS Genomewide Association Study 02). J Infect Dis 199,
419–426 (2009).
28. Ferreira, M.A., et  al. Quantitative trait loci for CD4:CD8 lym-
phocyte ratio are associated with risk of type 1 diabetes and HIV-1
immune control. Am J Hum Genet 86, 88–92 (2010).
29. Fellay, J., et al. Common genetic variation and the control of HIV-1
in humans. PLoS Genet 5, e1000791 (2009).
30. Kulski, J.K., & Dawkins, R.L. The P5 multicopy gene family in the
MHC is related in sequence to human endogenous retroviruses
HERV-L and HERV-16. Immunogenetics 49, 404–412 (1999).
31. Migueles, S.A., et  al. HLA B*5701 is highly associated with
restriction of virus replication in a subgroup of HIV-infected long
term nonprogressors. Proc Natl Acad Sci U S A 97, 2709–2714
(2000).
32. Altfeld, M., et  al. Influence of HLA-B57 on clinical presenta-
tion and viral control during acute HIV-1 infection. AIDS 17,
2581–2591 (2003).
G enomics and I nfectious D iseases   •   5 8 5
33. Pelak, K., et  al. Host determinants of HIV-1 control in African
Americans. J Infect Dis 201, 1141–1149 (2010).
34. Pereyra, F., et al. The major genetic determinants of HIV-1 control
affect HLA class  I  peptide presentation. Science 330, 1551–1557
(2010).
35. Fellay, J., et al. Host genetic determinants of T cell responses to the
MRKAd5 HIV-1 gag/pol/nef vaccine in the step trial. J Infect Dis
203, 773–779 (2011).
36. Pelak, K., et  al. Copy number variation of KIR genes influences
HIV-1 control. PLoS Biol 9, e1001208 (2011).
37. Adnan, S., et al. Nef interference with HIV-1-specific CTL antiviral
activity is epitope specific. Blood 108, 3414–3419 (2006).
38. Cohen, G.B., et al. The selective downregulation of class I major his-
tocompatibility complex proteins by HIV-1 protects HIV-infected
cells from NK cells. Immunity 10, 661–671 (1999).
39. Goulder, P.J., Bunce, M., Luzzi, G., Phillips, R.E., & McMichael,
A.J. Potential underestimation of HLA-C-restricted cytotoxic
T-lymphocyte responses. AIDS 11, 1884–1886 (1997).
40. Kulkarni, S., et al. Differential microRNA regulation of HLA-C
expression and its association with HIV control. Nature 472,
495–498 (2011).
41. Le Clerc, S., et al. Genomewide association study of a rapid progres-
sion cohort identifies new susceptibility alleles for AIDS (ANRS
Genomewide Association Study 03). J Infect Dis 200, 1194–1201
(2009).
42. Chun, T.W., et  al. Gene expression and viral prodution in
latently infected, resting CD4+ T cells in viremic versus aviremic
HIV-infected individuals. Proc Natl Acad Sci U S A 100, 1908–
1913 (2003).
43. Lambotte, O., et  al. HIV controllers:  a homogeneous group of
HIV-1-infected patients with spontaneous control of viral replica-
tion. Clin Infect Dis 41, 1053–1056 (2005).
44. Siddiqui, R.A., et  al. X chromosomal variation is associated with
slow progression to AIDS in HIV-1-infected women. Am J Hum
Genet 85, 228–239 (2009).
45. Chantarangsu, S., et  al. Genome-wide association study identifies
variations in 6p21.3 associated with nevirapine-induced rash. Clin
Infect Dis 53, 341–348 (2011).
46. Lai, C.L., Ratziu, V., Yuen, M.F., & Poynard, T. Viral hepatitis B.
Lancet 362, 2089–2094 (2003).
47. Ganem, D., & Prince, A.M. Hepatitis B virus infection—natural
history and clinical consequences. N Engl J Med 350, 1118–1129
(2004).
48. Custer, B., et  al. Global epidemiology of hepatitis B virus. J Clin
Gastroenterol 38, S158–168 (2004).
49. Lin, T.M., et  al. Hepatitis B virus markers in Chinese twins.
Anticancer Res 9, 737–741 (1989).
50. Kamatani, Y., et al. A genome-wide association study identifies vari-
ants in the HLA-DP locus associated with chronic hepatitis B in
Asians. Nat Genet 41, 591–595 (2009).
51. Mbarek, H., et al. A genome-wide association study of chronic hepa-
titis B identified novel risk locus in a Japanese population. Hum Mol
Genet 20, 3884–3892 (2011).
52. Png, E., et al. A genome-wide association study of hepatitis B vaccine
response in an Indonesian population reveals multiple independent
risk variants in the HLA region. Hum Mol Genet 20, 3893–3898
(2011).
53. Nishida, N., et al. Genome-wide association study confirming asso-
ciation of HLA-DP with protection against chronic hepatitis B
and viral clearance in Japanese and Korean. PLoS One 7, e39175
(2012).
54. Ray Kim, W. Global epidemiology and burden of hepatitis C.
Microbes Infect 4, 1219–1225 (2002).
55. Liu, C.H., et  al. Pegylated interferon-alpha-2a plus ribavirin for
treatment-naïve Asian patients with hepatitis C virus genotype 1
infection: a multicenter, randomized controlled trial. Clin Infect Dis
47, 1260–1269 (2008).
 56. Yan, K.K., et al. Treatment responses in Asians and Caucasians with
chronic hepatitis C infection. World J Gastroenterol 14, 3416–3420
(2008).
57. McHutchison, J.G., et  al. Peginterferon alfa-2b or alfa-2a with
ribavirin for treatment of hepatitis C infection. N Engl J Med 361,
580–593 (2009).
58. Ge, D., et  al. Genetic variation in IL28B predicts hepatitis C
treatment-induced viral clearance. Nature 461, 399–401 (2009).
59. Wilson, J.F., et  al. Population genetic structure of variable drug
response. Nat Genet 29, 265–269 (2001).
60. Thomas, D.L., et  al. Genetic variation in IL28B and spontaneous
clearance of hepatitis C virus. Nature 461, 798–801 (2009).
61. Tanaka, Y., et al. Genome-wide association of IL28B with response
to pegylated interferon-alpha and ribavirin therapy for chronic hep-
atitis C. Nat Genet 41, 1105–1109 (2009).
62. Suppiah, V., et  al. IL28B is associated with response to chronic
hepatitis C interferon-alpha and ribavirin therapy. Nat Genet 41,
1100–1104 (2009).
63. Rauch, A., et  al. Genetic variation in IL28B is associated with
chronic hepatitis C and treatment failure: a genome-wide associa-
tion study. Gastroenterology 138, 1338–1345, 1345.e1331–1337
(2010).
64. Lange, C.M., et al. Serum ferritin levels are associated with a distinct
phenotype of chronic hepatitis C poorly responding to pegylated
interferon-alpha and ribavirin therapy. Hepatology 55, 1038–1047
(2012).
65. Clark, P.J., et al. Interleukin 28B polymorphisms are the only com-
mon genetic variants associated with low-density lipoprotein cho-
lesterol (LDL-C) in genotype-1 chronic hepatitis C and determine
the association between LDL-C and treatment response. J Viral
Hepat 19, 332–340 (2012).
66. Tanaka, Y., et al. Genome-wide association study identified ITPA/
DDRGK1 variants reflecting thrombocytopenia in pegylated inter-
feron and ribavirin therapy for chronic hepatitis C. Hum Mol Genet
20, 3507–3516 (2011).
67. Thompson, A.J., et  al. Genome-wide association study of
interferon-related cytopenia in chronic hepatitis C patients. J
Hepatol 56, 313–319 (2012).
68. Patin, E., et  al. Genome-wide association study identifies variants
associated with progression of liver fibrosis from HCV infection.
Gastroenterology 143, 1244–1252.e1212 (2012).
69. Ank, N., et al. Lambda interferon (IFN-lambda), a type III IFN, is
induced by viruses and IFNs and displays potent antiviral activity
against select virus infections in vivo. J Virol 80, 4501–4509 (2006).
70. Sirén, J., Pirhonen, J., Julkunen, I., & Matikainen, S. IFN-alpha
regulates TLR-dependent gene expression of IFN-alpha, IFN-beta,
IL-28, and IL-29. J Immunol 174, 1932–1937 (2005).
71. Ank, N., et  al. An important role for type III interferon
(IFN-lambda/IL-28) in TLR-induced antiviral activity. J Immunol
180, 2474–2485 (2008).
72. Sheppard, P., et al. IL-28, IL-29 and their class II cytokine receptor
IL-28R. Nat Immunol 4, 63–68 (2003).
73. Kotenko, S.V., et  al. IFN-lambdas mediate antiviral protection
through a distinct class II cytokine receptor complex. Nat Immunol
4, 69–77 (2003).
74. Marcello, T., et al. Interferons alpha and lambda inhibit hepatitis C
virus replication with distinct signal transduction and gene regula-
tion kinetics. Gastroenterology 131, 1887–1898 (2006).
75. Newport, M.J. Why hasn’t human genetics told us more about
tuberculosis? Int J Tuberc Lung Dis 13, 1049–1050 (2009).
76. Alter, A., Grant, A., Abel, L., Alcaïs, A., & Schurr, E. Leprosy as a
genetic disease. Mamm Genome 22, 19–31 (2011).
77. Thye, T., et al. Genome-wide association analyses identifies a suscep-
tibility locus for tuberculosis on chromosome 18q11.2. Nat Genet
42, 739–741 (2010).
78. Thye, T., et al. Common variants at 11p13 are associated with sus-
ceptibility to tuberculosis. Nat Genet 44, 257–259 (2012).
5 8 6   •   G enomics in C linical P ractice
79. Png, E., et  al. A genome wide association study of pulmonary
tuberculosis susceptibility in Indonesians. BMC Med Genet 13, 5
(2012).
80. Mahasirimongkol, S., et  al. Genome-wide association studies of
tuberculosis in Asians identify distinct at-risk locus for young
tuberculosis. J Hum Genet 57, 363–367 (2012).
81. Zhang, F.R., et  al. Genomewide association study of leprosy. N
Engl J Med 361, 2609–2618 (2009).
82. Zhang, F., et al. Identification of two new loci at IL23R and RAB32
that influence susceptibility to leprosy. Nat Genet 43, 1247–1251
(2011).
83. Wong, S.H., et al. Leprosy and the adaptation of human toll-like
receptor 1. PLoS Pathog 6, e1000979 (2010).
84. Wong, S.H., Hill, A.V., Vannberg, F.O., & I.-A.-U.K.L.G.
Consortium. Genomewide association study of leprosy. N Engl J
Med 362, 1446–1447; author reply 1447–1448 (2010).
85. Murray, C.J., et  al. Global malaria mortality between 1980 and
2010: a systematic analysis. Lancet 379, 413–431 (2012).
86. Mackinnon, M.J., Mwangi, T.W., Snow, R.W., Marsh, K., &
Williams, T.N. Heritability of malaria in Africa. PLoS Med 2, e340
(2005).
87. Malaria Genomic Epidemiology Network. A global network for
investigating the genomic epidemiology of malaria. Nature 456,
732–737 (2008).
88. Jallow, M., et  al. Genome-wide and fine-resolution association
analysis of malaria in West Africa. Nat Genet 41, 657–665 (2009).
89. Timmann, C., et  al. Genome-wide association study indicates
two novel resistance loci for severe malaria. Nature 489, 443–446
(2012).
90. Fakiola, M., et al. Common variants in the HLA-DRB1-HLA-DQA1
HLA class  II region are associated with susceptibility to visceral
leishmaniasis. Nat Genet 45(2), 208–213 (2013).
91. Davila, S., et al. Genome-wide association study identifies variants
in the CFH region associated with host susceptibility to meningo-
coccal disease. Nat Genet 42, 772–776 (2010).
92. Chen, D., et al. Genome-wide association study of HPV seroposi-
tivity. Hum Mol Genet 20, 4714–4723 (2011).
93. Kennedy, R.B., et al. Genome-wide genetic associations with IFNγ
response to smallpox vaccine. Hum Genet 131, 1433–1451 (2012).
94. Kennedy, R.B., et  al. Genome-wide analysis of polymorphisms
associated with cytokine responses in smallpox vaccine recipients.
Hum Genet 131, 1403–1421 (2012).
95. Ovsyannikova, I.G., et  al. Genome-wide association study of
antibody response to smallpox vaccine. Vaccine 30, 4182–4189
(2012).
96. Khor, C.C., et al. Genome-wide association study identifies suscep-
tibility loci for dengue shock syndrome at MICB and PLCE1. Nat
Genet 43, 1139–1141 (2011).
97. Zúñiga, J., et al. Genetic variants associated with severe pneumonia
in A/H1N1 influenza infection. Eur Respir J 39, 604–610 (2012).
98. Zhou, J., et al. A functional variation in CD55 increases the sever-
ity of 2009 pandemic H1N1 influenza A virus infection. J Infect
Dis 206, 495–503 (2012).
99. Zhang, H., et  al. Genome-wide association study identifies
1p36.22 as a new susceptibility locus for hepatocellular carci-
noma in chronic hepatitis B virus carriers. Nat Genet 42, 755–758
(2010).
100. Clifford, R.J., et  al. Genetic variations at loci involved in the
immune response are risk factors for hepatocellular carcinoma.
Hepatology 52, 2034–2043 (2010).
101. Liu, L., et al. A genome-wide association study with DNA pooling
identifies the variant rs11866328 in the GRIN2A gene that affects
disease progression of chronic HBV infection. Viral Immunol 24,
397–402 (2011).
102. Li, S., et al. GWAS identifies novel susceptibility loci on 6p21.32
and 21q21.3 for hepatocellular carcinoma in chronic hepatitis B
virus carriers. PLoS Genet 8, e1002791 (2012).
 103. Chan, K.Y., et al. Genome-wide association study of hepatocellular
carcinoma in Southern Chinese patients with chronic hepatitis B
virus infection. PLoS One 6, e28798 (2011).
104. Tse, K.P., et  al. Genome-wide association study reveals multiple
nasopharyngeal carcinoma-associated loci within the HLA region
at chromosome 6p21.3. Am J Hum Genet 85, 194–203 (2009).
105. Burns, J.C., & Glodé, M.P. Kawasaki syndrome. Lancet 364, 533–
544 (2004).
106. Rowley, A.H. Kawasaki disease:  novel insights into etiology and
genetic susceptibility. Annu Rev Med 62, 69–77 (2011).
107. Burgner, D., et al. A genome-wide association study identifies novel
and functionally related susceptibility loci for Kawasaki disease.
PLoS Genet 5, e1000319 (2009).
108. Tsai, F.J., et  al. Identification of novel susceptibility loci for
Kawasaki disease in a Han Chinese population by a genome-wide
association study. PLoS One 6, e16853 (2011).
109. Kim, J.J., et  al. A genome-wide association analysis reveals 1p31
and 2p13.3 as susceptibility loci for Kawasaki disease. Hum Genet
129, 487–495 (2011).
110. Khor, C.C., et  al. Genome-wide association study identifies
FCGR2A as a susceptibility locus for Kawasaki disease. Nat Genet
43, 1241–1246 (2011).
111. Foxman, E.F., & Iwasaki, A. Genome-virome interactions: examin-
ing the role of common viral infections in complex disease. Nat Rev
Microbiol 9, 254–264 (2011).
112. Smyth, D.J., et  al. A genome-wide association study of non-
synonymous SNPs identifies a type 1 diabetes locus in the
interferon-induced helicase (IFIH1) region. Nat Genet 38, 617–
619 (2006).
113. Nejentsev, S., Walker, N., Riches, D., Egholm, M., & Todd, J.A.
Rare variants of IFIH1, a gene implicated in antiviral responses,
protect against type 1 diabetes. Science 324, 387–389 (2009).
114. Barreiro, L.B., & Quintana-Murci, L. From evolutionary genetics
to human immunology: how selection shapes host defence genes.
Nat Rev Genet 11, 17–30 (2010).
115. Chen, H., et  al. Psoriasis patients are enriched for genetic vari-
ants that protect against HIV-1 disease. PLoS Genet 8, e1002514
(2012).
116. Liu, Y., et al. A genome-wide association study of psoriasis and pso-
riatic arthritis identifies new disease loci. PLoS Genet 4, e1000041
(2008).
117. Asumalahti, K., et al. A candidate gene for psoriasis near HLA-C,
HCR (Pg8), is highly polymorphic with a disease-associated sus-
ceptibility allele. Hum Mol Genet 9, 1533–1542 (2000).
118. Asumalahti, K., et al. Coding haplotype analysis supports HCR as
the putative susceptibility gene for psoriasis at the MHC PSORS1
locus. Hum Mol Genet 11, 589–597 (2002).
119. Barrett, J.C., et  al. Genome-wide association defines more than
30 distinct susceptibility loci for Crohn’s disease. Nat Genet 40,
955–962 (2008).
120. Anderson, C.A., et al. Meta-analysis identifies 29 additional ulcer-
ative colitis risk loci, increasing the number of confirmed associa-
tions to 47. Nat Genet 43, 246–252 (2011).
121. Jostins, L., et al. Host-microbe interactions have shaped the genetic
architecture of inflammatory bowel disease. Nature 491, 119–124
(2012).
122. Liu, H., et  al. Identification of IL18RAP/IL18R1 and IL12B
as leprosy risk genes demonstrates shared pathogenesis between
inflammation and infectious diseases. Am J Hum Genet 91,
935–941 (2012).
123. Lalande, J.D., & Behr, M.A. Mycobacteria in Crohn’s disease: how
innate immune deficiency may result in chronic inflammation.
Expert Rev Clin Immunol 6, 633–641 (2010).
124. Burgner, D., Jamieson, S.E., & Blackwell, J.M. Genetic susceptibil-
ity to infectious diseases: big is beautiful, but will bigger be even
better? Lancet Infect Dis 6, 653–663 (2006).
G enomics and I nfectious D iseases   •   5 8 7
125. Hill, A.V. Evolution, revolution and heresy in the genetics of infec-
tious disease susceptibility. Philos Trans R Soc Lond B Biol Sci 367,
840–849 (2012).
126. de Bakker, P.I., Neale, B.M., & Daly, M.J. Meta-analysis of
genome-wide association studies. Cold Spring Harb Protoc 2010,
pdb.top81 (2010).
127. Manolio, T.A., et al. Finding the missing heritability of complex
diseases. Nature 461, 747–753 (2009).
128. Price, A.L., et al. Principal components analysis corrects for stratifi-
cation in genome-wide association studies. Nat Genet 38, 904–909
(2006).
129. Ott, J., Kamatani, Y., & Lathrop, M. Family-based designs for
genome-wide association studies. Nat Rev Genet 12, 465–474
(2011).
130. Sawcer, S., et al. Genetic risk and a primary role for cell-mediated
immune mechanisms in multiple sclerosis. Nature 476, 214–219
(2011).
131. Teo, Y.Y., Small, K.S., & Kwiatkowski, D.P. Methodological chal-
lenges of genome-wide association analysis in Africa. Nat Rev
Genet 11, 149–160 (2010).
132. Newport, M.J., & Finan, C. Genome-wide association studies and
susceptibility to infectious diseases. Brief Funct Genomics 10, 98–
107 (2011).
133. Ioannidis, J.P., Thomas, G., & Daly, M.J. Validating, augmenting
and refining genome-wide association signals. Nat Rev Genet 10,
318–329 (2009).
134. Quintana-Murci, L., Alcais, A., Abel, L., & Casanova, J.L.
Immunology in natura: clinical, epidemiological and evolution-
ary genetics of infectious diseases. Nat Immunol 8, 1165–1171
(2007).
135. Sabeti, P.C., et  al. Genome-wide detection and characterization
of positive selection in human populations. Nature 449, 913–918
(2007).
136. de Bakker, P.I., & Telenti, A. Infectious diseases not immune to
genome-wide association. Nat Genet 42, 731–732 (2010).
137. Byun, M., et  al. Whole-exome sequencing-based discovery of
STIM1 deficiency in a child with fatal classic Kaposi sarcoma.
J Exp Med 207, 2307–2312 (2010).
138. Liu, L., et  al. Gain-of-function human STAT1 mutations impair
IL-17 immunity and underlie chronic mucocutaneous candidiasis.
J Exp Med 208, 1635–1648 (2011).
139. Bogunovic, D., et  al. Mycobacterial disease and impaired
IFN-gamma immunity in humans with inherited ISG15 defi-
ciency. Science 337, 1684–1688 (2012).
140. Emond, M.J., et al. Exome sequencing of extreme phenotypes iden-
tifies DCTN4 as a modifier of chronic Pseudomonas aeruginosa
infection in cystic fibrosis. Nat Genet 44, 886–889 (2012).
141. Ramilo, O., et  al. Gene expression patterns in blood leukocytes
discriminate patients with acute infections. Blood 109, 2066–2077
(2007).
142. Smith, M.W., et  al. Hepatitis C virus and liver disease:  global
transcriptional profiling and identification of potential markers.
Hepatology 38, 1458–1467 (2003).
143. Chaussabel, D., et  al. Analysis of significance patterns identi-
fies ubiquitous and disease-specific gene-expression signatures
in patient peripheral blood leukocytes. Ann N Y Acad Sci 1062,
146–154 (2005).
144. Haining, W.N., et al. Identification of an evolutionarily conserved
transcriptional signature of CD8 memory differentiation that is
shared by T and B cells. J Immunol 181, 1859–1868 (2008).
145. Monaco, A., et al. Molecular immune signatures of HIV-1 vaccines
in human PBMCs. FEBS Lett 583, 3004–3008 (2009).
146. Ockenhouse, C.F., Bernstein, W.B., Wang, Z., & Vahey, M.T.
Functional genomic relationships in HIV-1 disease revealed by
gene-expression profiling of primary human peripheral blood
mononuclear cells. J Infect Dis 191, 2064–2074 (2005).
 147. Rotger, M., et  al. Genome-wide mRNA expression correlates of
viral control in CD4+ T-cells from HIV-1-infected individuals.
PLoS Pathog 6, e1000781 (2010).
148. Fragoso-Ontiveros, V., et al. Gene expression profiles induced by
E6 from non-European HPV18 variants reveals a differential acti-
vation on cellular processes driving to carcinogenesis. Virology 432,
81–90 (2012).
149. Gao, G., et al. A microRNA expression signature for the prognosis
of oropharyngeal squamous cell carcinoma. Cancer 119(1), 72–80
(2013).
150. Janus, J.R., et  al. Linking expression of FOXM1, CEP55 and
HELLS to tumorigenesis in oropharyngeal squamous cell carci-
noma. Laryngoscope 121, 2598–2603 (2011).
151. Herfs, M., et al. A discrete population of squamocolumnar junc-
tion cells implicated in the pathogenesis of cervical cancer. Proc
Natl Acad Sci U S A 109, 10516–10521 (2012).
152. Schlecht, N.F., et  al. Gene expression profiles in HPV-infected
head and neck cancer. J Pathol 213, 283–293 (2007).
153. Rasmussen, A.L., et  al. Early transcriptional programming links
progression to hepatitis C virus-induced severe liver disease in
transplant patients. Hepatology 56, 17–27 (2012).
154. Sakai, Y., et  al. Common transcriptional signature of
tumor-infiltrating mononuclear inflammatory cells and periph-
eral blood mononuclear cells in hepatocellular carcinoma patients.
Cancer Res 68, 10267–10279 (2008).
155. Berry, M.P., et al. An interferon-inducible neutrophil-driven blood
transcriptional signature in human tuberculosis. Nature 466, 973–
977 (2010).
156. Lesho, E., et al. Transcriptional responses of host peripheral blood
cells to tuberculosis infection. Tuberculosis (Edinb) 91, 390–399
(2011).
157. Liang, J., et al. Genome-wide expression profiling of the response
to linezolid in Mycobacterium tuberculosis. Curr Microbiol 64, 530–
538 (2012).
158. Lu, C., et  al. Novel biomarkers distinguishing active tuberculo-
sis from latent infection identified by gene expression profile of
peripheral blood mononuclear cells. PLoS One 6, e24290 (2011).
159. Tomlinson, G.S., et al. Transcriptional profiling of innate and adap-
tive human immune responses to mycobacteria in the tuberculin
skin test. Eur J Immunol 41, 3253–3260 (2011).
160. Wang, C., et al. Comparative miRNA expression profiles in indi-
viduals with latent and active tuberculosis. PLoS One 6, e25832
(2011).
161. Koth, L.L., et  al. Sarcoidosis blood transcriptome reflects lung
inflammation and overlaps with tuberculosis. Am J Respir Crit
Care Med 184, 1153–1163 (2011).
162. Idaghdour, Y., et  al. Feature Article:  Evidence for additive and
interaction effects of host genotype and infection in malaria. Proc
Natl Acad Sci U S A 109, 16786–16793 (2012).
163. Ioannidis, I., et al. Plasticity and virus specificity of the airway epi-
thelial cell immune response during respiratory virus infection.
J Virol 86, 5422–5436 (2012).
164. Baas, T., et  al. Integrated molecular signature of disease:  analysis
of influenza virus-infected macaques through functional genomics
and proteomics. J Virol 80, 10813–10828 (2006).
165. Josset, L., et al. Gene expression signature-based screening identi-
fies new broadly effective influenza a antivirals. PLoS One 5(2010).
166. Huang, Y., et al. Temporal dynamics of host molecular responses
differentiate symptomatic and asymptomatic Influenza A  infec-
tion. PLoS Genet 7, e1002234 (2011).
167. Bloom, C.I., et al. Detectable changes in the blood transcriptome
are present after two weeks of antituberculosis therapy. PLoS One
7, e46191 (2012).
168. Ozsolak, F., & Milos, P.M. RNA sequencing: advances, challenges
and opportunities. Nat Rev Genet 12, 87–98 (2011).
169. Maretti-Mira, A.C., et  al. Transcriptome patterns from primary
cutaneous Leishmania braziliensis infections associate with
5 8 8   •   G enomics in C linical P ractice
eventual development of mucosal disease in humans. PLoS Neglect
Trop D 6, e1816 (2012).
170. Sanger, F., Nicklen, S., & Coulson, A.R. DNA sequencing with
chain-terminating inhibitors. Proc Natl Acad Sci U S A 74, 5463–
5467 (1977).
171. Fleischmann, R.D., et al. Whole-genome random sequencing and
assembly of Haemophilus influenzae Rd. Science 269, 496–512
(1995).
172. Didelot, X., Bowden, R., Wilson, D.J., Peto, T.E., & Crook,
D.W. Transforming clinical microbiology with bacterial genome
sequencing. Nat Rev Genet 13, 601–612 (2012).
173. Quail, M.A., et al. A tale of three next generation sequencing plat-
forms: comparison of Ion Torrent, Pacific Biosciences and Illumina
MiSeq sequencers. BMC Genomics 13, 341 (2012).
174. Heger, M. In pilot project, FDA places MiSeqs in state and federal
labs to track food-borne pathogens. (www.GenomeWeb.​com​).
175. Parkhill, J., & Wren, B.W. Bacterial epidemiology and biology—
lessons from genome sequencing. Genome Biol 12, 230 (2011).
176. Relman, D.A. Microbial genomics and infectious diseases. N Engl
J Med 365, 347–357 (2011).
177. Lewis, T., et al. High-throughput whole-genome sequencing to dis-
sect the epidemiology of Acinetobacter baumannii isolates from a
hospital outbreak. J Hosp Infect 75, 37–41 (2010).
178. Binnewies, T.T., et  al. Ten years of bacterial genome sequenc-
ing:  comparative-genomics-based discoveries. Funct Integr
Genomics 6, 165–185 (2006).
179. Rohde, H., et  al. Open-source genomic analysis of
Shiga-toxin-producing E. coli O104:H4. N Engl J Med 365, 718–
724 (2011).
180. Brzuszkiewicz, E., et al. Genome sequence analyses of two isolates
from the recent Escherichia coli outbreak in Germany reveal the
emergence of a new pathotype: Entero-Aggregative-Haemorrhagic
Escherichia Coli (EAHEC). Arch Microbiol 193, 883–891 (2011).
181. Hillemann, D., Weizenegger, M., Kubica, T., Richter, E., &
Niemann, S. Use of the genotype MTBDR assay for rapid detec-
tion of rifampin and isoniazid resistance in Mycobacterium tuber-
culosis complex isolates. J Clin Microbiol 43, 3699–3703 (2005).
182. Miotto, P., et  al. Use of genotype MTBDR assay for molecular
detection of rifampin and isoniazid resistance in Mycobacterium
tuberculosis clinical strains isolated in Italy. J Clin Microbiol 44,
2485–2491 (2006).
183. Stabler, R.A., et al. Development and application of the active sur-
veillance of pathogens microarray to monitor bacterial gene flux.
BMC Microbiol 8, 177 (2008).
184. Batchelor, M., et  al. Development of a miniaturised
microarray-based assay for the rapid identification of antimicrobial
resistance genes in gram-negative bacteria. Int J Antimicrob Agents
31, 440–451 (2008).
185. Cleven, B.E., et al. Identification and characterization of bacterial
pathogens causing bloodstream infections by DNA microarray.
J Clin Microbiol 44, 2389–2397 (2006).
186. Hazen, T.C., Rocha, A.M., & Techtmann, S.M. Advances in moni-
toring environmental microbes. Curr Opin Biotechnol 24(3), 526–
533 (2013).
187. Gardy, J.L., et al. Whole-genome sequencing and social-network anal-
ysis of a tuberculosis outbreak. N Engl J Med 364, 730–739 (2011).
188. Lienau, E.K., et  al. Identification of a salmonellosis outbreak by
means of molecular sequencing. N Engl J Med 364, 981–982
(2011).
189. Rasko, D.A., et al. Bacillus anthracis comparative genome analysis
in support of the Amerithrax investigation. Proc Natl Acad Sci U S
A 108, 5027–5032 (2011).
190. Snitkin, E.S., et  al. Tracking a hospital outbreak of
carbapenem-resistant Klebsiella pneumoniae with whole-genome
sequencing. Sci Transl Med 4, 148ra116 (2012).
191. Chin, C.S., et al. The origin of the Haitian cholera outbreak strain.
N Engl J Med 364, 33–42 (2011).
 192. Gilmour, M.W., et al. High-throughput genome sequencing of two
Listeria monocytogenes clinical isolates during a large foodborne
outbreak. BMC Genomics 11, 120 (2010).
193. Grimont, F., & Grimont, P.A. Ribosomal ribonucleic acid gene
restriction patterns as potential taxonomic tools. Ann Inst Pasteur
Microbiol 137B, 165–175 (1986).
194. Maiden, M.C., et  al. Multilocus sequence typing:  a portable
approach to the identification of clones within populations of
pathogenic microorganisms. Proc Natl Acad Sci U S A 95, 3140–
3145 (1998).
195. Cheung, M.K., & Kwan, H.S. Fighting outbreaks with bacte-
rial genomics:  case review and workflow proposal. Public Health
Genomics 15, 341–351 (2012).
196. Wang, D., et al. Viral discovery and sequence recovery using DNA
microarrays. PLoS Biol 1, E2 (2003).
197. Marra, M.A., et al. The genome sequence of the SARS-associated
coronavirus. Science 300, 1399–1404 (2003).
198. Beerenwinkel, N., Gunthard, H.F., Roth, V., & Metzner, K.J.
Challenges and opportunities in estimating viral genetic diversity
from next-generation sequencing data. Front Microbiol 3, 329
(2012).
199. Yang, J., et  al. Unbiased parallel detection of viral pathogens in
clinical samples by use of a metagenomic approach. J Clin Microbiol
49, 3463–3469 (2011).
200. Djikeng, A., et  al. Viral genome sequencing by random priming
methods. BMC Genomics 9, 5 (2008).
201. Quashie, P.K., Mesplede, T., & Wainberg, M.A. Evolution of HIV
integrase resistance mutations. Curr Opin Infect Dis 26, 43–49
(2013).
202. Kagan, R.M., et al. A genotypic test for HIV-1 tropism combin-
ing Sanger sequencing with ultradeep sequencing predicts virologic
response in treatment-experienced patients. PLoS One 7, e46334
(2012).
203. Guell, M., et  al. Transcriptome complexity in a genome-reduced
bacterium. Science 326, 1268–1271 (2009).
204. Sorek, R., & Cossart, P. Prokaryotic transcriptomics: a new view
on regulation, physiology and pathogenicity. Nat Rev Genet 11,
9–16 (2010).
205. Sharma, C.M., et  al. The primary transcriptome of the major
human pathogen Helicobacter pylori. Nature 464, 250–255 (2010).
doi:10.1038/nature08756.
206. Yoder-Himes, D.R., Konstantinidis, K.T., & Tiedje, J.M.
Identification of potential therapeutic targets for Burkholderia
cenocepacia by comparative transcriptomics. PLoS One 5, e8724
(2010).
207. Perkins, T.T., et al. A strand-specific RNA-Seq analysis of the tran-
scriptome of the typhoid bacillus Salmonella typhi. PLoS Genet 5,
e1000569 (2009).
208. Wurtzel, O., et al. Comparative transcriptomics of pathogenic and
non-pathogenic Listeria species. Mol Syst Biol 8, 583 (2012).
209. Cho, I., & Blaser, M.J. The human microbiome: at the interface of
health and disease. Nat Rev Genet 13, 260–270 (2012).
210. Yamaoka, Y. Helicobacter pylori typing as a tool for tracking human
migration. Clin Microbiol Infect 15, 829–834 (2009).
211. Wells, C. Deaths involving Clostridium difficile:  England and
Wales, 2011. (Office for National Statistics, London, 2012).
212. Khoruts, A., Dicksved, J., Jansson, J.K., & Sadowsky, M.J. Changes
in the composition of the human fecal microbiome after bacterio-
therapy for recurrent Clostridium difficile–associated diarrhea. J
Clin Gastroenterol 44, 354–360 (2010).
213. Lawley, T.D., et  al. Targeted restoration of the intestinal micro-
biota with a simple, defined bacteriotherapy resolves relapsing
Clostridium difficile disease in mice. PLoS Pathog 8, e1002995
(2012).
214. Wikoff, W.R., Kalisak, E., Trauger, S., Manchester, M., & Siuzdak,
G. Response and recovery in the plasma metabolome tracks the
G enomics and I nfectious D iseases   •   5 8 9
acute LCMV-induced immune response. J Proteome Res 8, 3578–
3587 (2009).
215. Saad, R., Rizkallah, M.R., & Aziz, R.K. Gut pharmacomicro-
biomics:  the tip of an iceberg of complex interactions between
drugs and gut-associated microbes. Gut Pathog 4, 16 (2012).
216. Qin, J., et al. A human gut microbial gene catalogue established by
metagenomic sequencing. Nature 464, 59–65 (2010).
217. Hugenholtz, P., Goebel, B.M., & Pace, N.R. Impact of
culture-independent studies on the emerging phylogenetic view of
bacterial diversity. J Bacteriol 180, 4765–4774 (1998).
218. Kosakovsky Pond, S., et al. Windshield splatter analysis with the
Galaxy metagenomic pipeline. Genome Res 19, 2144–2153 (2009).
219. Santamaria, M., et  al. Reference databases for taxonomic assign-
ment in metagenomics. Brief Bioinform 13, 682–695 (2012).
220. Acinas, S.G., Marcelino, L.A., Klepac-Ceraj, V., & Polz, M.F.
Divergence and redundancy of 16S rRNA sequences in genomes
with multiple rrn operons. J Bacteriol 186, 2629–2635 (2004).
221. Hong, S., Bunge, J., Leslin, C., Jeon, S., & Epstein, S.S. Polymerase
chain reaction primers miss half of rRNA microbial diversity.
ISME J 3, 1365–1373 (2009).
222. Backhed, F. Host responses to the human microbiome. Nutr Rev
70 Suppl 1, S14–17 (2012).
223. Bouskra, D., et al. Lymphoid tissue genesis induced by commen-
sals through NOD1 regulates intestinal homeostasis. Nature 456,
507–510 (2008).
224. Reinhardt, C., et  al. Tissue factor and PAR1 promote
microbiota-induced intestinal vascular remodelling. Nature 483,
627–631 (2012).
225. Liu, L., et al. The human microbiome: a hot spot of microbial hori-
zontal gene transfer. Genomics 100, 265–270 (2012).
226. Meehan, C.J., & Beiko, R.G. Lateral gene transfer of an ABC
transporter complex between major constituents of the human gut
microbiome. BMC Microbiol 12, 248 (2012).
227. Marshall, B.J., & Warren, J.R. Unidentified curved bacilli in the
stomach of patients with gastritis and peptic ulceration. Lancet 1,
1311–1315 (1984).
228. Parsonnet, J., et al. Helicobacter pylori infection in intestinal- and
diffuse-type gastric adenocarcinomas. J Natl Cancer Inst 83, 640–
643 (1991).
229. Chen, Y., & Blaser, M.J. Inverse associations of Helicobacter pylori
with asthma and allergy. Arch Intern Med 167, 821–827 (2007).
230. el-Serag, H.B., & Sonnenberg, A. Opposing time trends of peptic
ulcer and reflux disease. Gut 43, 327–333 (1998).
231. Andersson, A.F., et al. Comparative analysis of human gut micro-
biota by barcoded pyrosequencing. PLoS One 3, e2836 (2008).
232. Price, L.B., et al. Community analysis of chronic wound bacteria
using 16S rRNA gene-based pyrosequencing:  impact of diabetes
and antibiotics on chronic wound microbiota. PLoS One 4, e6462
(2009).
233. Wolcott, R.D., Gontcharova, V., Sun, Y., & Dowd, S.E. Evaluation
of the bacterial diversity among and within individual venous leg
ulcers using bacterial tag-encoded FLX and titanium amplicon
pyrosequencing and metagenomic approaches. BMC Microbiol 9,
226 (2009).
234. Wolcott, R.D., Gontcharova, V., Sun, Y., Zischakau, A., & Dowd,
S.E. Bacterial diversity in surgical site infections: not just aerobic
cocci any more. J Wound Care 18, 317–323 (2009).
235. Grice, E.A., et al. Longitudinal shift in diabetic wound microbiota
correlates with prolonged skin defense response. Proc Natl Acad Sci
U S A 107, 14799–14804 (2010).
236. Scales, B.S., & Huffnagle, G.B. The microbiome in wound repair
and tissue fibrosis. J Pathol 229, 323–331 (2013).
237. Gan, B.S., Kim, J., Reid, G., Cadieux, P., & Howard, J.C.
Lactobacillus fermentum RC-14 inhibits Staphylococcus aureus
infection of surgical implants in rats. J Infect Dis 185, 1369–1372
(2002).
 238. Kinross, J.M., et al. A meta-analysis of probiotic and synbiotic use
in elective surgery:  does nutrition modulation of the gut micro-
biome improve clinical outcome? JPEN J Parenter Enteral Nutr
(2012).
239. Dunham, I., et al. An integrated encyclopedia of DNA elements in
the human genome. Nature 489, 57–74 (2012).
240. Zong, C., Lu, S., Chapman, A.R., & Xie, X.S. Genome-wide detec-
tion of single-nucleotide and copy-number variations of a single
human cell. Science 338, 1622–1626 (2012).
5 9 0   •   G enomics in C linical P ractice
241. Westermann, A.J., Gorski, S.A., & Vogel, J. Dual RNA-seq of
pathogen and host. Nat Rev Microbiol 10, 618–630 (2012).
242. Azhikina, T., et al. A new technique for obtaining whole patho-
gen transcriptomes from infected host tissues. BioTechniques 48,
139–144 (2010).
243. DeFrancesco, L. Life Technologies promises $1,000 genome. Nat
Biotechnol 30, 126 (2012).
 38.
GENOMIC PER SPECTIVES OF CLINICAL IMMUNOLOGY
Cornelius L. Verweij
INTRODUCTION
The analysis of gene expression in tissues, cells, and biologi-
cal systems has evolved in the last decade from the analysis
of a selected set of genes to an efficient high-throughput,
whole-genome screening approach of (potentially) all genes
expressed in a tissue or cell sample. This development allows
an open-ended survey to identify comprehensively the frac-
tion of genes that defines the samples’ unique biology. This
discovery-based research provides the opportunity to char-
acterize either new genes with unknown function, or genes
not previously known to be involved in a biological process.
The latter category may hold surprises that will sometimes
urge us to redirect our thinking. Besides a more detailed
understanding of basic processes in immunology, microar-
ray studies have increased our insight into the immunologi-
cal mechanisms that contribute to chronic immune-related
diseases. Of particular interest is the overlap of genes
identified in these diseases with genes that are induced in
response to a microbial stimulus. Moreover, genome-wide
studies have provided evidence for the existence of geneti-
cally determined differences in immune activity that might
have consequences for susceptibility and outcome of
immune-related diseases.
Large-scale gene-expression and genome-wide analysis
is of great relevance in the field of immunology to gener-
ating a global view of how the immune systems attacks
invading microorganisms, maintains tolerance, or creates a
“memory” for past infections. The exciting part of microar-
ray studies is that the many data points that are generated
give rise to unpredictable and unexpected results, which
may lead to new insights in immunology. The present chap-
ter reviews genomic research to characterize gene expres-
sion in immunology to broaden our understanding of the
biology of immunological processes and immune-related
diseases.
591
I N N AT E I M MU N I T Y
The innate immune system is central to the type of the
immune response that is mounted in a given situation. This
system acts effectively to sense altered signals without pre-
vious exposure to a pathogen such as a bacterium or virus.
The innate immune system controls and assists the acquired
antigen-specific immune response represented by T and B
lymphocytes, which both carry an infinite number of recep-
tor specificities that need to be exploited when necessary.
The influence of the innate immune system on the adaptive
immune response has long been recognized and used for its
adjuvant effect of (dead) microbes in antibody responses.
Since this system is also intimately linked to inflammatory
and modulatory processes, many human diseases result
from an aberrant primary or secondary innate immunity.
The application of genomics technology provides answers
to basic questions related to the regulation of innate and
adaptive immunity. The extensive experience that immu-
nologists have with cellular aspects of immunology is
highly beneficial in the generation of homogeneous cell
populations for genomics analyses of specific cell subsets.
Within such a framework, we may gain an understanding
of immune deregulation in autoimmune diseases, inflam-
matory responses, and certain tumors.
R E S P O N S E P RO G R A M S O F T H E
I N NAT E I M MU N E S YS T E M TO
M I C RO O RG A N I S M S
The innate immune system distinguishes between “self ” and
“non-self ” by the creation of a variety of pattern-recognition
receptors, called “Toll-like receptors” (TLRs), that rec-
ognize conserved motifs on pathogens that are not found
in higher eukaryotes. Each TLR member recognizes and
responds to different microbial components. Presently,
10 different mammalian TLRs have been recognized,
and their number is likely to increase. For example, lipo-
polysaccharide (LPS) from gram-negative bacteria signals
through TLR4, while the constituents of the cell wall of
gram-negative bacteria, lipoteichoic acid (LTA) and mur-
amyl dipeptide (MDP), both activate TLR2. Unmethylated
CpG-containing bacterial DNA sequences are able to
activate TLR9. Double-stranded RNA (dsRNA) mol-
ecules that are produced during the replication process of
many RNA and DNA viruses mediate an activation signal
through TLR3. Signaling through different TLRs (TLR9,
2, and 4) leads to activation of the same transcription fac-
tors, such as JNK, p38, and NF-kB, which induces the pro-
duction of cytokines, such as TNFα, which, after binding
to its receptor in turn, induces more NF-κB activation and
induces many other genes as well (Beutler 2004).
To gain insight into the host response against different
pathogens and their products, a meta-analysis of 32 stud-
ies has been performed by Jenner and Young ( Jenner and
Young 2005), in which stimulation of macrophages, den-
dritic cells, whole blood, peripheral blood mononuclear
cells (PBMC), T and B cells, and other non-immune cell
types were compared. Several different activators, such as
cytokines, parasites, bacteria, viruses, yeast, and microbial
components that bind to specific TLRs, were used to stimu-
late the different cell types. Based on these comparisons, a
common host-transcriptional response was defined. Several
different pathogens induced a group of genes with a corre-
lated expression profile involved in inflammation, includ-
ing TNFα, IL-1β, IL-6, IL-8, NF-kB family members,
and several chemokines (CCL3/MIP1α, CCL4/MIP1β,
CCL20/MIP3α, CXCL1/GRO1, CXCL2/MIP2α/GRO2,
and CXCL3/MIP2β/GRO3). A  comparison of the differ-
ent components known to activate distinct TLRs, such as
TLR2 by LTA and MDP, TLR3 by Poly I:C, and TLR4
by LPS, revealed that all three TLRs activated this set of
inflammatory/chemotactic genes in macrophages ( Jenner
and Young 2005; Nau et al. 2002). These genes are expressed,
not only in macrophages, but also in dendritic cells (DCs),
leucocytes, PBMCs, and whole blood during acute infec-
tion with several bacteria. In the meta-analysis it was shown
that these genes are induced not only by bacteria, but also
by the protozoan Leishmania chagasi in macrophages, and
by several viruses in peripheral blood cells and other non–
immune cell types such as endothelial and epithelial cells.
This group of co-regulated genes was previously recog-
nized in PBMC stimulated with several bacteria and LPS
(Boldrick et  al. 2002). Because many of these genes are
known to be induced by TNFα, which increases NF-kB
activation, it has been referred to as the “TNF/NF-kB
5 9 2   •   G enomic s in C l inica l P ractice
response set.” The TNF/NF-kB regulon thus resembles
part of the common innate response to infection (Figure
38.1). Interferon-stimulated genes comprise another set of
co-regulated genes included in the common innate response
program. These genes include OAS1 (2′,5′-oligo-adenylate
synthetase 1), OAS2, OASL, MX1, MX2, G1P2/ISG15,
IFI16, ISG20/HEM45, IFIM1, IFN-inducible chemo-
kines CCL8/MCP2, CXCL9/MIG, CXCL10/IP10, and
CXCL11/I-TAC, and metallothioneins. Most of the genes
induced by viruses appear in this cluster. Many cell types
are able to induce this interferon-stimulated gene program,
including macrophages, fibroblasts, PBMC, DCs, B cells,
as well as endothelial and epithelial cells. However, it has
now become clear that other organisms such as bacteria
and fungi are also able to induce an interferon response
program, through activation of the Toll-like receptors and
induction of interferon β, a concept that has been high-
lighted by Hertzog and colleagues (Hertzog, O’Neill, and
Hamilton 2003). But unlike stimulation of TLR3 and
TLR4 by poly I:C and LPS, respectively, the bacterial prod-
ucts LTA and MDP from gram-negative bacteria were not
able to induce an IFN-response program (Nau et al. 2002;
Jenner and Young 2005). Thus type I interferons, IFNα and
IFNβ, are involved in the expression program of many cell
types, in response to parasites, fungi, and bacterial and viral
stimuli that activate different TLRs. Type II interferon,
IFNγ, is produced by T cells, natural killer (NK) cells, and
natural killer-T (NKT) cells in response to mitogens and
cytokines. Interferons regulate many important activities
of macrophages and DCs and are required for an effective
innate immune response, which in turn directs the adaptive
immune response.
A DA P T I VE I M MU N I T Y
Both B- and T-lymphocytes can be recognized as the key
players in the adaptive immune response as a consequence
of their molecular and functional adaptation to invading
pathogens. Within both types of lymphocytes, subtypes
and distinct stages of differentiation exist that are a conse-
quence of their functional adaptation to different antigens.
Differences between T and B cells are reflected by the
differential expression of many genes. Each lymphocyte
type has its specific gene-expression signature. The T cell
signature includes CD2, TCR, and genes encoding TCR
signaling proteins (LAT and fyn). The B cell signature con-
tains genes like CD20, immunoglobulin genes, and genes
encoding BCR components, such as CD79B. Although
these two cell types are clearly distinct, they also share
 Innate immune response Adaptive immune response

Virus, dsRNA Bacteria, LPS

TLR4

TLR3
MyD88 DC
IRAK cytoplasm
TRAM
TRIF TRAF6
B cell
? IKKγ
IKKα IKKβ T cell

NF-kB p50
IRF3/7 IkBs
p65 IFNs
–
IL-12
+
P P P
IRF3/7
Nucleus
p50 p65

IFNα IL-1 IL-6 Th1 Th2

IFNβ TNFa, IL-12
IFNγ/STAT1 IL-4
IRF1 and 7 IL-13
Perforin/GranzymB IL-10R
Differential gene expression after IgG IgE, IgG4
TLR stimulation by pathogens
anti-viral, anti-tumor growth Autoimmunity Allergy

Figure 38.1 
Ligation of TLR3 and TLR4 activate the IFN- and NF-κB pathway. TLR3 stimulation predominantly induces phosphorylation of
interferon regulatory factor 3 (IRF3) and IRF7, which activate type I interferon genes. TLR4 stimulation predominantly induces the release
of inhibitory IκBs from NF-κB, allowing phosphorylation and translocation of NF-κB to the nucleus, where target genes are activated. TLR3
and TLR4 both activate the IFN- and NF-κB pathways; however, TLR4 stimulation leads to a more profound activation of the NF-kB pathway
with production of proinflammatory cytokines ( Jenner and Young 2005). The innate immune response profoundly affects the adaptive/specific
immune response. In particular, the production of IFNs and IL-12 leads to differentiation of T helper cells to Th1 cells, while IL-4 leads to Th2
differentiation. Gene-expression profiling has identified specific gene-expression profiles of these differentiated T cells that explain their function in
autoimmunity and allergy (Rogge et al. 2000, Chtanova et al. 2001, Hamalainen et al. 2001).

remarkable similarities in their activation program. The
“activation gene-expression signature” of T lymphocytes
and B lymphocytes includes many genes that are either
induced or repressed upon activation (Alizadeh et  al.
2000, Teague et al. 1999a, Glynne et al. 2000). Central to
the activation status is the expression of the transcription
factor NF-kB. NF-kB factors play a central role in immu-
nity, and members of this family (NF-kB1, NF-kB2,
RelB, c-Rel, and IkBα) are included in the activation
gene-expression signature together with a prominent
group of target genes, consisting of cytokines, chemo-
kines, and anti-apoptotic genes like A1 (an anti-apoptotic
bcl-2 family member) and c-IAP2. Differences that exist
are at the level of, for instance, CCR7, CXCR5, and
BCL-2 expression, which are expressed at much higher
levels in activated B cells. The elevated expression of che-
mokine receptors is likely to be involved in lymphocyte
homing of activated B cells. One of the features of acti-
vated T cells is the repressed expression of the genes for
interferon receptors and for STAT-1, which is involved in
interferon signal transduction (Teague et al. 1999b).
T LY M P H O C Y T E S
A nice example of the way microarray studies may help
address basic questions in immunology is revealed by
a study that aimed to unravel the pathway of CD28
co-stimulation in T cells. Co-stimulation of T cells via
both the TCR and CD28 enables sustained activation
and cell cycle entry. Diehn and colleagues (Diehn et  al.
2002) have used human peripheral blood T cells that were
stimulated with antibodies to CD3 and/or CD28. The data
showed that signals induced by anti-CD3 were amplified
by co-stimulation with anti-CD28, with the most striking

G enomic P er s pective s o f C l inica l I mmuno l ogy   •   5 9 3

enhancement of transcriptional activity of genes that are
responsive to the transcription factors NFAT. This effect
could be blocked by FK506, an inhibitor of the phospha-
tase calcineurin, thereby interfering with the nuclear trans-
location of NFAT proteins. Conventional biochemical
techniques were then used to demonstrate that stimulation
of CD28 leads to reduced nuclear export of NFATc pro-
tein caused by increased phosphorylation (and thus inac-
tivation) of the nuclear export kinase glycogen synthase
kinase-3 (GSK3). Even in the absence of CD3 stimulation,
engagement of CD28 stimulated the phosphorylation of
GSK3, which becomes more effective when NFAT is acti-
vated by CD3 stimulation. Without the use of microar-
rays, the NFAT-dependent pathway induced by CD28
co-stimulation would not have been recognized easily.
Among effector/memory T helper (Th) cells, a distinc-
tion can be made based on the pattern of secreted proteins,
dividing them into Th1 cells (expressing IFNγ and TNFα)
involved in phagocyte-dependent immune response, and
Th2 cells (expressing IL-4, IL-5 and IL-13) that promote
IgE production and eosinophil function. A  dysbalance in
Th differentiation may lead to pathology. Th1 cells may
be held responsible for tissue damage in inflammatory dis-
eases such as rheumatoid arthritis and inflammatory bowel
disease. Th2 cells play a role in the pathogenesis of allergic
reactions and asthma (Figure 38.1). Studies to gain insight
into the program that effects the creation and intercon-
version of Th1 and Th2 cells are of importance to define
molecular markers to distinguish these subsets in a clini-
cal setting and to design strategies to selectively silence or
reverse a detrimental Th cell subset. Many genes were dis-
covered that were differentially expressed in Th1 and Th2
cells, including transcription factors, cytokines and recep-
tors, co-stimulatory molecules, apoptosis-related genes,
genes involved in migration and adhesion, and genes with
unknown functions (Rogge et  al. 2000, Chtanova et  al.
2001, Hamalainen et al. 2001). Studies performed on fully
polarized human Th1 and Th2 cells at different time points
of stimulation identified a set of approximately 100 dif-
ferentially expressed genes that encode for lytic proteins,
co-stimulatory membrane molecules, cytokines, chemo-
kines, and genes indicating differential signaling, transcrip-
tional control, and cell differentation. The identification of
these differentially expressed genes provides a starting point
for many future studies (Burska et al. 2014).
B LY M P H O C Y T E S
Gene-expression profiling studies have also provided
a molecular definition of different phases in B cell
5 9 4   •   G enomic s in C l inica l P ractice
differentiation. The characteristic signature of B cells in the
anergic state contains genes that have a negative regulatory
function (Glynne et  al. 2000). On the other hand, B cell
activation–specific genes, including the proto-oncogenes
LSIRF/IRF4 and c-myc, the anti-apoptotic molecule A1,
and the inhibitory transcription factor LKLF, implicated
in lymphocyte mitogenesis, were blocked in anergic cells.
This gene profile may explain how tolerance is maintained
in B cells.
Germinal-center B cells differentiate into plasma cells or
memory B cells, while undergoing affinity maturation and
somatic mutation. Germinal-center B cells and plasma cells
show distinct expression signatures (Shaffer et  al. 2001).
Plasma cells show a higher expression of plasmacytic differen-
tiation regulators blimp-1 and XBP-1. Germinal center B cells
show an expression signature that interestingly is retained in
lymphoma cell lines (Alizadeh et al. 2000; also, see below).
This latter profile can be explained by the high expression of
BCL-6, because the dominant negative form of BCL-6 intro-
duced into GC-derived B cell lines is able to block terminal
differentiation into plasma cells (Shaffer et al. 2001).
M I C R OA R R AY S T U D I E S I N
I M MU N O L O G I C A L D I S E A S E S
Information about the differences in gene expression
between cells or tissues of healthy and diseased individuals
provides markers to differentiate between health and dis-
ease, and disease states. In combination with biochemical
research, the expression profiles may provide clear insights
into the disease mechanisms.
I N FEC T I O US D I S E A S E S
Bacteria and Viruses
In vivo analyses of gene-expression changes during viral
infections have been mainly carried out in animal stud-
ies. The response of nonhuman primates to viruses such
as smallpox and SIV/HIV chimeric virus (SHIV) are
quite informative. Rubins and colleagues (Rubins et  al.
2004) found a prominent upregulation of genes that repre-
sent an IFN response, cell cycle/cell proliferation, and genes
that encode for immunoglobulins. Strikingly, a TNFα/
NF-kB–induced response was lacking, while these genes
were induced during infection with another fatal systemic
virus, Ebola. Possibly, this response was modulated by the
smallpox virus, as they are known to encode TNFα receptor
homologues such as CrmB.
 Bosinger and colleagues (Bosinger et  al. 2004)  used
SHIV, which is persistent and causes a permanent decrease
in CD4+ cell counts in macaques. The SHIV challenge
induced a suppression of genes regulating innate immu-
nity, including the LPS receptors, CD14 and TLR4. This
may impair the host response to opportunistic infections
such as gram-negative bacteria, cytomegalovirus (CMV),
and Candida albicans. Genes that were persistently and
most strongly upregulated during infection were the
IFN-response genes, such as MX1, MX2, OAS1, OAS2,
OAS3, ISG20, ISG15, STAT1, IFIT1, IFITM1, and
CXCL10. Like the smallpox virus, the SHIV virus also
failed to induce a TNFα/NF-kB response. Because, in addi-
tion to the IFN response, TNFα contributes importantly
to the antiviral response (Guidotti and Chisari 2001), these
viruses may evade the host immune response, which allows
them to persist.
During HIV infection, B cells are indirectly affected
as well, indicated by hypergammaglobulinemia, increased
expression of activation markers and terminal differentia-
tion, and increased susceptibility to apoptosis. Interestingly,
in HIV-viremic patients, B cells show an increased expres-
sion of IFN-responsive genes such as MX1, ISG15, and
OAS1 (Moir et  al. 2004). These genes partly overlapped
with IFN-response genes that are over-expressed in approxi-
mately 50% of systemic lupus erythematosus (SLE) patients
(Bennett et al. 2003), lupus being a disease in which B cell
perturbations are prominent as well. In addition, increased
expression of CD95, the Fas receptor, (known to be upregu-
lated by IFN) was associated with increased susceptibility
to CD95-mediated apoptosis. In line with this observation,
an inverse correlation of total numbers of B cells and HIV
plasma viremia was found.
In peripheral blood cells from chronic hepatitis C
patients, an enhanced expression of IFNα and IFNβ mRNA
has been found, and induction of the IFN-response gene
MX1 as well (Mihm et al. 2004). Unfortunately, microar-
ray studies on chronic in vivo bacterial infections are lack-
ing. It would be interesting to verify whether the in vitro
response to bacteria reflects the in vivo response. For HIV, it
is clear that the in vitro response (Izmailova et al. 2003) acti-
vates the same pathways as the in vivo response measured in
peripheral blood (Bosinger et al. 2004)
Parasites
Usage of a combination of murine cDNA microarrays,
cytokine-deficient mice, and confirmatory biological
assays led to the identification of gene-expression profiles
that associate with lethal type 1 (IL-12/IFN-γ) and type 2
G enomic P er s pective s o f C l inica l I mmuno l ogy   •   5 9 5
(IL-4) polarized immune responses during infection with
the parasitic trematode Schistosoma mansomi. These results
uncovered the contributions of previously unappreciated
disease mechanisms that contribute to parasite-induced
pathogenesis. The granuloma formation around the schis-
tosomal eggs in the livers differed, not in size, but in quali-
tative aspects revealed by gene-expression levels of infected
livers. In the type 2 polarized animals, increased expression
of genes involved in collagen deposition, wound repair,
and matrix remodeling genes, and fibrosis associated with
increased morbidity were observed. The type 1 polarized
animals showed increased expression of proinflammatory
cytokine genes, and genes involved in apoptosis in associa-
tion with increased mortality.
The reciprocal approach to host–pathogen interactions
is designed to gain a better understanding of pathogen
responses to a host contact upon infection. These experi-
ments require the genome sequence of the pathogen and
microarrays containing gene sequences of the respective
pathogenic microorganisms. The results of this pathogen
genome-mining approach may yield the identification of
potential vaccine candidates, as was exemplified by studies
using N. meningitidis group B bacteria (Pizza et al. 2000).
AU TO I M MU N E D I S E A S E S
Systemic Lupus Erythematosus (SLE)
Another study revealed remarkable results upon
gene-expression profiling within SLE. A characteristic fea-
ture of the inflammatory autoimmune disease SLE is the
heterogeneity in its clinical presentation, therapy respon-
siveness, and severity. Baechler and colleagues (Baechler
et  al. 2003)  determined the mRNA expression profiles
of unstimulated PBMC from 48 SLE patients and con-
trols. Distinct gene-expression patterns, consisting of 161
genes, were identified for patients and controls. Strikingly,
in about 50% of the patients, an upregulated expression
of IFN-response genes, including enhanced expression of
mRNA for STAT-1, MX1, and ISG15, was observed. In
addition, this pathway was associated with more severe
disease involving the kidneys, hematopoetic cells, and/or
the central nervous system. However, mRNA levels of the
IFNs themselves were not different between patients and
controls, which probably indicates that the IFN-response
genes are a more sensitive readout for the activation of this
pathway; or, alternatively, that another cytokine or media-
tor induces the expression of these genes.
The over-expression of IFN-response genes in SLE
patients has been confirmed by others (Bennett et al. 2003).
In addition, a granulocyte-specific gene response associ-
ated with increased numbers of immature granulocytes in
peripheral blood of SLE patients was found.
The IFN-response genes most likely reflect the activity
of type I interferons, (possibly IFNα) after comparison of
the altered gene-expression of PBMC in response to in vitro
IFNγ or IFNα treatment (Kirou et  al. 2004). Analysis of
a selection of three IFNα-induced genes by real-time poly-
merase chain reaction (PCR) revealed a higher expression
in a panel of SLE patients, compared to healthy controls
and rheumatoid arthritis (RA) patients (Kirou et al. 2005).
In the SLE patients, a higher expression of IFNα-induced
genes was associated with a higher prevalence of renal dis-
ease and disease activity scores; lower C3 levels, hemoglo-
bin, and albumin levels; and higher anti-double-stranded
DNA titers and erythrocyte sedimentation rate.
Rheumatoid Arthritis (RA)
To generate a molecular description of synovial tissue
from RA patients that would allow us to unravel novel
aspects of pathogenesis and identify different forms of dis-
ease, we profiled gene expression in synovial tissue from
affected joint tissues of patients with RA (n = 21) using the
“Lymphochip” and microarrays containing approximately
24,000 cDNA spots, manufactured at Stanford University
(van der Pouw Kraan TC et al. 2003a; van der Pouw Kraan
TC et  al. 2003b). Unsupervised hierarchical clustering of
the RA gene-expression signatures revealed considerable
variability within the RA tissues, resulting in the identifica-
tion of at least two molecularly distinct forms of RA tissues.
The first group revealed abundant expression of clusters of
genes indicative of an involvement of the adaptive immune
response, with evidence for a prominent role of an activated
STAT-1 pathway, including STAT-1 signaling receptors and
STAT-1 target genes, among these STAT-1 itself (Lehtonen,
Matikainen, and Julkunen 1997), MMPs, GBP1, ICSBP,
IP-10, Caspase-1, TAP-1, and IRF-1. Among the genes
overexpressed in this group of patients, a relative majority
of nine genes is located on chromosome 6p21.3, which har-
bors the HLA genes and is responsible for one-third of the
genetic influence on the development of rheumatoid arthri-
tis. The expression profiles of the second group of RA tis-
sues revealed an increased tissue-remodeling activity and a
low inflammatory gene-expression signature, which is asso-
ciated with fibroblast de-differentiation.
In addition, fibroblast-like synoviocytes (FLS) were
cultured from high and low inflammatory RA synovial tis-
sues, and their expression profile was examined on Stanford
microarrays with a complexity of 24,000 cDNA spots.
5 9 6   •   G enomic s in C l inica l P ractice
A most intriguing finding was that the expression profiles
of FLS show the imprint of the tissues they were derived
from. FLS that are related to tissues with a low inflamma-
tion profile are characterized by increased expression of
IGF2 and IGFBP5, indicative of an autocrine signaling of
the IGF receptor, which supports FLS proliferation and
survival. FLS from high-inflammatory tissues exhibit a
gene-expression program that is indicative of an activated
TGFβ/activin-induced pathway. A  hallmark of this path-
way is the induction of smooth-muscle actin, a myofibro-
blast marker. The presence of myofibroblasts in RA synovial
tissues was confirmed by immunohistochemical staining of
smooth-muscle actin. Myofibroblasts are involved in tissue
repair and wound healing, but also contribute to the inflam-
matory process by the secretion of cytokines and chemo-
kines e.g. IL-8, IL-6, and CXCL12. In conclusion, these
results confirmed the heterogeneous nature of rheumatoid
arthritis and suggested the existence of distinct pathogenic
mechanisms that contribute to RA, as was suggested by
Firestein and Zvaifler (Firestein and Zvaifler 2002).
Gene-expression profiling in whole blood cells from RA
patients revealed significant variation between RA patients
that allows stratification of patients on the basis of distinct
molecular characteristics (Burska et al. 2014). Interestingly,
in a subset of patients, an expression profile was induced
that showed similarities to a profile observed after micro-
bial infections, suggesting that an infectious agent may be
involved in the pathogenesis of RA.
Differences in gene expression were also observed
between early arthritis (disease duration less than two
years), and established RA patients (with an average dis-
ease duration of 10 years; Olsen et al. 2004). Analysis of the
expression of 4,300 genes in PBMC revealed that in early
RA genes, such as colony-stimulating factor 3 receptor,
cleavage stimulation factor, and TGFβ receptor II, which
affect B cell function, are expressed at higher levels than
during established disease. Genes involved in immune and
inflammatory processes and genes related to cell prolifera-
tion and neoplasia were expressed at late stages of disease.
Comparison between the genes that were expressed at
higher levels in early RA and influenza-induced genes indi-
cated that about a quarter of the early arthritis genes over-
lapped with the influenza-induced genes. This finding led
the authors to suggest that the early arthritis signature may
partly reflect the response to an unknown infectious agent.
The differences in expression profiles between sub-
groups of RA patients confirm the heterogeneous character
of the disease and provide opportunities to stratify patients
based on molecular criteria for clinical studies and evalua-
tion of targeted therapies (see also Chapter 29).
 Multiple Sclerosis (MS)
Only a limited number of studies are reported on micro-
array analysis of MS lesions obtained at autopsy. Analysis
of affected brain tissues from four MS patients compared
to two controls revealed increased expression of transcripts
encoding inflammatory cytokines such as IL-6, IL-17, and
IFN-γ, cytokine receptors IL-1R, IL-18Rtype2, IL-11Rα,
and TNFRp75, and downstream signaling pathways (Lock
et  al. 2002). Comparison of the gene-expression profiles
of active or acute lesions with silent lesions revealed many
genes whose transcripts were at least two-fold differentially
expressed. Among these genes were included those encoding
FGF-12, IL-17, and the interferon-induced 17-kD/15-kD
protein mRNA (see also Chapter 32).
Achiron and colleagues (2004) identified a unique
gene-expression signature of 1,109 genes in PBMC from
26 MS patients in comparison to 18 healthy subjects. The
signature contains genes that implicate processes such as
T-cell activation, inflammation, and apoptosis. Among
the upregulated genes were lymphotoxin β (LTB); sev-
eral integrins, which may play a role in adhesion to vas-
cular endothelial adhesion structures, permitting their
penetrance through the blood–brain barrier; cathepsin
K and B; and anti-apoptotic genes. Interestingly, genes
involved in the TNFα/ NF-κB pathway, including TNFα
itself, TNFα-induced protein 6, TNF receptor-associated
factor 6 (TRAF6), NF-κB1, NF-κB2, and NF-κBIA, were
reduced in multiple sclerosis (MS) patients.
Multi-Immune Disorders
Maas and colleagues (2002) compared the expression of
4,300 genes in PBMC of control individuals (n = 9) before
and after immunization with influenza vaccine to those of
individuals with four distinct autoimmune diseases: rheu-
matoid arthritis (RA, n  =  20), systemic lupus erythema-
tosus (SLE, n = 24), type I diabetes, and multiple sclerosis
(MS, n = 5; Maas et al. 2002). The induced gene-expression
profile of the normal immune response exhibited coordi-
nate changes in expression of genes with related functions
over time. Three days after immunization, genes encoding
proteins involved in key signal-transduction pathways (e.g.,
PKC, phospholipase C, 1,2-DAG kinase, MAPK, STATs
and STAT inhibitors, AP-1, IFN regulatory proteins, and
cell cycle proteins) were expressed. After three weeks, the
program had shifted toward functional activity (chemo-
kines, complement components and IFN-inducible pro-
teins, and leukocyte homing/adhesion molecules). During
the entire course of the immune response to influenza,
G enomic P er s pective s o f C l inica l I mmuno l ogy   •   5 9 7
genes encoding ribosomal proteins were suppressed, a pro-
cess that has been linked to differentiation of eukaryotic
cells. The autoimmune patients displayed a remarkably
similar gene-expression profile, which was unrelated to the
pattern of the immunized group. Two gene clusters were
differentially expressed between patients with an autoim-
mune disease and the controls. Surprisingly, genes with a
distinct expression pattern in autoimmunity were not nec-
essarily “immune response” genes, but were downregulated
genes that encode proteins involved in, for instance, apop-
tosis (such as TRADD, TRAP1, TRIP, TRAF2, CASP6,
and CASP8), ubiquitin/protesome function (UBE2M,
UBE2G2, and POH1), inhibitors of cell cycle progres-
sion (CDKN1B, CDKN2A, and BRCA1), and inducers
of cell differentiation (LIF and CD24). No genes could be
selected that distinguished among the different autoim-
mune diseases.
Cancer
Microarrays have been particularly useful in the classi-
fication of cancers. Alizadeh and colleagues (Alizadeh
et  al. 2000)  performed a systematic survey to define the
gene-expression programs of normal lymphocytes in dif-
ferent stages of differentiation and determined the extent
to which these programs are inherited by human lymphoid
malignancies. The idea was to extend markers for deter-
mining the cellular origins of lymphomas, which so far
had relied on the analysis of somatic hypermutation in the
variable-regions. Such a reference database provides a rich
framework allowing interpretation of the gene-expression
signature in lymphoid malignancies. The knowledge gained
from the assembly of a reference database was applied to
classify diffuse large B-cell lymphoma (DLBCL), a clini-
cally heterozygous disease in terms of treatment response
and subsequent survival, for which no good diagnostic
parameters are available. Unsupervised hierarchical cluster-
ing performed on expression data of lymph node biopsies
from DLBCL patients revealed two groups of patients: one
group showed characteristics of the gene-expression profile
of germinal center B cells, while the other showed expres-
sion profiles characteristic of activated peripheral blood B
cells. This distinction turned out to be a good predictor for
survival. The five-year survival probability for the germinal
center B–like group was 70%, whereas only 20% of the acti-
vated B–like DLBCL patients survived for five years.
Golub and colleagues (Golub et al. 1999) clustered 38
tumor samples by creating self-organizing maps (SOM)
with a preset number of two or four clusters (groups of
patients) for new class discovery. Class predictors were
subsequently identified from these SOMs. For testing the
class predictors consisting of various numbers of genes (10–
100), the one with the highest cross-validation accuracy
rate was used. A set of known acute lymphocytic leukemia
(ALL) and acute myeloid leukemia (AML) tumor samples
was used to test how well this class-prediction method per-
formed (Golub et  al. 1999). With two exceptions, all 38
samples were correctly grouped into AML and ALL, while
the distinction between B-cell and T-cell ALL was made as
well.
Others have identified sets of genes that predict the
clinical outcome of breast cancer patients. The author’s unit
has used another strategy to classify breast tumors accord-
ing to their clinical behavior (van ‘t Veer et al. 2002). A total
of 78 sporadic lymph-node–negative patients were selected
to identify a prognostic gene-expression signature for clini-
cal outcome:  metastasis or disease-free status within five
years. From the 25,000 genes on the arrays, a selection was
made of 5,000 genes that showed a set threshold of varia-
tion between the samples. A supervised method was then
applied to these genes for the identification of the genes
that are the best predictors for clinical outcome. A list of
70 genes turned out to be the best predictors for disease
development, with 83% accuracy. So only a small group
of genes within a large pool that showed inter-individual
variation in expression determines one important aspect
of disease; namely, metastasis. In a follow-up study (van de
Vijver et al. 2002), expression analysis of these 70 genes in
biopsies from 295 patients again proved to have prognostic
value. From the patients with a poor-prognosis profile, only
50.6% remained free of metastasis over a period of 10 years,
while 85.2 % of the patients with a good-prognosis profile
remained free of metastasis (see also Chapter 36).
G E N ET I C A N D G E N O M I C
VA R I AT I O N I N I M MU N E
AC T I VI T Y
Many immune-related diseases have a genetic component
that contributes to susceptibility and/or disease severity.
Part of the genetic basis of disease may be explained by
genetically determined alterations in the immune response
in individuals. In particular, the often observed disease asso-
ciation with HLA alleles supports this concept. Besides
HLA, there might be other, partly subtle genetic varia-
tions that could dramatically influence the outcome of an
immune response.
Since the innate immune system initiates the host’s
defense against pathogens and is required for an effective
5 9 8   •   G enomic s in C l inica l P ractice
adaptive immune response, genetically determined differ-
ences in host-effector programs against pathogens are likely
to affect the control of infectious diseases, autoimmune dis-
eases, and cancer. Accordingly, Westendorp and colleagues
have provided proof for genetically determined differences
in IL-10 and TNF production capacity that could have dra-
matic clinical consequences (Westendorp et al. 1997). The
assembly of a database containing expression signatures of
various immune cells in different phases of activation and
differentiation would be a valuable resource for elucidating
inter-individual differences in gene expression.
A detailed study on inter-individual differences in
unstimulated blood cell gene-expression profiles has been
reported by Whitney and colleagues (Whitney et al. 2003).
The genes that show intrinsic differences in gene expression
in 75 individuals were male- and female-specific genes, genes
involved in antigen presentation such as MHC class II genes
and TAP-2, and IFN-responsive genes. The most consistent
donor-dependent gene of all was DDX17, an RNA helicase
that is suggested to function as a pre-mRNA splicing regula-
tor (Lee 2002). Possibly, differential expression of DDX17
gives rise to inter-individual differences in splice variants.
The inter-individual variation in interferon-induced genes
was also observed by Radich and colleagues (Radich et al.
2004). In their study, gene-expression profiling of periph-
eral blood lymphocytes from unrelated individuals over a
period of several weeks to months also revealed a high level
of variation between different individuals, in particular in
interferon-type I  (IFNα and β) response genes, including
CIG5, IFIT1, IFIT4, MX1, and USP18. Approximately
10–20% of the individuals showed a higher baseline level.
Interestingly, the inducibility of these genes to in vitro inter-
feron stimulation was lower in those individuals.
These observations are in line with the inter-individual
differences in gene expression reported by Cobb and col-
leagues (Cobb et  al. 2005). Gene-expression levels of
unstimulated peripheral blood cells showed a remarkable
constant profile within the same individual over a 24-hour
period. The average difference from the mean expres-
sion values was less than 10%. In contrast, expression val-
ues between different donors (inter-individual variation)
showed a greater variation than the variation within the
same individual. As expected, traumatic injury induced
major changes in gene expression with a magnitude greater
than the inter-individual variance.
Accordingly, Boldrick and colleagues observed
inter-individual differences in response to some bacteria
and their products (Boldrick et al. 2002). Part of the induc-
tive response was repeatedly reduced in one donor. Most of
the genes that showed a poor induction in this donor fit in
a group of IFN/STAT-1–responsive genes, including some
genes that are important for antiviral responses, such as
OAS1 and MX1 (Figure 38.1). Taking into account the fact
that donors with an increased baseline IFN-response gene
program show a reduced induction of these genes upon in
vitro stimulation with IFN (Radich et al. 2004), the poor
response by the individual may reflect high baseline levels
of IFN-response genes.
The results from these studies indicate that the indi-
vidual gene-expression levels in peripheral blood cells are
a stable feature over time, while differences between indi-
viduals may reflect genetic variation, although subclinical
infections cannot be ruled out as a cause. In addition, dur-
ing clinically manifest disease, a disruption of immunologi-
cal homeostasis is detected in the gene-expression program
in peripheral blood cells with a variation that is larger than
the inter-individual differences, and indicates processes
induced by infection, malignancies, or trauma.
The assumption that gene-expression levels are (in
part) genetically determined has been verified by Cheung
and Morley (Cheung et  al. 2003, Morley et  al. 2004).
They established that the inter-individual differences in
gene-expression levels indeed carry a genetic component
(Cheung et  al. 2003). This conclusion was based on the
analysis of immortalized human B cells from unrelated
individuals, siblings from the same family, and between
monozygotic twins. In a follow-up study, this genetic com-
ponent was analyzed in more detail (Morley et  al. 2004).
Baseline gene-expression levels in immortalized B cell lines
from 94 unrelated individuals from 14 families from which
genotypes for genetic markers (SNP genotypes) were avail-
able. Linkage analysis was performed on genes that varied
more between individuals than between replicates from
the same individual (3,554 genes out of a total of 8,500
genes). This analysis resulted in the discovery of genes that
are linked to markers (potential regulators of transcription)
within 5 Mb of the gene (cis-acting transcriptional regula-
tors) and markers more distant than 5 Mb from the target
gene or markers on a different chromosome (trans-acting
transcriptional regulators). The majority of the genes were
regulated by trans-acting elements. The DDX17 gene, iden-
tified as individual-specific by Whitney and colleagues
(Whitney et al. 2003), was also identified in this study and
shown to be linked to a cis-acting transcriptional regula-
tor on chromosome 22q13. In addition, master regulators
of transcription, or hotspots of genetic determinants, were
identified that regulate transcription of many genes. These
master regulators are located within a 5 Mb region on chro-
mosome 14q32 and on chromosome 20q13. For the genes
with cis-acting linkages, individuals with allelic marker
G enomic P er s pective s o f C l inica l I mmuno l ogy   •   5 9 9
differences indeed showed different expression levels. These
results were confirmed by allele-specific real-time PCR in
heterozygous individuals, in which one allele is expressed at
higher levels than the other allele.
Individual-specific gene-expression signatures are thus,
at least in part, genetically determined. This knowledge
provides new opportunities to explore gene-expression pro-
files in pathological conditions. Expression profiles from
patients, but also from unrelated individuals and family
members, can be used to identify polymorphic genes that
influence the immune response and susceptibility to infec-
tious and (auto)immune diseases (Figure 38.2).
C O N C LU S I O N S —T H E F U T U R E
Gene-expression profiling will lead to a molecular definition
of disease. Consequently, autoimmune diseases, immune
deficiency diseases, and malignancies that are currently
defined by their clinical phenotypes will soon be classi-
fied into molecular subtypes based on functional genom-
ics experiments. Since most infectious processes result in an
ordered series of host regulatory responses, gene-expression
profiles of host cells could also be used to define the stage
of the host’s response in individual patients. In the ideal
situation, pathogen-specific profiles may be recognized that
might be particularly useful for rapid diagnosis, especially
in cases where it is technically difficult or impossible to cul-
ture the pathogen. In addition to diagnostic purposes, this
will also have a considerable impact on therapy for infec-
tious diseases.
In addition, the assembly of comprehensive
gene-expression databases that contain molecular signa-
tures of specific cell subsets and biological networks will
be useful to improve our understanding of the biologi-
cal basis of dysregulation in disease. Knowledge that will
be gained by identifying the genes that are expressed in a
cell or tissue will make the biology of the system more
comprehensible. While this is in part true, it is clear that
transcriptomics information does not address the regula-
tion that occurs downstream of transcription, involving
mRNA splicing, protein expression, and post-translational
modification. A  systems approach requires the integra-
tion of as many different data types as possible; including,
besides gene-expression data, information on the proteome
(the global inventory of all proteins in an organ, tissue of
individual cell type) and the metabolome (global inven-
tory of metabolites). Together, the data on the different
levels of information transfer from genome to phenotype
define a new approach to biology, which is termed “systems
 Innate immune response Adaptive immune response
Healthy
Endogenous Viruses
TLR ligands
Bacteria

Genetically determined differences in Directed specific immune response

expression programs at baseline and
after stimulation of the innate
immune response

Autoimmunity, cancer, (chronic) infections

Healthy
Non-infected infected RA SLE Cancer

IFN/NFκB pathways IFN/stat1pathway IFNpathway/Disease severity Outcome/survival

Figure 38.2 
Immune response in chronic diseases and infections. Gene-expression profiling of immune cells from healthy individuals at baseline and
challenged with different microorganisms identified, at least in part, genetically determined expression programs, which influence the adaptive
immune response. The genetically determined differences in the innate immune response play a key role in the capacity to clear infections. Disease
profiling identifies patient-specific and disease-specific expression programs that partly overlap with expression programs induced by pathogens,
suggesting the involvement of the innate immune system in the susceptibility to chronic diseases.

biology” (Kitano 2002, Hood 2003):  a computational
modeling approach aimed at understanding the structure
and dynamics of cellular and organismal functions. Systems
biology will certainly become a mainstay of drug develop-
ment, allowing the identification of novel drug targets,
followed by target-directed drug effects combined with tai-
lored therapy. It can further be anticipated that this infor-
mation will allow us to determine how the disease pathway
can be manipulated pharmacologically with greater speci-
ficity than is achieved by current drugs.
Furthermore, differences in responses between indi-
viduals might provide a lead to an explanation of how
genetic variation affects the immune system. Evidence
exists that these inter-individual differences are geneti-
cally determined. One of the distinctive gene-expression
programs that varies among individuals involves the IFN/
STAT signaling pathway (Radich et al. 2004, Whitney et al.
2003, Boldrick et al. 2002). These findings make it tempt-
ing to speculate that inter-individual differences that exist
in healthy controls might explain the heterogeneity that is
observed in immune-related diseases such as SLE, RA, and
MS (Figure 38.1). Because baseline and microbial-induced
alterations of transcript levels of immune-related genes dif-
fer among individuals, these expression profiles may inform
the prediction of an individual’s response to infectious
agents, and his susceptibility to cancer and autoimmune
diseases. The activation of the IFN pathway in SLE and RA
may therefore reflect an inherited genetic capacity for an
altered activity of the IFN-response program. Analogous
to these findings, the most consistent intrinsic variable
gene in healthy donors, DDX17 (Whitney et  al. 2003),
was selectively expressed in a subgroup of RA patients who
showed a high degree of inflammation and high expression
of IFN-responsive genes (van der Pouw Kraan TC et  al.
2003a).
In the near future, new analysis methods will evolve to
further improve the identification of relationships between
the expression levels of many genes, to identify new signal-
ing pathways, determine genetic differences, and further
improve prediction of disease susceptibility and subclas-
sification of patients within a disease—finally leading to a
more effective and personalized treatment.

6 0 0   •   G enomic s in C l inica l P ractice

 AC K N OW L E D G E M E N T S
Financial support was provided by the Netherlands
Organization of Pure Scientific Research; the Dutch
Arthritis Association; the European Union FP6 STREP
program; “AUTOROME” (www.AUTOROME.de); and
the Center for Medical Systems Biology (CMSB), a center of
excellence financed by the Netherlands Genomics Initiative.
REFERENCES
Achiron A, Gurevich M, Friedman N, Kaminski N, and Mandel M
(2004). Blood transcriptional signatures of multiple sclerosis: unique
gene expression of disease activity. Ann Neurol 55(3):410–417.
Alizadeh AA, Eisen MB, Davis RE, et al. (2000). Distinct types of dif-
fuse large B-cell lymphoma identified by gene expression profiling.
Nature 403(6769):503–511.
Baechler EC, Batliwalla FM, Karypis G, et al. (2003). Interferon-inducible
gene expression signature in peripheral blood cells of patients with
severe lupus. Proc Natl Acad Sci U S A 100(5):2610–2615.
Bennett L, Palucka AK, Arce E, et al. (2003). Interferon and granulo-
poiesis signatures in systemic lupus erythematosus blood. J Exp Med
197(6):711–723.
Beutler B (2004). Inferences, questions and possibilities in Toll-like
receptor signalling. Nature 430(6996):257–263.
Boldrick JC, Alizadeh AA, Diehn M, et al. (2002). Stereotyped and spe-
cific gene expression programs in human innate immune responses
to bacteria. Proc Natl Acad Sci U S A 99(2):972–977.
Bosinger SE, Hosiawa KA, Cameron MJ, et  al. (2004). Gene expres-
sion profiling of host response in models of acute HIV infection. J
Immunol 173(11):6858–6863.
Burska AN, et al. (2014). Gene expression analysis in RA: towards per-
sonalized medicine. The Pharmacogenomics Journal 14:93–106;
doi:10.1038/tpj.2013.48.
Cheung VG, Conlin LK, Weber TM, et al. (2003). Natural variation in
human gene expression assessed in lymphoblastoid cells. Nat Genet
33(3):422–425.
Chtanova T, Kemp RA, Sutherland AP, Ronchese F, and Mackay CR
(2001). Gene microarrays reveal extensive differential gene expres-
sion in both CD4(+) and CD8(+) type 1 and type 2 T cells. J
Immunol 167(6):3057–3063.
Cobb JP, Mindrinos MN, Miller-Graziano C, et al. (2005). Application
of genome-wide expression analysis to human health and disease.
Proc Natl Acad Sci U S A 102(13):4801–4806.
Diehn M, Alizadeh AA, Rando OJ, et al. (2002). Genomic expression
programs and the integration of the CD28 costimulatory signal in
T cell activation. Proc Natl Acad Sci U S A 99(18):11796–11801.
Firestein GS and Zvaifler NJ (2002). How important are T cells in
chronic rheumatoid synovitis?: II. T cell-independent mechanisms
from beginning to end. Arthritis Rheum 46(2):298–308.
Glynne R, Akkaraju S, Healy JI, Rayner J, Goodnow CC, and Mack
DH (2000). How self-tolerance and the immunosuppressive drug
FK506 prevent B-cell mitogenesis. Nature 403(6770):672–676.
Golub TR, Slonim DK, Tamayo P, et al. (1999). Molecular classification
of cancer:  class discovery and class prediction by gene expression
monitoring. Science 286(5439):531–537.
Guidotti LG and Chisari FV (2001). Noncytolytic control of viral
infections by the innate and adaptive immune response. Annu Rev
Immunol 19:65–91.:65–91.
Hamalainen H, Zhou H, Chou W, Hashizume H, Heller R, and
Lahesmaa R (2001). Distinct gene expression profiles of human type
1 and type 2 T helper cells. Genome Biol 2(7):RESEARCH0022.
G enomic P er s pective s o f C l inica l I mmuno l ogy   •   6 0 1
Hertzog PJ, O’Neill LA, and Hamilton JA (2003). The interferon
in TLR signaling:  more than just antiviral. Trends Immunol
24(10):534–539.
Hood L (2003). Systems biology: integrating technology, biology, and
computation. Mech Ageing Dev 124(1):9–16.
Izmailova E, Bertley FM, Huang Q, et al. (2003). HIV-1 Tat reprograms
immature dendritic cells to express chemoattractants for activated T
cells and macrophages. Nat Med 9(2):191–197.
Jenner RG and Young RA (2005). Insights into host responses against
pathogens from transcriptional profiling. Nat Rev Microbiol
3(4):281–294.
Kirou KA, Lee C, George S, et  al. (2004). Coordinate overexpression
of interferon-alpha-induced genes in systemic lupus erythematosus.
Arthritis Rheum 50(12):3958–3967.
Kirou KA, Lee C, George S, Louca K, Peterson MG, and Crow MK
(2005). Activation of the interferon-alpha pathway identifies a sub-
group of systemic lupus erythematosus patients with distinct sero-
logic features and active disease. Arthritis Rheum 52(5):1491–1503.
Kitano H (2002). Systems biology:  a brief overview. Science
295(5560):1662–1664.
Lee CG (2002). RH70, a bidirectional RNA helicase, co-purifies with
U1snRNP. J Biol Chem 277(42):39679–39683.
Lehtonen A, Matikainen S, and Julkunen I (1997). Interferons
up-regulate STAT1, STAT2, and IRF family transcription factor
gene expression in human peripheral blood mononuclear cells and
macrophages. J Immunol 159(2):794–803.
Lock C, Hermans G, Pedotti R, et al. (2002). Gene-microarray analysis
of multiple sclerosis lesions yields new targets validated in autoim-
mune encephalomyelitis. Nat Med 8(5):500–508.
Maas K, Chan S, Parker J, Slater A, Moore J, Olsen N, and Aune TM
(2002). Cutting edge:  molecular portrait of human autoimmune
disease. J Immunol 169(1):5–9.
Mihm S, Frese M, Meier V, et al. (2004). Interferon type I gene expres-
sion in chronic hepatitis C. Lab Invest 84(9):1148–1159.
Moir S, Malaspina A, Pickeral OK, et al. (2004). Decreased survival of
B cells of HIV-viremic patients mediated by altered expression of
receptors of the TNF superfamily. J Exp Med 200(7):587–599.
Morley M, Molony CM, Weber TM, et  al. (2004). Genetic analy-
sis of genome-wide variation in human gene expression. Nature
430(7001):743–747.
Nau GJ, Richmond JF, Schlesinger A, Jennings EG, Lander ES, and Young
RA (2002). Human macrophage activation programs induced by
bacterial pathogens. Proc Natl Acad Sci U S A 99(3):1503–1508.
Olsen N, Sokka T, Seehorn CL, et al. (2004). A gene expression signature
for recent onset rheumatoid arthritis in peripheral blood mononu-
clear cells. Ann Rheum Dis 63(11):1387–1392.
Pizza M, Scarlato V, Masignani V, et al. (2000). Identification of vaccine
candidates against serogroup B meningococcus by whole-genome
sequencing. Science 287(5459):1816–1820.
Radich JP, Mao M, Stepaniants S, et al. (2004). Individual-specific varia-
tion of gene expression in peripheral blood leukocytes. Genomics
83(6):980–988.
Rogge L, Bianchi E, Biffi M, et  al. (2000). Transcript imaging of the
development of human T helper cells using oligonucleotide arrays.
Nat Genet 25(1):96–101.
Rubins KH, Hensley LE, Jahrling PB, et al. (2004). The host response
to smallpox: analysis of the gene expression program in peripheral
blood cells in a nonhuman primate model. Proc Natl Acad Sci U S A
19;101(42):15190–15195.
Shaffer AL, Rosenwald A, Hurt EM, et  al. (2001). Signatures of the
immune response. Immunity 15(3):375–385.
Teague TK, Hildeman D, Kedl RM, et al. (1999a). Activation changes
the spectrum but not the diversity of genes expressed by T cells. Proc
Natl Acad Sci U S A 96(22):12691–12696.
Teague TK, Hildeman D, Kedl RM, et al. (1999b). Activation changes
the spectrum but not the diversity of genes expressed by T cells. Proc
Natl Acad Sci U S A 96(22):12691–12696.
van ‘t Veer LJ, Dai H, van de Vijver MJ, et  al. (2002). Gene expres-
sion profiling predicts clinical outcome of breast cancer. Nature
415(6871):530–536.
van de Vijver MJ, He YD, van’t Veer LJ, et al. (2002). A gene-expression
signature as a predictor of survival in breast cancer. N Engl J Med
19;347(25):1999–2009.
van der Pouw Kraan TC, Van Gaalen FA, Huizinga TW, et al. (2003a).
Discovery of distinctive gene expression profiles in rheumatoid
synovium using cDNA microarray technology:  evidence for the
existence of multiple pathways of tissue destruction and repair.
Genes Immun 4(3):187–196.
6 0 2   •   G enomic s in C l inica l P ractice
van der Pouw Kraan TC, Van Gaalen FA, Kasperkovitz PV, et al. (2003b).
Rheumatoid arthritis is a heterogeneous disease: evidence for differ-
ences in the activation of the STAT-1 pathway between rheumatoid
tissues. Arthritis Rheum 48(8):2132–2145.
Westendorp RG, Langermans JA, Huizinga TW, et al. (1997). Genetic
influence on cytokine production and fatal meningococcal disease.
Lancet 349(9046):170–173.
Whitney AR, Diehn M, Popper SJ, et al. (2003). Individuality and varia-
tion in gene expression patterns in human blood. Proc Natl Acad Sci
U S A 100(4):1896–1901.
 39.
GENOMIC APPLICATIONS IN CLINICAL PEDIATRICS
Vinod Cherian Varghese, Sian Morgan, and Ian Frayling
INTRODUCTION
Recent advances in human genetics and genomics, such as
the Human Genome Project, characterization of human
genomic variations (for example, single-nucleotide poly-
morphisms [SNPs] in the HapMap project; copy number
variations [CNVs]), comparative genomics, epigenetics/
epigenomics, and others are providing new insights and
suggesting new methods of applying genomics to the study
of human health and disease. Historically, the practice of
medical genetics, specifically clinical genetics, has strong
and close relationships with clinical pediatrics. It is widely
acknowledged that special consideration needs to be given
to the application of genetic and genomic advances in the
care of children. In this chapter, we will describe the ori-
gins of using genetic information in early diagnosis and
preventative intervention in the care of children (through
neonatal-screening programs), and then discuss how
this model informs the application of human genomic
approaches to the care of children in the current genom-
ics era. In addition, selected case scenarios will be included
to illustrate the use of current and next-generation genomic
technologies. Scientific and technical aspects are not
included, since these are discussed elsewhere in this book
(Chapters 2, 6, and 10).
G E N ET I C A N D G E N O M I C
S C R E E N I N G I N P E D I AT R I C S A N D
C H I L D H E A LT H
Large-scale newborn-screening began in September of 1962
when the Diagnostic Laboratories of the Massachusetts
Department of Public Health in the United States began
collecting blood specimens from all newborns in the state
to test them for phenylketonuria (PKU)[1]‌, a rare defect
in phenylalanine metabolism that results in mental retar-
dation in children if left untreated. This was followed by
603
specific dietary therapy, with overwhelming evidence for its
being effective in preventing mental retardation in children
with PKU if instituted early from the neonatal period[2, 3].
Dr. Guthrie adapted his phenylalanine assay to work with
blood dried on filter paper, which could be easily obtained
from newborns by means of a simple heel stick. Guthrie’s
phenylalanine assay was adapted by diagnostic laboratories
in many countries, which began screening all newborns for
PKU using dried blood collected on filter paper, now called
“Guthrie card” (Figure 39.1).
Building on the success of PKU screening, laboratories
adapted Guthrie’s assay to detect other amino acids and
sugars, which allowed screening for additional metabolic
disorders whose deleterious effects could be prevented by
early dietary intervention, including maple syrup urine
disease (leucine)[4]‌, homocystinuria (methionine), and
galactosemia (galactose)[5]. With the availability of excess
blood on the Guthrie cards, other disorders for which a
metabolite, hormone, or protein could be assayed for diag-
nosis using a screening blood test were added to the states’
newborn-screening panels. These included hemoglobinop-
athies[6], congenital hypothyroidism[7], congenital adrenal
hyperplasia (CAH)[8], and biotinidase deficiency (BD)[8].
Table 39.1 gives a list of the core conditions currently
included in newborn-screening programs. In each disorder,
it has been determined that clinical disease can be mitigated
or prevented by early dietary or medical intervention.
Since their inception in the United States,
newborn-screening programs have been created in many
countries and regions all over the world and have become
an indispensable part of the healthcare system and pre-
ventative medicine around the world[9]‌. Most programs
have focused on the detection of two diseases, namely
PKU and congenital hypothyroidism. Other disorders
have been included, depending on local prevalence, inter-
est of professionals, technological availability, and feasi-
bility of interventions. Not only do the number and types
of conditions tested differ widely between countries, but

Figure 39.1 
Example of a Guthrie card for collection of blood specimens
for newborn screening (from the Ohio Department of Health, USA).
Blood samples are collected within the dashed circles on filter paper
attached to the top of the form. (Adapted with permission from
‘Applications in Clinical Pediatrics’, by Michael R. Konikof and Michael
D. Bates, p. 425–436. In ‘Genomics and Clinical Medicine’, Eds. Kumar,
D & Weatherall, DJ. Oxford University Press, NY, 2008.)
the designs of these programs with respect to legislation,
funding mechanisms, and follow-up also vary. Although
the cost-effectiveness of newborn-screening for rare con-
ditions has been debated in the international community,
the creation of new programs and the expansion of exist-
ing programs continue[10].
Work in human genetics has elucidated the under-
lying molecular defects responsible for many inherited
congenital disorders. By applying molecular techniques
to newborn-screening, many genetic diseases can be
detected shortly after birth, before their full or irreversible
6 0 4   •   G enomics in C linical P ractice
phenotypical expression. Disorders with well-defined
genetics and common mutations represented in a signifi-
cant proportion of affected individuals are particularly well
suited to this form of analysis. Genomic DNA can be eas-
ily extracted from newborns’ blood specimens by inexpen-
sive, automated methods, further enhancing the feasibility
of this approach[11]. Traditional molecular methodologies,
however, analyze only one gene or mutation at a time,
requiring separate assays for each disease. Thus, while excel-
lent in a research setting, these procedures are not amenable
to population-based, high-throughput newborn-screening
programs. They are primarily used clinically as second-tier
confirmatory assays for such disorders as cystic fibro-
sis, sickle-cell anemia, and other hemoglobinopathies,
and medium-chain acyl-CoA dehydrogenase (MCAD)
deficiency[12].
The recent application of tandem mass spectrometry
(MS/MS) to the specimens on Guthrie cards has further
expanded the ability of newborn-screening programs to
identify congenital disorders. MS/MS recognizes amino
acids and acylcarnitines by their characteristic mass spec-
tra and is thus able to identify more than 20 inborn errors
of metabolism not previously a part of newborn-screening
programs[13,14]. MS/MS also has the advantage of being
able to measure concentrations of multiple compounds,
including amino acids and acylcarnitines, in the same
sample, greatly increasing the utility of this method. MS/
MS has revolutionized the field of newborn-screening by
allowing the detection of multiple metabolic disorders in
a single step. Previously, each screened disease required a
separate assay. This was true not only for older assays, like
the bacterial-inhibition assay for PKU, but also for newer
molecular assays for genetic disorders. Thus, MS/MS is a
straightforward, efficient, and cost-effective method for
the detection of diseases involving altered amino acid and
acylcarnitine metabolism. Additional work is expanding
the range of the utility of MS/MS to include identifica-
tion of hexose monophosphates in patients with galactose-
mia[14] and of steroids in patients with congenital adrenal
hypoplasia[15].
The ability to extract DNA from the newborn blood
specimens and recent advances in microarray technology
offer the opportunity for primary screening of a vast num-
ber of genetic disorders that could not previously be iden-
tified in the newborn. DNA microarrays, which contain
representations of thousands of genes, are fabricated either
by robotically spotting gene fragments onto a glass slide or
by synthesizing oligonucleotides onto glass by photolitho-
graphic techniques[16]. Microarrays can be used to measure
the expression of thousands of genes simultaneously by the
Table 39.1  LIST OF PEDIATRIC DISEASES INCLUDED IN THE NEWBORN-SCREENING PROGRAMS

DISORDER SCREENING METHOD TREATMENT

Phenylketonuria Guthrie bacterial inhibition assay Diet restricting phenylalanine
Fluorescence assay
Amino acid analyzer
Tandem mass spectrometry
Congenital hypothyroidism Measurement of thyroxine and thyrotropin Oral levothyroxine
Hemoglobinopathies Hemoglobin electrophoresis Prophylactic antibiotics
Isoelectric focusing Immunization against
High-performance liquid chromatography Streptococcus pneumoniae and Haemophilus influenzae
Follow-up DNA analysis
Galactosemia Beutler test Galactose-free diet
Paigen test
Maple syrup urine disease Guthrie bacterial inhibition assay Diet restricting intake of branched-chain amino acids
Tandem mass spectrometry
Homocystinuria Guthrie bacterial i­ nhibition assay Vitamin B12
Tandem mass spectrometry Diet restricting methionine and supplementing cystine
Biotinidase deficiency Calorimetric assay Oral biotin
Congenital adrenal hyperplasia Radio immunoassay Glucocorticoids
Enzyme immunoassay Mineralocorticoids
Salt
Cystic fibrosis Immunoreactive trypsinogen assay Improved nutrition
­followed by DNA testing
Sweat chloride testing Management of pulmonary symptoms
Fatty acid disorders a
Tandem mass spectrometry Various
Organic acid disordersb Tandem mass spectrometry Various
Other amino acid disorders Tandem mass spectrometry Various
Hearing screen Tandem mass spectrometry Hearing aids
Audiometry Cochlear implants
a
Carnitine uptake defect, very long-chain hydroxyacyl-CoA dehydrogenase deficiency, long-chain hydroxyacyl-CoA dehydrogenase deficiency, medium-chain
hydroxyacyl-CoA dehydrogenase deficiency, trifunctional protein deficiency.
b
Glutaric acidemia type 1,3-hydroxy 3-methylglutaric aciduria, isovaleric, acidemia, 3-methylcrotonyl-CoA carboxylase deficiency, methylmalonic acidemia, beta
ketothiolase deficiency, propionic acidemia, multiple carboxylase deficiency, argininosuccinate acidemia, citrullinemia type 1, tyrosinemia type 2.
SOURCE: National Newborn-screening and Genetics Resource Center (http://genes-r-us.uthscsa.edu/)
(Adapted with permission from ‘Applications in Clinical Pediatrics’, by Michael R. Konikof and Michael D. Bates, p. 425–436. In ‘Genomics and Clinical Medicine’,
Eds. Kumar, D & Weatherall, DJ. Oxford University Press, NY, 2008.)

specific hybridization of nucleic acid strands having com-
plementary nucleotide sequences. This stringent hybridiza-
tion can also be applied to the analysis of genomic DNA to
detect SNPs, point mutations, insertions, and deletions[17].
Microarrays have been designed to test these concepts
using well-characterized genetic disorders such as hemo-
globinopathies, α1-antitrypsin deficiency, and factor V
Leiden[18]. Oligonucleotides representing multiple disease
alleles in the above disorders were spotted onto a microar-
ray, and DNA with known mutations was extracted from
newborn blood specimens and hybridized to the array.
Hybridization signals were detected corresponding to the
various known mutations, indicating that the microarray
correctly identified the tested gene mutations. The entire
process, from DNA extraction to data analysis, can be auto-
mated, and with the addition of multiplex amplification,
several disorders may be assayed on a single microarray. This
will allow population-based first-tier molecular screening
for a large number of genetic disorders, which has not pre-
viously been possible.
Primary newborn-screening utilizing microarray tech-
nology offers several advantages over traditional methods

G enomic A pplications in C linical P ediatrics   •   6 0 5

Table 39.2  POTENTIAL ADVANTAGES AND LIMITATIONS OF USING DNA MICROARRAYS IN
NEWBORN-SCREENING

Advantages
Screen many genes in single assay.
Identify any mutations rather than only common or previously characterized mutations. Can be automated for high-throughput.
Potential for increased specificity with high sensitivity.
Can detect heterozygotes (disease-gene carriers).
Allows genotyping of modifying genes to improve phenotype prediction of single-gene disorders. Less affected by clinical factors
­( prematurity, diet, medications, etc.).
Limitations
Cost.
Amplification of genomic DNA by polymerase chain reaction (PCR) can result in artifacts or contamination. May identify
­polymorphisms—mutations that have no effect on gene or protein function.
Does not identify alterations in post-transcriptional or post-translational processes.
Does not identify non-Mendelian genetic events (e.g., gene duplications, epigenetic influences, and non-random X chromosome inactivation).
Not applicable to disorders with no known genetic basis (e.g., congenital hypothyroidism).

SOURCE: Adapted with permission from ‘Applications in Clinical Pediatrics’, by Michael R. Konikof and Michael D. Bates, pages 425–436. In ‘Genomics and
Clinical Medicine’, Eds. Kumar, D & Weatherall, DJ. Oxford University Press, NY, 2008.

(Table 39.2). Just as MS/MS can measure multiple analytes
from a single sample, DNA microarrays can screen many
genes for mutations in a single step, with little additional
cost for each disorder. The major limiters at present are the
feature density of the microarray and the resolution of laser
scanners. The completion of the sequencing of the human
genome facilitates the design of “resequencing” microar-
rays to screen for and identify any mutations of a specific
gene, rather than only common or previously character-
ized ones. Because the entire process can be automated with
high throughput, microarrays are well suited to population
screening. This is largely facilitated with the availability of
several next-generation sequencing platforms and com-
putational analytical tools. DNA microarrays are able to
screen for certain disorders that are not amenable to tradi-
tional protein-based methods, such as fragile-X syndrome
(through identification of untranslated, triplet-repeat DNA
sequences)[19], the various severe combined immunode-
ficiencies (through detection of excised T-cell receptor
sequences rather than of various T-cell related proteins)[20],
and cystic fibrosis. For example, the ΔF508 mutation of
the cystic fibrosis transmembrane conductance regulator
(CFTR), which accounts for 80% of mutant cystic fibrosis
alleles in the Caucasian population (but only 30% of mutant
alleles in Asians or other non-European individuals), is the
only mutation assayed in the newborn-screening program
of at least one state in the United States[21]. There are now
more than 1,000 known mutations in the CFTR gene, and
all of these could be assayed using one microarray, which
would improve detection of cystic fibrosis cases in ethnic
groups where uncommon mutations are over-represented.
Microarrays also have the potential to be more specific,
while maintaining sensitivity, than traditional metabolite-
or protein-based assays, which commonly use threshold val-
ues that yield ratios of false-to-true positives ranging from
10:1 to 75:1[22]. More-specific tests would reduce the rate of
false positives, thus decreasing the need for follow-up testing
and its resultant costs, not to mention the increased anxi-
ety for parents and medical caregivers. Microarrays are able
to distinguish disease-gene carriers (heterozygotes) by the
selective detection of hybridization signal intensity and the
simultaneous probing for wild-type alleles. Carriers could
then be excluded from follow-up testing and laboratory
reporting, which now diverts resources from detecting dis-
eased individuals, the primary mission of newborn-screening
programs. Alternatively, carriers could be easily identified
and reported, if this was appropriate from ethical and/or
legal perspectives. In addition, microarrays could be used to
improve phenotype prediction of Mendelian “single-gene”
disorders by genotyping known modifying or interacting
genes, which could improve prediction of disease severity
that impacts treatment strategies, planning, and counsel-
ing[23]. Finally, DNA-based analysis is not affected by clini-
cal factors that may interfere with analysis of blood samples
for metabolites, hormones, or proteins. Factors such as pre-
maturity, medications of the mother or child, or parenteral
nutrition will not affect DNA-based analysis.
DNA microarrays do have several limitations that must be
overcome before their widespread use in newborn-screening
programs will become practical (Table 39.2). Cost has

6 0 6   •   G enomics in C linical P ractice

been a major factor limiting their use, but newer manu-
facturing procedures have eased this constraint somewhat,
and economies of scale would be achieved if standardized
microarrays were developed for use in newborn-screening
programs. Still, the cost of single-use microarrays is much
higher than for standard assays used in public health labora-
tories, many of which cost a few U.S. dollars at the most. The
high-throughput, automated technology on which poly-
merase chain reaction (PCR)-driven microarrays depend
may also be another source of problems. Contamination
of samples and resulting PCR artifacts must be minimized,
and systems for organizing, storing, and accessing the vast
amounts of generated data will need to be developed. In
addition to cost and technical issues, the very nature of the
information derived from microarrays can be a limitation
in some cases. For example, if a previously uncharacterized
point mutation is identified, it may not be possible to pre-
dict whether the mutation is deleterious. This approach will
not detect post-transcriptional and post-translational mod-
ifications or impairments, nor will it find non-Mendelian
genetic events such as gene duplications, epigenetic influ-
ences, and non-random X-chromosome inactivation, and
it thus may lead to false-positive or false-negative screening
results[24]. Some disorders that have a final common pheno-
type may also be caused by mutations in different genes, and
all such genes would need to be included in the microar-
ray. For example, PKU is usually caused by a mutation in
the phenylalanine hydroxylase gene, but mutations in the
cofactor pathway of tetrahydrobiopterin synthesis may
cause a form of atypical PKU[22]. Yet other disorders have no
known genetic basis, such as congenital hypothyroidism, in
which only a small percentage of cases have a readily identi-
fiable genetic cause[25]. These disorders will not be amenable
to DNA-based microarray screening unless a genetic etiol-
ogy is elucidated.
With the advent of new technologies such as MS/MS
and DNA microarrays, newborn-screening programs are
poised to significantly expand the number and types of dis-
orders that are detectable (see Chapters 10 & 14). The prin-
ciples of population screening developed in 1968 by Wilson
and Jungner form a basis for applying these new technolo-
gies to newborn-screening[26]. In particular, the ability to
screen for a disorder using these new technologies does not
imply the obligation to do so. The early identification of a
genetic disorder is of no benefit (1) if treatment is not avail-
able or (2) if an abnormal finding determines only suscepti-
bility for, and not certainty of, disease.
An example of a genetic disorder for which effective
treatment is currently not available is Huntington’s disease,
a devastating neurological disorder of adult onset, about
G enomic A pplications in C linical P ediatrics   •   6 0 7
which there continues to be much debate regarding the
clinical and ethical merits of pre-symptomatic diagnosis[27].
Even for disorders that present in childhood, such as famil-
ial adenomatous polyposis and Duchenne muscular dystro-
phy, pre-symptomatic diagnosis is controversial because of
the lack of preventative treatments[28].
Potential examples for susceptibility screening include
predispositional screening for adult-onset, multifactorial
disorders such as type II diabetes mellitus or cardiovascular
disease. Screening for these disorders is often not medically
beneficial and may have negative psychological conse-
quences to both the affected individual and his or her fam-
ily. Currently, pilot studies have been undertaken to detect
predisposition to type I diabetes mellitus by typing multiple
HLA DQ alleles[29]. If successful, this effort would have
enormous public health implications, as type I diabetes mel-
litus affects over 300,000 people in the United States alone,
and generates huge direct and indirect healthcare costs.
However, no treatment is available to prevent the develop-
ment of diabetes in predisposed individuals, so early identi-
fication will not necessarily improve the outcome for these
children. This information may, in fact, become available
to insurance providers or employers, and could lead to dis-
crimination against identified children or their families (see
Chapter 17—Ethical, Legal, and Social Implications, ELSI).
Limiting newborn-screening to disorders currently
assayed by screening programs, though, will not take full
advantage of the public health potential of microarray tech-
nology. It will be necessary to carefully evaluate each new dis-
order added to the panel to determine immediate and future
ramifications of screening, as well as to develop methods
of obtaining informed consent from parents of newborns,
which is now not required in mandatory screening programs.
D I AG N O S T I C G E N ET I C A N D G E N O M I C
A P P L I C AT I O N S I N C H I L D R E N
K A RYOT Y P I N G
Ever since trisomy 21 was identified as the underlying cause
of Down syndrome by Jerome Lejune in 1959 using karyo-
typing[30], increasing numbers of new genetic technologies
have been introduced to detect chromosomal abnormalities
in pediatric practice. In particular, the past few years have
seen an explosion in newer genetic technologies, including
array-comparative genomic hybridization (array-CGH),
exome sequencing, and whole-genome sequencing. Patterns
and power of the diagnostic clinical pediatrics have dra-
matically changed with the rapid applications of genetic and
genomic technologies. Karyotype is the term used to describe
the number and appearance of chromosomes. Traditionally,
chromosomes have been analyzed using the light microscope.
The chromosome analysis takes into account their length and
banding patterns, including the position of the centromeres.
The basic number of chromosomes in the somatic cells of
an individual is 46. Therefore, in normal diploid organisms,
autosomal chromosomes are present in two copies. There
are also sex chromosomes. Polyploid cells have multiple cop-
ies of chromosomes, and haploid cells have single copies.
The study of karyotypes is made possible by staining, such
as G-banding. For humans, white blood cells are used most
frequently to analyze chromosomes because they are easily
induced to divide and grow in tissue culture.
CASE 1
A term, small-for-gestational-age female baby was admit-
ted to the neonatal unit in view of her poor feeding and
dysmorphic features. There was a history of polyhydram-
nios in the antenatal period. On examination, the baby was
noticed to have some dysmorphic features. These included
short palpebral fissures, micrognathia, overlapping fingers
(index finger overlapping the third, and the fifth finger
overlapping the fourth), talipes, and “rocker bottom” feet.
1 2
6 7 8 9
13 14 15
19 20
Figure 39.2  Karyotype showing three copies of chromosome 18, consistent with Edwards syndrome.
6 0 8   •   G enomics in C linical P ractice
An echocardiogram revealed a peri-membranous ventricu-
lar septal defect (VSD), and abdominal ultrasound revealed
“horseshoe” kidneys. In view of her clinical presentation,
the neonatologist requested karyotyping for the baby,
which showed aneuploidy for chromosome 18 (47, XX
+18). The baby’s clinical profile (Figure 39.2) was consis-
tent with the cytogenetic diagnosis of trisomy 18 (Edwards
syndrome)[31].
Most clinicians request cytogenetic analysis when one
of the common aneuploidies (trisomy 13, 18, or 21) is sus-
pected. Whilst the standard Giemsa banded chromosome
analysis is the routine method, this has now been replaced
with a rapid fluorescent in situ hybridization (FISH) test
or rapid quantitative fluorescent polymerase chain reaction
(QF-PCR) test, as these give a targeted result within 24–72
hours, while an urgent karyotype can take up to 7–10 days.
CASE 2
A 14-year-old girl was referred to the pediatric endocri-
nologist in view of her short stature. She was the firstborn
of a non-consanguineous relationship, and no significant
problems were noted in the antenatal or neonatal period.
Further history revealed that the child had not yet attained
menarche. On examination, the child’s height was on the
3 4 5
10 11 12
16 17 18
21 22 X
 1 2
6 7
13 14 15
19 20
Figure 39.3  Karyotype showing a single X chromosome, and no Y chromosome, consistent with Turner syndrome.
0.4th centile on the growth chart. She was noticed to have
a low posterior hairline, multiple skin nevi, and lack of
breast development. Cardiac examination was normal. The
pediatric endocrinologist requested a karyotype (45, XO),
which confirmed the clinical suspicion of Turner syndrome
(Figure 39.3). The possibility of Turner syndrome should be
considered in any girl who presents with short stature and
delayed puberty. This child had some additional clues in the
form of low posterior hairline and multiple skin nevi to sug-
gest this diagnosis[32].
CASE 3
A six-year-old male child was referred to the clinical geneti-
cist in view of his developmental delay and dysmorphic fea-
tures. Antenatal period was uneventful. There was history
of poor feeding in the neonatal period, which improved
with time. He had global developmental delay, especially in
the area of expressive speech. On examination in the genet-
ics clinic, the child was noticed to have upslanting palpe-
bral fissures, mid-face hypoplasia, unilateral single palmar
crease, and joint laxity. Cardiac examination was normal,
and an echocardiogram did not identify any abnormali-
ties. The clinical geneticist requested a karyotype with an
extended cell count (30 cells), which showed a mosaic
G enomic A pplications in C linical P ediatrics   •   6 0 9
3 4 5
8 9 10 11 12
16 17 18
21 22 X
trisomy 21 (mosaic Down syndrome) (47, XY +21/46, XY;
Figure 39.4). The clinical geneticist suspected a diagnosis
of mosaic Down syndrome from the history and clinical
examination findings[33]. In a “routine” karyotype, it is usual
to analyze up to 10 cells. If mosaicism is suspected, the clini-
cian can request the analysis of more cells.
FLU O R E S C E N T I N S IT U
H Y B R I D I Z AT I O N—FI S H
Fluorescence in situ hybridization (FISH) is the visualiza-
tion of specific DNA sequences in metaphase chromosomes
and interphase cells using a fluorescent microscope. Both
numerical and structural chromosomal rearrangements
can be detected in either dividing or non-dividing inter-
phase nuclei using probes ranging from whole chromo-
some “paints” to individual gene-specific probes. The target
sequence can range from one to several hundred kilobases[34].
CASE 4
A four-year-old girl was referred to the pediatric cardiolo-
gist in view of her cardiac murmur. In the history, she had
mild developmental delay. The child was noticed to be quite
friendly and had some facial dysmorphic features, including
 1 2 3 4 5

6 7 8 9 10 11 12

13 14 15 16 17 18

19 20 21 22 X Y
→

47, XY, +21

Figure 39.4  Karyotype showing the cell line with three copies of chromosome 21 consistent with mosaic Down syndrome.

The clinical features of this child are classical of

Williams syndrome, which is caused by a deletion in chro-
ELN mosome 7q11 region. The most common deletion causing
Deleted chromosome 7 Williams syndrome is about 1.55 Mb (million base pairs)
in size[35], and involves several genes (contiguous gene syn-
drome), including the elastin gene, one of the many genes
involved in vascular development. Microdeletions of this
Normal size could not be identified by conventional karyotyping,
chromosome 7 and the application of FISH test is now routine to confirm
a chromosome deletion of this size.

CASE 5

A six-year-old male child was referred to the pediatric endo-

Figure 39.5 
Metaphase showing a deletion of the ELN probe on one
chromosome 7 homologue using Fluorescent in situ Hybridisation
(FISH) studies.
peri-orbital fullness, jowly (full) cheeks, wide mouth with full
lips, and widely spaced teeth. An echocardiogram revealed
moderate supra-valvular aortic stenosis. The pediatric cardi-
ologist requested a FISH test, which confirmed the typical
microdeletion of the long-arm chromosome 7 (46,XX.ish
del(7)(q11.23q11.23) (ELN-) consistent with the clinical
suspicion of Williams syndrome (Figure 39.5).
crinologist in view of his short stature. There was history
of polyhydramnios in the antenatal period. The child was
admitted to the neonatal unit soon after birth in view of
his poor feeding and hypotonia, which gradually improved
with time. A karyotype performed in the neonatal period
did not identify any abnormalities. The child was noticed
to have unilateral undescended testes, which was corrected
at the age of two years. The parents were also concerned
about his developmental delay and excessive weight gain.
They recalled that the child was particularly difficult to

6 1 0   •   G enomics in C linical P ractice


SNRPN
Normal chromosome 15
Deleted chromosome 15
Figure 39.6 
Metaphase showing a deletion of the SNRPN probe on one
chromosome 15 homologue using Fluorescent in situ Hybridisation
(FISH) studies.
feed during infancy. However, following this period, he
had become obsessed with food. On examination in the
clinic, the child was noticed to have central obesity with
small hands and feet. The scrotum was hypoplastic with
decreased rugosity. The pediatric endocrinologist requested
a FISH test to check for Prader-Willi syndrome, which con-
firmed the typical microdeletion of the long arm of chro-
mosome 15 (Figure 39.6; 46,XY.ish del(15)(q11.2q11.2)
(SNRPN-, D15S10-), consistent with the clinical diagnosis
of Prader-Willi syndrome.
The clinical features of this child are suggestive of
Prader-Willi syndrome (PWS)[36]. There are three dif-
ferent genetic mechanisms that can cause PWS. These
include: (a) deletion involving the paternal copy of chromo-
some 15q11-q13, (b) maternal uniparental disomy (UPD)
of this chromosomal region, and (c) imprinting defects. The
majority (about 70%) are caused by paternal deletions, and
these can be identified by FISH analysis[36, 37].
MU LT I P L E X L I G AT I O N-D E P E N D E N T
P RO B E A M P L I FI C AT I O N—M L PA
Multiplex ligation-dependent probe amplification (MLPA)
is a method developed to determine the copy number of
genomic DNA sequences in a single multiplex PCR-based
reaction. Each MLPA probe consists of two oligonucle-
otides, which recognize adjacent target DNA for a successful
ligation. Only complete probe (ligated probes), which has a
unique length, can be exponentially amplified by PCR. The
relative number of fragments can be determined after the
PCR reaction by comparing the peak pattern obtained on a
given sample with that obtained from reference samples[38].
G enomic A pplications in C linical P ediatrics   •   6 1 1
CASE 6
A term newborn boy was admitted to the neonatal unit
soon after birth in view of his cyanosis. Antenatal period
was uneventful, and the antenatal scans did not pick up
any abnormalities. On examination, apart from a cardiac
murmur, the baby was also noticed to have long fingers and
notched ear helices. An echocardiogram revealed tetralogy
of Fallot, and there was absence of thymic shadow on the
chest X-ray. The neonatologist requested an MLPA analy-
sis to check for 22q11 deletion. The report confirmed such
deletion (Figure 39.7; rsa 22q11.2 (P250)x1, consistent
with the 22q11 deletion syndrome.
The clinical features of this baby were suggestive of
22q11.2-deletion syndrome[39]. Although many laborato-
ries use FISH analysis to detect this deletion, MLPA is being
increasingly used for the same purpose, as it also detects
the reciprocal duplication. Most patients with a hemizy-
gous deletion of 22q11.2 have outflow tract heart defects,
immune deficiency, transient neonatal hypocalcemia, velo-
pharyngeal insufficiency, and a distinctive facial appearance.
Most also have learning disabilities and behavioral anoma-
lies[40]. In view of these findings, it would be advisable to
carry out MLPA studies on both parents to exclude the pos-
sibility that the deletion is familial. The child himself is at a
50% risk of passing the deletion to any offspring.
A R R AY- C O M PA R AT I VE G E N O M I C
H Y B R I D I Z AT I O N (A R R AY- C G H )
Array-CGH (see glossary) is a genome-wide screen-
ing tool that has replaced conventional karyotyping for
the detection of structural chromosomal aberrations[41].
Conventional karyotyping can generally detect dosage
imbalances down to 5 million base pairs in size. However,
many newly identified recurrent imbalances below this
resolution have now been described. This new technique
has allowed the investigation of the whole genome to iden-
tify these newly described imbalances. Array-CGH allows
the detection of dosage imbalances down to 200,000 bases
in size. Array-CGH offers a much higher diagnostic yield
(15–20%) than a G-banded karyotype (5%) in patients pre-
senting with developmental delay, with or without congeni-
tal abnormalities[42].
Array-CGH is a technique based on the co-hybridization
of sample extracted DNA and control genomic extracted
DNA. DNA is fluorescently labelled with different dyes and
co-hybridized with mapped DNA sequences that are spotted
onto a glass slide surface (Figure 39.8; see also Chapter 10).
Amongst many potential clinical applications,
array-CGH has revolutionized cytogenetic investigations
 1.60

1.40

1.20

1.00

0.80

EHMT1(i)
SMARCB1(ii)
RAB36(ii)
SLC25A19

SMARCB1(i)

EHMT1(ii)
MICAL3

DKFZp566
ATDA1(il)
ATDA1(i)

AI651963
SHANK3

PPP1A3B
USP19

CUGBP2
RAB36(i)

SNAPD3

GATA4
GATA3

APH3AL(i)
KLKB1
GNAZ
BID

0.60

MSRA
NEBL
PPIL2

ARSA

CRK
APH3AL(ii)
TOP3B

FRG1

GEMIN4
HIC2
0.40
TXNRD2
CDC45L
CLTCL1

TBX1(ii)

KLHL22

SNAP29
TBX1(i)

DGCR9
CLDN5

PCQAP

LZTA1
GP1BB

ZNF74
HIRA

0.20

0.00

Figure 39.7 
Multiplex Ligation Dependent Probe Amplification (MLPA) using the P250 probe mix showed a deletion at 22q11.2 region consistent
with the 22q11.2 deletion syndrome.

in children with developmental delay and dysmorphic fea- stature. However, dysmorphic features did not suggest any
tures. A  number of dedicated systematic databases have specific genetic diagnosis. An MRI brain scan was normal.
now been set up that largely depend on array-CGH; for A  karyotype done previously showed an apparently bal-
example, the University of California, Santa Cruz (UCSC; anced de novo translocation between chromosomes 7 and
www.UCSC.ed) and the Wellcome Trust Sanger Institute 13 (46,XX,t(7;13)(q21.2;q12.3); Figure 39.8a). This did
in Hinxton, England (www.sanger.ac.uk). In the United not fully explain developmental delay and mild dysmorphic
Kingdom, the DECIPHER, an acronym of DatabasE of features.
Chromosomal Imbalance and Phenotype in Humans using Following discussion with a clinical geneticist, the
Ensembl Resources, was initiated in 2004 at the Sanger pediatrician requested an array-CGH analysis for the child.
Institute and funded by the Wellcome Trust. It docu- Array-CGH analysis showed that, in the regions of the
ments submicroscopic chromosome abnormalities and breakpoints on chromosomes 7 and 13, there had actually
maps them to the human genome using Ensembl or the been a loss of genetic material (DNA) (Figure  39.8a:  loss
UCSC Genome Browser. Its aims are to increase medical of genetic material indicated in red on the diagram of the
and scientific knowledge about chromosomal imbalances, chromosomes); thus the translocation was not balanced. It
improve medical care and genetic advice for affected indi- was agreed that the loss of genetic material from the long
viduals/families, and help facilitate research into the study arms of chromosome 13 and chromosome 7 was causally
of genes that influence human development and health. It is linked with the phenotype and its severity.
a database that is widely utilized to interpret chromosomal Array-CGH analysis of DNA at approximately 1
imbalances detected by array-CGH in routine diagnostic Mb resolution using the BlueGnome CytoChip (v1.01)
laboratories. array showed two interstitial deletions (Figure 39.8b).
An 8.42 Mb deletion of the long arm of chromosome 7
was detected. A 0.2 Mb deletion of the long arm of chro-
CASE 7
mosome 13 was also detected. These deletions appears to
A three-year-old boy was referred to the pediatrician in view be at, or close to, the breakpoints of the apparently bal-
of his developmental delay and some dysmorphic features. anced de novo translocation t(7;13)(q21.2;q12.3) previ-
On examination, the child had microcephaly and short ously reported in this child. It appears therefore that this

6 1 2   •   G enomics in C linical P ractice

 A

1 2 3 4 5

6 7 8 9 10 11 12

13 14 15 16 17 18

19 20 21 22 X Y

B 0
16
32
48
63
79
95
i) Chromosome 7
111
127
143
159
2.00 1.60 1.20 0.80 0.40 0.00 0.40 0.80 1.20 1.60 2.00
Chromosome 7
Log2 Ratio Ch1/Ch2

01100200_top-01100200_bottom.bsn - 14/06/2006

0
11
23
34
46
57
68
ii) Chromosome 13
80
91
103
114
2.00 1.60 1.20 0.80 0.40 0.00 0.40 0.80 1.20 1.60 2.00
Chromosome 13
Log2 Ratio Ch1/Ch2

01100200_top-01100200_bottom.bsn - 14/06/2006

Figure 39.8 
(A) Karyotype showing an apparently balanced reciprocal translocation between the long arm of chromosome 7, and the long arm of
chromosome 13, t(7;13)(q21.2;q12.3). (B) Array CGH analysis using the BlueGnome CytoChip 1Mb array chowing a 8.42 Mb deletion of
chromosome 7 and a 0.2 Mb deletion of chromosome 13.
patient has two cryptic deletions at, or near, both break-
points of the apparently balanced de novo translocation,
t(7;13)(q21.2;q12.3). Submicroscopic imbalances in
patients with de novo apparently balanced translocations
presenting with an abnormal phenotype have recently
been revealed at or close to the breakpoints by array-
CGH in about 50% cases[43, 44].
CASE 8
A five-year-old child was referred to the clinical geneticist
in view of developmental delay and dysmorphic features.
The child was born at 39 weeks’ gestation, and there were
no concerns during the antenatal or neonatal period. The
parents first became concerned when the child was not sit-
ting without support even by one year of age. On follow-up,
the child had delay in most of her motor milestones, as well
as speech. An MRI brain scan was arranged when she was
three years of age, which was normal. A conventional karyo-
type was requested by the pediatrician, which did not iden-
tify any abnormality.
On examination, the child had normal growth param-
eters, including head circumference, for her age. The geneti-
cist noticed facial dysmorphic features, including high
anterior hairline, prominent forehead, and epicanthic folds.
The systemic examination was normal. In view of the devel-
opmental delay and dysmorphic features, and the absence
of any definite clues to suggest a single-gene disorder, the
geneticist requested array-CGH analysis.
The array-CGH analysis of DNA from the above
child using the BlueGnome CytoChip Oligo ISCA
(v2.0) 8x60k array showed a deletion of the sequences
detected by the oligonucleotide probes located from
74,461,225 bp to 78,205,204 bp on chromosome 15
(Figure 39.9a,b). This equates to a 3.7 Mb interstitial
deletion of the long arm of chromosome 15 at q24.1 to
q24.3. In DECIPHER, 47 genes, of which 37 are listed as
Online Mendelian Inheritance in Man (OMIM) genes,
map within the deleted region. The 15q24 recurrent
microdeletion syndrome maps within the imbalance. The
deletion found in this child overlaps the newly identified
recurrent 15q24 recurrent microdeletion syndrome[45,
46]
. Several similar deletions have also been reported in
DECIPHER. It appears that deletions of this region are
associated with clinical problems such as intellectual dis-
ability, growth retardation, and dysmorphism.
In view of this finding, the next step will be to offer
testing to both parents of this child in order to determine
whether the abnormality has arisen de novo or has been
inherited. The clinical significance and recurrence risks can
6 1 4   •   G enomics in C linical P ractice
then be more fully assessed. This child is at 50% risk of pass-
ing on the abnormal chromosome 15 to any offspring she
may have.
CASE 9
A four-year-old boy was seen in the pediatric clinic in view
of his global developmental delay, hypotonia, and dysto-
nia. He was born at term, and there were no concerns in
the antenatal or neonatal period. There was no significant
family history. On examination, the child had mild muscu-
lar hypotonia with normal reflexes. There were no signifi-
cant dysmorphic features. The pediatrician had previously
arranged a karyotype, which did not identify any abnor-
malities. An array-CGH was requested.
Array-CGH analysis of DNA from this child using the
BlueGnome CytoChip Oligo ISCA (v2.0) 8x60k array
showed a copy number gain of the sequences detected by
the oligonucleotide probes located from 130,671,707 bp
to 134,868,378 bp on chromosome 11, and a copy number
loss of the sequences detected by the oligonucleotide probes
located from 230,451 bp to 4,077,584bp on chromosome
12 (Figure 39.10a,b,c). This equates to a 4.2 Mb terminal
copy number gain of the long arm of chromosome 11 from
11q24.3 to 11qter, and a 3.8 Mb terminal copy number
loss of the short arm of chromosome 12 from 12pter to
12p13.32. In DECIPHER, 13 HGNC genes, 10 of which
are OMIM genes, are listed in the duplicated region of chro-
mosome 11, and 28 HGNC genes, 23 of which are OMIM
genes, are listed in the deleted region of chromosome 12.
The abnormality detected in this child therefore
appears to consist of the loss of a small distal segment of
chromosome 12 short-arm material, and its replacement
by a small distal segment of chromosome 11 long-arm
material, probably as the result of an unbalanced translo-
cation. The child is therefore trisomic for the distal seg-
ment of the long arm of chromosome 11, 11q24.3->qter,
and monosomic for the distal segment of the short arm
of chromosome 12, 12p13.32->pter. The phenotypical
abnormalities noted in this child are almost certainly the
result of the chromosome imbalances revealed by aCGH.
Although there are no patients in DECIPHER with
similar copy number gains of chromosome 11, there are
multiple patients with similar deletions of chromosome
12. Phenotypical features noted in these patients include
intellectual disability, autism, muscular hypotonia, dysto-
nia, and delayed speech and language development, some
of which have also been noted in this patient. This child is
at 50% risk of passing on the derived chromosome 12 to
any future offspring.
 (A) 15:0kb

p13
p12
p11.2
p11.1
q12 q11.2
q14
q21.2 q21.1
q21.3
q22.2
q23
q25.3 q25.2 q25.1
q26.1
q26.2
q26.3

15:102, 531kb −2.00 −1.60 −1.20 −0.80 −0.40 −0.00 0.40 0.80 1.20 1.60 2.00
Log2 ratio Ch1/Ch2

Figure 39.9 
(A) Array CGH analysis using the CytoChip Oligo 8x60k (v2.0) array showing a 3.7Mb interstitial deletion at the long arm of
chromosome 15.
 (B)
RP11-414J4
Normal chromosome 15
Figure 39.9 
(B) Metaphase showing a deletion of the sequences recognized by the RP11-414J4 probe on the long arm of the chromosome 15
homologue using FISH, therefore confirming the array CGH analysis.
In view of this finding, the next step will be to offer
testing to both parents of this child in order to determine
whether one of the parents is a carrier of the balanced
chromosomal rearrangement. The clinical significance and
recurrence risks can then be more fully assessed.
R E C E N T N EW G E N O M I C
A P P L I C AT I O N S
There are several new genomic techniques that are being
investigated and rapidly introduced in clinical pediat-
ric practice. These include exome-sequencing focusing
all exomes of an individual’s genome, deep-sequencing
of specific genomic regions, and whole-genome sequenc-
ing (see also Chapter  10). Out of these, exome sequenc-
ing has some clinical utility when the putative gene can be
located and subsequently identified. This method has led
to the delineation and identification of a few genes for rare
genetic dysmorphic syndromes[47,48]. All these methods are
now vastly improved at significantly lower costs, with the
availability of new-generation sequencing platforms and
computational analysis tools. (A detailed discussion of the
techniques and specific disorders is beyond the scope of
this chapter.)
Genetic analyses for the identification and character-
ization of genetic disorders are an obvious application of
genomics in pediatrics. Most notable development has been
6 1 6   •   G enomics in C linical P ractice
Deleted
chromosome 15
in specific panels of microarrays or specific gene mutations
in tumors[49]. DNA microarrays can be used to identify
mutations and polymorphisms in either single large genes
or panels of genes whose mutation can result in clinically
similar phenotypes[50]. Such disease-specific microarrays
or gene panels are now commercially available, assisting
in rapid clinical diagnosis of rare disorders and infectious
diseases[51].
Microarray analyses using genomic DNA may also be
useful for testing panels of polymorphisms or mutations
that act as modifiers of risk in genetic disorders, or for com-
binations of genes that result in multigenic disorders. For
such analyses, panels of genes might be re-sequenced, or
profiles of large numbers of SNPs or copy number varia-
tions linked to disease genes or modifier genes could be
analyzed[17]. However, no such specific information could
be obtained for use in the promotion of early intervention
in cases of increased risk.
G E N O M I C C L A S S I FI C AT I O N
O F P E D I AT R I C D I S O R D E R S
The genomics-based tools developed following the com-
pletion of the Human Genome Project are being applied
to determine the genetic basis of pediatric disorders,
including congenital malformations. Genomics tech-
niques are being used to classify many different kinds of
diseases (see Chapters  10  & 20)  and to determine the
(A) (B)

(C)
11q subtelomere probe 12p subtelomere probe

Derivative chromosome 12

Normal chromosome 12

Normal chromosome 11s

Figure 39.10 
(A) & (B) Array CGH analysis using the CytoChip Oligo 8x60k (v2.0) array showing a 4.2Mb copy number gain of the long arm of
chromosome 11, and a 3.8 Mb copy number loss of the short arm of chromosome 12. (C) Metaphase showing a derived chromosome with a partial loss of
the terminal segment of one chromosome 12 long arm material and replaced with a chromosome segment from the short arm of chromosome 11 material.
risk associated with the various classifications in order
to refine therapeutic approaches. Among the pediatric
disorders reported to have been studied to date are psy-
chiatric and developmental disorders, rheumatological
and immune disorders, digestive system disorders (such
as celiac disease, inflammatory bowel disease, and bili-
ary atresia), and renal disorders (Table 39.3). However,
most of the work has been done in oncology, in order
to identify prospectively the prognoses associated with
histologically similar tumors, based on significant differ-
ences in their gene-expression signatures. The first exam-
ple of this approach in pediatric oncology demonstrated
that medulloblastoma gene-expression profiles could be
used, not only to distinguish between different types of
tumors, but also to predict survival[52]. Since then, studies
have been reported for a variety of leukemias, lymphomas,
and solid-tumor malignancies, particularly brain tumors.
From the standpoint of cancer biology and pathophysiol-
ogy, the identification of genes differentially expressed in
tumors with a poor prognosis might suggest approaches
to the development of novel and potentially more effec-
tive therapies for patients with such tumors. Clinically,
such gene-expression profiling provides a potentially
more objective approach to disease classification than tra-
ditional histopathological approaches.
I N FEC T I O US D I S E A S E S
A key challenge in patients with infectious diseases is to
detect and characterize pathogens, particularly for organ-
isms that are difficult to culture, such as Mycobacterium
tuberculosis, so that therapy can be initiated in a timely
manner. Genomics approaches can also be applied to such
challenges once the relevant pathogen genome(s) have
been sequenced (see Chapters 11 and 37). This sequence
information is necessary for the preparation of microar-
rays that can detect genomic DNA from the pathogen(s)
of interest[53]. The use of microarrays can potentially
decrease the time necessary to identify a central nervous
system pathogen from many days currently, to just a few
hours[54], which would allow targeted therapy to begin
sooner. These assays can also be used to identify patho-
gen subtypes and the presence of antimicrobial-resistance
markers[55]. Examples of such approaches include the use
of microarrays to detect, subtype, and determine rifampin
resistance for various Mycobacterium species[56], and for
the sensitive detection of E.  coli O157:H7, an impor-
tant food-borne cause of hemorrhagic colitis and the
hemolytic-uremic syndrome[57].
6 1 8   •   G enomics in C linical P ractice
P H A R M AC O G E N ET I C S
A N D P H A R M AC O G E N O M I C S
An exciting current frontier for genomic medicine is to use
an individual’s genomic information to “personalize” drug
therapy through the approaches of pharmacogenetics and
pharmacogenomics (see also Chapter 7). Such personaliza-
tion will require paying attention to both the patient’s age
and genome, because the targets and toxicities of medica-
tions change with age[58]. Thus, investigation of developmen-
tal pharmacogenetics and pharmacogenomics is needed to
understand relevant age-related changes in the expression
of drug targets and in drug metabolism, so that the most
efficacious drugs with the least risk for adverse reactions for
a patient at a given age can be identified.
ET H I C A L , L E G A L , A N D S O C I A L
I M P L I C AT I O N S U N I Q U E TO
P E D I AT R I C S
No discussion of the use of genomics in clinical pediatrics
would be complete without a thorough consideration of
the ethical, legal, and social implications of this emerg-
ing technology (see also Chapter 17). Traditional medical
ethics has been governed by four principles:  beneficence
(doing good), non-maleficence (doing no harm), autonomy
(the right to choose), and justice (being fair and equitable).
These principles also apply to the use of genomic informa-
tion in pediatrics, although the “usual” ethics may need
to be modified to deal with the unique ethical quandaries
found in genomic medicine[59].
Genomic medicine, by generating vast amounts of
personal genetic data, poses significant ethical dilemmas,
both for individuals and for society, and the potential for
abuse of this information is real. A major concern is that
the existence of such detailed genetic or genomic infor-
mation could result in “genetic discrimination” to deny
access to health and life insurance or employment[60]. In
fact, public pressure in the United States has resulted in
the passage of laws prohibiting genetic discrimination and
regulating the use of genetic information, and U.S.  fed-
eral legislation has been debated in many recent sessions
of Congress[61]. Although genetic discrimination may
be a widespread public concern, there are few examples
of such discrimination and, as yet, no evidence that it is
common[62]. Most lawsuits that have been filed involve
cases of actual disease identified by genetic testing, not
pre-symptomatic or predispositional conditions. There
are almost no well-documented cases of health insurers’
Table 39.3  GENOMIC APPROACHES IN SELECTED PEDIATRIC DISORDERS

Cardiology
Kawasaki disease Nomura et al., 2005
Kawasaki disease (response to therapy) Abe et al., 2005
Right ventricular outflow obstructive lesions Konstantinov et al., 2004
Dentistry
Dental caries Paakkonen et al., 2005
Dermatology
Giant hairy nevi Dasu et al., 2004
Hypertrophic and normal scars Tsou et al., 2000
Gastroenterology/Hepatology
Biliary atresia Bezerra et al., 2002
Celiac disease Diasclado et al., 2004
Eosinophilic esophagitis Blanchard et al., 2006
Inflammatory bowel disease Dieckgraefe et al., 2000
Genetics
Duchenne muscular dystrophy Noguchi et al., 2003
Mitochondrial disorders Crimi et al., 2005
Rett syndrome Colantuoni et al., 2001
X-linked myotubular myopathy Noguchi et al., 2005
Infectious diseases
Neisseria meningitides Sun et al., 2000
Parvo virus B19 infection Kerr et al., 2005
Vaccine-derived polio virus Laassri et al., 2005
Neonatology/Obstetrics
Premature rupture of membranes Tromp et at, 2004
Nephrology
Acute renal allograft rejection Sarwal et al., 2003
Focal segmental glomerulosclerosis Schwab et al., 2004
Neurology/Psychiatry
Brain arteriovenous malformations Hashimoto et al., 2004
Neurofibromatosis type 1 Tang et al., 2004
Neurological diseases Tang et al., 2005
Tardive dyskinesia in therapy of schizophrenia Nikoloff et al., 2002
Oncology
Acute megakaryoblastic leukemia in Down syndrome Lightfoot et al., 2004
Acute myelogenous leukemia Ross, M. E. et al., 2004
Acute myeloid leukemia Yagi et al., 2003
Ependymoma Suarez-Merino et al., 2005
Ewing’s sarcoma Okali et al., 2004
Juvenile pilocytic astrocytomas Wong et al., 2005
Large cell lymphomas Thompson et al., 2005
Leukemias Moos et al., 2002
Medulloblastoma Pomeroy et al, 2002; Fernandez-Teijeiro et al., 2004

(continued)
 Table 39.3 CONTINUED
Neuroblastoma
Retinoblastoma
Wilms’ tumor
Rheumatology/immunology
Hyper-IgE syndrome
Rheumatoid diseases
Systemic lupus erythematosus
Other
Muscle-wasting following burns
Recurrent aphthous ulceration
Unexplained drowning (molecular autopsy)
SOURCE: Adapted with permission from ‘Applications in Clinical Pediatrics’, by Michael R. Konikof and Michael D. Bates, p. 425–436. In
‘Genomics and Clinical Medicine’, Eds. Kumar, D & Weatherall, DJ. Oxford University Press, NY, 2008.
collecting or using pre-symptomatic genetic test results
in their underwriting decisions, and little evidence that
such testing affects a person’s ability to obtain health
insurance[63]. Despite this apparent lack of improper use
of genetic information, as our knowledge of genomics
increases, these dilemmas are likely to increase in both fre-
quency and importance. Deciding what to do about such
predispositions will depend on the likelihood of occur-
rence in at-risk individuals and the natural history of the
disease. The challenge for clinicians will be to discuss with
their patients the possible adverse social consequences of
testing so that patients can make informed decisions about
whether or not to proceed with the testing[64].
Genetic testing of children presents unique ethical con-
siderations. Such testing may be diagnostic or predictive (see
also Chapter 14). Diagnostic genetic testing refers to testing
of a symptomatic child’s genetic material to establish a med-
ical diagnosis. This type of testing, with direct benefit to the
child who is symptomatic, is generally non-controversial.
Predictive genetic testing, on the other hand, refers to the
testing of asymptomatic children for genetic conditions
that may present months to years in the future, if at all.
Predictive testing can be pre-symptomatic (i.e., virtually all
children with the specific genotype will develop the disease)
or predispositional (i.e., children with the specific genotype
are at risk of developing the disease, but not certain to do
so). Predictive genetic testing can be performed for diseases
with either childhood- or adult-onset, and this testing may
offer direct, indirect, or no benefit to the child. If there are
treatments or preventative measures available for the con-
dition being tested, predictive genetic testing of children
should be offered, and, in some cases, strongly advised, simi-
lar to considerations given to neonatal screening. If there
6 2 0   •   G enomics in C linical P ractice
Hiyarna et al., 2004
Gratias et al., 2005
Williams et al., 2004
Tanaka et al., 2005
Barnes et al., 2004
Bennett et al., 2003
Barrow et al., 2003; Dasu et al., 2005
Borra et al., 2004
Tester et al., 2005
are no treatments or preventative measures available, or the
condition is adult-onset, testing of children becomes more
controversial, as is the testing of children for carrier status.
Here, the need-to-know of the child and parents must be
weighed against the adverse effects of a harmful diagnosis
on family relationships and the child’s self-esteem and devel-
oping social identity. In addition, mature adolescents who
desire genetic testing for future reproductive planning or
other reasons must be considered differently than younger
children who have not yet reached the stage of formal
operational thought and thus are unable to give assent. In
any case, thorough genetic counseling is a necessity for any
genetic testing of children, and should include the child, the
parents, and the siblings. In most of these situations, genetic
testing of children should be deferred until the child reaches
adulthood or can make reproductive decisions[65].
S U M M A RY
The application of genomic medicine approaches to the
care of children is not as advanced as for adults because
of the relatively small numbers of affected individu-
als for pediatric as opposed to adult disorders, as well as
because of the unique ethical and social implications for
such applications in the care of children. However, clini-
cal pediatrics offers unique and challenging opportunities
where historically all conventional and new genetic and
genomic laboratory methods are increasingly applied.
This chapter has highlighted a few classic case scenarios
wherein both conventional and newly emerging genomic
methods of array comparative genomic hybridization are
demonstrated to have contributed to clinical diagnosis
and interpretation of underlying causal and biological
mechanisms. The chapter also points to the successes of
neonatal-screening programs that utilize genomic and
proteomic methods in rapid screening of high-risk neo-
nates who could be selectively investigated and supported
by dietary and other medical interventions. This approach
is by far the best example of public health applications
of genetics and genomics. Finally, as the experience with
genomic medicine is gained by its application in adults,
and as genomic medicine in general becomes acceptable
to society, the approaches of genomic medicine would be
expected to play an increasing role in the care of children,
as for adults.
AC K N OW L E D G E M E N T S
The authors are grateful to all clinicians for their permission
to include selected cases in this chapter. Special thanks are
due to Professor Dhavendra Kumar, the main Editor of this
edition, for providing access to the first edition, Genomics
and Clinica Medicine, and allowing to adapt limited text
material from some sections of the chapter ‘Applications in
Clinical Pediatrics’ (by Michael R.  Konikof and Michael
D. Bates, pages 425–436).
REFERENCES
1. Bodamer, O.A., Screening for phenylketonuria. Annales Nestlé
(English ed.). 2010. 68(2): pp. 53–57.
2. Horner, F.A., et  al., Termination of dietary treatment of phenyl-
ketonuria. New England Journal of Medicine, 1962. 266(2): pp.
79–81.
3. Allen, R.J., The detection and diagnosis of phenylketonuria. American
Journal of Public Health and the Nation’s Health, 1960. 50(11): pp.
1662–1666.
4. Naylor, E.W., and R. Guthrie, Newborn-screening for maple syrup
urine disease (branched-chain ketoaciduria). Pediatrics, 1978.
61(2): pp. 262–266.
5. Schweitzer-Krantz, S., Early diagnosis of inherited metabolic disorders
towards improving outcome: the controversial issue of galactosaemia.
European Journal of Pediatrics, 2003. 162(1): pp. S50–S53.
6. Garrick, M.D., P. Dembure, and R. Guthrie, Sickle-cell anemia
and other hemoglobinopathies:  procedures and strategy for screening
employing spots of blood on filter paper as specimens. New England
Journal of Medicine, 1973. 288(24): pp. 1265–1268.
7. Dussault, J.H., et  al., Preliminary report on a mass screening pro-
gram for neonatal hypothyroidism. The Journal of pediatrics, 1975.
86(5): pp. 670–674.
8. White, P.C., Neonatal screening for congenital adrenal hyperplasia.
Nature Reviews Endocrinology, 2009. 5(9): pp. 490–498.
9. Loeber, G., D. Webster, and A. Aznarez, Quality evaluation of
newborn-screening programs. Acta Paediatrica, 1999. 88(s432):
pp. 3–6.
10. Pandor, A., et  al., Clinical effectiveness and cost-effectiveness of neo-
natal screening for inborn errors of metabolism using tandem mass
G enomic A pplications in C linical P ediatrics   •   6 2 1
spectrometry:  a systematic review. Health Technol Assess, 2004.
8(12): pp. iii, 1–121.
11. Lin, Z., et al., A simple automated DNA extraction method for dried
blood specimens collected on filter paper. Journal of the Association for
Laboratory Automation, 2005. 10(5): pp. 310–314.
12. Dobrowolski, S.F., et  al., DNA microarray technology for neonatal
screening. Acta Paediatrica, 1999. 88(s432): pp. 61–64.
13. Chace, D.H., T.A. Kalas, and E.W. Naylor, Use of tandem mass spec-
trometry for multianalyte screening of dried blood specimens from new-
borns. Clinical Chemistry, 2003. 49(11): pp. 1797–1817.
14. Waisbren, S.E., Newborn-screening for metabolic disorders.
JAMA:  The Journal of the American Medical Association, 2006.
296(8): pp. 993–995.
15. Minutti, C.Z., et al., Steroid profiling by tandem mass spectrometry
improves the positive predictive value of newborn-screening for con-
genital adrenal hyperplasia. Journal of Clinical Endocrinology and
Metabolism, 2004. 89(8): pp. 3687–3693.
16. Fan, J.-B., et  al., Parallel genotyping of human SNPs using generic
high-density oligonucleotide tag arrays. Genome research, 2000.
10(6): pp. 853–860.
17. Hacia, J.G., and F.S. Collins, Mutational analysis using oligonucleotide
microarrays. Journal of medical genetics, 1999. 36(10): pp. 730–736.
18. Toder, R., DNA arrays as diagnostic tools in human healthcare.
Expert review of molecular diagnostics, 2002. 2(5): pp. 422–428.
19. Bailey, D.B., et al., Ethical, legal, and social concerns about expanded
newborn screening:  fragile X syndrome as a prototype for emerging
issues. Pediatrics, 2008. 121(3): pp. e693–e704.
20. Kalman, L., et  al., Mutations in genes required for T-cell develop-
ment:  IL7R, CD45, IL2RG, JAK3, RAG1, RAG2, ARTEMIS,
and ADA and severe combined immunodeficiency:  HuGE review.
Genetics in Medicine, 2004. 6(1): pp. 16–26.
21. Green, N.S., S.M. Dolan, and M. Oinuma, Implementation of
newborn-screening for cystic fibrosis varies widely between states.
Pediatrics, 2004. 114(2): pp. 515–516.
22. Green, N.S., and K.A. Pass, Neonatal screening by DNA microar-
ray: spots and chips. Nature reviews genetics, 2005. 6(2): pp. 147–151.
23. Jain, S., and P. Heutink, From single genes to gene networks:
high-throughput-high-content screening for neurological disease.
Neuron. 68(2): pp. 207–217.
24. Mendell, J.T., and H.C. Dietz, When the message goes awry:
disease-producing mutations that influence mRNA content and per-
formance. Cell, 2001. 107(4): pp. 411–414.
25. Gruters, A., and Krude, H, Update on the management of congenital
hypothyroidism. Horm Res, 2007. 68(Suppl 5): p. 107.
26. Andermann, A., et al., Revisiting Wilson and Jungner in the genomic
age: a review of screening criteria over the past 40 years. Bulletin of the
World Health Organization, 2008. 86(4): pp. 317–319.
27. Wusthoff, C., The dilemma of confidentiality in Huntington disease.
JAMA:  The Journal of the American Medical Association, 2003.
290(9): pp. 1219–1220.
28. Ross, L.F., Predictive genetic testing for conditions that present in
childhood. Kennedy Institute of Ethics Journal, 2002. 12(3):
pp. 225–244.
29. Barker, J.M., et al., Prediction of autoantibody positivity and progres-
sion to type 1 diabetes:  Diabetes Autoimmunity Study in the Young
(DAISY). Journal of Clinical Endocrinology and Metabolism,
2004. 89(8): pp. 3896–3902.
30. Lejeune, J.R.M., Les caryotypes de la trisomie 21. La Revue du
Praticien, 1964. 14: pp. 57–68.
31. Cereda, A., and J.C. Carey, The trisomy 18 syndrome. Orphanet jour-
nal of rare diseases. 7(1): pp. 81.
32. Saenger, P., et  al., Recommendations for the diagnosis and manage-
ment of Turner syndrome. Journal of Clinical Endocrinology and
Metabolism, 2001. 86(7): pp. 3061–3069.
33. Devlin, L., and P. Morrison, Mosaic Down’s syndrome prevalence in
a complete population study. Archives of disease in childhood, 2004.
89(12): pp. 1177.
 34. Klinger, K., et  al., Rapid detection of chromosome aneuploidies in
uncultured amniocytes by using fluorescence in situ hybridization
(FISH). American journal of human genetics, 1992. 51(1): pp. 55.
35. Game, X., et al., Williams Beuren Syndrome. New England Journal
of Medicine. 362(15): pp. 1449.
36. Cassidy, S.B., and D.J. Driscoll, Prader-Willi syndrome. European
Journal of Human Genetics, 2008. 17(1): pp. 3–13.
37. Buiting, K. Prader-Willi syndrome and Angelman syndrome. in
American Journal of Medical Genetics Part C: Seminars in Medical
Genetics, 2010. 154C(3): pp. 365–376.
38. Sellner, L.N., and G.R. Taylor, MLPA and MAPH: new techniques
for detection of gene deletions. Human mutation, 2004. 23(5):  pp.
413–419.
39. Kapadia, R.K., and A.S. Bassett, Recognizing a common genetic syn-
drome:  22q11.2-deletion syndrome. Canadian Medical Association
Journal, 2008. 178(4): pp. 391–393.
40. Bassett, A.S., et  al., Practical guidelines for managing patients with
22q11.2-deletion syndrome. Journal of pediatrics. 159(2): pp. 332.
41. Rickman, L., et  al., Prenatal detection of unbalanced chromosomal
rearrangements by array-CGH. Journal of medical genetics, 2006.
43(4): pp. 353–361.
42. Stankiewicz, P., and A.L. Beaudet, Use of array-CGH in the evalu-
ation of dysmorphology, malformations, developmental delay, and
idiopathic mental retardation. Current opinion in genetics and
development, 2007. 17(3): pp. 182–192.
43. Baptista, J., et  al., Breakpoint mapping and array-CGH in trans-
locations:  comparison of a phenotypically normal and an abnor-
mal cohort. The American Journal of Human Genetics, 2008.
82(4): pp. 927–936.
44. De Gregori, M., et al., Cryptic deletions are a common finding in
‘balanced’ reciprocal and complex chromosome rearrangements:
a study of 59 patients. Journal of medical genetics, 2007. 44(12):
pp. 750–762.
45. Magoulas, P.L., and A.W. El-Hattab, Chromosome 15q24 microdele-
tion syndrome. Orphanet journal of rare diseases. 7(1): pp. 2.
46. Mefford, H.C., et  al., Further clinical and molecular delineation of
the 15q24 microdeletion syndrome. Journal of medical genetics.
49(2): pp. 110–118.
47. Puffenberger, E.G., et  al., Genetic mapping and exome sequenc-
ing identify variants associated with five novel diseases. PLoS One.
7(1): pp. e28936.
48. Poot, M., and R. Hochstenbach, A three-step workflow procedure for
the interpretation of array-based comparative genome hybridization
results in patients with idiopathic mental retardation and congenital
anomalies. Genetics in Medicine. 12(8): pp. 478–485.
49. Amatschek, S., et  al., Tissue-wide expression profiling using cDNA
subtraction and microarrays to identify tumor-specific genes. Cancer
research, 2004. 64(3): pp. 844–856.
6 2 2   •   G enomics in C linical P ractice
50. Schork, N.J., D. Fallin, and J.S. Lanchbury, Single nucleotide poly-
morphisms and the future of genetic epidemiology. Clinical genetics,
2000. 58(4): pp. 250–264.
51. Chen, B., et  al., Developing a sustainable process to provide quality
control materials for genetic testing. Genetics in Medicine, 2005.
7(8): pp. 534–549.
52. Pomeroy, S.L., et  al., Prediction of central nervous system embryo-
nal tumour outcome based on gene expression. Nature, 2002.
415(6870): pp. 436–442.
53. Call, D.R., M.K. Borucki, and F.J. Loge, Detection of bacterial
pathogens in environmental samples using DNA microarrays.
Journal of microbiological methods, 2003. 53(2): pp. 235–243.
54. Hanson, E.H., et al., Potential use of microarray technology for rapid
identification of central nervous system pathogens. Military medicine,
2004. 169(8): pp. 594–599.
55. Bodrossy, L., and A. Sessitsch, Oligonucleotide microarrays in micro-
bial diagnostics. Current opinion in microbiology, 2004. 7(3):
pp. 245–254.
56. Kato-Maeda, M., et  al., Comparing genomes within the species
Mycobacterium tuberculosis. Genome research, 2001. 11(4):
pp. 547–554.
57. Call, D.R., F.J. Brockman, and D.P. Chandler, Detecting and geno-
typing Escherichia coli O157: H7 using multiplexed PCR and nucleic
acid microarrays. International Journal of Food Microbiology, 2001.
67(1): pp. 71–80.
58. Russo, R., et  al., Pediatric pharmacogenetic and pharmacogenomic
studies: the current state and future perspectives. European journal of
clinical pharmacology. 67(1): pp. 17–27.
59. O’Lonergan, T.A., and H. Milgrom, Ethical considerations in
research involving children. Current allergy and asthma reports,
2005. 5(6): pp. 451–458.
60. Parthasarathy, S., Regulating risk:  defining genetic privacy in the
United States and Britain. Science, Technology and Human Values,
2004. 29(3): pp. 332–352.
61. Morrow, J.S., Insuring fairness: the popular creation of genetic antidis-
crimination. The Georgetown Law Journal, 2009. 98: pp. 215.
62. Everett, M., The social life of genes:  privacy, property and the new
genetics. Social science and medicine, 2003. 56(1): pp. 53–65.
63. Hall, M.A., and S.S. Rich, Patients’ fear of genetic discrimination by
health insurers: the impact of legal protections. Genetics in Medicine,
2000. 2(4): pp. 214–221.
64. Guttmacher, A.E., F.S. Collins, and E.W. Clayton, Ethical, legal,
and social implications of genomic medicine. New England Journal of
Medicine, 2003. 349(6): pp. 562–569.
65. Wade, C.H., B.A. Tarini, and B.S. Wilfond, Growing up in the
genomic era: implications of whole-genome sequencing for children,
families, and pediatric practice. Annual review of genomics and
human genetics. 2013. 14: pp. 535–555.
 40.
GENETICS AND GENOMICS IN CLINICAL
OPHTHALMOLOGY, I: THE SPECTRUM OF GENETIC
EYE DISEASE
Graeme Charles M. Black
I N T R O D U C T I O N :  H OW MU C H O F
C L I N I C A L O P H T H A L M O L O GY I S
G E N ET I C ?
The overall contribution of genetics to the etiology of all
aspects of ophthalmic disease is broader than is generally
accepted. Of course, when faced in the clinic with a patient
with retinitis pigmentosa, the ophthalmologist is immedi-
ately aware of the inherited basis of the presenting condi-
tion. No less significant, however, are the genetic etiologies
that are recognized for rare sporadic diseases such as anoph-
thalmia, microphthalmia, or optic nerve hypoplasia; for
common early-onset phenotypes such as squint or myopia;
and for later-onset phenotypes such as primary open-angle
glaucoma, senile cataract, and, in particular, for macular
degeneration.
Human Mendelian conditions associated with ophthal-
mic phenotypes account for perhaps a third of the single-gene
disorders documented in the Online Mendelian Inheritance
in Man database (OMIM: http://www.ncbi.nlm.nih. gov/
entrez/query.​fcgi?db=OMIM​), while the retina’s central
nervous system origins bestow it with a broad range of gene
expression. The single-gene (Mendelian) group of disorders
can affect any tissue of the eye, encompass all age groups,
and affect every ophthalmic subspecialty. Those that lack
systemic associations have little impact upon reproductive
fitness and have represented a powerful resource of exten-
sively characterized conditions for gene-identification pro-
grams. As a result of their relative ease of study—initially
through linkage studies, later via autozygosity analysis, and
most recently as a result of next-generation sequencing—
this group of disorders has received considerable attention.
This is demonstrated by the number of mapped and cloned
genes known to underlie retinal disease, which in the ten
years preceding 2013 increased to 206 and 246 as described
623
in the Retnet database (http://www.sph.​uth.tmc.edu/
RetNet/home.htm​).
The study of these rare phenotypes may be deemed to be
apparently disproportionate, but it is perhaps worthwhile
to reflect that rare disorders as a collective grouping (“rare”
defined as those affecting less than one individual in 2,000)
affect around one in 17 individuals in the United Kingdom,
to some degree explaining and justifying such an intensive
study. Many such conditions are of early-onset and, for the
pediatric ophthalmologist, constitute an important group
that includes perhaps as many as half of those children reg-
istered as blind or partial-sighted in industrialized countries
(Rahi et  al., 2003; Gilbert and Foster 2001; Alagaratnam
et al., 2002).
Monogenic conditions often also represent excellent
models for complex phenotypes. This is demonstrated by
macular dystrophies such as Stargardt disease as a model for
age-related macular degeneration (AMD; Westeneng van
Haaften et al., 2012). They are a powerful source for novel
drugs and genetic medicines (Antonarakis and Beckmann,
2006; Brinkman et  al., 2006; O’Connor and Crystal,
2006)  and some, such as corneal dystrophies (e.g., Fuchs’
dystrophy, brittle cornea syndrome) represent important
models of quantitative traits such as central corneal thick-
ness (Rohrbach et  al., 2013; Vithana et  al., 2011). Of
course, it is these commoner ophthalmic conditions that
both contribute most significantly to world blindness and
are encountered most frequently by the general ophthal-
mologist. This group includes early-onset (myopia, squint)
and late-onset disorders (age-related macular degeneration
[see Chapter  46], senile cataract, and primary open-angle
glaucoma [POAG; see Chapter  45]) and many have a
strong genetic component.
For age-related cataract, segregation analyses suggest
the existence of a major genetic component that accounts
for about half of cataract variability, while twin studies also
show significant heritability for both cortical (53–58%)
and nuclear cataracts (48%) (Heiba et al., 1995; Hammond
et al., 2000, 2001). POAG also has a strong genetic com-
ponent, which is known thanks to family and twin concor-
dance studies and a range of approaches including the study
of single-gene disorders and association studies. For a long
time, the study of common disorders lagged behind that of
monogenic disorders. However, successful genome-wide
searches have identified many potential loci—for example,
for POAG (9q22, 20p12), cataract (6p12-q12), and AMD
(1q31, 10q26 and 17q25) (Wiggs et al., 2004; Iyengar et al.,
2004; Weeks et al., 2004; Ozel et al., 2014); and have identi-
fied predisposing variations or polymorphisms in a number
of genes—such as for age-related cataract in EPHA2 and
SLC16A12, and for POAG in Col8A2 (Yang et al., 2013;
Zuercher et al., 2010; Desronvil et al., 2010). The increasing
impact of genome-wide searches in the study of common
disease is illustrated by the recent success of a combined
positional/candidate-gene approach that has identified
genes that make an important contribution to AMD (see
Chapter 46). Overall, as a consequence, there is now a real-
istic prospect that we are beginning to better understand
groups of conditions whose complex multigenic nature has
previously hindered progress. Identification of the comple-
ment system in AMD pathogenesis, for example, now
presents novel therapeutic targets for a disease that was pre-
viously difficult to study and treat.
The need for a working knowledge of genetics is
becoming increasingly mainstream as that science
actively contributes to clinical practice. For conditions
such as retinoblastoma, inherited retinal dystrophies,
and congenital cataract, genetic testing already has a part
to play in the diagnosis, counselling, and management
of patients and their families. Such services are becom-
ing more comprehensive and widely available with the
advent of novel technologies. As genetic testing con-
tributes to the clinician’s ability to diagnose and has a
potentially profound impact upon decision-making for
patients and families, the power to alter disease progres-
sion or management has seldom been so tangible. This
is exemplified in the progress that has been made in the
development of novel gene-based therapies. Viral-based
gene-delivery approaches to the retina and/or retinal
pigment epithelium diseases have now been attempted
in rodents and larger animals such as the RPE65- and
RPGRIPL1-deficient dogs (Acland et al., 2001; Lhériteau
et  al., 2014). Most recently, promise has been shown in
trials in humans (e.g., Maguire et  al., 2009; Bainbridge
et al., 2008). Such gene-therapy studies demonstrate that
6 2 4   •   G enomics in C linical P ractice
retinal morphology and function can be improved and
that disease progression can also be altered.
MO N O G E N I C D I S E A S E A N D
O P HT H A L MO L O GY
The impact and importance of classical Mendelian genetics
in ophthalmology has been profound and is well illustrated
using the examples of ocular patterning defects, congenital
cataract, and retinitis pigmentosa.
Defects of Ocular Patterning
Defects of early ocular development are often lumped into
a single heterogeneous group termed microphthalmia. This
represents not one, but a spectrum of clinical entities which
can be referred to as MAC (microphthalmia, anophthal-
mia, coloboma). These may be unilateral or bilateral and
may range from complete anophthalmia, through microph-
thalmia (defined as an eye with an axial length of <19.5 mm
at the age of one year) to nanophthalmia (defined as a short
axial length, high hypermetropia, and shallow anterior seg-
ment, but normal visual function). Such defects result from
interruptions in the earliest processes of ocular develop-
ment, including the formation of the optic vesicle (which
forms as neuroectodermal evaginations of the forebrain),
the induction of the lens placode/vesicle, and the differen-
tiation of the neural retina and retinal pigment epithelium
(RPE). Since, in many cases, there may be an associated fail-
ure of the optic fissure to close (resulting in the formation
of a colobomatous defect), this, too, is seen as a key event.
The underlying cause of ocular developmental abnor-
malities is varied and may be associated with a large num-
ber of single-gene disorders as well as developmental
abnormalities of a broad etiology, including chromosomal
abnormalities, copy number variation, maternal infection
or toxin ingestion (e.g., anti-epileptic drugs or alcohol),
environmental influences, and fetal disruptions. There has
been considerable progress in understanding the effects
of copy number variation on ocular developmental dis-
orders (Bardakjian et  al., 2009; Raca et  al., 2011). An
increasing number of single-gene defects are recognized
as underlying these conditions, both in humans and in
other mammals. CHX10 and RAX (neural retina-specific
homeodomain-containing transcription factors) were
identified using a candidate-gene approach and have been
defined in only a very small number of patients (Voronina
et al., 2004; Percin et al., 2000). Homozygosity approaches
recognized a number of other contributing genes, including
ALDH1A3 (Fares-Taie et  al., 2013)  and STRA6 (Pasutto
et  al 2007), while next-generation sequencing has now
shown the power of identifying de novo dominant causes of
ocular developmental disorders (Srour et al., 2013). These
observations increase understanding of the interconnecting
pathways and cell-specific defects that can lead to defects
in organogenesis—in the case of RAX, the evidence from
human, rodent, and other vertebrate models suggests that it
acts as a transcriptional activator that determines retinal cell
fate (Furukawa et al., 1997; Giannaccini et al., 2013).
As in many ocular genetic conditions, there is consider-
able intrafamilial phenotypical variability amongst ocular
developmental disorders—for example, amongst patients
with congenital cataract, optic fissure closure defects, and
anterior segment dysgenesis. This suggests, as well as envi-
ronmental influences, the existence of genetic modifiers.
There has been limited progress in the developmental area,
with the identification of digenic effects amongst patients
with developmental glaucoma phenotypes (Kelberman
et al., 2011). Many of the genes that control ocular devel-
opment have key extraocular roles and, consequently,
mutations in these genes underlie multisystemic disease.
For example, patients carrying PAX6 mutations may also
have impaired olfaction, cognitive dysfunction, or autism
(Heyman et al., 1999; Malandrini et al., 2001; Chao et al.,
2003). MRI scans of patients with OTX2 and SOX2 muta-
tions have been described with neurodevelopmental pheno-
types (Ragge et al., 2005a, 2005b). There is recognition of
significant degrees of non-penetrance; for example, amongst
patients carrying mutations in OTX2 (Ragge et al., 2005),
patients with unilateral disease carrying familial SOX2
mutations, and also of phenotypical heterogeneity amongst
patients with mutations in the same gene. This understand-
ing has allowed an extension of the phenotypical range for
disease-causing genes—for example, the identification of
patients with early-onset severe retinal dystrophies who
carry mutations in OTX2 (Henderson et al., 2009).
Genetic testing for patients with ocular developmental
defects can facilitate recurrent risk estimation and accurate
diagnosis. Currently, the majority of patients with MAC
lack a recognized molecular defect, which suggests that a
range of genetic defects/mechanisms remain to be identi-
fied. Nonetheless, such findings are of increasing clini-
cal value—for example, pathogenic mutations SOX2 and
OTX2 have been found in significant numbers (~15–20%)
of patients with severe bilateral anophthalmia and microph-
thalmia (Ragge et  al., 2005a, 2005b; Zhou et  al., 2008),
presenting a valuable means for the genetic screening of
patients with the severest forms of bilateral ocular devel-
opmental disorders. In the near future, it is likely that sys-
tematic high-throughput sequencing programs will identify
G enetics and G enomics in C linical O phthalmology, I   •   6 2 5
novel underlying genetic causes where they exist, and that
this will lead the way to utilizing exome sequencing for
diagnosis (Yang et al., 2013). Since many of these molecular
defects are likely to be rare and also de novo, this will require
a trio-sequencing analysis (Dyment et  al., 2013)  and will
present considerable bioinformatic difficulties. Since in
around half of cases microphthalmia has multisystemic
associations, careful clinical phenotyping will be criti-
cal to enable the interpretation of gene-specific structural
abnormalities.
Congenital Cataract
Cataract is defined as any opacification of the crystalline eye
lens, and congenital forms have an incidence of around 2–5
per 10,000 live births in the United Kingdom (Rahi et al.,
2001). Around half of these cases are inherited, either as an
isolated defect or as part of a multisystemic condition.
As with other developmental ocular disorders, the
detailed understanding and diagnosis of congenital cataract
are hindered by the condition’s considerable genetic hetero-
geneity. There are well over 100 genes known to be mutated
in human childhood cataract, and delayed diagnosis com-
plicates both diagnosis and management. The intimate
interactions of the lens placode, neuroectoderm, and ocular
mesenchyme during development of the globe, anterior seg-
ment, and lens mean that there is a considerable phenotypi-
cal range of disorders due to altered lens development. In the
past, a detailed understanding of genotype–phenotype cor-
relation was limited by the strong reliances on single-family
analysis and on conventional sequencing technologies(e.g.,
Jia et  al., 2013; Dave et  al., 2013). High-throughput
next-generation sequencing technologies will consider-
ably accelerate the understanding of the detailed molecular
epidemiology of this condition (Aldahmesh et  al., 2012).
Prompt identification of congenital or infantile cataracts is
key to successful management.
Over a half of bilateral isolated congenital cataract
patients have a genetic basis for their opacity, most com-
monly inherited in an autosomal dominant fashion,
although recessive forms are increasingly recognized, with
more than 25 genes associated with autosomal dominant
forms (Table 40.1).
Cataracts are also recognized with many multisystemic
conditions, including Lowe, Cockayne, and Micro syn-
dromes, as well as a number of metabolic conditions. Here
an example one such metabolic condition is the rare sterol
storage disorder cerebrotendinous xanthomatosis (CTX),
which is caused by mutations in sterol 27-hydroxylase
(the CYP27A1 gene), and for which prompt diagnosis is
Table 40.1  ISOLATED LENS AND ANTERIOR SEGMENT ABNORMALITIES: A PARADIGM FOR GENETIC
HETEROGENEITY

GENE DISEASE/PHENOTYPE GENE/LOCUS (OMIM#)

ADAMTSL4 Ectopia lentis, isolated, autosomal recessive 610113
BFSP1 Cataract, cortical juvenile-onset 603307
BFSP2 Cataract, autosomal dominant multiple types 1 603212
Cataract, congenital
Cataract, juvenile-onset
CHMP4B Cataract, posterior polar 3 610897
CRYAA Cataract, autosomal dominant nuclear 123580
Cataract, congenital, autosomal recessive
CRYAB Cataract, posterior polar 2 123590
Myopathy, myofibrillar alpha-B crystallin-related
CRYBA1 Cataract, congenital zonular with sutural opacities 123610
CRYBA4 Microphthalmia with cataract 4 123631
Cataract, lamellar 2
CRYBB1 Cataract, congenital nuclear autosomal recessive 3 600929
CRYBB2 Cataract, sutural with punctate and cerulean opacities 123620
Cataract, Coppock-like
Cataract, cerulean type 2
CRYBB3 Cataract, congenital nuclear 2 MIMID 123630
CRYGC variable zonular pulverulent 123680
Cataract, Coppock-like
CRYGD Cataract, nonnuclear polymorphic congenital MIMID 123690
Cataract, congenital cerulean type 3
Cataract, crystalline aculeiform
CRYGS Cataract, progressive polymorphic cortical 123730
EPHA2 Cataract, posterior polar 1 176946
FOXC1 Axenfeld-Rieger syndrome type 3 601090
FOXD3 Anterior segment dysgenesis 611539
Peter’s anomaly
FOXE3 Aphakia congenital primary 601094
Anterior segment mesenchymal dysgenesis
GJA3 Cataract, zonular pulverulent-3 MIMID 121015
GJA8 Cataract-microcornea syndrome MIMID 600897
Cataract, zonular pulverulent-1
HSF4 Cataract, lamellar MIMID 602438
Cataract, Marner type
MAF Cataract, pulverulent or cerulean with or without microcornea MIMID 177075
MIR184 Endothelial dystrophy, iris hypoplasia, congenital cataract, and stromal thinning (EDICT) 613146
syndrome MIMID
OPA3 Optic atrophy 3 with cataract MIMID 606580
PAX6 Cataract, with late-onset corneal dystrophy 607108
Peters anomaly
PITX2 Peters anomaly MIMID 601542
Iridogoniodysgenesis type 2

(continued)

6 2 6   •   G enomics in C linical P ractice

Table 40.1 CONTINUED

GENE DISEASE/PHENOTYPE GENE/LOCUS (OMIM#)

PITX3 Cataract, posterior polar 4 602669
Anterior segment mesenchymal dysgenesis
PVRL3 Severe bilateral congenital cataract 607147
PXDN Congenital cataract, corneal opacity, developmental glaucoma 605158
RNLS Congenital cataract, autosomal recessive 609360
SOX2 Microphthalmia syndromic 3 MIMID 184429
TDRD7 Cataract, autosomal recessive congenital 4 MIMID 611258
TMEM114 Cataract and microphthalmia 611573
VIM Cataract, pulverulent, autosomal dominant 193060
VSX2 Microphthalmia, cataracts and iris abnormalities 610092

key (Khan et  al., 2013). The precise delineation of bilat-
eral congenital cataract in infants therefore requires an
interdisciplinary approach involving the ophthalmologist,
the pediatrician, and the clinical geneticist. In the future
diagnosis will be enhanced through the introduction of
next-generation sequencing.
T H E I N H E R IT E D R ET I NA L
DYS T RO P H I E S , I N C LU D I N G
R ET I N IT I S P I G M E N TO S A
In the field of ophthalmology, no one group of conditions
exemplifies the progress that has been made using human
molecular genetic analyses better than retinitis pigmentosa.
Retinitis pigmentosa (RP) is a collective description for a
group of inherited retinal degenerations. The condition has
an incidence estimated to be between one in 3,000–5,000
(Heckenlively, 1989; Dryja and Li, 1995). RP presents with
an initial, progressive degeneration of the rod photorecep-
tors that starts in the peripheral retina, leading to tunnel
vision and night blindness (nyctalopia). As the disease pro-
gresses, the central retina and cone photoreceptors become
involved, there is a concomitant reduction in central vision,
loss of visual acuity, and color perception. As the neural
retina degenerates, retinal arterioles become attenuated,
and a characteristically waxy optic disc pallor develops. In
addition, pigment cells derived from the underlying RPE
migrate into the retina and lead to the formation of bone
spicules, one of the most characteristic features of the dis-
ease. An initial reduction of the dark-adapted rod (scotopic)
electroretinogram (ERG) is followed by a reduction of the
light-adapted, cone-mediated (photopic) ERG. Ultimately,
the ERG becomes unrecordable.
Clinically, RP is highly heterogeneous in its onset
and severity. There is high genetic heterogeneity with the
recognition of isolated autosomal dominant (AD), autoso-
mal recessive (AR), X-linked (XL) and digenic forms. RP
can also be part of a group of syndromic disorders. Fifty-five
loci have been described for non-syndromic RP, of which
48 genes have been identified. A further 158 genes or loci
have been described that are associated with retinal disease
(RetNet, http://​www.sph.uth.tmc.edu/RetNet/home.htm​,
accessed October 2013).
Cloning of Single Genes Causing Retinitis
Pigmentosa: A Powerful Approach
Rhodopsin was identified as the first gene mutated in
retinitis pigmentosa, underlying one form of ADRP.
Identification relied on a positional candidate-gene
approach in 1990 (Dryja et al., 1990), and null mutations
were also found to be responsible for a form of ARRP two
years later (Rosenfeld et  al., 1992). Since that time, the
molecular bases of a vast number of monogenic retinal dys-
trophies have been identified by using techniques that dem-
onstrate the extraordinary power of genetic approaches for
dissecting genetically heterogeneous disorders.
Positional approaches have been used extremely suc-
cessfully in linked families, initially focusing on standard
techniques for autosomal dominant and X-linked families.
More recently, the use of homozygosity-mapping in mul-
tiply consanguineous families has successfully identified
many mutated genes in recessive forms of retinal disease.
These technologies have collectively identified mutations
in genes of a wide range of functions but, crucially, have
allowed identification of genes whose function could not
have been identified using hypothesis-driven approaches—
often these provide the most interesting and unsuspected
areas for subsequent investigation. Examples include the
ubiquitous splicing factors PRPF3, 8, 31, and PAP-1;

G enetics and G enomics in C linical O phthalmology, I   •   6 2 7

SNRNP200 (Towns et al., 2010; Zhao et al., 2009); and the
retina-specific protein C2ORF71, which has subsequently
been shown to be a cilium-specific protein (Nishimura
et al., 2010).
Candidate gene approaches have been highly successful
in particular in a number of the metabolic pathways that
have been repeatedly implicated in human retinal degen-
erative phenotypes, including photoreceptor-specific
functions such as phototransduction, visual cycle, and
photoreceptor-specific transcription factors. Importantly,
as a consequence, proteins whose alteration (in structure
or expression) is responsible for RP can increasingly be
grouped according to their normal function (e.g., Rivolta
et  al., 2002; Hims et  al., 2003; Kennann et  al., 2005).
These proteins include structural photoreceptor proteins,
RNA-splicing regulators and transcription factors, pro-
teins involved in the photoreceptor’s ciliary apparatus or
intracellular transport, and so on. Such groupings are use-
ful for our understanding of RP and for the possible iden-
tification of further genes. While not an infallible method
for predicting mode of inheritance or prognosis/effect on
photoreceptor function, such a functional understanding is
nevertheless critical for understanding the molecular mech-
anisms underlying specific mutations and for developing
therapeutic strategies. Two such pathways/structures are
expanded upon below; namely, the RNA-splicing factors
and the photoreceptor cilium apparatus.
Most recently, the advent of next-generation sequencing
to allow high-throughput parallel interrogation of whole
exomes in patients with inherited retinal disease has mas-
sively accelerated the identification of novel disease-causing
genes, including NEK2 (Nishiguchi et  al., 2013)  and
ITM2B (Audo et al., 2014).
Defects of RNA-Splicing Factors
Intron removal from precursor messenger RNA
(pre-mRNA) transcripts occurs via the nuclear-located spli-
ceosome complex, whose assembly requires small nuclear
ribonucleoproteins (snRNP), pre-mRNA, and other
nuclear proteins. Amongst the genes causing ADRP, four
(RP13, RP11, RP18, and RP9) are caused by mutations
in genes that encode pre-mRNA-splicing factors (McKie
et  al., 2001; Vithana et  al., 2001; Chakarova et  al., 2002;
Keen et al., 2002; Maita et al., 2004). In addition, one form
of ARRP is caused by mutations in SNRNP200 (Towns
et  al., 2010). Such splicing genes function, not only in
the retina, but also throughout the human body, and the
precise molecular reasons why a defect affecting a ubiqui-
tous process such as pre-mRNA splicing should confer a
6 2 8   •   G enomics in C linical P ractice
tissue-specific pathogenesis are unclear. Recent evidence
has shown that these splicing factors are expressed highly
in the retina, although whether this is relevant to this tissue
specificity is unknown (Cao et al., 2011). Another hypoth-
esis suggests that, because the rods have such a huge demand
for protein synthesis (the outer segment is turned over every
10 days), any compromise of the splicing machinery will be
deleterious. The retina also has one of the highest meta-
bolic rates of any tissue in the body, which may add further
demands to an already “stressed” splicing system.
D E FEC TS O F T H E P H OTO R EC E P TO R
C I L IUM
The word “ciliopathy,” first coined for Bardet-Biedl syn-
drome (Ansley et al., 2003), is now used to label an expand-
ing group of disorders that share a number of clinical
phenotypical overlaps and that are caused by mutations of
genes that encode proteins, which, when absent or defec-
tive, result in the abnormal formation or function of cilia.
The cilium is a membrane-bound projection assembled, on
virtually all cell types, from mature centrioles or basal bod-
ies. Primary cilia are important for extracellular signaling.
The analysis of molecules expressed in the photoreceptor
cilium has been the object of a considerable degree of inter-
est, and many such molecules have been implicated in forms
of syndromic RP, non-syndromic RP, as well as other retinal
degenerations. In the photoreceptor, the connecting cilium
is the conduit between the inner and outer segments, and it
is key to intracellular photoreceptor transport. Alteration
of ciliary function can result in multisystemic clinical mani-
festations and, in addition to retinal degeneration, particu-
lar features include renal disease, cerebral anomalies, situs
abnormalities, and limb abnormalities such as polydactyly.
Non-syndromic RP is an extensively covered topic, with
many excellent reviews (Anasagasti et al., 2012; Zobor et al.,
2012). Amongst the multitude of genetic causes, RPGR,
the gene mutated in the RP3 form of XLRP, was an impor-
tant early molecule to be localized to the cilium (reviewed
by Rachel et  al., 2012). In addition to XLRP, mutations
in RPGR are responsible for cone dystrophy (Yang et  al.,
2002), cone-rod dystrophy (Ayyagari et  al., 2002), and
atrophic macular dystrophy (Demirci et al., 2002). A pos-
sible role for RPGR in the cilium has been suggested for
many years by the clinical observations of a small number
of XLRP families with abnormal nasal cilia and axonomes
(Arden and Fox, 1979; Fox et al., 1980; Hunter et al., 1988).
A  putative role for RPGR in syndromic RP was derived
from the description of an XLRP family who were suscep-
tible to respiratory infections (recurrent bronchitis and
sinusitis) (van Dorp, 1992) and were later found to have an
RPGR splice site mutation (Dry et al., 1999). The condition
was similar to immotile cilia syndrome (ICS1), except for
an absence of deafness or sterility. Since then, other families
have been described with RPGR mutations that have RP,
hearing loss (sensorineural or conductive), and recurring
sinorespiratory infections (Zito et  al., 2003; Iannaccone
et al., 2003; Iannaccone et al., 2004).
RPGR interacts with a number of other proteins that
participate in intraflagellar transport, including RPGRIP-1
(mutated in Leber congenital amurosis, LCA; Dryja et al.,
2001), CEP290 (LCA and syndromic ciliopathies; Brancati
et al., 2007; den Hollander et al., 2006), and NPHP5 (neph-
rocyctin 5 or IQCB1, mutated in LCA and Senior Loken
syndrome; Otto et al., 2005; Estrada-Cuzcano et al., 2011).
Syndromic RP and the Connecting Cilium
Bardet-Biedl syndrome:  Bardet-Biedl syndrome (BBS) is
a rare recessive condition with a range of features, includ-
ing RP, obesity, polydactyly, kidney abnormalities, learning
difficulties, heart defects, and genital anomalies (Mykytyn
and Sheffield, 2004). Renal malformations and abnormal
renal function leading to end-stage renal disease (ESRD)
are a major cause of morbidity. Sixteen genes are known to
be associated with BBS, many of which may be associated
with other ciliopathy phenotypes, including:  McKusick–
Kaufman syndrome (MKKS), an autosomal recessive
condition characterized by hydrometrocolpos, post-axial
polydactyly, and congenital heart disease; Meckel–Gruber
syndrome (MKS), which is associated with occipital
encephalocele and posterior fossa defects, cystic dysplastic
kidneys, hepatic abnormalities, and polydactyly; Joubert
syndrome ( JBTS), another rare recessive disorder char-
acterized by hypotonia, ataxia, developmental delay, and
irregular breathing patterns. BBS proteins exist in a mac-
romolecular complex, underlining the importance of the
cilium in the development of multisystemic syndromic reti-
nal dystrophies.
Senior-Loken syndrome: Nephronophthisis (NPHP) is
associated with renal tubular basement membrane disinte-
gration, tubular atrophy, interstitial fibrosis, and cyst forma-
tion. It results in progressive renal failure during early life.
Nephronophthisis-associated ciliopathies comprise a range
of conditions: NPHP, Senior Loken syndrome (SLSN), as
well JBTS and MKS. Twelve genes have now been shown
to be mutated in NPHP/SLSN: NPHP1, INVS, NPHP3,
NPHP4, IQCB1, CEP290, GLIS2, RPGRIP1L, NEK8,
TMEM67, TTC21B, and XPNPEP3. There is a strong
overlap of mutations and clinical phenotypes amongst
G enetics and G enomics in C linical O phthalmology, I   •   6 2 9
patients within this spectrum—exemplified by the obser-
vation that mutations in IQCB1 can cause both LCA and
SLSN.
Usher syndrome:  Usher syndrome (USH) is an auto-
somal recessive disorder characterized by bilateral sen-
sorineural hearing loss and RP. Usher syndrome occurs
at a frequency of one in 16–50,000 (Keats and Corey,
1999) and has been estimated to account for 2% of all RP
cases (Boughman, Vernon, & Shaver, 1983). Three subtypes
(USH1, USH2, and USH3), are recognized, according to
the severity of hearing impairment, presence of vestibular
dysfunction, and age at onset of RP. There are five USH1
genes, of which one gene (MYO7A or USH1B) is the most
important and accounts for the majority of USH1 clinical
cases. MYO7A, encoding unconventional myosin VIIa, in
conjunction with vezatin strengthens the adhesion between
cells and between stereocilia (Petit, 2009). There are three
USH2 genes, while the USH3A gene encodes the trans-
membrane protein clarin-1 (Petit et al., 2009; Richardson
et al., 2011). The cochlear hair cells are connected with the
tectorial membrane via stereocilia. Sound sensing is under-
taken via the hair bundle that composes these stereocilia,
which are connected to form a “U” or “V” and act to sense
membrane displacement. The study of Usher syndrome has
provided insights into hair cell differentiation, maintenance,
and structure, initially through a developing understanding
of USH1 and USH2 proteins and the organization of their
protein complexes.
Translating the Benefits of Molecular Genetic
Advances into Clinical Practice
Ophthalmology has been at the vanguard of developments
in genomic medicine for both monogenic and complex dis-
orders. As in many other branches of medicine, the transla-
tion of these benefits to the clinic lagged somewhat behind
the research progress. This situation is now changing rapidly
in the provision of genetic testing for Mendelian disorders,
and in the development of novel therapies.
C L I N I C A L LY R E L EVA N T G E N ET I C T E S T I N G
Once the genetic basis of an inherited disorder has been
identified, genetic testing can potentially provide valuable
information regarding diagnoses, prognoses, and repro-
ductive risks, and such developments are now increasingly
translated into routinely available genetic tests. The major-
ity of such analyses test for mutations in known genes and
rely on the identification of a sequence variation that alters
the expression or function of a gene or protein, and can be
predicted to be pathogenic or disease-causing. There several
areas of relevance, which will be discussed next.
Truly Single-Gene Disorders
In cases where there is a strong relationship between a
phenotype and genotype, genetic identification of patho-
genic mutations is relatively straightforward. Such a link
exists amongst some of the monogenic macular dystro-
phies such as Sorsby and Doyne honeycomb (Malattia
Leventinese) dystrophies, where the vast majority of cases
result from single-point mutations in the TIMP3 and
EFEMP1 genes, respectively (Li et al., 2005; Stone et al.,
1999). The same is also true for the corneal dystrophies
linked to chromosome 5q31 that result from alterations
in the TGFBI gene, in which specific mutations cause rec-
ognizable phenotypes such as granular, lattice type I, and
Bowman’s layer (Thiel-Behnke and Reis-Buckler) dystro-
phies (Munier et al., 2002). Another area of success in the
delivery of genetic testing has been in the diagnosis and
counselling of families with retinoblastoma, a condition
resulting almost exclusively from mutations in a single
large gene.
Genetically Heterogeneous Single-Gene
Disorders
The process of mutation-screening for conditions such
as RP, in which mutations in many different genes can
cause an identical phenotype, has been transformed by
high-throughput next-generation screening (NGS).
Selective testing through direct sequencing of genes
that cause common forms of RP remains a valuable and
achievable goal in certain circumstances. For example,
since RPGR mutations account for 70–75% of XLRP,
and the majority (~60%) of the mutations are present in
the ORF15 exon (Vervoort et  al., 2000), this is a com-
mon cause of RP, accounting for up to ~17% of all cases
(Vervoort and Wright, 2002). This gene fragment is
highly repetitive and is refractive to sequencing via NGS
(O’Sullivan et al., 2012).
The advent of sensitive high-density oligonucleotide
arrays (Arrayed Primer Extension, or APEX, array) pro-
vided one solution for the diagnosis of heterogeneous
eye disease by testing simultaneously for all mutations
associated with a given disease. These arrays have been
used to test for known disease-causing variants associated
with LCA. However, these are unable to detect novel
disease-causing mutations (Koenekoop et al., 2007; Kurg
et al., 2010).
6 3 0   •   G enomics in C linical P ractice
NGS technologies permit the simultaneous analysis
of numerous DNA templates in one reaction. This hugely
enhances the output of single experiments (Metzker, 2010;
Shendure, 2008) and allows parallelization on a level pre-
viously unattainable. Techniques can extend to whole
genomes, although currently the targeted sequencing (or
analysis) of coding regions implicated in disease remains the
most practical solution. NGS techniques are a cost-effective
alternative to conventional genetic testing methodologies
and have recently been applied to the diagnosis of hetero-
geneous ocular diseases such as RP. This augurs a transfor-
mation of genetic testing within clinical services, although,
while the cost of NGS per base of sequencing is dramati-
cally reduced compared to that of conventional methodolo-
gies, a concomitant increase in demand for testing is also
likely (O’Sullivan et al., 2012).
Common Disorders Such as Age-Related Macular
Degeneration
Complex disorders have a significant heritable compo-
nent but do not follow classical Mendelian inheritance.
Identifying contributing genetic factors for AMD has been
extraordinarily successful, and it has been predicted that
this will lead to the implementation of personalized medi-
cine (Clark et al., 2013; Wiggs, 2008). However, for AMD
patients, detailed clinical examination remains the most
sensitive estimator of the disease’s effects on vision, and
investigating the genetic contributions to AMD is currently
limited to understanding disease mechanisms (Stone et al.,
2012).
O P HT H A L M I C G E N ET I C S A N D T H E
D EVE L O PM E N T O F N OVE L T H E R A P I E S
In the context of treating genetic disease, it is tempting
solely to think of gene-replacement strategies, or “gene ther-
apy,” in which a faulty gene is replaced by a normal copy. It is
now clear, however, that there are many promising avenues
for attempting to treat inherited ocular disorders—some of
which we turn to next.
Conventional Therapies for Inherited Ocular Disease
It is worth emphasizing, when considering the development
of novel therapies to treat inherited disease, the opportuni-
ties provided by conventional therapies. An absent protein
function that results from a genetic mutation can in occa-
sional circumstances be nullified by dietary modification.
Examples include the use of arginine-restricted diets to slow
visual loss in patients with gyrate atrophy (Kaiser-Kupfer
et al., 2004). Another example has been the use of isotreti-
noin as a potential treatment for Stargardt disease, caused
by mutation in ABCA4 (Travis et al., 2007). Isotretinoin, or
13-cis-retinoic acid, is widely used to treat acne and reduces
the light-induced accumulation of A2E-containing lipo-
fuscin in abca4-deficient mice, suggesting the therapeutic
potential of retinoids for this disease (Radu et  al., 2003,
2005).
Gene Replacement Strategies
Although conceptually simple, gene-replacement strategies
were fraught with early technical difficulties and suffered
a number of well-publicized setbacks (Hacein-Bey-Abina
et al., 2003; Kohn et al., 2003). The eye, and in particular
the retina and RPE, are nevertheless ideally placed to act as
a model system in which to develop novel gene-based thera-
pies, thanks to their anatomically and immunologically
privileged status. A variety of viral-based vectors in a wide
range of animal models have allowed a broad-based assess-
ment of novel strategies. As a result, it is now clear that it is
possible to deliver biologically active molecules to the retina
and RPE in a manner that makes them genuinely effective.
Leber congenital amaurosis (LCA2) is associated with
mutations in RPE65, which encodes the retinal pigment
epithelium (RPE)–specific 65 kDa protein. LCA2 is rare
(<1 in 106 births), but the existence of large animal mod-
els in particular has led to this being an exemplar of retinal
gene therapy. RPE65 functions as an isomerase to convert
all-trans retinoid to 11-cis retinal within the visual cycle
( Jin et al., 2005; Moiseyev et al., 2005). Reduced or absent
RPE65 function results in absent 11-cis retinal production,
accumulation of all-trans-retinyl esters in the RPE, and later
photoreceptor degeneration. Naturally occurring Rpe65
mutant murine and dog models result in retinal degenera-
tion. In 2001, successful gene therapy was reported using
recombinant adeno-associated virus (AAV2) vector con-
taining Rpe65 cDNA to treat Rpe65 mutant Briard dogs
(Acland et  al., 2001). Gains in visual function remained
stable over time, with measures of visual behavior and corti-
cal responses all improved (Le Meur et al., 2007; Bennicelli
et al., 2008).
Such studies led to human phase I and II gene therapy
trials, whose results have been both published and well pub-
licized (e.g., Maguire et al., 2008; Bainbridge et al., 2008).
These demonstrated that AAV subretinal gene delivery is
safe and can even lead to improvements in light sensitivity
and retinal function, for example, as measured by the ability
to navigate a walking maze ( Jacobson et al., 2012). RPE65
G enetics and G enomics in C linical O phthalmology, I   •   6 3 1
gene therapy has thus been shown to be safe, and clinical and
visually effective. There have been promising results from
early-phase trials for other progressive retinal dystrophies
such as those caused by mutations in MERTK, USH1A, and
REP-1 (which causes choroideremia). An important proof
of concept has also been achieved for the static cone dys-
trophy achromatopsia where there is loss of cone function,
but not of cone cells. In one study, stable improvement for
2–3 years was achieved in canine models using recombinant
adeno-associated virus (rAAV)-mediated gene replacement
Cngb3 therapy (Komáromy et al., 2010).
Delivery of Growth Factors
There is a large body of research covering the delivery of
growth factors, often via viral transduction, into the eye
in an attempt to improve, in particular, the health of the
retina in retinal degeneration: either alone or as an adjunct
to gene-replacement strategies. For example, photorecep-
tor apoptosis was prevented in rats with an induced retinal
detachment after the subretinal injection of AAV-expressing
glial cell line–derived neurotrophic factor (GDNF)
(Wu et  al., 2002). Ciliary neurotrophic factor (CNTF)
expressed from encapsulated human retina-derived cells
preserved retinal structure in a canine model of RP (Pde6B-
/-
) (Tao et al., 2002) and slowed photoreceptor apoptosis in
a dose-dependent manner in the naturally occurring retinal
degeneration slow (RDS) mouse model of RP, caused by
mutation of RDS peripherin (Liang et  al., 2001). CNTF
is a promising therapeutic that is proposed in clinical trials
for RP (Sieving et al., 2006; Raz-Prag et al., 2009; Talcott
et al., 2011).
Allele-Specific Silencing as a Treatment for
Dominant Ocular Disease
The opportunity to use the cornea as a model tissue for
treating inherited disease has shown promise. The oppor-
tunities provided by easy access, available stem-cells and
avascularity are significant, and this has been coupled to
the study of rare families with epithelial corneal dystro-
phies, such as Meesmann corneal dystrophy (MECD). This
keratin disorder is caused by dominant-negative mutations
in either K3 or K12, which are specific to anterior corneal
epithelial keratinocytes (Irvine et  al., 1997). Like many
autosomal dominant disorders, MECD is not caused by
haploinsufficiency but rather results from mutations—usu-
ally missense—that act in a dominant-negative fashion.
A methodology is required to remove or silence the mutant
allele while not altering wild-type allele expression. RNA
interference, such as by using short-interfering RNA (siR-
NAs, which are double-stranded RNA molecules), is one
potential strategy for allele-specific gene-silencing.
MECD, like other epithelial keratin disorders, has been
demonstrated to be an attractive target for siRNA therapy
(Liao et al., 2011; Allen et al., 2013). siRNAs directed at
a mutant allele may be designed and then tested, using
high-throughput methods employing reporter genes, for
their ability to knockdown a mutant allele. It has been shown
that it is possible to design siRNAs of both high potency
and allele-specificity. Indeed, for another keratin disorder,
pachyonychia congenita, caused by dominant-negative mis-
sense mutations in one of keratins K6a, K6b, K6c, K16,
or K17, a small and successful Phase Ib clinical trial was
recently carried out to treat hyperkeratotic lesions with
siRNA (Leachman et  al., 2010). This was a first-in-man
study that has given proof of concept in vivo for siRNA
therapy. For MECD and many other conditions, a poten-
tial therapeutic siRNA would have to be designed for each
individual mutation, many of which are family-specific.
However, other epithelial corneal dystrophies, such as those
caused by phenotype-specific mutations in TGFBI, would
be more ideally suited to such technologies.
For dominant retinal dystrophies, similar methodolo-
gies have been employed to attempt to remove dominant
alleles in an allele-specific fashion. Ribozymes are synthetic
catalytic RNA molecules that can target and cleave mutated
mRNA molecules. Hammerhead ribozymes were used to
knockdown rod γPDE in mice to generate a model of RP.
The authors suggested that this approach could be used as a
therapy for dominant RP. AAV-mediated delivery of ribo-
zymes has also been used in an autosomal dominant RP rat
model carrying the P23H rhodopsin mutation (Drenser
et  al., 1998). A  more general ribozyme could be used to
knockdown the mutant and wild-type alleles and be supple-
mented with a “hardened” replacement gene that is resistant
to the ribozyme (Liu et al., 2005).
Retinal Implants
Visual prostheses that attempt to allow light-sensing through
the stimulation of visual pathway neurons using electronic
implants have been explored in clinical trials using implants
lying under or on top of the retina (Humayun et al., 2012;
Zrenner et al., 2011). The development of the field of visual
prostheses has progressed through clinical trials, and they
have been shown to be relatively safe. Although to date these
implants can only produce rudimentary levels of “vision,”
such processes show promise for the future and indeed have
6 3 2   •   G enomics in C linical P ractice
already achieved federal Food and Drug Administration
approval (Ahuja et al., 2011; Stingl et al., 2013).
C O N C LU S I O N S
The management of inherited ophthalmic disease, and in
particular the introduction of powerful genetic testing strat-
egies and novel therapeutic strategies, is an exciting field. It
is one that raises strong emotions and has been the subject
of considerable publicity. It is an arena that also raises ethi-
cal questions for clinicians and scientists alike. While much
has been written on the implications of the new genetic
knowledge for the diagnosis of genetic conditions, little has
been done to apply these to ophthalmology. In the future,
it will be important to recognize that, as these new tech-
nologies will have to be applied across many specialties,
communication between those working in ophthalmology
and other fields will be key. Nonetheless, the hope must be
that ophthalmology, which has been at the vanguard of the
molecular era, can maintain this preeminence in translating
these findings for clinical benefit.
REFERENCES
Acland GM, et al. (2001). Gene therapy restores vision in a canine model
of childhood blindness. Nat Genet. 28:92–95.
Ahuja AK, Dorn JD, Caspi A, et al., and Argus II Study Group (2011).
Blind subjects implanted with the Argus II retinal prosthesis are able
to improve performance in a spatial-motor task. Br J Ophthalmol.
95:539–543.
Alagaratnam J, Sharma TK, Lim CS, Fleck BW (2002). A survey of
visual impairment in children attending the Royal Blind School,
Edinburgh using the WHO childhood visual impairment database.
Eye. Sep;16(5):557–561.
Aldahmesh MA, Khan AO, Mohamed JY, et al. (2012). Genomic analy-
sis of pediatric cataract in Saudi Arabia reveals novel candidate dis-
ease genes. Genet Med. 14:955–962.
Allen EH, Atkinson SD, Liao H, et  al. (2013). Allele-specific siRNA
silencing for the common keratin 12 founder mutation in Meesmann
epithelial corneal dystrophy. Invest Ophthalmol Vis Sci. 54:494–502.
Anasagasti A, Irigoyen C, Barandika O, López de Munain A, Ruiz-Ederra
J. (2012). Current mutation discovery approaches in Retinitis
Pigmentosa. Vision Res. 75:117–129.
Ansley SJ, Badano JL, Blacque OE, et  al. (2003). Basal body dysfunc-
tion is a likely cause of pleiotropic Bardet-Biedl syndrome. Nature.
425:628–633.
Antonarakis SE, Beckmann JS (2006). Mendelian disorders deserve
more attention. Nat Rev Genet. 7, 277–282.
Arden GB, Fox B (1979). Increased incidence of abnormal nasal cilia in
patients with retinitis pigmentosa. Nature. 279:534–536.
Audo I, Bujakowska K, Orhan E, et  al. (2014). The familial demen-
tia gene revisited:  a missense mutation revealed by whole-exome
sequencing identifies ITM2B as a candidate gene underlying a novel
autosomal dominant retinal dystrophy in a large family. Hum Mol
Genet. 23(2):491–501.
Ayyagari R, Demirci FY, Liu J, et  al. (2002). X-linked recessive atro-
phic macular degeneration from RPGR mutation. Genomics.
80:166–171.
Bainbridge JW, Smith AJ, Barker SS, Robbie S, Henderson R, Balaggan
K (2008). Effect of gene therapy on visual function in Leber’s con-
genital amaurosis. N Engl J Med. 358:2231–2239.
Bardakjian TM, Schneider AS, Ng D, Johnston JJ, Biesecker LG (2009).
Association of a de novo 16q copy number variant with a pheno-
type that overlaps with Lenz microphthalmia and Townes-Brocks
syndromes. BMC Med Genet. 10:137.
Bennicelli J, Wright JF, Komaromy A, Jacobs JB, Hauck B, Zelenaia O
(2008). Reversal of blindness in animal models of Leber congenital
amaurosis using optimized AAV2-mediated gene transfer. Mol Ther.
16:458–465.
Boughman JA, Vernon M, Shaver KA. (1983). Usher syndrome: defini-
tion and estimate of prevalence from two high-risk populations. J
Chronic Dis. 36(8):595–603.
Brancati F, Barrano G, Silhavy JL, et  al. (2007). CEP290 muta-
tions are frequently identified in the oculo-renal form of Joubert
syndrome-related disorders. Am J Hum Genet. 81:104–113.
Brinkman RR, Dube MP, Rouleau GA, Orr AC, Samuels ME (2006).
Human monogenic disorders—a source of novel drug targets. Nat
Rev Genet. 7:249–260.
Cao H, Wu J, Lam S, et al. (2011). Temporal and tissue specific regu-
lation of RP-associated splicing factor genes PRPF3, PRPF31
and PRPC8—implications in the pathogenesis of RP. PLoS One.
6:e15860.
Chakarova CF, Hims MM, Bolz H, et al. (2002). Mutations in HPRP3,
a third member of pre-mRNA splicing factor genes, implicated
in autosomal dominant retinitis pigmentosa. Hum Mol Genet.
11:87–92.
Chao LY, et al. (2003). Missense mutations in the DNA-binding region
and termination codon in PAX6. Hum Mutat. 21:138–145
Clark SJ, Ridge LA, Herbert AP, et  al. (2013). Tissue-specific host
recognition by complement factor H is mediated by differential
activities of its glycosaminoglycan-binding regions. J Immunol.
190:2049–2057.
Dave A, Laurie K, Staffieri SE, et al. (2013). Mutations in the EPHA2
gene are a major contributor to inherited cataracts in South-Eastern
Australia. PLoS One. 8:e72518.
Demirci FY, Rigatti BW, Wen G, et al. (2002). X-linked cone-rod dys-
trophy (locus COD1): identification of mutations in RPGR exon
ORF15. Am J Hum Genet. 70:1049–1053.
den Hollander AI, Koenekoop RK, Yzer S, et al. (2006). Mutations in the
CEP290 (NPHP6) gene are a frequent cause of Leber congenital
amaurosis. Am J Hum Genet.79:556–561.
Desronvil T, Logan-Wyatt D, Abdrabou W, et al. (2010). Distribution of
COL8A2 and COL8A1 gene variants in Caucasian primary open
angle glaucoma patients with thin central corneal thickness. Mol Vis.
16:2185–2191.
Drenser KA, Timmers AM, Hauswirth WW, Lewin AS. (1998).
Ribozyme-targeted destruction of RNA associated with autoso-
mal dominant retinitis pigmentosa. Invest Ophthalmol Vis Sci.
139:681–689.
Dry KL, Manson FD, Lennon A, Bergen AA, Van Dorp DB, Wright AF.
(1999). Identification of a 5' splice site mutation in the RPGR gene
in a family with X-linked retinitis pigmentosa (RP3). Hum Mutat.
13(2):141–145.
Dryja TP, McGee TL, Reichel E, et  al. (1990). A point mutation of
the rhodopsin gene in one form of retinitis pigmentosa. Nature.
343:364–366.
Dryja TP, Li T (1995). Molecular genetics of retinitis pigmentosa. Hum
Mol Genet. 4:1739−1743.
Dryja TP, Adams SM, Grimsby JL, et al. (2001). Null RPGRIP1 alleles
in patients with Leber congenital amaurosis. Am J Hum Genet.
68:1295–1298.
G enetics and G enomics in C linical O phthalmology, I   •   6 3 3
Dyment DA, Smith AC, Alcantara D, et  al. (2013). Mutations
in PIK3R1 cause SHORT syndrome. Am J Hum Genet.
93:158–166.
Estrada-Cuzcano A, Koenekoop RK, Coppieters F, et al. (2011). IQCB1
mutations in patients with Leber congenital amaurosis. Invest
Ophthalmol Vis Sci. 52:834–839.
Fares-Taie L, Gerber S, Chassaing N, et  al. (2013). ALDH1A3 muta-
tions cause recessive anophthalmia and microphthalmia. Am J Hum
Genet. 92:265–270.
Fox B, Bull TB, Arden GB (1980). Variations in the ultrastructure of
human nasal cilia including abnormalities found in retinitis pigmen-
tosa. J Clin Pathol. 33:327–335.
Furukawa T, Kozak CA and Cepko CL. (1997). Rax, a novel paired-type
homeobox gene, shows expression in the anterior neural fold and
developing retina. Proc Natl Acad Sci U S A. 94:3088–3093.
Giannaccini M, Giudetti G, Biasci D, et al. (2013). Rx1 defines retinal
precursor identity by repressing alternative fates through the activa-
tion of TLE2 and Hes4. Stem Cells. 31(12):2842–2847.
Gilbert C, Foster A.(2001). Childhood blindness in the context of
VISION 2020—the right to sight. Bull World Health Organ.
79:227–232.
Hacein-Bey-Abina S, Von Kalle C, Schmidt M, et  al. (2003).
LMO2-associated clonal T cell proliferation in two patients after
gene therapy for SCID-X1. Science. 302:415–419.
Hammond CJ, et  al. (2000). Genetic and environmental factors in
age-related nuclear cataracts in monozygotic and dizygotic twins.
New England J. Med. 342:1786–1790.
Hammond CJ, et al. (2001). The heritability of age-related cortical cata-
ract: the twin eye study. Invest Ophthalmol Vis Sci. 42:601–605.
Heiba IM, et al. (1995). Evidence for a major gene for cortical cataract.
Invest. Ophthalmol Vis Sci. 36:227–235.
Heckenlively JR (1989). Retinitis Pigmentosa. Philadelphia,
PA: Lippincott Company.
Henderson RH, Williamson KA, Kennedy JS, et al. (2009). A rare de
novo nonsense mutation in OTX2 causes early onset retinal dystro-
phy and pituitary dysfunction. Mol Vis.15:2442–2447.
Heyman I, et al. (1999). Psychiatric disorder and cognitive function in a
family with an inherited novel mutation of the developmental con-
trol gene PAX6. Psychiatr Genet. 9:85–90.
Hims MM, Diager SP, Inglehearn CF (2003). Retinitis pigmen-
tosa: genes, proteins and prospects. Dev Ophthalmol. 37:109–125.
Humayun MS, Dorn JD, da Cruz L, Dagnelie G, Sahel JA, et al. 2012.
Interim results from the international trial of Second Sight’s visual
prosthesis. Ophthalmology. 119:779–788.
Hunter DG, Fishman GA, Kretzer FL (1988). Abnormal axonemes in
X-linked retinitis pigmentosa. Arch Ophthalmol. 106:362–368.
Iannaccone A, Breuer DK, Wang XF, et al. (2003). Clinical and immu-
nohistochemical evidence for an X linked retinitis pigmentosa syn-
drome with recurrent infections and hearing loss in association with
an RPGR mutation. J Med Genet. 40:e118.
Iannaccone A, Wang X, Jablonski MM, et al. (2004). Increasing evidence
for syndromic phenotypes associated with RPGR mutations. Am J
Ophthalmol. 137:785–786.
Irvine AD, Corden LD, Swensson O, et  al. (1997). Mutations in
cornea-specific keratin K3 or K12 genes cause Meesmann’s corneal
dystrophy. Nat Genet. 16:184–187.
Iyengar SK, et al. (2004). Identification of a major locus for age-related
cortical cataract on chromosome 6p12-q12 in the Beaver Dam Eye
Study. Proc Natl Acad Sci U S A. 101:14485–14490.
Jacobson SG, Cideciyan AV, Ratnakaram R, Heon E, Schwartz SB,
Roman AJ (2012). Gene therapy for Leber congenital amaurosis
caused by RPE65 mutations: safety and efficacy in 15 children and
adults followed up to 3 years. Arch Ophthalmol. 130:9–24.
Jia X, Zhang F, Bai J, et  al. (2013). Combinational analysis of linkage
and exome sequencing identifies the causative mutation in a Chinese
family with congenital cataract. BMC Med Genet. 14:107.
Jin M, Li S, Moghrabi WN, Sun H and Travis GH (2005). Rpe65 is
the retinoid isomerase in bovine retinal pigment epithelium. Cell.
122:449–459.
Kaiser-Kupfer MI, Caruso RC, Valle D, Reed GF (2004). Use of an
arginine-restricted diet to slow progression of visual loss in patients
with gyrate atrophy. Arch Ophthalmol. 122:982–984.
Keats BJ, Corey DP (1999). The Usher syndromes. Am J Med Genet.
89:158–166.
Keen TJ, Hims MM, McKie AB, et al. (2002). Mutations in a protein
target of the Pim-1 kinase associated with the RP9 form of autoso-
mal dominant retinitis pigmentosa. Eur J Hum Genet. 10:245–249.
Kelberman D, Islam L, Holder SE, et al. (2011). Digenic inheritance of
mutations in FOXC1 and PITX2:  correlating transcription fac-
tor function and Axenfeld-Rieger disease severity. Hum Mutat.
32:1144–1152.
Kennan A, Aherne A, Humphries P (2005). Light in retinitis pigmen-
tosa. Trends Genet. 21:103–110.
Kimura A, Singh D, Wawrousek EF, Kikuchi M, Nakamura M,
Shinohara T (2000). Both PCE-1/RX and OTX/CRX interac-
tions are necessary for photoreceptor-specific gene expression. J Biol
Chem. 275:1152–1160.
Khan AO, Aldahmesh MA, Mohamed JY, Alkuraya FS (2013). Juvenile
cataract morphology in 3 siblings not yet diagnosed with cerebro-
tendinous xanthomatosis. Ophthalmology. 120:956–960.
Koenekoop RK, Lopez I, den Hollander AI, Allikmets R, Cremers
FPM (2007). Genetic testing for retinal dystrophies and dys-
functions:  benefits, dilemmas, solutions. Clin Exp Ophthalmol.
35:473–485.
Kohn DB, et al. (2003). Occurrence of leukaemia following gene therapy
of X-linked SCID. Nat Rev Cancer. 3:477–488.
Komáromy AM, Alexander JJ, Rowlan JS, et  al. (2010). Gene therapy
rescues cone function in congenital achromatopsia. Hum Mol
Genet. 19:2581–2593.
Kurg A, Tõnisson N, Georgiou I, Shumaker J, Tollett J, Metspalu A
(2010). Arrayed primer extension:  solid-phase four-color DNA
resequencing and mutation detection technology. Genet Testing.
4:1–7.
Leachman SA, Hickerson RP, Schwartz ME, et  al. (2010). First-in-
human mutation-targeted siRNA phase Ib trial of an inherited skin
disorder. Mol Ther. 18:442–446.
Le Meur G, Stieger K, Smith AJ, Weber M, Deschamps JY, Nivard D
(2007). Restoration of vision in RPE65-deficient Briard dogs using
an AAV serotype 4 vector that specifically targets the retinal pig-
mented epithelium. Gene Ther. 14:292–303.
Lhériteau E, Petit L, Weber M, et al. (2014). Successful gene therapy in
the RPGRIP1-deficient dog, a large model of cone-rod dystrophy.
Mol Ther. 22(2):265–277.
Li Z, Clarke MP Barker MD, McKie N. (2005). TIMP3 mutation in
Sorsby’s fundus dystrophy: molecular insights. Expert Rev Mol Med.
7:1–15.
Liang FQ, Aleman TS, Dejneka NS (2001). Long-term protection of
retinal structure but not function using RAAV.CNTF in animal
models of retinitis pigmentosa. Mol Ther. 4:461–472.
Liao H, Irvine AD, Macewen CJ, et  al. (2011). Development of
allele-specific therapeutic siRNA in Meesmann epithelial corneal
dystrophy. PLoS One. 6:e28582.
Liu J, et al. (2005). Ribozyme knockdown of the gamma-subunit of rod
cGMP phosphodiesterase alters the ERG and retinal morphology
in wild-type mice. Invest Ophthalmol Vis Sci. 46:3836–3844.
Maguire AM, High KA, Auricchio A, et  al. (2009). Age-dependent
effects of RPE65 gene therapy for Leber’s congenital amaurosis: a
phase 1 dose-escalation trial. Lancet. 374:1597–1605.
Maita H, Kitaura H, Keen TJ, Inglehearn CF, Ariga H, Iguchi-Ariga
SM (2004). PAP-1, the mutated gene underlying the RP9 form of
dominant retinitis pigmentosa, is a splicing factor. Exp Cell Res.
300:283–296.
6 3 4   •   G enomics in C linical P ractice
Malandrini A, et al. (2001). PAX6 mutation in a family with aniridia,
congenital ptosis, and mental retardation. Clin Genet. 60:151–154.
McKie AB, McHale JC, Keen TJ, et  al. (2001). Mutations in the
pre-mRNA splicing factor gene PRPC8 in autosomal dominant
retinitis pigmentosa (RP13). Hum Mol Genet. 10:1555–1562.
Maguire AM, Simonelli F, Pierce EA, Pugh EN Jr, Mingozzi F, Bennicelli
J (2008). Safety and efficacy of gene transfer for Leber’s congenital
amaurosis. N Engl J Med. 358:2240–2248.
Metzker ML (2010). Sequencing technologies—the next generation.
Nat Rev Genet. 11:31–46.
Moiseyev G, Chen Y, Takahashi Y, Wu BX and Ma JX (2005). RPE65 is
the isomerohydrolase in the retinoid visual cycle. Proc Natl Acad Sci
U S A. 102:12413–12418.
Munier FL, Frueh BE, Othenin-Girard P, et al. (2002). BIGH3 muta-
tion spectrum in corneal dystrophies. Invest Ophthalmol Vis Sci.
43:949–954.
Mykytyn K, Sheffield VC (2004). Establishing a connection between
cilia and Bardet-Biedl Syndrome. Trends Mol Med. 10:106–109.
Narfström K, Seeliger M, Lai CM, Vaegan Katz M, Rakoczy EP (2008).
Morphological aspects related to long-term functional improvement
of the retina in the 4 years following rAAV-mediated gene transfer
in the RPE65 null mutation dog. Adv Exp Med Biol. 613:139–146.
Nishiguchi KM, Tearle RG, Liu YP, et  al. (2013). Whole genome
sequencing in patients with retinitis pigmentosa reveals pathogenic
DNA structural changes and NEK2 as a new disease gene. Proc Natl
Acad Sci U S A. 110:16139–16144.
Nishimura DY, Baye LM, Perveen R, et al. (2010). Discovery and func-
tional analysis of a retinitis pigmentosa gene, C2ORF71. Am J Hum
Genet. 86:686–695.
O’Connor TP, Crystal RG (2006). Genetic medicines: treatment strate-
gies for hereditary disorders. Nat Rev Genet. 7:261–276.
O’Sullivan J, Mullaney BG, Bhaskar SS, et al. (2012). A paradigm shift
in the delivery of services for diagnosis of inherited retinal disease. J
Med Genet. 49:322–326.
Otto EA, Loeys B, Khanna H, et al. (2005). Nephrocystin-5, a ciliary IQ
domain protein, is mutated in Senior-Loken syndrome and interacts
with RPGR and calmodulin. Nat Genet. 37, 282–288.
Ozel AB, Moroi SE, Reed DM, et  al. (2014). Genome-wide associa-
tion study and meta-analysis of intraocular pressure. Hum Genet.
133(1):41–57.
Pasutto F, Sticht H, Hammersen G, et al. (2007). Mutations in STRA6
cause a broad spectrum of malformations including anophthalmia,
congenital heart defects, diaphragmatic hernia, alveolar capillary
dysplasia, lung hypoplasia, and mental retardation. Am J Hum
Genet. 80:550–560.
Percin EF, et  al. (2000). Human microphthalmia associated with
mutations in the retinal homeobox gene CHX10. Nat Genet.
25:397–401.
Petit C, Richardson GP. Linking genes underlying deafness to hair-bundle
development and function. Nat Neurosci. 2009; 12:703–710.
Rachel RA, Li T, Swaroop A. (2012). Photoreceptor sensory cilia and cil-
iopathies: focus on CEP290, RPGR and their interacting proteins.
Cilia.1:22.
Raca G, Jackson CA, Kucinskas L, et  al. (2011). Array comparative
genomic hybridization analysis in patients with anophthalmia,
microphthalmia, and coloboma. Genet Med. 13:437–442.
Radu RA, Han Y, Bui TV, et al. (2005). Reductions in serum vitamin
A  arrest accumulation of toxic retinal fluorophores:  a potential
therapy for treatment of lipofuscin-based retinal diseases. Invest
Ophthalmol Vis Sci. 46:4393–4401.
Radu RA, Mata NL, Nusinowitz S, Liu X, Sieving PA, Travis GH
(2003). Treatment with isotretinoin inhibits lipofuscin accumula-
tion in a mouse model of recessive Stargardt’s macular degeneration.
Proc Natl Acad Sci U S A. 100:4742–4747.
Ragge NK, et  al. (2005a). Heterozygous mutations of OTX2 cause
severe ocular malformations. Am J Hum Genet. 76:1008–1022.
Ragge NK, Lorenz B, Schnieder A, et al. (2005b). SOX2 anophthalmia
syndrome. Am J Med Genet. 135:1–7.
Rahi JS, Dezateux C, British Congenital Cataract Interest Group
(2001). Measuring and interpreting the incidence of congenital ocu-
lar anomalies: lessons from a national study of congenital cataract in
the UK. Invest Ophthalmol Vis Sci. 42:1444–1448.
Rahi JS, Cable N (2003). British Childhood Visual Impairment Study
Group. Severe visual impairment and blindness in children in the
UK. Lancet. 362:1359–1365.
Raz-Prag D, Zeng Y, Sieving PA, Bush RA (2009). Photoreceptor pro-
tection by adeno-associated virus-mediated LEDGF expression in
the RCS rat model of retinal degeneration: probing the mechanism.
Invest Ophthalmol Vis Sci. 50:3897–3906.
Richardson GP, de Monvel JB, Petit C (2011). How the genetics of deaf-
ness illuminates auditory physiology. Annu Rev Physiol. 73:311–334.
Rivolta C, Sharon D, DeAngelis MM, Dryja TP (2002). Retinitis pig-
mentosa and allied diseases: numerous diseases, genes, and inheri-
tance patterns. Hum Mol Genet. 11:1219–1227.
Rohrbach M, Spencer HL, Porter LF, et al. (2013). ZNF469 frequently
mutated in the brittle cornea syndrome (BCS) is a single exon gene
possibly regulating the expression of several extracellular matrix
components. Mol Genet Metab. 109:289–295.
Rosenfeld PJ, Cowley GS, McGee TL, Sandberg MA, Berson EL, Dryja
TP (1992). A null mutation in the rhodopsin gene causes rod pho-
toreceptor dysfunction and autosomal recessive retinitis pigmen-
tosa. Nat Genet. 1:209–213.
Shendure J, Ji H (2008). Next-generation DNA sequencing. Nat Biotech.
26:1135–1145.
Sieving PA, Caruso RC, Tao W, et al. (2006). Ciliary neurotrophic fac-
tor (CNTF) for human retinal degeneration: phase I trial of CNTF
delivered by encapsulated cell intraocular implants. Proc Natl Acad
Sci U S A. 103:3896–3901.
Srour M, Chitayat D, Caron V, et  al. (2013). Recessive and dominant
mutations in retinoic acid receptor Beta in cases with microphthal-
mia and diaphragmatic hernia. Am J Hum Genet. 93:765–772.
Stingl K, Bach M, Bartz-Schmidt KU, et al. (2013). Safety and efficacy of
subretinal visual implants in humans: methodological aspects. Clin
Exp Optom. 96:4–13.
Stone EM, Lotery AJ, Munier FL, et al. (1999). A single EFEMP1 muta-
tion associated with both Malattia Leventinese and Doyne honey-
comb retinal dystrophy. Nat Genet. 22:199–202.
Stone EM, Aldave AJ, Drack AV. Recommendations for the genetic test-
ing of inherited eye diseases:  report of the American Academy of
Ophthalmology Task Force on Genetic Testing. Ophthalmology.
2012; 119:2408–2410.
Talcott KE, Ratnam K, Sundquist SM, et al. (2011). Longitudinal study
of cone photoreceptors during retinal degeneration and in response
to ciliary neurotrophic factor treatment. Invest Ophthalmol Vis Sci.
52:2219–2226.
Tao W, et al. (2002). Encapsulated cell-based delivery of CNTF reduces
photoreceptor degeneration in animal models of retinitis pigmen-
tosa. Invest Ophthalmol Vis Sci. 43:3292–3298.
Towns KV, Kipioti A, Long V, et  al. (2010). Prognosis for splicing
factor PRPF8 retinitis pigmentosa, novel mutations and cor-
relation between human and yeast phenotypes. Hum Mutat.
31:E1361–E1376.
Travis GH, Golczak M, Moise AR, Palczewski K (2007). Diseases caused
by defects in the visual cycle:  retinoids as potential therapeutic
agents. Annu Rev Pharmacol Toxicol. 47:469–512.
van Dorp DB, Wright AF, Carothers AD, Bleeker-Wagemakers EM
(1992). A family with RP3 type of X-linked retinitis pigmentosa: an
association with ciliary abnormalities. Hum Genet. 88:331–334.
G enetics and G enomics in C linical O phthalmology, I   •   6 3 5
Vervoort R, Lennon A, Bird AC, et  al. (2000). Mutational hot spot
within a new RPGR exon in X-linked retinitis pigmentosa. Nat
Genet. 25:462–466.
Vervoort R, Wright AF (2002). Mutations of RPGR in X-linked retinitis
pigmentosa (RP3). Hum Mutat. 19:486–500.
Vithana EN, Abu-Safieh L, Allen MJ, et al. (2001). A human homolog of
yeast pre-mRNA splicing gene, PRP31, underlies autosomal domi-
nant retinitis pigmentosa on chromosome 19q13.4 (RP11). Mol
Cell. 8:375–381.
Vithana EN, Aung T, Khor CC, et  al. (2011). Collagen-related genes
influence the glaucoma risk factor, central corneal thickness. Hum
Mol Genet. 20:649–658.
Voronina VA, Kozhemyakina EA, O’Kernick CM, et  al. (2004).
Mutations in the human RAX homeobox gene in a patient
with anophthalmia and sclerocornea. Hum Mol Genet. Feb
1;13(3):315–322.
Weeks DE, et al. (2004). Age-related maculopathy: a genomewide scan
with continued evidence of susceptibility loci within the 1q31,
10q26, and 17q25 regions. Am J Hum Genet. 75:174–189.
Westeneng-van Haaften SC, Boon CJ, Cremers FP, Hoefsloot LH,
den Hollander AI, Hoyng CB. (2012). Clinical and genetic
characteristics of late-onset Stargardt’s disease. Ophthalmology.
119:1199–1210.
Wiggs JL, et al. (2004). A genomewide scan identifies novel early-onset
primary open-angle glaucoma loci on 9q22 and 20p12. Am J Hum
Genet. 74:1314–1320.
Wiggs JL. (2008). Genomic promise:  personalized medicine for oph-
thalmology. Arch Ophthalmol. 126(3):422–423.
Wu WC, Lai CC, Chen SL, et al. (2002). Gene therapy for detached
retina by adeno-associated virus vector expressing glial cell
line-derived neurotrophic factor. Invest Ophthalmol Vis Sci.
43(11):3480–3488.
Yang Z, Peachey NS, Moshfeghi DM, et  al. (2002). Mutations in the
RPGR gene cause X-linked cone dystrophy. Hum Mol Genet.
11:605–611.
Yang J, Luo J, Zhou P, Fan Q, Luo Y, Lu Y (2013). Association of the
ephreceptor tyrosinekinase-type A2 (EPHA2) gene polymorphism
rs3754334 with age-related cataract risk: a meta-analysis. PLoS One.
8:e71003.
Yang Y, Muzny DM, Reid JG, et  al. (2013). Clinical whole-exome
sequencing for the diagnosis of Mendelian disorders. N Engl J Med.
369:1502–1511.
Zhao C, Bellur DL, Lu S, et  al. (2009). Autosomal-dominant reti-
nitis pigmentosa caused by a mutation in SNRNP200, a gene
required for unwinding of U4/U6 snRNAs. Am J Hum Genet.
85:617–627.
Zhou J, Kherani F, Bardakjian TM, et  al. (2008). Identification of
novel mutations and sequence variants in the SOX2 and CHX10
genes in patients with anophthalmia/microphthalmia. Mol Vis.
14:583–592.
Zito I, Downes SM, Patel RJ, et al. (2003). RPGR mutation associated
with retinitis pigmentosa, impaired hearing, and sinorespiratory
infections. J Med Genet. 40:609–615.
Zobor D, Zrenner E (2012). Retinitis pigmentosa—a review.
Pathogenesis, guidelines for diagnostics and perspectives.
Ophthalmologe. 109:501–514.
Zrenner E, Bartz-Schmidt KU, Benav H, Besch D, Bruckmann A (2011).
Subretinal electronic chips allow blind patients to read letters and
combine them to words. Proc Biol Sci. 278:1489–1497.
Zuercher J, Neidhardt J, Magyar I, et  al. (2010). Alterations of the 5′
untranslated region of SLC16A12 lead to age-related cataract.
Invest Ophthalmol Vis Sci. 51:3354–3361.
 41.
GENETICS AND GENOMICS IN CLINICAL
OPHTHALMOLOGY, II: GLAUCOMA
Roshanak Sharafieh, Anne H. Child, and Mansoor Sarfarazi
INTRODUCTION
Glaucoma is a group of neurodegenerative ocular condi-
tions and a leading cause of irreversible blindness. This
disorder afflicts nearly 67 million people worldwide and is
the second most prevalent cause of bilateral blindness after
cataracts, affecting over 6.7  million sufferers worldwide.1,2
By 2020, it is predicted that there will be 79.6  million
people suffering from glaucoma, increasing the number of
bilaterally blind individuals due to glaucoma to 11.1  mil-
lion.2 Glaucoma is a large and complex group of disorders
that are recognized by their widely diverse clinical and
histopathological manifestations. The common charac-
teristics of glaucoma are its distinctive optic neuropathy,
increased intraocular pressure (IOP), the appearance of
the optic nerve head, and certain visual field abnormali-
ties.1,3 Although elevated IOP is clearly the most frequent
contributory risk factor for glaucomatous optic atrophy,
it is not the only feature. Nevertheless, IOP and aqueous
humor dynamics (pressure regulators) are critical to our
understanding of glaucoma, not only because they are the
most common and best understood of the causative risk fac-
tors for glaucoma, but also because they are the only factors
that can be controlled to slow down the progression of optic
neuropathy.
Glaucoma is delineated according to anatomy of the
anterior chamber (open- and closed-angle), etiology (pri-
mary or secondary), and age-of-onset (infantile, juvenile,
and adult).4 Herein, the primary and secondary forms of
glaucoma are briefly covered.
P R I M A RY O P E N -A N G L E
G L AU C O M A ( P OAG )
This is the most common form of glaucoma in the
Western Hemisphere, with characteristic manifestations of
636
glaucomatous optic neuropathy and visual field loss.5 This
group of disorders consist of three distinctive subsets includ-
ing, juvenile-onset primary open-angle glaucoma ( JOAG)
that develop before the age of 40 in approximately one in
50,000 people with highly elevated IOPs in both eyes,6,7
normal tension glaucoma (NTG) with IOPs of typically
≤21 mmHg,8 and adult-onset POAG that typically occurs
after 40 years of age. For these three POAG subtypes, cases
in the majority of published families are inherited as auto-
somal dominant.
P OAG G E N ET I C S
Currently, 17 official gene loci (GLC1A to GLC1Q) have
been designated by the HUGO Genome Nomenclature
Committee (HGNC) for different subtypes of POAG
(Table 41.1).9–24 At least 13 additional loci have also been
reported for various POAG subtypes (1p32, 2q31-q34,
5q14.3, 6q27, 7p12.3, 10p13-p11, 10q22, 12q21.33,
13q31.1, 14q11-q12, 14q22, 15q12, and 19p13.2), though
these have not been confirmed by other groups or given an
official locus designation.15,25–33
Several of the above-mentioned loci were iden-
tified through classical genetic linkage analysis of
population-specific JOAG (i.e., GLC1A,9 GLC1J,17
GLC1K,17 GLC1M,18 and GLC1N20), NTG (i.e., GLC1E,13
GLC1P23) or POAG families (i.e., GLC1B,10 GLC1C,11
GLC1D,12 GLC1F,14 GLC1G,19 GLC1H,21 GLC1I,20
GLC1L,16 GLC1O,22 GLC1Q24). So far, six defective genes
have been identified for JOAG (MYOC), NTG (OPTN)
and adult-onset POAG (WDR36, NTF4, ASB10, and
TBK1). These are briefly discussed below.
MYO C I L I N (MYO C)
Myocilin is the only gene identified so far in JOAG famil-
ial and sporadic cases. This gene is located at the GLC1A
 Table 41.1  GLAUCOMA-RELATED GENES AND LOCI

GENE HGNC LOCUS CLASSIFICATION CHROMOSOME OMIM NO.

IDENTIFIED DESIGNATION LOCATION
Myocilin GLC1A JOAG 1q24.3 601652; 137750
- GLC1B POAG 2cen-q13 %606689
- GLC1C POAG 3q21-q24 %601682
- GLC1D POAG 8q23 %602429
Optineurin GLC1E NTG 10p13 602432; 137760
ASB10 GLC1F POAG 7q35-q36 615054; %603383
WDR36 GLC1G POAG 5q22.1 609669; 609887
- GLC1H POAG 2p16.2-p15 %611276
- GLC1I POAG 15q11-q13 %609745
- GLC1J JOAG 9q22-q23 %608695
- GLC1K JOAG 20p12 %608696
- GLC1L POAG 3p21-p22 -
- GLC1M JOAG 5q22.1-q32 %610535
- GLC1N JOAG 15q22.2-q25.1 %611274
NTF4 GLC1O POAG, NTG, JOAG 19q13.33 %613100
TBK1 GLC1P NTG 12q14 604834
- GLC1Q POAG 4q35.1-q35.2 -
- GLC2A ACG 10q -
CYP1B1 GLC3A PCG 2p22.2 601771; 231300
- GLC3B PCG 1p36.2-p36.1 %600975
- GLC3C PCG 14q24.3-q31.1 %613085
LTBP2 GLC3D PCG 14q24.3 602091; 613086
- GLC3E PCG 19p13.2 -

locus (1q24.3) and is also known as Trabecular meshwork
Inducible Glucocorticoid Response (TIGR).34 Since the ini-
tial discovery of MYOC, over 98 different glaucoma-causing
mutations have been identified throughout the world in
juvenile- and adult-onset POAG families and sporadic
cases according to the Myocilin Allele-Specific Phenotype
Database (http://myocilin.com/variants.php). Taken
together, MYOC mutations are responsible for 2–4% of
glaucoma cases (Table 41.2). At times, Cytochrome P450,
family 1, subfamily B, polypeptide 1 (CYP1B1), a causative
primary congenital glaucoma (PCG) gene, has also been
observed to play a role in JOAG patients, thus indicating
an underlying genetic relationship between the JOAG and
PCG phenotypes through a digenic interaction between
MYOC and CYP1B1 genes.17,34–52 These studies empha-
size the genetic heterogeneity of juvenile glaucoma, and
furthermore suggest that congenital glaucoma and juvenile
glaucoma may be allelic variants, indicating that the range
of expression of MYOC and CYP1B1 mutations may be
greater than previously appreciated.39,40,43,53–59 It has also
been proposed that MYOC may interact with OPTN,60
APOE,61 Flotillin-1,62 and α1-Syntrophin.63 Myocilin
transgenic mice show a moderate elevation of intraocular
pressure, loss of retinal ganglion cells in the peripheral ret-
ina, and axonal degeneration in the optic nerve, a pheno-
type that is similar to those observed in human glaucoma
patients.64
O P T I N EU R I N (O P T I C N EU RO PAT H Y
I N D U C I N G P ROT E I N; O PTN )
Optineurin at the GLC1E locus (10p13) was identified
using a large British normal tension glaucoma family.65
Other groups screened the OPTN gene and reported new
mutations in their NTG, POAG, and JOAG cases.66–81 The
most common OPTN-mutation (E50K) is observed in
NTG subjects of a younger age, with more advanced optic
disc cupping, smaller neuroretinal rim area, and progres-
sive visual fields requiring filtration surgeries (Table 41.2).82
Optineurin has also been observed to co-segregate with

G enetic s and G enomic s in C linical O p h t h almology, I I :   G laucoma   •   6 3 7

Table 41.2  A REPRESENTATIVE GENOTYPE–PHENOTYPE CORRELATION IN KNOWN GLAUCOMA GENES

GENE MUTATION PHENOTYPE TRAITS

MYOC –1000C/G (promoter) POAG High IOP; more damaged visual field; more females; surgery
required
MYOC G252R POAG Intermediate form
MYOC G326S POAG Late age-of-onset; high IOP
MYOC T353I POAG In Asians; high IOP
MYOC A363T POAG Late age-of-onset; high IOP
MYOC Q368X POAG Milder; late age-of-onset; high IOP; common in Westerners
MYOC P370L JOAG Severe; early-onset; high IOP; surgery required
MYOC Y371D JOAG Aggressive form; early onset; surgery required
MYOC T377M POAG Intermediate risk; high IOP; high vertical cup/disc ratio
MYOC D380A JOAG Severe visual impairment; surgery required
MYOC K423E JOAG Variable clinical presentation
MYOC F430L POAG Severe glaucomatous damage
MYOC C433R JOAG Varied ages; high IOP; high vertical cup/disc ratio
MYOC Y437H JOAG Severe; early onset; high IOP
MYOC I477N JOAG Severe; early onset; high IOP
OPTN T34T POAG Allele ‘A’ risk for adult Asians
OPTN E50K POAG Severe NTG; earlier onset
OPTN M98K POAG Prevalent in Asian NTGs
OPTN IVS7+24G>A POAG Prevalent in HTG; high IOP
OPTN 691_692InsAG POAG Modestly high IOP
WDR36 rs10038177 POAG Prevalent in high tension glaucoma (HTG)
CYP1B1 G61E PCG Moderate to severe form
CYP1B1 E173K PCG Severe form
CYP1B1 R368H PCG Better visual progress; later age-of-onset
CYP1B1 R390H PCG Better visual outcomes with early surgery
CYP1B1 L432V POAG Risk allele GG
CYP1B1 N498D PCG Milder form
CYP1B1 rs162562 POAG Risk for adult onset

MYOC,83–87 APOE,61 and TNF-α88 in several familial and
sporadic cases. The OPTN protein is known to interact
with at least 14 other proteins, one of which (TBK1) has
already shown to be equally involved in NTG patients.89
It is anticipated that other OPTN-interacting proteins
may be equally involved in NTG or other glaucoma phe-
notypes. OPTN transgenic mice and over-expression stud-
ies of OPTN-E50K mutations provided evidence for its
contribution in neuronal cell processing, impaired protein
trafficking, and retinal ganglion cell apoptosis.90–96 Further
ongoing protein and cellular studies will add new insight
into the role of this molecule and the etiology of glaucoma
cases. In addition to its role in glaucoma, OPTN is also
reported to cause another neurodegenerative condition,
amyotrophic lateral sclerosis (ALS), in certain familial and
sporadic cases.97–105 Different spectrums of OPTN muta-
tions have been reported in ALS (autosomal recessive) com-
pared to NTG (autosomal dominant) subjects.
WD R E P E AT D O M A I N 36 ( WD R36)
Gene mutations at the GLC1G locus (5q22.1) were
observed in patients with high IOP, though there were
few NTG individuals who possessed mutations as well.19
Since then, many groups worldwide have screened and
confirmed mutations in WDR36 gene.106–112 Four different
studies suggest that WDR36 may account for 10–17% of
glaucoma subjects,19,107,110,111,113 while others found no muta-
tions in their POAG families.32,114–116 It has been shown
that WDR36 may act cooperatively with certain MYOC

6 3 8   •   G enomic s in C linical P ractice

mutations to reduce the age-of-onset by almost 10 years.111
One study suggested possible genetic interaction between
the WDR36 and P53 gene variations in Spanish POAG
subjects.117 Another group speculated that mutations in
WDR36 may not be sufficient to cause POAG, and this
gene may act as a modifier locus.107 However, contrary to
this suggestion, it has already been shown that mutations
in this gene directly affect axon growth of retinal gan-
glion cells and lead to progressive retinal degeneration,118
pre-implantation embryonic lethality in mice,119 or may
alter cell proliferation in a glaucoma model.120
N EU ROT RO P H I N-4 (N TF4)
The NTF4 gene at the GLC1O locus (19q13.33) was iden-
tified using European POAG, NTG, and JOAG subjects
(n = 892): approximately 1.68% of all their cases contained
mutations in this gene.22 Since the initial finding, four
other groups screened their glaucoma subjects, with mixed
results.121–126 Over time, additional studies will clarify the
role NTF4 plays in glaucoma.
A N K Y R I N R E P E AT A N D S O C S
B OX- C O N TA I N I N G 10 (A S B10)
The ASB10 gene at the GLC1F locus (7q36.1) contained
a number of glaucoma-causing mutations in American and
German subjects. Initially, a synonymous change (T255T)
was observed in a large American family, possibly affecting
an exon splice-enhancer site and changing mRNA-splicing
in the lymphoblasts of glaucoma individuals. Further popu-
lation screening of the ASB10 gene identified additional
missense alterations accounting for 6% of the American and
German population groups.127 However, no disease-causing
mutations were identified in a group of 158 POAG subjects
from Iowa.128 No further confirmations have been made as
yet, and the exact role of this gene in the pathophysiology of
glaucoma still awaits further investigation.
TA N K-B I N D I N G K I NA S E 1 (TBK1)
The TANK-binding Kinase 1 gene at the GLC1P locus
(12q14.2) was singled out after a 780-base pair duplica-
tion containing four genes observed within the region. One
African-American NTG patient had a small duplication
spanning the TBK1 gene. Although the study could not
correlate glaucoma alone or as a group with any of the vari-
ants, the role of TBK1 and the other three genes need to be
determined by screening additional samples. To date, this is
the first copy number variation (CNV) mutation associated
G enetic s and G enomic s in C linical O p h t h almology, I I :   G laucoma   •   6 3 9
with NTG,89 though other CNVs have previously been
reported in POAG129,130 and PCG cases.131 As mentioned
above, TBK1 is one of the known interacting proteins for
OPTN, and it is very interesting that both of these proteins
are involved in the etiology of the same type of glaucoma,
namely NTG.
P OAG G E N E A S S O C I AT I O N S T U D I E S
In addition to the six above-mentioned POAG-causing
genes that were initially identified in their respective, multi-
ply affected, linked families, many other genes have also been
thought to be involved in various forms of glaucoma. Most
of these genes were identified either through genome-wide
association studies, by cross-species functional relationship,
or candidacy of specific genes. Representative lists of genes
associated with various forms of glaucoma are presented
in Table 41.3. For a number of these glaucoma-associated
genes, subsequent replication studies were unable to con-
firm the original study, or such associations were reported
in different subtypes of glaucoma than in the original study.
There are also numerous conflicting reports about specific
gene associations in different ethnical population with vari-
ous forms of glaucoma. For a number of newly reported
glaucoma–gene associations, there are still no confirma-
tions; therefore it is too early to speculate about the clinical
utility of such reports. Nevertheless, there is a controversy
regarding the overall value of such association studies for
this particular neurodegenerative disorder with its antici-
pated involvement of several hundred causative genes.
Perhaps the only exception is the reported association of
pseudoexfoliation glaucoma (XFS) with the LOXL1 gene
polymorphisms that have been confirmed in different pop-
ulations (see XFS section below).
A N G L E - C L O S U R E G L AU C O M A
One of the principal causes of irreversible blindness in the
Asian population is angle-closure glaucoma (ACG).2,132,133
ACG has been categorized as acute, sub-acute, intermit-
tent, or chronic, depending on appearance of the angle.134–137
There is a three- to five-fold risk increase in ACG develop-
ment in subjects with first-degree affected relatives.138–141
Several groups have studied ACG cases,141–153 but to
date, only one locus (GLC2A on 10q) has been officially
documented in a large autosomal dominant Singaporean
ACG family (Table  41.1).144 In the same study, seven
other chromosomal loci (1q, 3p, 7p, 10p, 11q, 13q, and
18p) were identified as potentially linked in other ACG
Table 41.3  A REPRESENTATIVE LIST OF GENES ASSOCIATED WITH GLAUCOMA

GENE CHROMOSOME OMIM NO.

LOCATION
β-1-adrenergic receptor ADRB1 10q25.3 109630
Angiotensin II receptor, type 1 AGTR1 3q24 106165
Angiotensin II receptor type 2 AGTR2 Xq23 300034
Apolipoprotein E APOE 19q13.32 107741
Atonal homolog 7 ATOH7 10q21.3 609875
Caveolin 1 CAV1 7q31.2 601047
Caveolin 2 CAV2 7q31.2 601048
Cadherin 1 CDH1 16q22.1 192090
Cyclin-dependent kinase inhibitor 1A CDKN1A 6p21.2 116899
Cyclin-dependent kinase inhibitor 2A CDKN2A 9p21.3 600160
Cyclin-dependent kinase inhibitor 2B CDKN2B 9p21.3 600431
Serine/threonine-protein kinase Chk2 CHEK2 22q12.1 604373
Collagen alpha-1(V) chain preproprotein COL5A1 9q34.3 120215
Collagen alpha-2(VIII) chain precursor COL8A2 1p34.3 120252
Collagen alpha-1(XI) chain (associated with ACG) COL11A1 1p21.1 120280
Endothelin receptor type A EDNRA 4q31.22 131243
Elongation of very long chain fatty acids protein 5 ELOVL5 6p12.1 611805
Mitochondrial enolase superfamily member 1 eNOSF1 18p11.32 607427
Glutathione S-transferase mu 1 GSTM1 1p13.3 138350
Hexokinase-2 HK2 2p12 601125
Heat Shock 70kDa Protein 1A HSPA1A 6p21.33 140550
Insulin-like growth factor II IGF2 11p15.5 147470
Interleukin 1, Alpha IL-1α 2q13 147760
Interleukin 1, Beta IL-1β 2q13 147720
Lysyl oxidase-like 1 LOXL1 15q24.1 153456
Mitofusin 1 MFN1 3q26.33 608506
Mitofusin 2 MTF2 1p36.22 608507
Methylenetetrahydrofolate reductase MTHFR 1p36.22 607093
NCK adaptor protein 2 NCK2 2q12.2 604930
Nitric oxide synthase 2A NOS2A 17q11.2 163730
Nitric oxide synthase 3 NOS3 7q36.1 163729
Neurotrophin 4 NTF4 19q13.33 162662
Oculomedin OCLM 1q31.1 604301
Noelin 2 OLFM2 19p13.2 -
Optic atrophy 1 OPA1 3q29 605290
Opticin OPTC 1q32.1 605127
Presenilin associated, rhomboid-like PARL 3q27.1 607858
Protein-L-isoaspartate O-methyltransferase PCMTD1 8q11.23 -
domain-containing protein 1 (associated with ACG)
Pleckstrin homology domain-containing family A member 7 PLEKHA7 11p15.1 612686
(associated with ACG)
Plexin domain containing 2 PLXDC2 10p12.31 606827
Serum paraoxonase/arylesterase 1 precursor PON1 7q21.3 168820

(continued)

6 4 0   •   G enomic s in C linical P ractice

 Table 41.3 CONTINUED

GENE CHROMOSOME OMIM NO.

LOCATION
Homeobox protein SIX1 SIX1 14q23.1 601205
Homeobox protein SIX6 SIX6 14q23.1 606326
S1 RNA-binding domain-containing protein 1 SRBD1 2p21 -
Sushi-repeat-containing protein, X-linked SRPX Xp11.4 300187
Suppression of tumorigenicity 18 protein (associated with ST18 8q11.23 -
ACG)
Transporter 1, ATP-binding cassette, sub-family B TAP1 6p21.32 170260
Trabecular meshwork inducible stretch response TISR 1q31.1 -
Toll-like receptor 4 TLR4 9q33.1 603030
Transmembrane and tetratricopeptide repeat containing 2 TMTC2 12q21.31 -
Tumor necrosis factor Alpha TNFα 6p21.33 191160
Cellular tumor antigen p53 TP53 17p13.1 191170
Zinc finger protein 469 ZNF469 16q24.2 612078
Zona pellucida glycoprotein 4 ZP4 1q43 613514

families.144 Recently, another provisional locus was reported P R I M A RY C O N G E N I TA L

on chromosome 3q27.1, though it has not been officially
designated.154
ACG has a few common clinical manifestations with
microphthalmia and nanophthalmia phenotypes, includ-
ing a short axial length, shallow anterior chamber depth,
small eyes with larger lenses, and elevated intraocular
pressure. Due to these clinical similarities, a number of
microphthalmia- and nanophthalmia-causing genes such as
VSX2/CHX10 (Visual System homeobox 2),155,156 MFRP
(Membrane Frizzled-Related Protein),157 and PRSS56
(Serine Protease)158 have been screened in ACG subjects.
However, these studies have been unable to demonstrate
that these genes play a significant role in ACG, although
several polymorphisms were noted.159–162 As yet, no single
causative angle-closure glaucoma gene has been identi-
fied. However, in one Chinese family, a compound het-
erozygote mutation in MYOC (Arg46Stop) and CYP1B1
(Leu432Val) has been described.163
Genome-wide association studies in several differ-
ent populations led to the discovery of numerous candi-
date genes for ACG. However, most of these were either
not reproduced in other populations, or their conclud-
ing results were controversial. The most recent study in
the Asian population identified three single-nucleotide
polymorphisms (SNPs) of rs11024102 in PLEKHA7
(11p15.1), rs3753841 in COL11A1 (1p21.1), and
rs1015213 (8q11.23), which are highly associated with
the ACG phenotype.164
G L AU C O M A ( P C G )
This condition, also known as buphthalmos or infantile glau-
coma, usually manifests itself at birth or within the first three
years of life.165 It is generally agreed that PCG is a heteroge-
neous condition, based on both clinical and genetic data.166–
169
To date, there are five chromosomal loci designated for
PCG: GLC3A to GLC3E, which show linkage in familial
primary congenital glaucoma cases169–171,172 (Table 41.1). At
present, the Cytochrome P4501B1 (CYP1B1) gene at the
GLC3A locus appears to account for 20%–100% of sporadic
and familial PCG cases worldwide,35,173–193 a variety of other
glaucoma forms such as JOAG and POAG,37,39,40,43,194–201
or  in digenic association with other genes such as
MYOC.43,57–59,178,202–205 As more studies are completed, the
importance of CYP1B1 in the etiology of various glaucoma
subtypes will help us understand the role this gene plays in
the degeneration of retinal ganglion cells that lead to visual
field loss. Since the discovery of CYP1B1 as the first gene
for PCG in 1997,126 more than 150 different mutations
have been reported in over 500 PCG patients worldwide
(Table  41.2). Therefore, CYP1B1 mutations are the most
identifiable cause of different forms of glaucoma, followed
by 98 mutations in the MYOC gene. These two genes are the
most clinically relevant glaucoma-causing genes that have
been identified so far.
There has been debate over whether another
gene-namely, Latent Transforming Growth Factor

G enetic s and G enomic s in C linical O p h t h almology, I I :   G laucoma   •   6 4 1

(TGF)-beta Binding Protein 2 (LTBP2) is a causative PCG
gene at the GLC3D locus (14q24.3). Initially, four consan-
guineous PCG families from Pakistan and of Gypsy ori-
gin were identified as having mutations within this gene.206
Soon after, additional mutations were confirmed in two
consanguineous Iranian PCG families.207 Interestingly,
in addition to the PCG phenotype, many affected sub-
jects used in these studies had other notable clinical fea-
tures (i.e., ectopia lentis and joint hypermobility) that are
typically observed in Marfan syndrome. Sometimes lens
displacement can lead to pupillary block and secondary
angle-closure glaucoma, which can be incorrectly diag-
nosed as primary congenital glaucoma in infants and young
children (reviewed here208). Further assessment of LTBP2
in different populations will determine its role in the etiol-
ogy of primary congenital glaucoma and other associated
ocular syndromes. So far, screening of the LTBP2 gene in
PCG subjects with isolated trabeculodysgenesis from the
United States,209 Britain,208 Saudi Arabia,210 and North
India211 has not revealed any definitive PCG-causing muta-
tions in this gene. Therefore, it is likely that LTBP2 muta-
tions are only involved in other complex ocular syndromes
such as megalocornea with zonular weakness, microsphe-
rophakia, isolated ectopia lentis, and Weill-Marchesani
syndrome, all with or without manifestation of glaucoma
as a secondary phenotype.
S E C O N DA RY G L AU C O M A ( S G )
Forms of glaucoma that are acquired as a consequence of
other ocular conditions are considered “secondary,” since no
specific cause for their glaucoma phenotype is distinguish-
able. Ocular conditions known to produce subsequent
glaucoma include, but are not limited to, exfoliation syn-
drome (XFS), pigment dispersion syndrome (PDS) and
Rieger syndrome (RS).
E X F O L I AT I O N SY N D RO M E ( X F S ) O R
P S EU D O E X F O L I AT I O N SY N D RO M E
Exfoliation syndrome is a systemic abnormality that mainly
affects the eye.212,213 XFS is the most commonly identified
form of glaucoma, but the pathogenesis and composition
of exfoliation material are still poorly understood. This
age-related disorder triggers an aggressive secondary glau-
coma, with formation and deposition of a white fibrillar
material in the anterior segment, conjunctiva, and orbital
tissues of the eye, as well as in tissues throughout the body.214
Despite an open-angle, IOP rises as a result of physical
6 4 2   •   G enomic s in C linical P ractice
blockage of the trabecular meshwork and injury to the
trabecular endothelial cells.215 As the majority of affected
subjects are diagnosed at very late stages of life, presence
of multiple cases in large families have been extremely
rare. This in turn has been a major drawback for classical
linkage-analysis and mapping of the defective gene(s) for
this disorder. Family history can play an important role
in exfoliation syndrome,216–218 though it is still unclear
whether XFS is transmitted maternally (i.e., whether there
is a possibility of mitochondrial inheritance), or is inher-
ited as an X-linked or autosomal inheritance with genomic
imprinting.216 Nonetheless, there are cases of male-to-male
transmission, suggesting that XFS can occur as an autoso-
mal dominant condition.219
Currently, no official XFS loci have been designated.
However, loss of heterozygosity for two DNA markers of
D7S479 and D13S175,220 and two other regions on chro-
mosomes 18q221 and 2q222,223 have been suggested as pos-
sible areas of interest for this condition.
The best-known genetic contribution to XFS was
identified through a genome-wide association study
in the Icelandic population. Several SNPs in the Lysyl
Oxidase-like 1 (LOXL1) gene (15q24.1) were shown to
be highly associated with the exfoliation phenotype.224
Replication studies in the Swedish population confirmed
genetic susceptibility of LOXL1 polymorphisms to
exfoliation with (XFG) or without glaucoma (XFS).224
However, no genetic association was observed in a group
of unrelated POAG patients for either the Icelandic or
Swedish populations.224 Since then, associations of the
LOXL1 gene polymorphisms have been confirmed in
a large group of XFG and XFS subjects from different
populations.224–243 Despite indisputable association of
XFS with LOXL1, the low penetrance and variable dis-
ease incidence within diverse populations suggest that
involvement of environmental factors alter the pheno-
typical expression of XFS.234 The association of LOXL1
polymorphisms with exfoliation (with or without glau-
coma) was confirmed in all of the populations studied.
However, researchers were unable to show a strong genetic
association between the LOXL1 gene polymorphisms and
the glaucoma phenotype. This suggests that these poly-
morphisms are exclusively associated with the presence of
exfoliation material in these patients and that the second-
ary presentation of glaucoma in XFG subjects is not asso-
ciated with LOXL1 gene polymorphisms. It is anticipated
that next-generation exome sequencing of familial and
sporadic XFS and XFG cases, currently underway, will
identify new disease-causing genes and shed new light on
the etiology of this phenotype.
 P I G M E N T D I S P E R S I O N SY N D RO M E ( P D S ) root cause of ocular disorders like glaucoma, particularly
the primary forms. There has been significant progress in
In the Western world, pigment dispersion syndrome (PDS)
identifying several causative glaucoma genes, studying their
accounts for 1–1.5% of all glaucoma cases and appears to be
proteins, and developing mouse models with specific gene
inherited in an autosomal dominant fashion.244–246 Pigment
mutations seen in humans. However, according to popula-
granules that normally remain in the back of the iris break
tion studies, the contributing genes thus detected, account
off and are carried by the aqueous humor to anterior segment
for only a fraction of the glaucoma incidence worldwide,
structures such as the trabecular meshwork, lens surface, cor-
meaning that this disorder is more complicated than origi-
neal endothelium, lens zonule, and iris surface. As granules
nally anticipated. We now know that variability in age-of-
accumulate in these areas, normal filtration and outflow of
onset, ethnicity, inheritance, and subtype of glaucoma play
aqueous humor are reduced significantly, putting pressure
a larger role in identifying common patterns. For this rea-
on the optic nerve; hence leading to the development of pig-
son, accurate patient diagnosis of the type of glaucoma is
mentary glaucoma.247–249 A linkage study indicated a possible
critical for preventing disease progression and selecting the
region on chromosome 7q35-36,250 though no confirmation
correct medical treatments for management. As glaucoma is
has been made by other groups. There has been debate over
considered a neurodegenerative condition rather than a sin-
whether there is an association between the LOXL1 gene
gle clear-cut ocular entity, the lack of an exact diagnosis and
polymorphism (rs1048661) and age-of-onset for PDS.243,251
definition of different subtypes is still a major hindrance in
Currently, there is a mouse model with secondary glaucoma
proper recognition and categorization of this group of dis-
due to pigmentary change252,253 being studied for gene and
orders. As many other systemic, clinical, and ocular diseases
protein involvements, but no human gene(s) have been iden-
are manifested with glaucoma as a secondary condition, a
tified to be disease-causing in PDS as yet.
proper distinction between primary and secondary forms
of this condition is of utmost importance. Therefore, it is
R I EG E R SY N D RO M E ( R S ) possible that several hundred genes act individually, or in
a polygenic or multifactorial fashion, to influence develop-
Rieger syndrome is also known as Axenfeld’s posterior
ment of the glaucoma phenotype.
embryotoxon-juvenile glaucoma, Rieger’s anomaly, Rieger’s
Previously, researchers had to sequence each individual
disease, or Rieger’s malformation, which is characterized by
gene, one fragment at a time. But new developments in
systemic issues such as maxillary hypoplasia, dental hypo-
next-generation sequencing (NGS) with massive, parallel,
plasia, failure of involution of periumbilical skin, and mal-
high-throughput enrichment technologies have provided
formation of the anterior segment of the eye.254–257 Axenfeld
opportunities for studying Mendelian disorders through
syndrome consists of two of the three major eye malforma-
direct-sequencing of the entire human genome within
tions found in Rieger syndrome, and it is suggested that the
weeks, rather than years. Although this advancement is in
two syndromes result from different expressions in the same
its relative infancy, it has rapidly become the standard for
gene.258 Nearly half the children with Axenfeld or Rieger
searching for genetic defects in familial and sporadic cases,
develop secondary glaucoma with an increased risk of
with or without prior information linking conditions to a
vision loss and blindness.256,259–263 RS is an autosomal domi-
given region of the genome. NGS has been successfully uti-
nant disorder with incomplete penetrance. Currently, two
lized for disease-gene discovery projects concerning several
loci have been identified: RIEG1, located on chromosome
Mendelian and complex diseases, leading to the identifica-
4q25-q26,264 and RIEG2, located on chromosome 13q14.265
tion of hundreds of genes in different disorders worldwide.
Presently, two genes encoding transcription factors (PITX2
With this new technology, one can not only identify a caus-
on 4q25 and FOXC1 on 6p25.3) have been identified as
ative gene, but also identify other genes in the genome that
disease-causing in Axenfeld-Rieger syndrome.257,266,267
collectively can contribute to the genetic makeup of the
disease phenotype. Since the genetics of glaucoma are more
complicated than once anticipated, NGS is likely to shed
G L AU C O M A G E N ET I C S A N D additional light on the genetic composition of the glaucoma
G E N O M I C S —W H AT D O W E D O N OW ? phenotypes, leading to a better understanding of these dis-
orders. In turn, more precise individualized genomic testing
Although this question may appear to be easy to answer, (on gene chips) can accurately diagnose patients with gene
Nature has its own way of throwing a curve ball. Over the mutations, and further guide a clinician to treat the patient
last three decades, we have attempted to understand the with the correct medications or appropriate surgeries, if

G enetic s and G enomic s in C linical O p h t h almology, I I :   G laucoma   •   6 4 3

needed. In due time, once the mechanism of glaucoma is
better understood, we will be able to prevent the disease
from beginning or possibly reverse its damaging effect on
the optic nerve by using targeted stem-cell therapy, as is
being used in treatment for retinal disorders like retinitis
pigmentosa (RP).268–271
G E N O M I C A P P L I C AT I O N
I N   G L AU C O M A
Genetic analysis of a group of glaucoma subtypes provided
substantial information on the complexity of this large
group of eye disorders, as well as new insights into genetic
contributions to the development of this devastating ocular
phenotype. So far, we have mapped over 30 different genetic
loci for various forms of glaucoma subtypes and identified
at least six of their causative molecules. Transgenic mice
have been generated for at least four of these glaucoma-
causing genes; thus helping researchers dissect the complex
cellular processes that lead to the development of glaucoma.
Furthermore, such studies are being used to determine
physical interactions between each of these individual glau-
coma-causing genes and/or with a number of other pro-
teins. Although we have come a long way in understanding
the genetic contribution of this phenotype, the application
of whole-genomic analysis to worldwide glaucoma sufferers
is still actively in progress, and it may take another five years
before we have a comprehensive genetic appreciation of
this particular group of eye disorders. It is anticipated that
a large number of new genes will be identified for glaucoma
in a relatively short period of time. However, as of this writ-
ing, the existing glaucoma genes and numerous mutations
identified by various groups have already provided the first
opportunity for prenatal and pre-symptomatic diagnosis,
as well as high-risk, gene carrier identification for differ-
ent glaucoma subtypes in families all over the world. This
initial molecular diagnosis has subsequently provided clini-
cians an enhanced ability to make an early diagnosis at any
stage of the life cycle; determine the patient’s predisposition
to congenital, juvenile, or adult-onset glaucoma; make an
accurate prognosis of the disease in late-life, and offer bet-
ter clinical treatment based on individualized gene infor-
mation. Prior genotype–phenotype correlation studies can
be used to provide new patients with a complete gene-risk
assessment. This, in turn, could have substantial clinical and
psychological benefits for the patients and their high-risk
relatives. However, even with this prospect at hand, there is
still a great deal of work that needs to be done before we can
provide life-changing, individualized genomic information
6 4 4   •   G enomic s in C linical P ractice
to these patients. We still need to identify several hundred
new glaucoma-causing genes through next-generation
genome sequencing and study the complicated relationship
between them and other interacting genes and proteins that
are not expected to be directly involved in the etiology of
glaucoma. Examining the gene-to-gene and protein-to-pro-
tein interactions of these hundreds of genes will enable us to
have a better understanding of the pathophysiology of this
neurodegenerative disorder and to use such information for
designing specific drugs or medical therapies for victims
affected by glaucoma worldwide.
Even after the majority of the glaucoma-causing genes
have been identified by NGS, the next step will be to use
such information for the development of new and effective
drug therapies. However, this will be an intricate process that
will require additional information from other scientific
disciplines, including the biochemical and pharmaceutical
professions, as well as require comprehensive genetic infor-
mation on other species and/or glaucoma-animal models.
As most of the genes are highly conserved across species,
studying these basic organisms (i.e., mouse, rat, nematode,
fruit fly, zebra fish, yeast, fungal and bacterial genomes) can
yield valuable insight for drug development and targeted
therapeutics in higher organisms. The complexities of mod-
ern drug discovery are in the development of technologies
that can integrate bioinformatics, disease modeling, protein
discovery, protein expression, gene chips, target validation,
and high-throughput screening for genes of unknown func-
tion. Incorporating such complex processes are key steps
for discovering effective drugs and determining their subse-
quent therapeutic application to various glaucoma subtypes.
The knowledge obtained through ongoing next-generation
genome sequencing of different glaucoma subtypes will not
only accelerate gene identification, genetic counseling, and
more objective clinical diagnosis, but will also be another
contributory component of the much-anticipated personal-
ized genomic medicine in the near future.
REFERENCES
1. Allingham RR, Damji KF, Freedman S, Moroi SE, Shafranov G,
Shields MB. Clinical Epidemiology of Glaucoma In:  Damji KF,
Freedman SF, Moroi SE, Rhee DJ, Shields MB, eds. Shields Textbook
of Glaucoma (5th Ed) (p. 149). Philadelphia, PA:  Lippincott
Williams & Wilkins; 2005.
2. Quigley HA, Broman AT. The number of people with glaucoma
worldwide in 2010 and 2020. The British Journal of Ophthalmology.
Mar 2006;90(3):262–267.
3. Van Buskirk E, Cioffi G. Glaucomatous optic neuropathy. American
Journal of Ophthalmology. 1992;113:447.
4. Shields MB, Krupin T. Classifications of the Glaucomas In: Ritch
R, Shields MB, and Krupin T, eds. The Glaucomas (2nd Ed; Vol. 2).
St. Louis, MO: Mosby; 1996.
 5. Sharts-Hopko NC, Glynn-Milley C. Primary open-angle glaucoma.
American Journal of Nursing. Feb 2009;109(2):40–47; Quiz 48.
6. Johnson AT, Drack AV, Kwitek AE, et al. Clinical features and link-
age analysis of a family with autosomal dominant juvenile glaucoma.
Ophthalmology. Apr 1993;100(4):524–529.
7. Hitchings RA. What is primary open angle glaucoma? In: Hitchings
RA, Lightman S, eds. Glaucoma. London: BMJ Publishing Group;
2000.
8. Hitchings RA. Low tension glaucoma-its place in modern
glaucoma practice. The British Journal of Ophthalmology. Aug
1992;76(8):494–496.
9. Sheffield VC, Stone EM, Alward WL, et  al. Genetic linkage of
familial open angle glaucoma to chromosome 1q21-q31. Nature
Genetics. May 1993;4(1):47–50.
10. Stoilova D, Child A, Trifan OC, Crick RP, Coakes RL, Sarfarazi
M. Localization of a locus (GLC1B) for adult-onset primary
open angle glaucoma to the 2cen-q13 region. Genomics. Aug 15,
1996;36(1):142–150.
11. Wirtz MK, Samples JR, Kramer PL, et  al. Mapping a gene for
adult-onset primary open-angle glaucoma to chromosome 3q.
American Journal of Human Genetics. Feb 1997;60(2):296–304.
12. Trifan OC, Traboulsi EI, Stoilova D, et al. A third locus (GLC1D)
for adult-onset primary open-angle glaucoma maps to the 8q23
region. American Journal of Ophthalmology. Jul 1998;126(1):17–28.
13. Sarfarazi M, Child A, Stoilova D, et al. Localization of the fourth
locus (GLC1E) for adult-onset primary open-angle glaucoma to
the 10p15-p14 region. American Journal of Human Genetics. Mar
1998;62(3):641–652.
14. Wirtz MK, Samples JR, Rust K, et  al. GLC1F, a new primary
open-angle glaucoma locus, maps to 7q35-q36. Archives of
Ophthalmology. Feb 1999;117(2):237–241.
15. Wiggs JL, Allingham RR, Hossain A, et  al. Genome-wide scan
for adult onset primary open angle glaucoma. Human Molecular
Genetics. Apr 12, 2000;9(7):1109–1117.
16. Baird PN, Foote SJ, Mackey DA, et al. Evidence for a novel glau-
coma locus at chromosome 3p21–22. Human Genetics. Jul
2005;117(2–3):249–257.
17. Wiggs JL, Lynch S, Ynagi G, et al. A genomewide scan identifies novel
early-onset primary open-angle glaucoma loci on 9q22 and 20p12.
American Journal of Human Genetics. Jun 2004;74(6):1314–1320.
18. Wang DY, Fan BJ, Canlas O, et al. Absence of myocilin and optineu-
rin mutations in a large Philippine family with juvenile onset primary
open angle glaucoma. Molecular Vision. Nov 9, 2004;10:851–856.
19. Monemi S, Spaeth G, DaSilva A, et  al. Identification of a novel
adult-onset primary open-angle glaucoma (POAG) gene on 5q22.1.
Human Molecular Genetics. Mar 15, 2005;14(6):725–733.
20. Wang DY, Fan BJ, Chua JK, et  al. A genome-wide scan maps
a novel juvenile-onset primary open-angle glaucoma locus
to 15q. Investigative Ophthalmology and Visual Science. Dec
2006;47(12):5315–5321.
21. Suriyapperuma SP, Child A, Desai T, et al. A new locus (GLC1H)
for adult-onset primary open-angle glaucoma maps to the 2p15-p16
region. Archives of Ophthalmology. Jan 2007;125(1):86–92.
22. Pasutto F, Matsumoto T, Mardin CY, et  al. Heterozygous NTF4
mutations impairing neurotrophin-4 signaling in patients with pri-
mary open-angle glaucoma. American Journal of Human Genetics.
Oct 2009;85(4):447–456.
23. Fingert JH, Robin AL, Stone JL, et  al. Identification of a Novel
Glaucoma Locus. Investigative Ophthalmology and Visual Science.
2010;51:E-Abstract 2159.
24. Porter LF, Urquhart JE, O’Donoghue E, et  al. Identification of a
novel locus for autosomal dominant primary open angle glaucoma
on 4q35.1-q35.2. Investigative Ophthalmology and Visual Science.
Oct 4, 2011;52(11):7859–7865.
25. Charlesworth JC, Dyer TD, Stankovich JM, et  al. Linkage to
10q22 for maximum intraocular pressure and 1p32 for maxi-
mum cup-to-disc ratio in an extended primary open-angle
G enetic s and G enomic s in C linical O p h t h almology, I I :   G laucoma   •   6 4 5
glaucoma pedigree. Investigative Ophthalmology and Visual Science.
2005;46:3723–3739.
26. Duggal P, Klein AP, Lee KE, et  al. A genetic contribution to
intraocular pressure:  the Beaver Dam Eye Study. Investigative
Ophthalmology and Visual Science. 2005;46:555–560.
27. Duggal P, Klein AP, Lee KE, et al. Identification of novel genetic loci
for intraocular pressure: a genomewide scan of the Beaver Dam Eye
Study. Archives of Ophthalmology. Jan 2007;125(1):74–79.
28. Hauser MA, Schmidt S, Nemesure B, et  al. Replication of the
Chromosome 2 Locus for POAG in an Independent Dataset of
African Origin. Investigative Ophthalmology and Visual Science.
2007;48:E-Abstract 4037.
29. Nemesure B, Jiao X, He Q, et  al. A genome-wide scan for pri-
mary open-angle glaucoma (POAG):  the Barbados Family
Study of Open-Angle Glaucoma. Human Genetics. May
2003;112(5–6):600–609.
30. Ramchand KV, Palin W, DelBono E, et  al. Fine Mapping of the
Chromosome 14 POAG Region and Exclusion of COCH as a
Candidate Gene. Investigative Ophthalmology and Visual Science.
2007;48:E-Abstract 5604.
31. Rice T, Rankinen T, Province MA, et  al. Genome-wide linkage
analysis of systolic and diastolic blood pressure: the Québec Family
Study. Circulation. 2000;102:1956–1963.
32. Rotimi CN, Chen G, Adeyemo AA, et al. Genomewide scan and
fine mapping of quantitative trait loci for intraocular pressure on 5q
and 14q in West Africans. Investigative Ophthalmology and Visual
Science. Aug 2006;47(8):3262–3267.
33. Wiggs JL, Allingham RR, Auguste J, et  al. Genomic Screen for
Adult-Onset Primary Open Angle Glaucoma:  Follow-up Studies
Suggest Loci on 14q11 and 15q. American Society of Human
Genetics; 2003:E-Abstract 1818.
34. Stone EM, Fingert JH, Alward WL, et al. Identification of a gene
that causes primary open angle glaucoma. Science (New York, N.Y.).
Jan 31, 1997; 275(5300):668–670.
35. Stoilov I, Akarsu AN, Sarfarazi M. Identification of three different
truncating mutations in cytochrome P4501B1 (CYP1B1) as the
principal cause of primary congenital glaucoma (buphthalmos) in
families linked to the GLC3A locus on chromosome 2p21. Human
Molecular Genetics. Apr 1997;6(4):641–647.
36. Stoilov IR, Costa VP, Vasconcellos JP, et al. Molecular genetics of
primary congenital glaucoma in Brazil. Investigative Ophthalmology
and Visual Science. Jun 2002;43(6):1820–1827.
37. Acharya M, Mookherjee S, Bhattacharjee A, et  al. Primary role
of CYP1B1 in Indian juvenile-onset POAG patients. Molecular
Vision. 2006;12:399–404.
38. Alward W. The genetics of open-angle glaucoma:  the story of
GLC1A and myocilin. Eye. 2000;14(Pt B):429–436.
39. Bayat B, Yazdani S, Alavi A, et  al. Contributions of MYOC
and CYP1B1 mutations to JOAG. Molecular Vision.
2008;14:508–517.
40. Melki R, Colomb E, Lefort N, Brezin AP, Garchon HJ. CYP1B1
mutations in French patients with early-onset primary open-angle
glaucoma. Journal of Medical Genetics. Sep 2004;41(9):647–651.
41. Melki R, Idhajji A, Driouiche S, et  al. Mutational analysis of the
Myocilin gene in patients with primary open-angle glaucoma in
Morocco. Ophthalmic Genetics. Sep 2003;24(3):153–160.
42. Michels-Rautenstrauss KG, Mardin CY, Budde WM, et  al.
Juvenile open angle glaucoma:  fine mapping of the TIGR gene
to 1q24.3-q25.2 and mutation analysis. Human Genetics. Jan
1998;102(1):103–106.
43. Vincent AL, Billingsley G, Buys Y, et  al. Digenic inheritance of
early-onset glaucoma: CYP1B1, a potential modifier gene. American
Journal of Human Genetics. Feb 2002;70(2):448–460.
44. Alward WL, Fingert JH, Coote MA, et  al. Clinical features asso-
ciated with mutations in the chromosome 1 open-angle glaucoma
gene (GLC1A). The New England Journal of medicine. Apr 9,
1998;338(15):1022–1027.
 45. Bruttini M, Longo I, Frezzotti P, et  al. Mutations in the myo-
cilin gene in families with primary open-angle glaucoma and
juvenile open-angle glaucoma. Archives of Ophthalmology. Jul
2003;121(7):1034–1038.
46. Puska P, Lemmela S, Kristo P, Sankila EM, Jarvela I. Penetrance and
phenotype of the Thr377Met Myocilin mutation in a large Finnish
family with juvenile- and adult-onset primary open-angle glaucoma.
Ophthalmic Genetics. Mar 2005;26(1):17–23.
47. Tamm ER, Russell P. The role of Myocilin/TIGR in glau-
coma:  Results of the Glaucoma Research Foundation Catalyst
Meeting in Berkeley, California, March 2000. Journal of Glaucoma.
2001;10(4):329–339.
48. Tsai JC, Chang HW, Kao CN, Lai IC, Teng MC. Trabeculectomy
with mitomycin C versus trabeculectomy alone for juvenile pri-
mary open-angle glaucoma. Ophthalmologica. 2003;217(1):
24–30.
49. Wiggs JL, Del Bono EA, Schuman JS, Hutchinson BT, Walton
DS. Clinical features of five pedigrees genetically linked to the juve-
nile glaucoma locus on chromosome 1q21-q31. Ophthalmology.
1995;102(12):1782–1789.
50. Wiggs JL, Allingham RR, Vollrath D, et  al. Prevalence of muta-
tions in TIGR/Myocilin in patients with adult and juvenile primary
open-angle glaucoma. American Journal of Human Genetics. Nov
1998;63(5):1549–1552.
51. Wiggs JL, Haines JL, Paglinauan C, et al. Genetic linkage of auto-
somal dominant juvenile glaucoma to 1q21-q31 in three affected
pedigrees. Genomics. May 15, 1994;21(2):299–303.
52. Wiggs JL, Vollrath D. Molecular and clinical evaluation of a patient
hemizygous for TIGR/MYOC. Archives of Ophthalmology. Nov
2001;119(11):1674–1678.
53. Brockmoller J, Cascorbi I, Kerb R, Sachse C, Roots I. Polymorphisms
in xenobiotic conjugation and disease predisposition. Toxicology
Letters. Dec 28, 1998;102–103:173–183.
54. Kim YM, Bombeck CA, Billiar TR. Nitric oxide as a bifunctional
regulator of apoptosis. Circulation Research. 1999;84:253–256.
55. Klett CP, Bonner TI. Identification and characterization of the rat
M1 muscarinic receptor promoter. Journal of Neurochemistry. Mar
1999;72(3):900–909.
56. Gong G, Kosoko-Lasaki O, Haynatzki GR, Wilson MR. Genetic
dissection of myocilin glaucoma. Human Molecular Genetics. Apr 1,
2004;13(1):R91–R102.
57. Kaur K, Reddy AB, Mukhopadhyay A, et al. Myocilin gene impli-
cated in primary congenital glaucoma. Clinical Genetics. Apr
2005;67(4):335–340.
58. Zhuo YH, Wang M, Wei YT, et al. Analysis of MYOC gene muta-
tion in a Chinese glaucoma family with primary open-angle glau-
coma and primary congenital glaucoma. Chinese Medical Journal.
2006;119(14):1210–1214.
59. Chen Y, Jiang D, Yu L, et  al. CYP1B1 and MYOC mutations in
116 Chinese patients with primary congenital glaucoma. Archives of
Ophthalmology. Oct 2008;126(10):1443–1447.
60. Park BC, Tibudan M, Samaraweera M, Shen X, Yue BY. Interaction
between two glaucoma genes, optineurin and myocilin. Genes Cells.
Aug 2007;12(8):969–979.
61. Fan BJ, Wang DY, Fan DS, et al. SNPs and interaction analyses of
myocilin, optineurin, and apolipoprotein E in primary open angle
glaucoma patients. Molecular Vision. 2005;11:625–631.
62. Joe MK, Sohn S, Choi YR, Park H, Kee C. Identification of
flotillin-1 as a protein interacting with myocilin: implications for the
pathogenesis of primary open-angle glaucoma. Biochemical and bio-
physical research communications. Nov 4, 2005;336(4):1201–1206.
63. Joe MK, Kee C, Tomarev SI. Myocilin interacts with syntrophins and
is member of dystrophin-associated protein complex. The Journal of
Biological Chemistry. Apr 13, 2012;287(16):13216–13227.
64. Senatorov V, Malyukova I, Fariss R, et  al. Expression of mutated
mouse myocilin induces open-angle glaucoma in transgenic mice.
Journal of Neuroscience. Nov 15, 2006;26(46):11903–11914.
6 4 6   •   G enomic s in C linical P ractice
65. Rezaie T, Child A, Hitchings R, et  al. Adult-onset primary
open-angle glaucoma caused by mutations in optineurin. Science
(New York, N.Y.). Feb 8, 2002;295(5557):1077–1079.
66. Alward WL, Kwon YH, Kawase K, et al. Evaluation of optineu-
rin sequence variations in 1,048 patients with open-angle glau-
coma. American Journal of Ophthalmology. Nov 2003;136(5):
904–910.
67. Ariani F, Longo I, Frezzotti P, et al. Optineurin gene is not involved
in the common high-tension form of primary open-angle glau-
coma. Graefe’s Archive for Clinical and Experimental Ophthalmology
[Albrecht von Graefes Archiv fur klinische und experimentelle
Ophthalmologie]. Sep 2006;244(9):1077–1082.
68. Aung T, Ebenezer ND, Brice G, et  al. Prevalence of optineurin
sequence variants in adult primary open angle glaucoma:  impli-
cations for diagnostic testing. Journal of Medical Genetics. Aug
2003;40(8):e101.
69. Baird PN, Richardson AJ, Craig JE, et  al. Analysis of optineu-
rin (OPTN) gene mutations in subjects with and without glau-
coma:  the Blue Mountains Eye Study. Clinical and Experimental
Ophthalmology. Oct 2004;32(5):518–522.
70. Chen JH, Xu L, Li Y. [Study on the optic neuropathy induced
response protein gene mutation in Chinese patients with pri-
mary open-angle glaucoma]. Zhonghua Yi Xue Za Zhi. Jul 2,
2004;84(13):1098–1102.
71. Fuse N, Takahashi K, Akiyama H, et  al. Molecular genetic analy-
sis of optineurin gene for primary open-angle and normal tension
glaucoma in the Japanese population. Journal of Glaucoma. Aug
2004;13(4):299–303.
72. Jansson M, Wadelius C, Rezaie T, Sarfarazi M. Analysis of rare vari-
ants and common haplotypes in the optineurin gene in Swedish
glaucoma cases. Ophthalmic Genetics. Jun 2005;26(2):85–89.
73. Leung YF, Fan BJ, Lam DS, et al. Different optineurin mutation pat-
tern in primary open-angle glaucoma. Investigative Ophthalmology
and Visual Science. Sep 2003;44(9):3880–3884.
74. Melki R, Belmouden A, Akhayat O, Brezin A, Garchon HJ. The
M98K variant of the Optineurin (OPTN) gene modifies initial
intraocular pressure in patients with primary open angle glaucoma.
Journal of Medical Genetics. Nov 2003; 40(11):842–844.
75. Mukhopadhyay A, Komatireddy S, Acharya M, et  al. Evaluation
of Optineurin as a candidate gene in Indian patients with primary
open angle glaucoma. Molecular Vision. 2005;11:792–797.
76. Sripriya S, Nirmaladevi J, George R, et  al. OPTN gene:  pro-
file of patients with glaucoma from India. Molecular Vision.
2006;12:816–820.
77. Tang S, Toda Y, Kashiwagi K, et  al. The association between
Japanese primary open-angle glaucoma and normal tension glau-
coma patients and the optineurin gene. Human Genetics. Aug
2003;113(3):276–279.
78. Toda Y, Tang S, Kashiwagi K, et al. Mutations in the optineurin gene
in Japanese patients with primary open-angle glaucoma and normal
tension glaucoma. American Journal of Medical Genetics. Feb 15,
2004;125A(1):1–4.
79. Umeda T, Matsuo T, Tanabe Y, et  al. Optineurin gene polymor-
phisms in Japanese glaucoma patients and normal individuals.
Investigative Ophthalmology and Visual Science. 2003;44:E-Abstract
1111.
80. Umeda T, Matsuo T, Nagayama M, et  al. Clinical relevance of
optineurin sequence alterations in Japanese glaucoma patients.
Ophthalmic Genetics. Jun 2004;25(2):91–99.
81. Yasuda N, Nakamoto K, Funayama T, Mashima Y. [Low penetrance
of His26Asp mutation in the optineurin gene in a Japanese family
with normal-tension glaucoma]. Nippon Ganka Gakkai Zasshi. Aug
2006;110(8):594–600.
82. Aung T, Rezaie T, Okada K, et al. Clinical features and course of
patients with glaucoma with the E50K mutation in the optineu-
rin gene. Investigative Ophthalmology and Visual Science. Aug
2005;46(8):2816–2822.
 83. Forsman E, Lemmela S, Varilo T, et  al. The role of TIGR and
OPTN in Finnish glaucoma families:  a clinical and molecular
genetic study. Molecular Vision. May 30, 2003;9:217–222.
84. Rakhmanov VV, Nikitina N, Zakharova FM, et  al. [Mutations
and polymorphisms in the genes for myocilin and optineurin as
the risk factors of primary open-angle glaucoma]. Genetika. Nov
2005;41(11):1567–1574.
85. Weisschuh N, Neumann D, Wolf C, Wissinger B, Gramer E.
Prevalence of myocilin and optineurin sequence variants in
German normal tension glaucoma patients. Molecular Vision.
2005;11:284–287.
86. Willoughby CE, Chan LL, Herd S, et  al. Defining the pathoge-
nicity of optineurin in juvenile open-angle glaucoma. Investigative
Ophthalmology and Visual Science. Sep 2004;45(9):3122–3130.
87. Yao HY, Cheng CY, Fan BJ, et al. [Polymorphisms of myocilin and
optineurin in primary open angle glaucoma patients]. Zhonghua Yi
Xue Za Zhi. Feb 28, 2006; 86(8):554–559.
88. Funayama T, Ishikawa K, Ohtake Y, et  al. Variants in optineu-
rin gene and their association with tumor necrosis factor-alpha
polymorphisms in Japanese patients with glaucoma. Investigative
Ophthalmology and Visual Science. Dec 2004; 45(12):4359–4367.
89. Fingert JH, Robin AL, Stone JL, et  al. Copy number variations
on chromosome 12q14 in patients with normal tension glaucoma.
Human Molecular Genetics. Jun 15, 2011;20(12):2482–2494.
90. Chalasani ML, Radha V, Gupta V, et  al. A glaucoma-associated
mutant of optineurin selectively induces death of retinal gan-
glion cells which is inhibited by antioxidants. Investigative
Ophthalmology and Visual Science. Apr 2007;48(4):1607–1614.
91. Chi ZL, Akahori M, Obazawa M, et  al. Overexpression of opti-
neurin E50K disrupts Rab8 interaction and leads to a progressive
retinal degeneration in mice. Human Molecular Genetics. Jul 1,
2010;19(13):2606–2615.
92. De Marco N, Buono M, Troise F, Diez-Roux G. Optineurin
increases cell survival and translocates to the nucleus in a
Rab8-dependent manner upon an apoptotic stimulus. The Journal
of Biological Chemistry. Jun 9, 2006;281(23):16147–16156.
93. Meng Q, Lv J, Ge H, et al. Overexpressed mutant optineurin (E50K)
induces retinal ganglion cells apoptosis via the mitochondrial path-
way. Molecular Biology Reports. May 2012;39(5):5867–5873.
94. Meng Q, Xiao Z, Yuan H, et al. Transgenic mice with overexpres-
sion of mutated human optineurin (E50K) in the retina. Molecular
biology reports. Feb 2012; 39(2):1119–1124.
95. Park B, Ying H, Shen X, et  al. Impairment of protein traffick-
ing upon overexpression and mutation of optineurin. PloS One.
2010;5(7):e11547.
96. Shen X, Ying H, Qiu Y, et al. Processing of optineurin in neuro-
nal cells. The Journal of Biological Chemistry. Feb 4, 2011;286(5):
3618–3629.
97. Johnson JO, Mandrioli J, Benatar M, et  al. Exome sequencing
reveals VCP mutations as a cause of familial ALS. Neuron. Dec 9,
2010;68(5):857–864.
98. Maruyama H, Morino H, Ito H, et al. Mutations of optineurin in
amyotrophic lateral sclerosis. Nature. May 13, 2010;465(7295):
223–226.
99. Belzil VV, Daoud H, Desjarlais A, et  al. Analysis of OPTN as a
causative gene for amyotrophic lateral sclerosis. Neurobiology of
Aging. Mar 2011;32(3):555 e13–4.
100. Del Bo R, Tiloca C, Pensato V, et  al. Novel optineurin muta-
tions in patients with familial and sporadic amyotrophic lateral
sclerosis. Journal of Neurology, Neurosurgery, and Psychiatry. Nov
2011;82(11):1239–1243.
101. Osawa T, Mizuno Y, Fujita Y, et al. Optineurin in neurodegenera-
tive diseases. Neuropathology. Dec 2011;31(6):569–574.
102. Sugihara K, Maruyama H, Kamada M, Morino H, Kawakami H.
Screening for OPTN mutations in amyotrophic lateral sclerosis
in a mainly Caucasian population. Neurobiology of Aging. Oct
2011;32(10):1923 e9–10.
G enetic s and G enomic s in C linical O p h t h almology, I I :   G laucoma   •   6 4 7
103. Chio A, Calvo A, Mazzini L, et  al. Extensive genetics of
ALS:  a population-based study in Italy. Neurology. Nov 6,
2012;79(19):1983–1989.
104. Johnson L, Miller JW, Gkazi AS, et al. Screening for OPTN muta-
tions in a cohort of British amyotrophic lateral sclerosis patients.
Neurobiology of Aging. Dec 2012;33(12):2948 e15–7.
105. van Blitterswijk M, van Vught PW, van Es MA, et al. Novel opti-
neurin mutations in sporadic amyotrophic lateral sclerosis patients.
Neurobiology of Aging. May 2012;33(5):1016 e1–7.
106. Fan BJ, Wang DY, Cheng CY, et al. Different WDR36 mutation
pattern in Chinese patients with primary open-angle glaucoma.
Molecular Vision. 2009;15:646–653.
107. Hauser MA, Allingham RR, Linkroum K, et  al. Distribution
of WDR36 DNA sequence variants in patients with primary
open-angle glaucoma. Investigative Ophthalmology and Visual
Science. Jun 2006;47(6):2542–2546.
108. Hewitt AW, Dimasi DP, Mackey DA, Craig JE. A glaucoma
case-control study of the WDR36 gene D658G sequence variant.
American Journal of Ophthalmology. Aug 2006;142(2):324–325.
109. Miyazawa A, Fuse N, Mengkegale M, et al. Association between
primary open-angle glaucoma and WDR36 DNA sequence vari-
ants in Japanese. Molecular Vision. 2007;13:1912–1919.
110. Pasutto F, Mardin CY, Michels-Rautenstrauss K, et  al. Profiling
of WDR36 missense variants in German patients with glau-
coma. Investigative Ophthalmology and Visual Science. Jan
2008;49(1):270–274.
111. Raymond V, Dubois S, Marquis A, et  al. Large scale mutation
analysis of the third glaucoma causing gene, WDR36, at GLC1G,
in French-Canadian population of Quebec. American Society of
Human Genetics; 2005:E-Abstract 1950.
112. Weisschuh N, Wolf C, Wissinger B, Gramer E. Variations in the
WDR36 gene in German patients with normal tension glaucoma.
Molecular Vision. 2007;13:724–729.
113. Sarfarazi M, Monemi S, Choudhary D, Rezaie T, Schenkman JB
(2008). Roles of CYP1B1, optineurin and WDR36 gene muta-
tions in glaucoma. In: Tombran-Tink J, Barnstable CJ, Shields MB,
eds. Mechanisms of the Glaucomas: Disease Processes and Therapeutic
Modalities. Totowa, NJ: Humana Press; 233–273.
114. Fingert JH, Alward WL, Kwon YH, et al. No association between
variations in the WDR36 gene and primary open-angle glaucoma.
Archives of Ophthalmology. Mar 2007;125(3):434–436.
115. Kramer PL, Samples JR, Monemi S, Sykes R, Sarfarazi M, Wirtz
MK. The role of the WDR36 gene on chromosome 5q22.1
in a large family with primary open-angle glaucoma mapped
to this region. Archives of Ophthalmology. Sep 2006; 124(9):
1328–1331.
116. Pang CP, Fan BJ, Canlas O, et al. A genome-wide scan maps a novel
juvenile-onset primary open angle glaucoma locus to chromosome
5q. Molecular Vision. 2006;12:85–92.
117. Blanco-Marchite C, Sanchez-Sanchez F, Lopez-Garrido MP,
et  al. WDR36 and P53 gene variants and susceptibility to pri-
mary open-angle glaucoma:  analysis of gene–gene interac-
tions. Investigative Ophthalmology and Visual Science. Oct
2011;52(11):8467–8478.
118. Chi ZL, Yasumoto F, Sergeev Y, et  al. Mutant WDR36 directly
affects axon growth of retinal ganglion cells leading to progressive
retinal degeneration in mice. Human Molecular Genetics. Oct 1,
2010;19(19):3806–3815.
119. Gallenberger M, Meinel DM, Kroeber M, et al. Lack of WDR36
leads to preimplantation embryonic lethality in mice and delays
the formation of small subunit ribosomal RNA in human cells in
vitro. Human Molecular Genetics. 2011 Feb 1;20(3):422–435.
120. Footz T, Dubois S, Sarfarazi M, Raymond V, Walter MA.
Co-variation of STI1 and WDR36/UTP21 alters cell prolifera-
tion in a glaucoma model. Molecular Vision. 2011;17:1957–1969.
121. Crooks K, Liu Y, Lie W, Schmidt S, Allingham R, Hauser M.
POAG Is Not Associated with NTF4 Variants in a Southeastern
 US Caucasian Population. Investigative Ophthalmology and Visual
Science. 2010;51:E-Abstract 2180.
122. Liu Y, Liu W, Crooks K, Schmidt S, Allingham RR, Hauser MA.
No evidence of association of heterozygous NTF4 mutations in
patients with primary open-angle glaucoma. American Journal
of Human Genetics. Mar 12, 2010;86(3):498–499; author reply,
500.
123. Zhang M, Wang J, Chen H, et  al. Evaluation of NTF4 as a
Candidate Gene for Primary Open Angle Glaucoma. Investigative
Ophthalmology and Visual Science. 2010;51:E-Abstract 1641.
124. Vithana EN, Nongpiur ME, Venkataraman D, et al. Identification
of a novel mutation in the NTF4 gene that causes primary
open-angle glaucoma in a Chinese population. Molecular Vision.
2010;16:1640–1645.
125. Rao KN, Kaur I, Parikh RS, et al. Evaluation of NTF4, VAV2 and
VAV3 Genes in Primary Open Angle and Primary Angle Closure
Glaucomas in an Indian Population. Investigative Ophthalmology
and Visual Science. 2010;51:E-Abstract 6436.
126. Rao KN, Kaur I, Parikh RS, et  al. Variations in NTF4, VAV2
and VAV3 genes are not involved with primary open angle
and primary angle closure glaucomas in an Indian popula-
tion. Investigative Ophthalmology and Visual Science. Oct
2010;51(10):4937–4941.
127. Pasutto F, Keller KE, Weisschuh N, et al. Variants in ASB10 are
associated with open-angle glaucoma. Human Molecular Genetics.
Mar 15, 2011;21(6):1336–1349.
128. Fingert JH, Roos BR, Solivan-Timpe F, et al. Analysis of ASB10
variants in open angle glaucoma. Human Molecular Genetics. Jul
30, 2012;21(20):4543–4548.
129. Davis LK, Meyer KJ, Schindler EI, et al. Copy number variations
and primary open-angle glaucoma. Investigative Ophthalmology
and Visual Science. 2011;52(10):7122–7133.
130. Liu Y, Gibson J, Wheeler J, et al. GALC deletions increase the risk
of primary open-angle glaucoma: the role of Mendelian variants in
complex disease. PloS one. 2011;6(11):e27134.
131. Lee JH, Ki CS, Kim HJ, et  al. Analysis of copy number varia-
tion using whole genome exon-focused array CGH in Korean
patients with primary congenital glaucoma. Molecular Vision.
2011;17:3583–3590.
132. Foster PJ, Johnson GJ. Primary angle closure:  classification and
clinical features. In:  Hitchings RA, ed. Fundamentals of Clinical
Ophthalmology:  Glaucoma. London:  BMJ Publishing Group;
2000:145–152.
133. Resnikoff S, Pascolini D, Etya’ale D, et  al. Global data on visual
impairment in the year 2002. Bulletin of the World Health
Organization. Nov 2004;82(11):844–851.
134. Congdon NG, Quigley HA, Hung PT, Wang TH, Ho TC.
Screening techniques for angle-closure glaucoma in rural Taiwan.
Acta Ophthalmologika Scandinavika. Apr 1996;74(2):113–119.
135. Kim YY, Jung HR. Clarifying the nomenclature for primary
angle-closure glaucoma. Survey of Ophthalmology. Sep–Oct
1997;42(2):125–136.
136. Leighton DA, Phillips CI, Tsukahara S. Profile of presenting
states of eyes in angle-closure glaucoma. The British Journal of
Ophthalmology. Sep 1971; 55(9):577–584.
137. Lowe RF. Clinical types of primary angle closure glaucoma.
Australia and New Zealand Journal of Ophthalmology. Aug
1988;16(3):245–250.
138. Alsbirk PH. Anterior chamber depth and primary angle-closure
glaucoma. II. A genetic study. Acta Ophthalmologika (Copenhagen).
Jun 1975;53(3):436–449.
139. Leighton DA. Survey of the first-degree relatives of glaucoma
patients. Transactions of the Ophthalmological Societies of the United
Kingdom. Apr 1976;96(1):28–32.
140. Lowe RF. Primary angle-closure glaucoma. Family histories and
anterior chamber depths. The British Journal of Ophthalmology.
Apr 1964;48:191–195.
6 4 8   •   G enomic s in C linical P ractice
141. Lowe RF. Primary angle-closure glaucoma. Inheritance and
environment. The British Journal of Ophthalmology. Jan
1972;56(1):13–20.
142. Abu-Amero KK, Morales J, Mohamed GH, Osman MN, Bosley
TM. Glutathione S-transferase M1 and T1 polymorphisms in
Arab glaucoma patients. Molecular Vision. 2008;14:425–430.
143. Abu-Amero KK, Morales J, Osman MN, Bosley TM. Nuclear
and mitochondrial analysis of patients with primary angle-closure
glaucoma. Investigative Ophthalmology and Visual Science. Dec
2007;48(12):5591–5596.
144. Aung T, Bowman R, Chew PT, et al. Genome-Wide Linkage Scan
for Primary Angle Closure Glaucoma. Investigative Ophthalmology
and Visual Science. 2003;44:E-Abstract 3224.
145. Aung T, Yong VH, Chew PT, et al. Molecular analysis of the myo-
cilin gene in Chinese subjects with chronic primary-angle closure
glaucoma. Investigative Ophthalmology and Visual Science. Apr
2005;46(4):1303–1306.
146. Aung T, Yong VH, Lim MC, et al. Lack of association between the
rs2664538 polymorphism in the MMP-9 gene and primary angle
closure glaucoma in Singaporean subjects. Journal of Glaucoma.
Jun–Jul 2008;17(4):257–258.
147. Dandona L, Dandona R, Mandal P, et al. Angle-closure glaucoma
in an urban population in southern India. The Andhra Pradesh eye
disease study. Ophthalmology. Sep 2000;107(9):1710–1716.
148. Lowe RF. Primary angle-closure glaucoma:  changing concepts of
inheritance and environment. Transactions of the Australian College
of Ophthalmology. 1971;3:11–17.
149. Othman MI, Sullivan SA, Skuta GL, et al. Autosomal dominant
nanophthalmos (NNO1) with high hyperopia and angle-closure
glaucoma maps to chromosome 11. American Journal of Human
Genetics. Nov 1998;63(5):1411–1418.
150. Seah SK, Foster PJ, Chew PT, et  al. Incidence of acute primary
angle-closure glaucoma in Singapore. An island-wide survey.
Archives of Ophthalmology. Nov 1997;115(11):1436–1440.
151. Tomlinson A, Leighton DA. Ocular dimensions in the heredity of
angle-closure glaucoma. The British Journal of Ophthalmology. Jul
1973;57(7):475–486.
152. Wang IJ, Chiang TH, Shih YF, et  al. The association of single
nucleotide polymorphisms in the MMP-9 genes with susceptibil-
ity to acute primary angle closure glaucoma in Taiwanese patients.
Molecular Vision. 2006;12:1223–1232.
153. Yuan HP, Xiao Z, Yang BB. A study on the association of apolipo-
protein E genotypes with primary open-angle glaucoma and pri-
mary angle-closure glaucoma in northeast of China. Zhonghua Yan
Ke Za Zhi. 2007;43:416–420.
154. Vithana EN, Khor CC, Cornes BK, et  al. Association analysis
identifies a susceptibility locus on chromosome 3q27 for primary
angle closure glaucoma. Investigative Ophthalmology and Visual
Science. 2012;53:E-Abstract 5290.
155. Ferda Percin E, Ploder LA, Yu JJ, et  al. Human microphthalmia
associated with mutations in the retinal homeobox gene CHX10.
Nature Genetics. Aug 2000;25(4):397–401.
156. Bar-Yosef U, Abuelaish I, Harel T, Hendler N, Ofir R, Birk OS.
CHX10 mutations cause non-syndromic microphthalmia/
anophthalmia in Arab and Jewish kindreds. Human Genetics. Sep
2004;115(4):302–309.
157. Sundin OH, Leppert GS, Silva ED, et al. Extreme hyperopia is the
result of null mutations in MFRP, which encodes a Frizzled-related
protein. Proceedings of the National Academy of Sciences of the
United States of America. Jul 5, 2005; 102(27):9553–9558.
158. Gal A, Rau I, El Matri L, et  al. Autosomal-recessive posterior
microphthalmos is caused by mutations in PRSS56, a gene encod-
ing a trypsin-like serine protease. American Journal of Human
Genetics. Mar 11, 2011;88(3):382–390.
159. Aung T, Lim MC, Wong TT, et al. Molecular analysis of CHX10
and MFRP in Chinese subjects with primary angle closure glaucoma
and short axial length eyes. Molecular Vision. 2008;14:1313–1318.
 160. Awadalla MS, Burdon KP, Thapa SS, Hewitt AW, Craig JE.
A cross-ethnicity investigation of genes previously impli-
cated in primary angle closure glaucoma. Molecular Vision.
2008;18:2247–2254.
161. Nair KS, Hmani-Aifa M, Ali Z, et  al. Alteration of the serine
protease PRSS56 causes angle-closure glaucoma in mice and pos-
terior microphthalmia in humans and mice. Nature Genetics. Jun
2011;43(6):579–584.
162. Wang IJ, Lin S, Chiang TH, et  al. The association of membrane
frizzled-related protein (MFRP) gene with acute angle-closure
glaucoma-a pilot study. Molecular Vision. 2008;14:1673–1679.
163. Dai X, Nie S, Ke T, Liu J, Wang Q, Liu M. [Two variants in
MYOC and CYP1B1 genes in a Chinese family with primary
angle-closure glaucoma]. Zhonghua Yi Xue Yi Chuan Xue Za Zhi.
Oct 2008;25(5):493–496.
164. Vithana EN, Khor CC, Qiao C, et al. Genome-wide association
analyses identify three new susceptibility loci for primary angle clo-
sure glaucoma. Nature Genetics. Oct 2012;44(10):1142–1146.
165. Rubin SE, Marcus CH. Glaucoma in childhood. Pediatric
Ophthalmology, Ophthalmology Clinics of North America.
1996;9:215–216.
166. Jaffar MS. Care of the Infantile Glaucoma Patient. In:  Reinecke
RD, ed. Ophthalmology Annual. Vol 7.  New  York:  Raven Press,
1988:15–37.
167. Lichter PR. Genetic clues to glaucoma’s secrets. The L Edward
Jackson Memorial Lecture. Part  2. American Journal of
Ophthalmology. Jun 15, 1994;117(6):706–727.
168. Lichter PR. Visual field validity and the search for glaucoma’s Holy
Grail. Ophthalmology. Sep 1994;101(9):1479–1480.
169. Sarfarazi M, Akarsu AN, Hossain A, et al. Assignment of a locus
(GLC3A) for primary congenital glaucoma (Buphthalmos) to
2p21 and evidence for genetic heterogeneity. Genomics. Nov 20,
1995;30(2):171–177.
170. Akarsu AN, Turacli ME, Aktan SG, et al. A second locus (GLC3B)
for primary congenital glaucoma (Buphthalmos) maps to the 1p36
region. Human Molecular Genetics. Aug 1996;5(8):1199–1203.
171. Sarfarazi M, Stoilov I. The third genetic locus (GLC3C) for pri-
mary congenital glaucoma (PCG) maps to chromosome 14q24.3.
American Journal of Human Genetics (Supplement). 2002;71(4).
172. Verdin H, Leroy BP, D’haene B, et al. Identity-by-descent mapping
reveals a new locus for primary congenital glaucoma, GLC3E, on
chromosome 19p13.2. Investigative Ophthalmology and Visual
Science. 2012;53:E-Abstract 5289.
173. Aklillu E, Oscarson M, Hidestrand M, Leidvik B, Otter C,
Ingelman-Sundberg M. Functional analysis of six different poly-
morphic CYP1B1 enzyme variants found in an Ethiopian popula-
tion. Molecular Pharmacology. Mar 2002;61(3):586–594.
174. Bejjani BA, Lewis RA, Tomey KF, et  al. Mutations in CYP1B1,
the gene for cytochrome P4501B1 are the predominant cause of
primary congenital glaucoma in Saudi Arabia. American Journal of
Human Genetics. Feb 1998;62(2):325–333.
175. Belmouden A, Melki R, Hamdani M, et  al. A novel frameshift
founder mutation in the cytochrome P450 1B1 (CYP1B1) gene is
associated with primary congenital glaucoma in Morocco. Clinical
Genetics. Oct 2002;62(4):334–339.
176. Chitsazian F, Tusi BK, Elahi E, et al. CYP1B1 mutation profile of
Iranian primary congenital glaucoma patients and associated hap-
lotypes. Journal of Molecular Diagnosis. Jul 2007;9(3):382–393.
177. Colomb E, Kaplan J, Garchon HJ. Novel cytochrome P450 1B1
(CYP1B1) mutations in patients with primary congenital glau-
coma in France. Human Mutation. Dec 2003;22(6):496.
178. El-Ashry MF, Abd El-Aziz MM, Bhattacharya SS. A clinical and
molecular genetic study of Egyptian and Saudi Arabian patients
with primary congenital glaucoma (PCG). Journal of Glaucoma.
Jan 2007;16(1):104–111.
179. Jansson I, Stoilov I, Sarfarazi M, Schenkman JB. Effect of two
mutations of human CYP1B1, G61E and R469W, on stability and
G enetic s and G enomic s in C linical O p h t h almology, I I :   G laucoma   •   6 4 9
endogenous steroid substrate metabolism. Pharmacogenetics. Dec
2001;11(9):793–801.
180. Kakiuchi-Matsumoto T, Isashiki Y, Ohba N, Kimura K, Sonoda
S, Unoki K. Cytochrome P450 1B1 gene mutations in Japanese
patients with primary congenital glaucoma. American Journal of
Ophthalmology. Mar 2001; 131(3):345–350.
181. Kubota R, Mashima Y, Ohtake Y, et  al. Novel mutations in the
myocilin gene in Japanese glaucoma patients. Human Mutation.
Sep 2000;16(3):270.
182. Martin SN, Sutherland J, Levin AV, Klose R, Priston M, Heon
E. Molecular characterisation of congenital glaucoma in a con-
sanguineous Canadian community:  a step towards prevent-
ing glaucoma related blindness. Journal of Medical Genetics. Jun
2000;37(6):422–427.
183. Messina-Baas OM, Gonzalez-Huerta LM, Chima-Galan C, et al.
Molecular analysis of the CYP1B1 gene:  Identification of novel
truncating mutations in patients with primary congenital glau-
coma. Ophthalmic Research. 2007; 39(1):17–23.
184. Michels-Rautenstrauss KG, Mardin CY, Zenker M, et al. Primary
congenital glaucoma:  three case reports on novel mutations and
combinations of mutations in the GLC3A (CYP1B1) gene.
Journal of Glaucoma. Aug 2001;10(4):354–357.
185. Ohtake Y, Kubota R, Tanino T, Miyata H, Mashima Y. Novel com-
pound heterozygous mutations in the cytochrome P4501B1 gene
(CYP1B1) in a Japanese patient with primary congenital glau-
coma. Ophthalmic Genetics. Sep 2000;21(3):191–193.
186. Ohtake Y, Tanino T, Suzuki Y, et  al. Phenotype of cytochrome
P4501B1 gene (CYP1B1) mutations in Japanese patients with pri-
mary congenital glaucoma. The British Journal of Ophthalmology.
Mar 2003;87(3):302–304.
187. Panicker SG, Reddy AB, Mandal AK, et al. Identification of novel
mutations causing familial primary congenital glaucoma in Indian
pedigrees. Investigative Ophthalmology and Visual Science. May
2002;43(5):1358–1366.
188. Plasilova M, Stoilov I, Sarfarazi M, et al. Identification of a single
ancestral CYP1B1 mutation in Slovak Gypsies (Roms) affected
with primary congenital glaucoma. Journal of Medical Genetics.
Apr 1999;36(4):290–294.
189. Reddy AB, Panicker SG, Mandal AK, et  al. Identification of
R368H as a predominant CYP1B1 allele causing primary congeni-
tal glaucoma in Indian patients. Investigative Ophthalmology and
Visual Science. Oct 2003; 44(10):4200–4203.
190. Sena DF, Finzi S, Rodgers K, Del Bono E, Haines JL, Wiggs JL.
Founder mutations of CYP1B1 gene in patients with congenital
glaucoma from the United States and Brazil. Journal of Medical
Genetics. Jan 2004;41(1):e6.
191. Sivadorai P, Cherninkova S, Bouwer S, et  al. Genetic heteroge-
neity and minor CYP1B1 involvement in the molecular basis of
primary congenital glaucoma in Gypsies. Clinical Genetics. Jul
2008;74(1):82–87.
192. Tang YM, Green BL, Chen GF, et al. Human CYP1B1 Leu432Val
gene polymorphism:  ethnic distribution in African-Americans,
Caucasians and Chinese; oestradiol hydroxylase activity; and dis-
tribution in prostate cancer cases and controls. Pharmacogenetics.
Dec 2000;10(9):761–766.
193. Weisschuh N, Wolf C, Wissinger B, Gramer E. A clinical and
molecular genetic study of German patients with primary con-
genital glaucoma. American Journal of Ophthalmology. Apr
2009;147(4):744–753.
194. Bhattacharjee A, Banerjee D, Mookherjee S, et  al. Leu432Val
polymorphism in CYP1B1 as a susceptible factor towards pre-
disposition to primary open-angle glaucoma. Molecular Vision.
2008;14:841–850.
195. Chakrabarti S, Devi KR, Komatireddy S, et al. Glaucoma-associated
CYP1B1 mutations share similar haplotype backgrounds in
POAG and PACG phenotypes. Investigative Ophthalmology and
Visual Science. Dec 2007;48(12):5439–5444.
 196. Chakrabarti S, Kaur K, Komatireddy S, et  al. Gln48His is the
prevalent myocilin mutation in primary open angle and primary
congenital glaucoma phenotypes in India. Molecular Vision. Feb 4,
2005;11:111–113.
197. Kass MA, Heuer DK, Higginbotham EJ, et  al. The Ocular
Hypertension Treatment Study:  a randomized trial determines
that topical ocular hypotensive medication delays or prevents the
onset of primary open-angle glaucoma. Archives of Ophthalmology.
Jun 2002;120(6):701–713; discussion, 829–730.
198. Kumar A, Basavaraj MG, Gupta SK, et  al. Role of CYP1B1,
MYOC, OPTN, and OPTC genes in adult-onset primary
open-angle glaucoma:  predominance of CYP1B1 mutations in
Indian patients. Molecular Vision. 2007;13:667–676.
199. Lopez-Garrido MP, Sanchez-Sanchez F, Lopez-Martinez
F, et  al. Heterozygous CYP1B1 gene mutations in Spanish
patients with primary open-angle glaucoma. Molecular Vision.
2006;12:748–755.
200. Reddy AB, Kaur K, Mandal AK, et al. Mutation spectrum of the
CYP1B1 gene in Indian primary congenital glaucoma patients.
Molecular Vision. Sep 30, 2004;10:696–702.
201. Suri F, Kalhor R, Zargar SJ, et al. Screening of common CYP1B1
mutations in Iranian POAG patients using a microarray-based
PrASE protocol. Molecular Vision. 2008;14:2349–2356.
202. Bagiyeva S, Marfany G, Gonzalez-Angulo O, Gonzalez-Duarte R.
Mutational screening of CYP1B1 in Turkish PCG families and
functional analyses of newly detected mutations. Molecular Vision.
2007;13:1458–1468.
203. Chakrabarti S, Kaur K, Rao KN, et  al. The transcription fac-
tor gene FOXC1 exhibits a limited role in primary congenital
glaucoma. Investigative Ophthalmology and Visual Science. Jan
2009;50(1):75–83.
204. Melki R, Lefort N, Brezin AP, Garchon HJ. Association of a com-
mon coding polymorphism (N453S) of the cytochrome P450 1B1
(CYP1B1) gene with optic disc cupping and visual field alteration
in French patients with primary open-angle glaucoma. Molecular
Vision. 2005;11:1012–1017.
205. Soley GC, Bosse KA, Flikier D, et  al. Primary congenital glau-
coma: a novel single-nucleotide deletion and varying phenotypic
expression for the 1,546–1,555dup mutation in the GLC3A
(CYP1B1) gene in 2 families of different ethnic origin. Journal of
Glaucoma. Feb 2003;12(1):27–30.
206. Ali M, McKibbin M, Booth A, et  al. Null mutations in LTBP2
cause primary congenital glaucoma. American Journal of Human
Genetics. May 2009; 84(5):664–671.
207. Narooie-Nejad M, Paylakhi SH, Shojaee S, et al. Loss of function
mutations in the gene encoding latent transforming growth factor
beta binding protein 2, LTBP2, cause primary congenital glaucoma.
Human Molecular Genetics. Oct 15, 2009;18(20):3969–3977.
208. Sharafieh R, Child AH, Khaw PT, Fleck B and Sarfarazi M.
LTBP2 gene analysis in the GLC3C-linked family and 94
CYP1B1-negative cases with primary congenital glaucoma.
Ophthalmic Genetics. 2013 Mar-Jun;34(1–2):14–20.
209. Lim SH, Tran-Viet KN, Yanovitch TL, et al. CYP1B1, MYOC,
and LTBP2 mutations in primary congenital glaucoma patients
in the United States. American Journal of Ophthalmology. Mar
2013;155(3):508–517 e5.
210. Abu-Amero KK, Osman EA, Mousa A, et  al. Screening of
CYP1B1 and LTBP2 genes in Saudi families with primary congen-
ital glaucoma: genotype-phenotype correlation. Molecular Vision.
2011;17:2911–2919.
211. Mohanty K, Tanwar M, Dada R, Dada T. Screening of the LTBP2
gene in a north Indian population with primary congenital glau-
coma. Molecular Vision. 2013; 19:78–84.
212. Nenciu A. [Pseudo-exfoliative syndrome-etiology, clinical aspects,
diagnosis]. Oftalmologia. 2007;51(4):34–40.
213. Stefan C, Tebeanu E, Nenciu A. [Pseudoexfoliative glaucoma].
Oftalmologia. 2007;51(3):50–52.
6 5 0   •   G enomic s in C linical P ractice
214. Ritch R. Exfoliation syndrome and occludable angles. Transactions
of the American Ophthalmological Society. 1994;92:845–944.
215. Sowka J. Pseudoexfoliation syndrome and pseudoexfoliative glau-
coma. Optometry. Apr 2004;75(4):245–250.
216. Damji KF, Bains HS, Stefansson E, et al. Is pseudoexfoliation syn-
drome inherited? A review of genetic and nongenetic factors and a
new observation. Ophthalmic Genetics. Dec 1998;19(4):175–185.
217. Gifford H, Jr. A clinical and pathological study of exfoliation
of the lens capsule. American Journal of Ophthalmology. Oct
1958;46(4):508–524.
218. Jerndal T, Svedbergh B. Goniodysgenesis in exfoliation glaucoma.
Advances in Ophthalmology. 1978;35:45–64.
219. Sarfarazi M. Discussion regarding PEX. 2004.
220. Kozobolis VP, Detorakis ET, Sourvinos G, Pallikaris IG,
Spandidos DA. Loss of heterozygosity in pseudoexfoliation syn-
drome. Investigative Ophthalmology and Visual Science. May
1999;40(6):1255–1260.
221. Lemmela S, Forsman E, Sistonen P, Eriksson A, Forsius H, Jarvela
I. Genome-wide scan of exfoliation syndrome. Investigative
Ophthalmology and Visual Science. Sep 2007;48(9):4136–4142.
222. Sotirova V, Irkec M, Caudill M, Orhan M, Akarsu AN, Sarfarazi
M. Ascertainment and molecular genetic study of families with
pseudoexfoliation syndrome (PEX). Investigative Ophthalmology
and Visual Science. 1998;39:E-Abstract 139.
223. Wiggs JL, Anderson JS, Stefansson E. A genomic screen suggests
a locus on chromosome 2p16 for pseudoexfoliation syndrome.
American Society of Human Genetics; 1998:E-Abstract 1818.
224. Thorleifsson G, Magnusson KP, Sulem P, et  al. Common
sequence variants in the LOXL1 gene confer susceptibil-
ity to exfoliation glaucoma. Science (New  York, N.Y.). Sep 7,
2007;317(5843):1397–1400.
225. Aragon-Martin JA, Ritch R, Liebmann J, et  al. Evaluation of
LOXL1 gene polymorphisms in exfoliation syndrome and exfolia-
tion glaucoma. Molecular Vision. 2008;14:533–541.
226. Yang X, Zabriskie NA, Hau VS, et  al. Genetic association of
LOXL1 gene variants and exfoliation glaucoma in a Utah cohort.
Cell Cycle. Feb 15, 2008;7(4):521–524.
227. Challa P, Schmidt S, Liu Y, et  al. Analysis of LOXL1 polymor-
phisms in a United States population with pseudoexfoliation glau-
coma. Molecular Vision. 2008;14:146–149.
228. Fan BJ, Pasquale L, Grosskreutz CL, et al. DNA sequence variants
in the LOXL1 gene are associated with pseudoexfoliation glau-
coma in a U.S. clinic-based population with broad ethnic diversity.
BMC Medical Genetics. 2008;9:5.
229. Pasutto F, Krumbiegel M, Mardin CY, et  al. Association of
LOXL1 common sequence variants in German and Italian
patients with pseudoexfoliation syndrome and pseudoexfoliation
glaucoma. Investigative Ophthalmology and Visual Science. Apr
2008;49(4):1459–1463.
230. Fuse N, Miyazawa A, Nakazawa T, Mengkegale M, Otomo
T, Nishida K. Evaluation of LOXL1 polymorphisms in eyes
with exfoliation glaucoma in Japanese. Molecular Vision.
2008;14:1338–1343.
231. Ozaki M, Lee KY, Vithana EN, et  al. Association of
LOXL1 gene polymorphisms with pseudoexfoliation in the
Japanese. Investigative Ophthalmology and Visual Science. Sep
2008;49(9):3976–3980.
232. Tanito M, Minami M, Akahori M, et al. LOXL1 variants in elderly
Japanese patients with exfoliation syndrome/glaucoma, primary
open-angle glaucoma, normal tension glaucoma, and cataract.
Molecular Vision. 2008;14:1898–1905.
233. Fingert JH, Alward WL, Kwon YH, et  al. LOXL1 mutations
are associated with exfoliation syndrome in patients from the
Midwestern United States. American Journal of Ophthalmology.
Dec 2007;144(6):974–975.
234. Hewitt AW, Sharma S, Burdon KP, et al. Ancestral LOXL1 variants
are associated with pseudoexfoliation in Caucasian Australians but
 with markedly lower penetrance than in Nordic people. Human
Molecular Genetics. Mar 1, 2008;17(5):710–716.
235. Traboulsi EI, Sarfarazi M. The use of microarray technology in
deciphering the cause of genetic eye diseases:  LOXL1 and exfo-
liation syndrome. American Journal of Ophthalmology. Mar
2008;145(3):391–393.
236. Hayashi H, Gotoh N, Ueda Y, Nakanishi H, Yoshimura N. Lysyl
oxidase-like 1 polymorphisms and exfoliation syndrome in the
Japanese population. American Journal of Ophthalmology. Mar
2008;145(3):582–585.
237. Chen L, Jia L, Wang N, et al. Evaluation of LOXL1 polymorphisms
in exfoliation syndrome in a Chinese population. Molecular Vision.
2009;15:2349–2357.
238. Lee KY, Ho SL, Thalamuthu A, et  al. Association of LOXL1
polymorphisms with pseudoexfoliation in the Chinese. Molecular
Vision. 2009;15:1120–1126.
239. Lemmela S, Forsman E, Onkamo P, et al. Association of LOXL1
gene with Finnish exfoliation syndrome patients. Journal of
Human Genetics. May 2009;54(5):289–297.
240. Mori K, Imai K, Matsuda A, et al. LOXL1 genetic polymorphisms
are associated with exfoliation glaucoma in the Japanese popula-
tion. Molecular Vision. 2008;14:1037–1040.
241. Mossbock G, Renner W, Faschinger C, Schmut O, Wedrich A,
Weger M. Lysyl oxidase-like protein 1 (LOXL1) gene polymor-
phisms and exfoliation glaucoma in a Central European popula-
tion. Molecular Vision. 2008;14:857–861.
242. Ramprasad VL, George R, Soumittra N, et  al. Association of
non-synonymous single nucleotide polymorphisms in the LOXL1
gene with pseudoexfoliation syndrome in India. Molecular Vision.
2008;14:318–322.
243. Wolf C, Gramer E, Muller-Myhsok B, et  al. Lysyl oxidase-like 1
gene polymorphisms in German patients with normal tension
glaucoma, pigmentary glaucoma and exfoliation glaucoma. Journal
of Glaucoma. Feb 2010;19(2):136–141.
244. Bovell AM, Damji KF, Dohadwala AA, et  al. Familial occur-
rence of pigment dispersion syndrome. Canadian Journal of
Ophthalmology. Feb 2001;36(1):11–17.
245. Mandelkorn RM, Hoffman ME, Olander KW, et  al. Inheritance
and the pigmentary dispersion syndrome. Annals of Ophthalmology.
Jun 1983;15(6):577–582.
246. Olander KW, Mandelkorn RM, Hoffman ME, Zimmerman TJ.
The pigment dispersion syndrome and open-angle glaucoma.
Annals of Ophthalmology. Sep 1982;14(9):809–810.
247. Ritch R, Steinberger D, Liebmann JM. Prevalence of pigment dis-
persion syndrome in a population undergoing glaucoma screening.
American Journal of Ophthalmology. Jun 15, 1993;115(6):707–710.
248. Qing G, Wang N. Clinical signs and characteristics of pigmentary
glaucoma in Chinese. Japanese Journal of Ophthalmology. May-Jun
2008;52(3):162–166.
249. Qing G, Wang N, Tang X, et al. Clinical characteristics of pigment
dispersion syndrome in Chinese patients. Eye (London, England).
Aug 2009; 23(8):1641–1646.
250. Andersen JS, Pralea AM, DelBono EA, et al. A gene responsible for
the pigment dispersion syndrome maps to chromosome 7q35-q36.
Archives of Ophthalmology. Mar 1997;115(3):384–388.
251. Rao KN, Ritch R, Dorairaj SK, et  al. Exfoliation syndrome
and exfoliation glaucoma-associated LOXL1 variations are not
involved in pigment dispersion syndrome and pigmentary glau-
coma. Molecular Vision. 2008;14:1254–1262.
252. Chang B, Smith RS, Hawes NL, et al. Interacting loci cause severe
iris atrophy and glaucoma in DBA/2J mice. Nature Genetics. Apr
1999;21(4):405–409.
G enetic s and G enomic s in C linical O p h t h almology, I I :   G laucoma   •   6 5 1
253. Mo JS, Anderson MG, Gregory M, et al. By altering ocular immune
privilege, bone marrow-derived cells pathogenically contribute
to DBA/2J pigmentary glaucoma. The Journal of Experimental
Medicine. May 19, 2003;197(10):1335–1344.
254. Amendt BA, Semina EV, Alward WL. Rieger syndrome: a clinical,
molecular, and biochemical analysis. Cellular and Molecular Life
Sciences. Oct 2000;57(11):1652–1666.
255. Alward WL. Axenfeld-Rieger syndrome in the age of molecular
genetics. American Journal of Ophthalmology. Jul 2000;130(1):
107–115.
256. O’Dwyer EM, Jones DC. Dental anomalies in Axenfeld-Rieger
syndrome. International Journal of Paediatric Dentistry. Nov
2005;15(6):459–463.
257. Martinez-Glez V, Lorda-Sanchez I, Ramirez JM, et al. Clinical pre-
sentation of a variant of Axenfeld-Rieger syndrome associated with
subtelomeric 6p deletion. European Journal of Medical Genetics.
Mar–Apr 2007;50(2):120–127.
258. Hoskins HD, Shaffer RN. Rieger’s syndrome: a form of iridocor-
neal mesodermal dysgenesis. Journal of Pediatric Ophthalmology.
1972;9:26–30.
259. Firth HV, Hurst JA, Hall JG. Part  2:  Anterior segment eye
malformations. In:  Firth JA and Hall JG, eds. Oxford Desk
Reference:  Clinical Genetics. Oxford:  Oxford University Press;
2005.
260. Fitch N, Kaback M. The Axenfeld syndrome and the Rieger syn-
drome. Journal of Medical Genetics. Feb 1978;15(1):30–34.
261. Harley RD, Nelson LB, Olitsky SE. Harley’s Pediatric
Ophthalmology (5th Ed). Philadelphia, PA:  Lippincott Williams
& Wilkins; 2005.
262. Shields MB, Buckley E, Klintworth GK, Thresher R.
Axenfeld-Rieger syndrome. A  spectrum of developmental disor-
ders. Survey of Ophthalmology. May–Jun 1985;29(6):387–409.
263. Hjalt TA, Semina EV. Current molecular understanding of
Axenfeld-Rieger syndrome. Expert Reviews in Molecular Medicine.
Nov 8, 2005;7(25):1–17.
264. Murray JC, Bennett SR, Kwitek AE, et al. Linkage of Rieger syn-
drome to the region of the epidermal growth factor gene on chro-
mosome 4. Nature Genetics. Sep 1992;2(1):46–49.
265. Phillips JC, del Bono EA, Haines JL, et  al. A second locus for
Rieger syndrome maps to chromosome 13q14. American Journal
of Human Genetics. Sep 1996;59(3):613–619.
266. Kulak SC, Kozlowski K, Semina EV, Pearce WG, Walter MA.
Mutation in the RIEG1 gene in patients with iridogoniodys-
genesis syndrome. Human Molecular Genetics. 1998;7(7):
1113–1117.
267. Mirzayans F, Gould DB, Heon E, et al. Axenfeld-Rieger syndrome
resulting from mutation of the FKHL7 gene on chromosome
6p25. European Journal of Human Genetics. 2000;8(1):71–74.
268. Mooney I, LaMotte J. A review of the potential to restore vision
with stem cells. Clinical and Experimental Optometry. Jan
2008;91(1):78–84.
269. Huang Y, Enzmann V, Ildstad ST. Stem cell-based therapeutic
applications in retinal degenerative diseases. Stem Cell Reviews. Jun
2011;7(2):434–445.
270. Jin ZB, Okamoto S, Osakada F, et al. Modeling retinal degenera-
tion using patient-specific induced pluripotent stem cells. PloS
One. 2011;10;6(2):e17084.
271. Rowland TJ, Buchholz DE, Clegg DO. Pluripotent human
stem cells for the treatment of retinal disease. Journal of Cellular
Physiology. Feb 2012;227(2):457–466.
 42.
GENETICS AND GENOMICS IN CLINICAL
OPHTHALMOLOGY, III: AGE-RELATED
MACULAR DEGENERATION
Mark E. Kleinman and Jayakrishna Ambati
INTRODUCTION
Age-related macular degeneration (AMD) is the most com-
mon cause of irreversible blindness in the developed world.
Currently, the epidemic is estimated to affect between 30
and 50 million people across the globe and is imposing mas-
sive health-care expenses while significantly reducing qual-
ity of life. The growth estimates are startling especially in
the United States, where an aging Baby Boomer generation
is now in its seventh decade of life; a situation that will most
certainly be complicated by limited financial resources.
Approximately 1.75 million Americans are affected with an
advanced form of the disease, a prevalence that is expected
nearly double to 3 million by 2020, with another 7 million
at risk.1 Western Europe is estimated to have over 3 million
with advanced disease.
AMD is characterized as a progressive retinal degen-
eration with hallmark features such as the accumulation
of sub- and intra-retinal lipoproteinaceous deposits called
drusen, abnormal pigment clumping, subretinal fluid or
hemorrhage, and atrophy of the retinal pigment epithelium
(RPE). Severe vision loss from AMD results from choroi-
dal neovascularization (CNV), the invasion of the retina
by abnormal choroidal blood vessels; or from geographic
atrophy (GA), the apoptotic loss of RPE, photoreceptors,
and choriocapillaris (Figure 42.1). The pathology most
often occurs in the central visual field, leading to distortion
(metamorphopsias), loss of contrast, or black spots (scoto-
mas) experienced by the patient. The disease’s unfortunate
predilection for the macula, an area that is required for
high-quality binocular visual acuity, has intrigued scien-
tists and clinicians since the disease was first identified in
the late nineteenth century. Theories related to the expo-
sure of focused light rays, high metabolic demand, oxida-
tive stress overload, and spatial heterogeneity of cellular
652
architecture have all been advanced, but in the last decade,
the field of genetics has offered a truly unique perspective
on AMD. Our fundamental knowledge of its pathogenesis
has matured considerably in the past decade, and the model
of AMD now serves as a prime example of how to improve
our translational approaches to human genetics to develop
effective therapies. Dozens of genes have now been asso-
ciated with this condition, with many relevant pathways
being targeted in clinical trials. Therefore, we will focus this
chapter on the polygenic nature of AMD and the charted
development of potential therapeutic approaches that have
been directly related to the robust genetic associations dis-
covered thus far.
Dissecting the multifactorial contributions of AMD
pathogenesis has proven to be a daunting challenge. Several
large epidemiological studies, including the Beaver Dam
Eye Study, Blue Mountain Eye Study, and the Age-Related
Eye Disease Study (AREDS), have acquired extensive data-
sets that identified significant risk factors such as age and
smoking, as well as phenotypical features associated with
disease progression, including drusen size and retinal pig-
ment abnormalities.2–4 Yet the true significance of genetic
susceptibility to this disease was not realized until twin
concordance, small family linkage, and segregation analysis
studies suggested that upwards of 70% of AMD may be due
to genetic predisposition.5–9 Shortly after the millennium,
the application of higher resolution genome-wide associa-
tion studies to the field resulted in a rush of reports with
multiple single-nucleotide polymorphisms (SNPs) iden-
tified as disease cohorts. As more studies continued to be
published, specific pathways began to emerge that were
subsequently studied in animal models and then targeted
for translation. The strongest genetic signature identified
to date involves SNPs located in immune-related genes
belonging to the complement activation system.
 (A) (B) (E)

RPE

(F)

200 µm

(C) (D)

(G)

PED
* *

Multimodal imaging of advanced dry or wet (neovascular) age-related macular degeneration. A. Color fundus photograph of patient
Figure 42.1 
with advanced dry AMD with large drusen (black arrowheads) and areas of geographic atrophy of the retinal pigment epithelium (RPE, black
arrow). B. Fundus autofluorescent imaging (488nm blue light) clearly revealing focal area of RPE loss (white arrow) and adjacent abnormal
hyperfluorescence indicative of RPE dysfunction in advanced dry AMD. C. Color fundus photograph of patient with acute decrease in vision shows
neovascular AMD with a large CNV membrane, sub-retinal heme, and fluid. D. Fluorescein angiogram of same eye with neovascular AMD at late
time-point (10 minutes) exhibited significant leakage due to CNV formation. E. Spectral domain optical coherence tomography (SD-OCT) image
of a normal retina displaying clear delineation of RPE layer and neural retina. F. SD-OCT of the eye with advanced dry AMD shown in panel A,
revealing severe loss of the RPE and outer retinal layers where photoreceptors are normally located (white arrowheads). G. SD-OCT of the eye
with neovascular AMD shown in panel C, confirming the presence of sub-retinal fluid (asterisks) and significant disruption of the RPE, known as
pigment epithelial detachment (PED).

G E N ET I C S I G N AT U R ES O F AMD but at an earlier age.12 CFH is expressed in the RPE

I MMU N E PAT H WAY AC T I VAT I O N
I N AM D
Over a decade ago, scientists performing compositional
analyses of drusen deposits discovered that they contain
terminal complement complexes and immunoglobulins;
prompting targeted studies of regulators of complement
activation in AMD-affected individuals. Immune activa-
tion is regulated by two intertwined pathways: the innate
and the adaptive (or acquired) systems. In traditional
models, complement activation regulates the opsonization
of pathogens via the classical, alternative, and lectin path-
ways, which all channel into a common terminus. Multiple
immunovascular diseases such as glomerulonephritis and
rheumatoid arthritis10 are associated with disorders of the
alternative complement pathway, and several are specifi-
cally due to deficiencies with the endogenous complement
inhibitor, complement factor H (CFH).11 Interestingly,
one of these disorders—membranoproliferative glomerulo-
nephritis (MPGN)—is also associated with drusen forma-
tion that appears in a spatial distribution nearly identical to
and choroid and inactivates C3b through factor I binding,
thus preventing complement-mediated host cell death in
tissues.13,14
Following these observations, a group of independent
studies published results on several SNPs within the human
CFH gene (1q31) that significantly increase AMD risk.13,15–17
One SNP alone, the Y402H variant (rs1061170), which
is located in the short consensus repeat domain 7 (SCR7),
increased AMD risk by approximately 2.7-fold in patients
of European descent. Y402H has been linked to sev-
eral downstream molecular events, including:  lowered
heparin-binding affinity to the Y402H CFH variant, caus-
ing reduced C3b inhibition and elevated alternative com-
plement pathway activation;18 decreased neutralization of
the pro-inflammatory lipid peroxidation product malondi-
aldehyde (MDA), leading to IL-8 and TNF-α induction;19
and disrupted interactions with C-reactive protein (CRP),20
resulting in increased concentrations of free unbound CRP
in the choroid.21 Recent evidence now indicates that SCR7
also biochemically interacts with Fibulin-3 (EFEMP-1,
2p16), a known component in drusen and a locus for an

G e n etic s a nd G e no m ic s in C l inic a l O p h t h a l m o l ogy, I I I   •   6 5 3

inherited form of familial drusen.22 The functional asso-
ciation of CFH with AMD is further complicated by its
alternative splicing to factor H-like 1 (FHL-1), which also
encodes SCR723 as well as flanking related genes, CFHR1
and CFHR3, which are linked to AMD pathogenesis.24
Interestingly, mouse models of Cfh deficiency did not
reveal a robust phenotype and showed limited structural
abnormalities, which also occur in the human condition of
AMD.25
As of 2013, the CFH data had been validated in numer-
ous cohorts, building a significant case for the potential
of therapeutic blockade of complement activation in the
treatment of AMD. In efforts to fully map alternative
complement-factor-pathway–related genetics and func-
tional alterations in AMD, additional work has contin-
ued to identify other SNP associations, with confirmatory
data now existing for complement factor 2 (C2), 3 (C3), B
(CFB), and I (CFI) (Figure 42.2).26–28 Multivariate analy-
ses of C3 and CFH SNPs in AMD cohorts have estab-
lished very strong environmental–genetic interactions,
with active smokers’ odds ratios for AMD risk ballooning
to 10 if they are homozygous for an associated C3 SNP
(rs2230199, GG) and even higher for those with Y402H.29
These data suggest that such factors may be multiplicative
and account for a very high percentage of the incidence of
advanced AMD. Another regulator of alternative pathway
activation, complement factor D (CFD), was also linked
to AMD in one study group but has not been confirmed
Cfh
Cf2 Cfi
Immunity
Cf3 Cfd
Cfb
Vegfa Tgfbr1
Adamts9 Fib6 Htra1
Angiogenesis/ECM
Timp3 Col8a1 Arms2
Efemp1 (Fib3)
Lipc Lpl
Abca1 Lipid Metabolism Cetp
Abca4 Apoe
Fads1-3
Figure 42.2 
Pathway classification of confirmed genes associated with
AMD. The array of genes associated with AMD can be broadly
classified into the following categories: immunity, angiogenesis and
extracellular matrix (ECM), and lipid metabolism.
6 5 4   •   G e no m ic s in C l inic a l Pr actic e
independently.30 More detailed scientific investigations of
alternative regulators of complement activation are now
being explored, with some known factors, including CD59,
exhibiting altered expression levels in animal models and
human eyes with AMD, but there is no evidence yet of SNP
associations.31–33 Most importantly, multiple pharmacologi-
cal approaches to complement pathway inhibition that are
founded on the strong genetic signature of AMD are now
being tested in clinic trials, including small molecule and
antibody-mediated neutralization of C3, CFD, and C5.
Aside from complement-induced pathways, the
innate immune system is also amplified via recognition
of pathogen-associated molecular patterns (PAMPs)
through conserved trans-membrane proteins called
Toll-like receptors (TLRs).34 Ten human TLRs have been
identified in humans, many of which are expressed in the
RPE.35 Several of these are now known to be involved in
ocular inflammatory processes36:  TLR2, which detects
peptidoglycan motifs present in gram-positive bacteria
cell walls; TLR3, which is activated by double-stranded
RNA (dsRNA) originating from viral genomes and rep-
licate intermediates; and TLR4, which binds to lipo-
polysaccharide (LPS), a lethal component of the outer
membrane of gram-negative bacteria.37 With activation,
these receptors induce nuclear translocation of various
factors involved in interferon induction, cytokine pro-
duction, and apoptosis.38 Initial genome-wide mapping
studies of extended pedigrees identified an association
between the 9q33 locus, where TLR4 resides, and AMD,
suggesting that the locus may be partially contributory
in some populations.8,9 Another investigation revealed a
link between AMD and the 4q32 locus,7 which the TLR2
gene maps to. From here, several groups have set out to
determine the genetic contribution of multiple TLRs on
AMD, given the strong immune-regulatory link to this
disease, but results have been controversial. There are cur-
rently no reports on TLR2 SNPs and AMD, but animal
models show evidence for the induction of neovascular
lesions similar to those in AMD via TLR2 signaling39
and its role in chemokine expression.40 The TLR4 SNP,
D299G,41 was observed to significantly increased risk of
AMD in one American cohort,42 but other studies with
international groups did not observe this genetic asso-
ciation.43,44 TLR4 may be important in the clearance and
metabolism of photoreceptor cellular material,45 a process
involved in early AMD pathogenesis.46 More recently,
an association was found between AMD and two other
TLR polymorphisms:  the TLR3 SNP, L412F, located
at 4q35; and the TLR7 SNP, Q11L, located at Xp22;47
however, these data became statistically insignificant after
researchers corrected for multiple comparisons. Further
studies of the hypomorphic L412F SNP suggested a pro-
tective effect for advanced dry AMD compared to normal
controls,48,49 yet this conclusion was similarly contro-
versial, as the result was not found in other cohorts.50,51
Importantly, dsRNA-induced retinotoxicity was observed
in mouse models and exhibited many features that reca-
pitulated GA, including large focal areas of isolated RPE
cell loss consistent with the human disease as well as dis-
rupted ZO-1 cell–cell junctions and increased caspase-3
activity.49,52 Stemming from this discovery, downstream
analyses focused on determining whether dsRNAs may
be present and involved in RPE cell death in GA. In
fact, significantly increased levels of dsRNA sequences
were immuno-localized in drusen, Bruch’s membrane,
and surrounding RPE in affected eyes. Using unbiased
adaptor-based polymerase chain reaction followed by
Sanger sequencing, we identified the accumulation of a
specific ~300-base pair dsRNA transcript derived from
Alu sequences in eyes with GA. Alu sequences are short,
interspersed, repetitive retrotransposons that are spe-
cific to primates and account for approximately 10% of
the human genome.53 Interrogation of human tissues for
downregulation of RNA-processing enzymes capable of
degrading these molecules was performed for a compre-
hensive panel of such targets. Eventually, it was discovered
that that DICER1, an RNA processing enzyme critical
to micro-RNA biogenesis, was specifically downregu-
lated in GA eyes compared to age-matched controls.54
Subsequent studies in our laboratory determined that
Alu RNA-induced RPE cell death was dependent on the
immune receptor MyD88, and inflammasome activation
via caspase-1 cleavage.55 Interestingly, MyD88 serves as a
convergence pathway for most TLR signaling;56 however,
to date, Alu RNA has not been shown to interact with
TLR3 or with any other innate immune receptors. These
data highlight the importance of non-coding genetics as
an approach for deciphering AMD disease mechanisms.
With a new understanding of the widespread pathological
effects of transcribed Alu in the eye and in other disease
models, efforts are currently underway to develop bio-
informatics pipelines with massively parallel sequencing
data in order to spatially localize Alu inserts in humans
and determine sequence composition, as there are many
subfamilies of these inserts. High-throughput screen-
ing assays of these mobile genetic elements have been
reported57,58 and may be useful if applied to AMD cohorts
for analyses of altered levels or mapping of Alu compared
to normal, age-matched controls. With these intrigu-
ing yet inconclusive reports, further immuno-genetic
G e n etic s a nd G e no m ic s in C l inic a l O p h t h a l m o l ogy, I I I   •   6 5 5
association studies are warranted, especially in light of
expression patterns and newly discovered functions of
specific immune mediators in the retina,59,60 and the per-
sistent speculation on the potential infectious etiology of
AMD61 providing a source of potent ligands for this sig-
naling pathway.
A N G I O G E N ET I C S I N
N E O VAS C U LA R AM D
Neovascular or “wet” AMD results in upwards of 90% of the
vision loss associated with this disease, but it only is found
in 10% of those affected with this disease. As described, the
main pathological event leading to retinal dysfunction is the
ingrowth and leakage of choroidal blood vessels (or CNV)
in the sub-retinal and sub-RPE space. This causes leakage of
fluid and hemorrhage in the macula, and often beneath the
fovea, the point of maximum visual acuity in the center of
the macula, severely altering the anatomical layering of the
chorioretinal tissues as well as the optical properties of the
eye. Early treatment modalities were focused on mechanical
or thermal ablation of the CNV lesion using a sub-retinal
surgical excision or argon-laser photocoagulation, but these
approaches certainly did not improve vision, nor did they
prevent further decline in visual acuity compared to con-
trols. Laser-based approaches were optimized to include the
use of photodynamic therapy, effectively inducing vascular
thrombosis and cessation of leakage in abnormal vascula-
ture, did result in improvements over observation only, but
patients unfortunately still continued to lose their vision.62
In the first few years after the turn of the millennium,
investigators gained critical knowledge on essential growth
factors in CNV pathogenesis63–66 and initiated studies to
evaluate the preclinical efficacy of targeting a key molecule,
vascular endothelial growth factor-A (VEGF-A), with neu-
tralizing antibodies and RNA aptamers to suppress intra-
ocular neovascular growth in animal models.67–69 These
results prompted a rapid translation to clinical trials; and,
less than a decade later, the field of wet AMD treatment
has been transformed by the widespread deployment of
antibody-based therapies targeting VEGF and its signal-
ing mechanisms, with unsurpassed gains in visual acuity
and anatomical reorganization. While early linkage and
genome-wide association studies (GWAS) did not show
evidence of VEGF and related polymorphisms in disease
cohorts,7,70 once VEGFA (6p21) became a candidate gene
with the success of therapeutic targeting, multiple inde-
pendent studies identified significant association of cer-
tain SNPs, specifically in the promoter region, in disease
cohorts.71,72 On review of genome-wide scanning studies,
two studies did find nominally significant associations with
a region on chromosome 6p, but fine mapping was directed
towards known genes involved in inherited retinal degen-
erations, such as RDS (6p21).8,73 As in other independent
GWAS studies, conflicting datasets have emerged from
more-detailed mapping experiments, which have either sup-
ported the association of various VEGFA SNPs (rs3025039,
rs4711751) with AMD,74,75 or shown no statistical geno-
typical difference.76,77
While this pharmaceutical revolution has drastically
improved our clinical outcomes, this approach is certainly
not a uniform success in our attempts to rescue or preserve
vision in all affected individuals. Approximately 40% of
patients will gain three lines of visual acuity with monthly
treatment, whereas over 50% will not exhibit a significant
response, and 10% will go on to lose further vision. With
long-term data out over seven years, these results have
greatly amplified the search for other molecular targets in
the angiogenesis pathways that are abnormally active dur-
ing CNV formation, and enhanced our efforts to improve
our understanding of variable treatment responses to
anti-VEGF-A therapy. Follow-up studies have now demon-
strated that specific VEGFA SNPs (rs943080 and rs833069)
are associated with decreased resolution of sub-retinal fluid
after initiation of therapy, suggesting that the clinical man-
agement of AMD may benefit from the utilization of cus-
tomized pharmacogenomics.78,79 Combination treatment
to block additional vaso-proliferative factors is also being
investigated, with advanced clinical trials analyzing the
efficacy of platelet-derived growth factor-B (PDGF-B) neu-
tralization with DNA aptamer-based drugs.80 To date, no
genetic associations with PDFG-B (22q13) SNPs have been
identified in AMD cohorts.
Compartmentalization of the specialized retinal cell
types and vascular barriers is critical to maintaining the
transparency and function of the retina. Seepage of cyto-
kines, coupled with exposure of normally concealed extra-
cellular matrix (ECM)–bound proteins through areas
of membranous breakdown in this multilayered cellular
architecture could be a significant instigator of subsequent
vascular in-growth and vision loss. Thus, aside from angio-
genic cytokines, many research groups have focused on
resolving the molecular composition of the ECM in both
drusen and surrounding areas of CNV. This approach is
strongly founded on previous research linking a form of
hereditary macular dystrophy with CNV (i.e., Sorsby’s
macular dystrophy) with gene mutations in tissue-inhibitor
of metalloproteinase-3 (TIMP-3, 22q12).81,82 TIMP-3 has
multiple functions, including its role in ECM formation
6 5 6   •   G e no m ic s in C l inic a l Pr actic e
and breakdown, as well as angiogenesis. Specifically, VEGF
receptor 2 (VEGFR-2) is blocked by TIMP-3, which pro-
hibits activation by VEGF-A.83 Specific TIMP-3 mutations
are able to disrupt this inhibitory action thus allowing
for unimpeded VEGF-A signaling and endothelial cell
migration.84 Given the undeniable genetic link to Sorsby’s
macular dystrophy and its strong immuno-localization to
drusen, TIMP-3 was initially targeted as a candidate gene
in AMD. On first investigation, no significant linkage was
observed,85 but further high-resolution studies revealed
an association with a region 100 kilobases upstream from
TIMP-3 (rs9621532), which was confirmed in subsequent
studies.75,86 Molecular investigations on the functional con-
sequences of this SNP revealed that TIMP-3 expression was
downregulated in RPE cells, suggesting a possible mecha-
nism for CNV formation in affected patients.87 Another
important facet of TIMP-3 biology and its role in AMD
pathogenesis is that molecular mapping of its physical inter-
actions has revealed binding to other biological factors that
have been significantly associated with AMD, including
Fibulin-3 (EFEMP-1).88 EFEMP-1 is involved in base-
ment membrane formation and maintenance. A  specific
EFEMP-1 mutation, R345W, is associated with a famil-
ial AMD-like condition known as Malattia Leventinese,
characterized by accelerated drusen biogenesis and possible
CNV or GA.89 Transgenic mouse models that encode this
mutation similarly develop sub-RPE deposits and struc-
tural features consistent with late retinal degeneration.90,91
Interestingly, the same EFEMP-1 domain that binds
TIMP-3 also interacts with CFH,22 thus representing a
plausible signaling axis upon which neovascular dysregula-
tion in advanced AMD may depend (Figure 42.3).
Another major target in AMD involving ECM biology
that has emerged from genetics studies resides at a locus on
10q26, which was identified in early genome-wide studies.
This region spans several hundred kilobases, with multiple
SNPs observed in linkage disequilibrium in AMD cohorts.7,70
This region encompasses three genes:  LOC387715/
age-related maculopathy susceptibility-2 (ARMS2),
HTRA1, and PLEKHA1. Dissection of the genetic con-
tributors of this association has been controversial and
time-consuming. Early reports demonstrated HTRA1 as
a major component of the disease risk,92,93 and functional
studies were extended to an animal model that showed
features of an AMD subtype.94 However, independent
studies mapped the high-risk allele to the ARMS2 locus
where the SNP rs10490924 (A69S) homozygous TT
genotype was associated with up to a risk of 4.1-fold com-
pared to wild-type.95,96 Studies analyzing expression data
of either HTRA1 or ARMS2 transcripts or protein have
 CRP

FIB3 CNV

Alternative SCR7 RPE

Complement CFH TIMP3 BM
Cascade CC
ECM Remodeling
Neovascularization
VEGF-A

Figure 42.3 
Central role of CFH-signaling in neovascular AMD progression. There is a strong genetic association between a specific CFH genotype,
402H, and significantly increased risk for progression to neovascular AMD. Mapping of this SNP on the CFH protein localized the SNP to the
short consensus repeat domain 7 (SCR7), which is a critical binding region for C-reactive protein (CRP), a well-established pro-inflammatory
and angiogenic molecule. CFH is an important auto-regulator of alternative complement-pathway activation and primarily serves as a direct
inhibitor of activated C3b, preventing formation of the C3 convertase. Alternative complement activation also induces the secretion of VEGF-A,
thus accelerating CNV progression. SCR7 dually interacts with TIMP3 and EFEMP1, both of which have been implicated in CNV pathogenesis
and are endogenous inhibitors of angiogenesis. This signaling axis reveals a central role for CFH and SCR7 domain in regulating pathological
angiogenesis during AMD progression.

been contradictory,97,98 but a consensus that suggests lev-
els are unchanged in the retina and RPE is beginning to
emerge.99,100 Currently, over 75% of all cases of AMD are
attributed to this genotype in combination with CFH
Y402H and smoking (which alone increases the odds ratio
for AMD risk to 4)  with ARMS2 carriers significantly
more likely to develop progressive, bilateral, neovascular
AMD.101 Subsequent studies have also observed this multi-
plicative effect, including data from a Finnish AMD cohort
in which carriers of all three major AMD-associated SNP
alleles (CFH, C3, and ARMS2) had an 18-fold higher
risk of disease incidence.102 Further investigations of the
ARMS2 locus resulted in the discovery of a protein that
localized to the outer membrane of mitochondria, which
may interact with specific components of the extracellular
matrix, including fibulin-1 and -6, COL1A1, COL4A2, and
fibronectin,103 but the function of this molecule remains
unknown and an area of intense investigation. ARMS-2
affinity for fibulin-6 has attracted considerable interest,
as the confirmed AMD locus on 1q31 mapped to associ-
ated SNPs in fibulin-6 (also known as hemecentin-1).104,105
Further cohorts tested have not revealed significant associa-
tions with these SNPs,9,106 raising the question of whether
this specific interaction in the ECM is an important con-
tributor to ARMS-2 involvement in AMD pathogenesis.
Recently, seven new AMD-associated loci were discov-
ered in a single study, with regions located near or in other
genomic regions also involved in ECM-regulation includ-
ing COL8A1, TGFBR1, and ADAMTS9.107 This impres-
sive finding fortifies that notion that ECM dysfunction in
AMD is probably multifactorial, with an array of potential
targets that interact with other known associated gene
products, of which ARMS-2 and CFH are likely to be the
most critical. Undoubtedly, the improved resolution of
signaling networks, coupled with modern bioinformatics
analyses that link angiogenesis, immune-mediated inflam-
mation, and ECM dysregulation, will considerably aid our
efforts to develop a complete molecular map of this disease
from which to expand our studies and identify therapeutic
targets.
G E N ET I C D ET E R M I N A N T S O F ALT E R E D
L I P I D META B O L I SM I N AM D
Nearly 50 years ago, investigations into disease associations
between hyperlipidemia and AMD were first reported. Since
then, significant evidence has accumulated supporting the
hypothesis that changes in lipid content and levels may be
involved in early age-related eye diseases. Histological stud-
ies of aged eyes demonstrated increased concentrations of
lipids in Bruch’s membrane, with subsequent studies focused
specifically on cholesterol accumulation.108,109 High-quality
ultra-structural imaging on well-preserved eyes suggested
that lipid particles concentrating in the outer retina were
derived from membrane debris released by nearby cells.110
An animal model of CNV-formation has also been devel-
oped based on subretinal lipid injections with pathological
features similar to the human condition.111 However, there
are conflicting data that increased high-density lipoproteins
are either protective of, or associated with, disease progres-
sion.112 The role of statins in AMD progression has been

G e n etic s a nd G e no m ic s in C l inic a l O p h t h a l m o l ogy, I I I   •   6 5 7

investigated, with results that suggest that patients with
elevated lipids on statin drugs are at increased risk for the
development of neovascular AMD.113 Several studies sup-
port the notion that modified intake of balance omega
poly-unsaturated fatty acids (ω-PUFAs) does offer a pro-
tective benefit for AMD progression.114 Recently, however,
results from the AREDS2 study were published, suggesting
no significant benefit from supplementation with balanced
ω-PUFAs over a median follow-up time of five years, fur-
thering the controversy over lipid composition and metabo-
lism in AMD progression and whether diet modification
can alleviate lipid deposition in the retina.115 Nonetheless,
several groups have narrowed their genetics investigation
to study specific molecules critical to lipid metabolism and
trafficking.
One of the most compelling genetic associations
between lipid regulation and AMD is ATP-binding cas-
sette, sub-family A  (ABCA4, 1p13) which is expressed in
cone and rod photoreceptors and involved in retinoid and
lipid transport.116 ABCA4 is best known for its strong asso-
ciation with Stargardt’s maculopathy, but it has also been
linked to AMD.117 An array of other genes encoding lipid
transporters and interacting proteins has also been inves-
tigated. ABCA1, a cholesterol efflux transporter involved
in regulation of circulating HDL, which is expressed
in the retina, has previously been linked to Tangier dis-
ease118 and low HDL levels. GWAS studies demonstrated
association of a SNP (rs1883025) in AMD cohorts that
exhibited increased risk for advanced phenotype.86 These
data were supported by subsequent analyses showing that
a normal allele for ABCA1 (TT) was observed to protect
subjects from the progression of intermediate to large
drusen.119,120 Animal models revealed that mice deficient
in ABCA1 harbor increased free cholesterol in choroid
macrophages, and that individuals with AMD expressed
significantly less ABCA1 in circulating monocytes com-
pared to age-matched controls.121 ABCA1 is also known
to promote the secretion of apolipoprotein E (ApoE,
19q13) from macrophages,122 and ApoE itself, in conjunc-
tion with ABCA1, has also been implicated as an impor-
tant regulator of cholesterol efflux.123,124 ApoE is found in
chylomicrons and intermediate-density lipoproteins, which
may also contain ApoB (2p24). Both ApoB/E localize to
basal laminar deposits and drusen associated with AMD
in human eyes, in established animal models, and in vitro
correlates of drusen biogenesis.125–127 There are three major
isoforms of ApoE (ApoE2/3/4), with the ApoE4 allele best
known for its association with increased risk and earlier
time of onset of Alzheimer’s disease.128 Multiple indepen-
dent groups have also found significant linkage analyses of
6 5 8   •   G e no m ic s in C l inic a l Pr actic e
ApoE4 association but with a protective benefit of AMD
incidence, and the ApoE2 allele demonstrating association
with increased AMD risk.129–133 This curious finding that
ApoE4 offered protection was further investigated in ani-
mal models, which demonstrated AMD-like features with
enforced expression of both ApoE2 and ApoE4 alleles.134
This discrepancy was attributed to the strict homozygos-
ity of the mouse model and the highly controlled high-fat
dietary intake. Certainly, the plausible association between
Alzheimer’s and AMD has been suspected and previously
studied, but further work is required in order to draw a
firm conclusion.135 These mechanistic insights suggest that
exploring approaches to improve lipid trafficking, and
most specifically cholesterol efflux from pro-inflammatory
ocular and circulating macrophages, may offer a potential
therapy to decrease drusen accumulation and slow AMD
progression. Several SNPs, both common and rare, in genes
for other mediators of HDL formation and trafficking are
associated with AMD risk with varying odds ratios between
studies, including hepatic lipase C (LIPC, 15q21, rs493258
and rs10468017), cholesteryl ester transfer protein (CETP,
16q21, rs3764261), lipoprotein lipase (LPL, 8p22), and
fatty acid desaturases 1-3 (FADS1-3, 11q12).75,86,107,136
A  follow-up study identified increased serum HDL levels
and decreased AMD risk with the rare protective allele in
LIPC.137 Significant confusion remains as to whether HDL
levels are associated with progression,138 and the formation
and clearance of these particles are determined by multiple
genetic and environmental inputs, and it is therefore com-
plicated to control for biological variability. Nonetheless,
these data provide substantial genetic evidence supporting
the pathological link between faulty lipoprotein metabo-
lism and AMD progression.
As past and future efforts in the fields of molecular
genetics and AMD continue to delineate this fascinating
interplay of signaling networks that co-regulate angiogen-
esis, lipid metabolism, and innate immunity, our funda-
mental knowledge of AMD pathogenesis will be markedly
improved, with the hope that early detection and interven-
tion may be utilized to prevent vision loss from this dev-
astating disease. The field has already undergone a major
revolution in patient care in the last 10 years with the rapid
advance of anti-VEGF-A pharmacologics to the forefront of
management of neovascular AMD. Yet, the scourge of dry
AMD and GA remain a completely unmet medical need,
with millions of affected individuals throughout the world
succumbing to blindness each year from the progressive loss
of RPE and overlying photoreceptors. With the skyrock-
eting use and acceptance of intraocular injections for the
treatment of this disease, clinical trials assessing the safety
and efficacy of locally delivered injectable small-molecule
and biological inhibitors of many of the targets and path-
ways detailed in this chapter are underway, in addition to
several other approaches. To date, clinical trials have been
completed or are underway, evaluating therapeutics based
on antibodies and aptamers to neutralize select comple-
ment components such as C3, C5aR, and CFD; novel and
rediscovered drugs that target the visual cycle to mitigate
toxic byproducts of retinal metabolism; and stem-cell-based
therapies. While all of these exciting avenues of clinical
investigation have immeasurable potential, it is the massive
acquisition of genomic data in ocular disease, specifically
with respect to AMD, that has created a unique library, with
whose aid all future scientists and clinicians will be able to
improve on and establish new models of age-related biol-
ogy and disease. Such precious resources utilized in these
extreme and time-consuming efforts will provide the neces-
sary foundation for more detailed studies of inflammation,
cell death, and vascular biology, with a common goal to pro-
vide cures for those at risk of or suffering from age-related
forms of blindness.
R E F E R E N C ES
1. Tomany SC, Wang JJ, Van Leeuwen R, et al. Risk factors for inci-
dent age-related macular degeneration: pooled findings from 3 con-
tinents. Ophthalmology. Jul 2004;111(7):1280–1287.
2. Ferris FL, Davis MD, Clemons TE, et al. A simplified severity scale
for age-related macular degeneration: AREDS Report No. 18. Arch
Ophthalmol. Nov 2005;123(11):1570–1574.
3. Ferris FL 3rd, Wilkinson CP, Bird A, et  al. Clinical classifica-
tion of age-related macular degeneration. Ophthalmology. Apr
2013;120(4):844–851.
4. Klein R, Klein BE, Linton KL. Prevalence of age-related mac-
ulopathy. The Beaver Dam Eye Study. Ophthalmology. Jun
1992;99(6):933–943.
5. Seddon JM, Cote J, Page WF, Aggen SH, Neale MC. The US
twin study of age-related macular degeneration:  relative roles of
genetic and environmental influences. Arch Ophthalmol. Mar
2005;123(3):321–327.
6. Klein ML, Schultz DW, Edwards A, et  al. Age-related macular
degeneration. Clinical features in a large family and linkage to chro-
mosome 1q. Arch Ophthalmol. Aug 1998;116(8):1082–1088.
7. Majewski J, Schultz DW, Weleber RG, et  al. Age-related macular
degeneration—a genome scan in extended families. Am J Hum
Genet. Sep 2003;73(3):540–550.
8. Seddon JM, Santangelo SL, Book K, Chong S, Cote J. A genome-
wide scan for age-related macular degeneration provides evidence
for linkage to several chromosomal regions. Am J Hum Genet. Oct
2003;73(4):780–790.
9. Abecasis GR, Yashar BM, Zhao Y, et al. Age-related macular degen-
eration:  a high-resolution genome scan for susceptibility loci in a
population enriched for late-stage disease. Am J Hum Genet. Mar
2004;74(3):482–494.
10. Holers VM, Thurman JM. The alternative pathway of complement
in disease:  opportunities for therapeutic targeting. Mol Immunol.
Jun 2004;41(2–3):147–152.
G e n etic s a nd G e no m ic s in C l inic a l O p h t h a l m o l ogy, I I I   •   6 5 9
11. Dragon-Durey MA, Fremeaux-Bacchi V, Loirat C, et  al.
Heterozygous and homozygous factor H deficiencies associated
with hemolytic uremic syndrome or membranoproliferative glo-
merulonephritis: report and genetic analysis of 16 cases. J Am Soc
Nephrol: JASN. Mar 2004;15(3):787–795.
12. Mullins RF, Aptsiauri N, Hageman GS. Structure and composition
of drusen associated with glomerulonephritis: implications for the
role of complement activation in drusen biogenesis. Eye (London,
England). Jun 2001;15(Pt 3):390–395.
13. Hageman GS, Anderson DH, Johnson LV, et al. A common hap-
lotype in the complement regulatory gene factor H (HF1/CFH)
predisposes individuals to age-related macular degeneration. Proc
Natl Acad Sci U S A. May 17 2005;102(20):7227–7232.
14. Jozsi M, Oppermann M, Lambris JD, Zipfel PF. The C-terminus
of complement factor H is essential for host cell protection. Mol
Immunol. Apr 2007;44(10):2697–2706.
15. Klein RJ, Zeiss C, Chew EY, et  al. Complement factor H poly-
morphism in age-related macular degeneration. Science. Apr 15
2005;308(5720):385–389.
16. Haines JL, Hauser MA, Schmidt S, et al. Complement factor H vari-
ant increases the risk of age-related macular degeneration. Science.
Apr 15 2005;308(5720):419–421.
17. Edwards AO, Ritter R 3rd, Abel KJ, Manning A, Panhuysen C,
Farrer LA. Complement factor H polymorphism and age-related
macular degeneration. Science. Apr 15 2005;308(5720):421–424.
18. Clark SJ, Higman VA, Mulloy B, et al. H384 allotypic variant of fac-
tor H associated with age-related macular degeneration has different
heparin-binding properties from the non-disease-associated form. J
Biol Chem. August 2006;281(34):24713–24720.
19. Weismann D, Hartvigsen K, Lauer N, et al. Complement factor H
binds malondialdehyde epitopes and protects from oxidative stress.
Nature. Oct 6, 2011;478(7367):76–81.
20. Giannakis E, Jokiranta TS, Male DA, et al. A common site within fac-
tor H SCR 7 responsible for binding heparin, C-reactive protein and
streptococcal M protein. Eur J Immunol. Apr 2003;33(4):962–969.
21. Bhutto IA, Baba T, Merges C, Juriasinghani V, McLeod DS,
Lutty GA. C-reactive protein and complement factor H in aged
human eyes and eyes with age-related macular degeneration. Br J
Ophthalmol. Sep 2011;95(9):1323–1330.
22. Wyatt MK, Tsai JY, Mishra S, et al. Interaction of complement fac-
tor H and fibulin3 in age-related macular degeneration. PLoS ONE.
2013;8(6):e68088.
23. Estaller C, Schwaeble W, Dierich M, Weiss EH. Human comple-
ment factor H: two factor H proteins are derived from alternatively
spliced transcripts. Eur J Immunol. Mar 1991;21(3):799–802.
24. Ansari M, McKeigue PM, Skerka C, et  al. Genetic influences on
plasma CFH and CFHR1 concentrations and their role in suscep-
tibility to age-related macular degeneration. Hum Mol Genet. Dec
2013;22(23):4857–4869.
25. Coffey PJ, Gias C, McDermott CJ, et  al. Complement fac-
tor H deficiency in aged mice causes retinal abnormalities
and visual dysfunction. Proc Natl Acad Sci U S A. Oct 16
2007;104(42):16651–16656.
26. Gold B, Merriam JE, Zernant J, et al. Variation in factor B (BF) and
complement component 2 (C2) genes is associated with age-related
macular degeneration. Nat Genet. Apr 2006;38(4):458–462.
27. Yates JR, Sepp T, Matharu BK, et al. Complement C3 variant and
the risk of age-related macular degeneration. N Engl J Med. Aug 9,
2007;357(6):553–561.
28. van de Ven JP, Nilsson SC, Tan PL, et al. A functional variant in the
CFI gene confers a high risk of age-related macular degeneration.
Nat Genet. July 2013; 45(7): 813–817.
29. Seddon JM, Reynolds R, Maller J, Fagerness JA, Daly MJ, Rosner
B. Prediction model for prevalence and incidence of advanced
age-related macular degeneration based on genetic, demographic,
and environmental variables. Invest Ophthalmol Vis Sci. May
2009;50(5):2044–2053.
 30. Stanton CM, Yates JR, den Hollander AI, et al. Complement factor
D in age-related macular degeneration. Invest Ophthalmol Vis Sci.
2011;52(12):8828–8834.
31. Bora NS, Kaliappan S, Jha P, et al. CD59, a complement regulatory
protein, controls choroidal neovascularization in a mouse model
of wet-type age-related macular degeneration. J Immunol. Feb 1,
2007;178(3):1783–1790.
32. Cipriani V, Matharu BK, Khan JC, et al. Genetic variation in com-
plement regulators and susceptibility to age-related macular degen-
eration. Immunobiology. Feb 2012;217(2):158–161.
33. Ebrahimi KB, Fijalkowski N, Cano M, Handa JT. Decreased mem-
brane complement regulators in the retinal pigmented epithelium
contributes to age-related macular degeneration. J Pathol. Apr
2013;229(5):729–742.
34. Medzhitov R, Janeway C, Jr. The Toll receptor family and microbial
recognition. Trends Microbiol. Oct 2000;8(10):452–456.
35. Ranjith-Kumar CT, Miller W, Sun J, et al. Effects of single nucleo-
tide polymorphisms on Toll-like receptor 3 activity and expression
in cultured cells. J Biol Chem. Jun 15 2007;282(24):17696–17705.
36. Chang JH, McCluskey PJ, Wakefield D. Toll-like receptors in ocular
immunity and the immunopathogenesis of inflammatory eye dis-
ease. Br J Ophthalmol. Jan 2006;90(1):103–108.
37. Oda K, Kitano H. A comprehensive map of the toll-like receptor
signaling network. Mol Syst Biol. 2006;2:1–20.
38. Kawai T, Akira S. TLR signaling. Cell Death Differ. May
2006;13(5):816–825.
39. Fujimoto T, Sonoda KH, Hijioka K, et al. Choroidal neovascular-
ization enhanced by Chlamydia pneumoniae via Toll-like receptor 2
in the retinal pigment epithelium. Invest Ophthalmol Vis Sci. Sep
2010;51(9):4694–4702.
40. Yang X, Coriolan D, Schultz K, Golenbock DT, Beasley D. Toll-like
receptor 2 mediates persistent chemokine release by Chlamydia
pneumoniae–infected vascular smooth muscle cells. Arterioscler
Thromb Vasc Biol. Nov 2005;25(11):2308–2314.
41. Arbour NC, Lorenz E, Schutte BC, et al. TLR4 mutations are asso-
ciated with endotoxin hyporesponsiveness in humans. Nat Genet.
Jun 2000;25(2):187–191.
42. Zareparsi S, Buraczynska M, Branham KE, et al. Toll-like receptor 4
variant D299G is associated with susceptibility to age-related macu-
lar degeneration. Hum Mol Genet. Jun 1, 2005;14(11):1449–1455.
43. Despriet DD, Bergen AA, Merriam JE, et  al. Comprehensive
analysis of the candidate genes CCL2, CCR2, and TLR4 in
age-related macular degeneration. Invest Ophthalmol Vis Sci. Jan
2008;49(1):364–371.
44. Kaur I, Hussain A, Hussain N, et  al. Analysis of CFH, TLR4,
and APOE polymorphism in India suggests the Tyr402His
variant of CFH to be a global marker for age-related macular
degeneration. Invest Ophthalmol Vis Sci. September 1, 2006
2006;47(9):3729–3735.
45. Kindzelskii AL, Elner VM, Elner SG, Yang D, Hughes BA, Petty
HR. Toll-like receptor 4 (TLR4) of retinal pigment epithelial cells
participates in transmembrane signaling in response to photorecep-
tor outer segments. J Gen Physiol. Aug 2004;124(2):139–149.
46. Rakoczy PE, Zhang D, Robertson T, et  al. Progressive
age-related changes similar to age-related macular degenera-
tion in a transgenic mouse model. Am J Pathol. October 1, 2002
2002;161(4):1515–1524.
47. Edwards AO, Chen D, Fridley BL, et al. Toll-like receptor polymor-
phisms and age-related macular degeneration. Invest Ophthalmol
Vis Sci. Apr 2008;49(4):1652–1659.
48. Zhou P, Fan L, Yu KD, Zhao MW, Li XX. Toll-like receptor 3
C1234T may protect against geographic atrophy through decreased
dsRNA binding capacity. FASEB J. Oct 2011;25(10):3489–3495.
49. Yang Z, Stratton C, Francis PJ, et al. Toll-like receptor 3 and geo-
graphic atrophy in age-related macular degeneration. N Engl J Med.
Oct 2, 2008;359(14):1456–1463.
6 6 0   •   G e no m ic s in C l inic a l Pr actic e
50. Klein ML, Ferris FL 3rd, Francis PJ, et  al. Progression of geo-
graphic atrophy and genotype in age-related macular degeneration.
Ophthalmology. Aug 2010;117(8):1554–1559.
51. Cho Y, Wang JJ, Chew EY, et al. Toll-like receptor polymorphisms and
age-related macular degeneration: replication in three case-control
samples. Invest Ophthalmol Vis Sci. Dec 2009;50(12):5614–5618.
52. Kleinman ME, Kaneko H, Cho WG, et al. Short-interfering RNAs
induce retinal degeneration via TLR3 and IRF3. Mol Ther. Jan
2012;20(1):101–108.
53. Batzer MA, Deininger PL. Alu repeats and human genomic diver-
sity. Nat Rev Genet. May 2002;3(5):370–379.
54. Kaneko H, Dridi S, Tarallo V, et  al. DICER1 deficit induces Alu
RNA toxicity in age-related macular degeneration. Nature. Mar 17
2011;471(7338):325–330.
55. Tarallo V, Hirano Y, Gelfand BD, et al. DICER1 loss and Alu RNA
induce age-related macular degeneration via the NLRP3 inflamma-
some and MyD88. Cell. May 11 2012;149(4):847–859.
56. Akira S, Hoshino K, Kaisho T. The role of Toll-like receptors and
MyD88 in innate immune responses. J Endotox Res. 2000;6(5):
383–387.
57. David M, Mustafa H, Brudno M. Detecting Alu insertions
from high-throughput sequencing data. Nucl Acid Res. Sep
2013;41(17):e169. doi:10.1093/nar/gkt612.
58. Witherspoon DJ, Zhang Y, Xing J, et al. Mobile element scanning
(ME-Scan) identifies thousands of novel Alu insertions in diverse
human populations. Genome Res. Jul 2013;23(7):1170–1181.
59. Kohno H, Chen Y, Kevany BM, et  al. Photoreceptor proteins
initiate microglial activation via Toll-like receptor 4 in retinal
degeneration mediated by all-trans-retinal. J Biol Chem. May 24,
2013;288(21):15326–15341.
60. Maloney SC, Antecka E, Orellana ME, et  al. Choroidal neovas-
cular membranes express toll-like receptor 3. Ophthalmic Res.
2010;44(4):237–241.
61. Margolis TP, Lietman T, Strauss E. Infectious agents and ARMD: a
connection? Am J Ophthalmol. Sep 2004;138(3):468–470.
62. Bressler NM. Photodynamic therapy of subfoveal choroidal neo-
vascularization in age-related macular degeneration with vertepor-
fin: two-year results of 2 randomized clinical trials—TAP Report 2.
Arch Ophthalmol. Feb 2001;119(2):198–207.
63. Kwak N, Okamoto N, Wood JM, Campochiaro PA. VEGF is
major stimulator in model of choroidal neovascularization. Invest
Ophthalmol Vis Sci. Sep 2000;41(10):3158–3164.
64. Cui JZ, Kimura H, Spee C, Thumann G, Hinton DR, Ryan SJ.
Natural history of choroidal neovascularization induced by vascu-
lar endothelial growth factor in the primate. Graefes Arch Clin Exp
Ophthalmol. Apr 2000;238(4):326–333.
65. Shen WY, Yu MJ, Barry CJ, Constable IJ, Rakoczy PE. Expression
of cell adhesion molecules and vascular endothelial growth fac-
tor in experimental choroidal neovascularisation in the rat. Br J
Ophthalmol. Sep 1998;82(9):1063–1071.
66. Ishibashi T, Hata Y, Yoshikawa H, Nakagawa K, Sueishi K, Inomata
H. Expression of vascular endothelial growth factor in experimental
choroidal neovascularization. Graefes Arch Clin Exp Ophthalmol.
Mar 1997;235(3):159–167.
67. Eyetech Study Group. Preclinical and phase 1A clinical evaluation
of an anti-VEGF pegylated aptamer (EYE001) for the treatment of
exudative age-related macular degeneration. Retina (Philadelphia,
PA.). Apr 2002;22(2):143–152.
68. Krzystolik MG, Afshari MA, Adamis AP, et al. Prevention of experi-
mental choroidal neovascularization with intravitreal anti-vascular
endothelial growth factor antibody fragment. Arch Ophthalmol.
Mar 2002;120(3):338–346.
69. Saishin Y, Takahashi K, Lima e Silva R, et al. VEGF-TRAP(R1R2)
suppresses choroidal neovascularization and VEGF-induced
breakdown of the blood-retinal barrier. J Cell Physiol. May
2003;195(2):241–248.
 70. Weeks DE, Conley YP, Tsai HJ, et al. Age-related maculopathy: a
genomewide scan with continued evidence of susceptibility loci
within the 1q31, 10q26, and 17q25 regions. Am J Hum Genet. Aug
2004;75(2):174–189.
71. Churchill AJ, Carter JG, Lovell HC, et al. VEGF polymorphisms
are associated with neovascular age-related macular degeneration.
Hum Mol Genet. Oct 1, 2006;15(19):2955–2961.
72. Haines JL, Schnetz-Boutaud N, Schmidt S, et al. Functional candi-
date genes in age-related macular degeneration: significant associa-
tion with VEGF, VLDLR, and LRP6. Invest Ophthalmol Vis Sci.
Jan 2006;47(1):329–335.
73. Schick JH, Iyengar SK, Klein BE, et al. A whole-genome screen of a
quantitative trait of age-related maculopathy in sibships from the Beaver
Dam Eye Study. Am J Hum Genet. Jun 2003;72(6): 1412–1424.
74. Lin JM, Wan L, Tsai YY, et al. Vascular endothelial growth factor
gene polymorphisms in age-related macular degeneration. Am J
Ophthalmol. Jun 2008;145(6):1045–1051.
75. Yu Y, Bhangale TR, Fagerness J, et  al. Common variants near
FRK/COL10A1 and VEGFA are associated with advanced
age-related macular degeneration. Hum Mol Genet. Sep 15
2011;20(18):3699–3709.
76. Richardson AJ, Islam FM, Guymer RH, Cain M, Baird PN. A
tag-single nucleotide polymorphisms approach to the vascular
endothelial growth factor-A gene in age-related macular degenera-
tion. Mol Vis. 2007;13:2148–2152.
77. Boekhoorn SS, Isaacs A, Uitterlinden AG, et al. Polymorphisms in
the vascular endothelial growth factor gene and risk of age-related
macular degeneration: the Rotterdam Study. Ophthalmology. Nov
2008;115(11):1899–1903.
78. Chang W, Noh DH, Sagong M, Kim IT. Pharmacogenetic association
with early response to intravitreal ranibizumab for age-related macular
degeneration in a Korean population. Mol Vis. 2013;19:702–709.
79. Zhao L, Grob S, Avery R, et  al. Common variant in VEGFA and
response to anti-VEGF therapy for neovascular age-related macular
degeneration. Curr Mol Med. Jul 2013;13(6):929–934.
80. Green LS, Jellinek D, Jenison R, Ostman A, Heldin CH, Janjic N.
Inhibitory DNA ligands to platelet-derived growth factor B-chain.
Biochemistry. Nov 12 1996;35(45):14413–14424.
81. Weber BH, Vogt G, Pruett RC, Stohr H, Felbor U. Mutations in the
tissue inhibitor of metalloproteinases-3 (TIMP3) in patients with
Sorsby’s fundus dystrophy. Nat Genet. Dec 1994;8(4):352–356.
82. Jacobson SG, Cideciyan AV, Bennett J, Kingsley RM, Sheffield VC,
Stone EM. Novel mutation in the TIMP3 gene causes Sorsby fun-
dus dystrophy. Arch Ophthalmol. Mar 2002;120(3):376–379.
83. Qi JH, Ebrahem Q, Moore N, et  al. A novel function for tissue
inhibitor of metalloproteinases-3 (TIMP3):  inhibition of angio-
genesis by blockage of VEGF binding to VEGF receptor-2. Nat
Med. Apr 2003;9(4):407–415.
84. Qi JH, Dai G, Luthert P, et al. S156C mutation in tissue inhibitor of
metalloproteinases-3 induces increased angiogenesis. J Biol Chem.
Jul 24 2009;284(30):19927–19936.
85. De La Paz MA, Pericak-Vance MA, Lennon F, Haines JL, Seddon JM.
Exclusion of TIMP3 as a candidate locus in age-related macular degen-
eration. Invest Ophthalmol Vis Sci. May 1997;38(6):1060–1065.
86. Chen W, Stambolian D, Edwards AO, et al. Genetic variants near
TIMP3 and high-density lipoprotein-associated loci influence sus-
ceptibility to age-related macular degeneration. Proc Natl Acad Sci
U S A. Apr 20 2010;107(16):7401–7406.
87. Ardeljan D, Meyerle CB, Agron E, et al. Influence of TIMP3/SYN3
polymorphisms on the phenotypic presentation of age-related mac-
ular degeneration. Eur J Hum Genet. Oct 2013;21(10):1152–1157.
doi:10.1038/ejhg.2013.14.
88. Klenotic PA, Munier FL, Marmorstein LY, Anand-Apte B. Tissue
inhibitor of metalloproteinases-3 (TIMP-3) is a binding partner of
epithelial growth factor-containing fibulin-like extracellular matrix
protein 1 (EFEMP1). Implications for macular degenerations. J Biol
Chem. Jul 16 2004;279(29):30469–30473.
G e n etic s a nd G e no m ic s in C l inic a l O p h t h a l m o l ogy, I I I   •   6 6 1
89. Stone EM, Lotery AJ, Munier FL, et al. A single EFEMP1 muta-
tion associated with both Malattia Leventinese and Doyne honey-
comb retinal dystrophy. Nat Genet. Jun 1999;22(2):199–202.
90. Fu L, Garland D, Yang Z, et al. The R345W mutation in EFEMP1
is pathogenic and causes AMD-like deposits in mice. Hum Mol
Genet. Oct 15 2007;16(20):2411–2422.
91. Marmorstein LY, McLaughlin PJ, Peachey NS, Sasaki T,
Marmorstein AD. Formation and progression of sub-retinal pig-
ment epithelium deposits in Efemp1 mutation knock-in mice:  a
model for the early pathogenic course of macular degeneration.
Hum Mol Genet. Oct 15 2007;16(20):2423–2432.
92. Yang Z, Camp NJ, Sun H, et  al. A variant of the HTRA1 gene
increases susceptibility to age-related macular degeneration.
Science. Nov 10 2006;314(5801):992–993.
93. Dewan A, Liu M, Hartman S, et al. HTRA1 promoter polymor-
phism in wet age-related macular degeneration. Science. Nov 10
2006;314(5801):989–992.
94. Jones A, Kumar S, Zhang N, et al. Increased expression of multi-
functional serine protease, HTRA1, in retinal pigment epithelium
induces polypoidal choroidal vasculopathy in mice. Proc Natl Acad
Sci U S A. Aug 30 2011;108(35):14578–14583.
95. Seddon JM, Francis PJ, George S, Schultz DW, Rosner B, Klein
ML. Association of CFH Y402H and LOC387715 A69S with
progression of age-related macular degeneration. JAMA. Apr 25
2007;297(16):1793–1800.
96. Kanda A, Chen W, Othman M, et al. A variant of mitochondrial
protein LOC387715/ARMS2, not HTRA1, is strongly associated
with age-related macular degeneration. Proc Natl Acad Sci U S A.
Oct 9, 2007;104(41):16227–16232.
97. Friedrich U, Myers CA, Fritsche LG, et  al. Risk- and
non-risk-associated variants at the 10q26 AMD locus influence
ARMS2 mRNA expression but exclude pathogenic effects due to
protein deficiency. Hum Mol Genet. April 1, 2011 2011;20(7):
1387–1399.
98. Yang Z, Tong Z, Chen Y, et al. Genetic and functional dissection
of HTRA1 and LOC387715 in age-related macular degeneration.
PLoS Genet. Feb 2010;6(2):e1000836.
99. Kanda A, Stambolian D, Chen W, Curcio CA, Abecasis GR,
Swaroop A. Age-related macular degeneration-associated vari-
ants at chromosome 10q26 do not significantly alter ARMS2
and HTRA1 transcript levels in the human retina. Mol Vis.
2010;16:1317–1323.
100. Wang G, Dubovy SR, Kovach JL, et  al. Variants at chromosome
10q26 locus and the expression of HTRA1 in the retina. Exp Eye
Res. May 3, 2013;112C:102–105.
101. Schmidt S, Hauser MA, Scott WK, et al. Cigarette smoking strongly
modifies the association of LOC387715 and age-related macular
degeneration. Am J Hum Genet. May 2006;78(5):852–864.
102. Seitsonen SP, Onkamo P, Peng G, et  al. Multifactor effects and
evidence of potential interaction between complement factor H
Y402H and LOC387715 A69S in age-related macular degenera-
tion. PLoS One. 2008;3(12):e3833.
103. Kortvely E, Hauck SM, Duetsch G, et al. ARMS2 is a constituent
of the extracellular matrix providing a link between familial and
sporadic age-related macular degenerations. Invest Ophthalmol
Vis Sci. Jan 2010;51(1):79–88.
104. Thompson CL, Klein BE, Klein R, et  al. Complement factor H
and hemicentin-1 in age-related macular degeneration and renal
phenotypes. Hum Mol Genet. Sep 1, 2007;16(17):2135–2148.
105. Schultz DW, Klein ML, Humpert AJ, et  al. Analysis of the
ARMD1 locus: evidence that a mutation in HEMICENTIN-1 is
associated with age-related macular degeneration in a large family.
Hum Mol Genet. Dec 15 2003;12(24):3315–3323.
106. Bojanowski CM, Tuo J, Chew EY, Csaky KG, Chan CC. Analysis
of Hemicentin-1, hOgg1, and E-selectin single nucleotide poly-
morphisms in age-related macular degeneration. Trans Am
Ophthalmol Soc. 2005;103:37–44; discussion 44–35.
 107. Fritsche LG, Chen W, Schu M, et  al. Seven new loci associ-
ated with age-related macular degeneration. Nat Genet. Apr
2013;45(4):433–439, 439e1–2. doi:10.1038/ng.2578.
108. Pauleikhoff D, Harper CA, Marshall J, Bird AC. Aging changes
in Bruch’s membrane. A  histochemical and morphologic study.
Ophthalmology. Feb 1990;97(2):171–178.
109. Curcio CA, Millican CL, Bailey T, Kruth HS. Accumulation
of cholesterol with age in human Bruch’s membrane. Invest
Ophthalmol Vis Sci. Jan 2001;42(1):265–274.
110. Curcio CA, Presley JB, Millican CL, Medeiros NE. Basal depos-
its and drusen in eyes with age-related maculopathy: evidence for
solid lipid particles. Exp Eye Res. Jun 2005;80(6):761–775.
111. Baba T, Bhutto IA, Merges C, et al. A rat model for choroidal neo-
vascularization using subretinal lipid hydroperoxide injection. Am
J Pathol. Jun 2010;176(6):3085–3097.
112. Klein R, Cruickshanks KJ, Nash SD, et  al. The prevalence of
age-related macular degeneration and associated risk factors. Arch
Ophthalmol. Jun 2010;128(6):750–758.
113. VanderBeek BL, Zacks DN, Talwar N, Nan B, Stein JD. Role of
statins in the development and progression of age-related macular
degeneration. Retina (Philadelphia, Pa.). Feb 2013;33(2):414–422.
114. Reynolds R, Rosner B, Seddon JM. Dietary omega-3 fatty acids,
other fat intake, genetic susceptibility, and progression to incident
geographic atrophy. Ophthalmology. May 2013;120(5):1020–
1028. doi:10.1016/j.ophtha.2012.10.020.
115. Age-Related Eye Disease Study 2 Research Group. Lutein + zea-
xanthin and omega-3 fatty acids for age-related macular degenera-
tion: the Age-Related Eye Disease Study 2 (AREDS2) randomized
clinical trial. JAMA. May 15 2013;309(19):2005–2015.
116. Molday RS, Zhong M, Quazi F. The role of the photoreceptor
ABC transporter ABCA4 in lipid transport and Stargardt macular
degeneration. Biochim Biophys Acta. Jul 2009;1791(7):573–583.
117. Fritsche LG, Fleckenstein M, Fiebig BS, et  al. A subgroup of
age-related macular degeneration is associated with mono-allelic
sequence variants in the ABCA4 gene. Invest Ophthalmol Vis Sci.
Apr 2012;53(4):2112–2118.
118. Bodzioch M, Orso E, Klucken J, et  al. The gene encoding
ATP-binding cassette transporter 1 is mutated in Tangier disease.
Nat Genet. Aug 1999;22(4):347–351.
119. Yu Y, Reynolds R, Rosner B, Daly MJ, Seddon JM. Prospective
assessment of genetic effects on progression to different stages of
age-related macular degeneration using multistate Markov models.
Invest Ophthalmol Vis Sci. Mar 2012;53(3):1548–1556.
120. Yu Y, Reynolds R, Fagerness J, Rosner B, Daly MJ, Seddon JM.
Association of variants in the LIPC and ABCA1 genes with inter-
mediate and large drusen and advanced age-related macular degen-
eration. Invest Ophthalmol Vis Sci. 2011;52(7):4663–4670.
121. Sene A, Khan AA, Cox D, et  al. Impaired cholesterol efflux in
senescent macrophages promotes age-related macular degenera-
tion. Cell Metab. Apr 2, 2013;17(4):549–561.
122. Von Eckardstein A, Langer C, Engel T, et  al. ATP binding cas-
sette transporter ABCA1 modulates the secretion of apolipopro-
tein E from human monocyte-derived macrophages. FASEB J. Jul
2001;15(9):1555–1561.
123. Chroni A, Nieland TJ, Kypreos KE, Krieger M, Zannis VI. SR-BI
mediates cholesterol efflux via its interactions with lipid-bound
ApoE. Structural mutations in SR-BI diminish cholesterol efflux.
Biochemistry. Oct 4, 2005;44(39):13132–13143.
6 6 2   •   G e no m ic s in C l inic a l Pr actic e
124. Yancey PG, Yu H, Linton MF, Fazio S. A pathway-dependent
on apoE, ApoAI, and ABCA1 determines formation of buoyant
high-density lipoprotein by macrophage foam cells. Arterioscler
Thromb Vasc Biol. May 2007;27(5):1123–1131.
125. Johnson LV, Forest DL, Banna CD, et al. Cell culture model that
mimics drusen formation and triggers complement activation asso-
ciated with age-related macular degeneration. Proc Natl Acad Sci
U S A. Nov 8, 2011;108(45):18277–18282.
126. Malek G, Li CM, Guidry C, Medeiros NE, Curcio CA.
Apolipoprotein B in cholesterol-containing drusen and basal
deposits of human eyes with age-related maculopathy. Am J Pathol.
Feb 2003;162(2):413–425.
127. Anderson DH, Ozaki S, Nealon M, et al. Local cellular sources of
apolipoprotein E in the human retina and retinal pigmented epi-
thelium:  implications for the process of drusen formation. Am J
Ophthalmol. Jun 2001;131(6):767–781.
128. Strittmatter WJ, Saunders AM, Schmechel D, et al. Apolipoprotein
E: high-avidity binding to beta-amyloid and increased frequency
of type 4 allele in late-onset familial Alzheimer disease. Proc Natl
Acad Sci U S A. Mar 1 1993;90(5):1977–1981.
129. Zareparsi S, Reddick AC, Branham KEH, et  al. Association of
apolipoprotein E alleles with susceptibility to age-related macu-
lar degeneration in a large cohort from a single center. Invest
Ophthalmol Vis Sci. May 1, 2004 2004;45(5):1306–1310.
130. Baird PN, Guida E, Chu DT, Vu HT, Guymer RH. The epsilon2
and epsilon4 alleles of the apolipoprotein gene are associated with
age-related macular degeneration. Invest Ophthalmol Vis Sci. May
2004;45(5):1311–1315.
131. Klaver CC, Kliffen M, van Duijn CM, et al. Genetic association
of apolipoprotein E with age-related macular degeneration. Am J
Hum Genet. Jul 1998;63(1):200–206.
132. Souied EH, Benlian P, Amouyel P, et al. The epsilon4 allele of the
apolipoprotein E gene as a potential protective factor for exuda-
tive age-related macular degeneration. Am J Ophthalmol. Mar
1998;125(3):353–359.
133. McKay GJ, Patterson CC, Chakravarthy U, et al. Evidence of asso-
ciation of APOE with age-related macular degeneration: a pooled
analysis of 15 studies. Hum Mutat. Dec 2011;32(12):1407–1416.
134. Malek G, Johnson LV, Mace BE, et  al. Apolipoprotein E
allele-dependent pathogenesis:  a model for age-related retinal
degeneration. Proc Natl Acad Sci U S A. Aug 16 2005;102(33):
11900–11905.
135. Klaver CC, Ott A, Hofman A, Assink JJ, Breteler MM, de Jong
PT. Is age-related maculopathy associated with Alzheimer’s dis-
ease? The Rotterdam Study. Am J Epidemiol. Nov 1 1999;150(9):
963–968.
136. Neale BM, Fagerness J, Reynolds R, et al. Genome-wide associa-
tion study of advanced age-related macular degeneration identifies
a role of the hepatic lipase gene (LIPC). Proc Natl Acad Sci U S A.
Apr 20 2010;107(16):7395–7400.
137. Reynolds R, Rosner B, Seddon JM. Serum lipid biomarkers and
hepatic lipase gene associations with age-related macular degenera-
tion. Ophthalmology. Oct 2010;117(10):1989–1995.
138. Fauser S, Smailhodzic D, Caramoy A, et al. Evaluation of serum
lipid concentrations and genetic variants at high-density lipo-
protein metabolism loci and TIMP3 in age-related macular
degeneration. Invest Ophthalmol Vis Sci. July 1, 2011;52(8):
5525–5528.
 43.
GENOMIC APPLICATIONS IN
AUDIOLOGICAL MEDICINE
Daphne Karfunkel-Doron, Zippora Brownstein, and Karen B. Avraham
INTRODUCTION
Hearing loss (HL) is the most prevalent sensorineural disor-
der, with an estimated genetic etiology of between 60–70%
worldwide[1]‌. Audiological or audiovestibular medicine is
the medical specialty concerned with the diagnosis and
management of audiovestibular and communication dis-
orders, affecting individuals of all ages. Auditory disorders
comprise a broad scope of etiologies that may be inherited,
infectious, inflammatory, vascular, traumatic, or metabolic.
This chapter will specifically address the genetic aspects of
HL.
The field of genetics has rapidly transformed in the last
century since the rediscovery of Mendel’s plant-breeding
experiments[2]‌. Following advances in recombinant DNA
technology and the polymerase chain reaction (PCR), the
mapping and identification of genes involved in hereditary
HL began, and by 1997, the first human deafness gene,
GJB2, was identified[3]. After the completion of the Human
Genome Project (HGP) in 2001[4,5], the “post-genomic era”
began, and along with the rapid developments in genomic
technologies and the field of bioinformatics, new deafness
genes have been discovered at unprecedented rates. These
discoveries have contributed to an exponential increase in
the understanding of the underlying genetic and molecular
mechanisms of HL, developed over the last two decades.
The accelerated discoveries and developments seen
in the post-genomic era have had an immense impact on
both the research and the clinical arenas of audiology[6]‌.
A  metamorphosis in the scope of practice of audiologists
and otolaryngologists is occurring, and as more genes are
discovered, the availability of clinical, molecular, genetic
diagnostic testing will grow. Clinicians will be required
to understand how these tests can be best used to guide
patient management, and staying current with such infor-
mation will require familiarity with online resources. Close
663
collaborations with genetic professionals who can assist
with the management of such patients can also serve as a
valuable resource for clinicians.
In this chapter, we will present current genomic tools
available to health practitioners today, and suggest ways
to incorporate these technologies into a practical clinical
approach for patients with hereditary HL. This approach
will facilitate more precise clinical genetic diagnosis and
improved patient care through rapid diagnosis, appropriate
genetic counseling, and proper medical management for
auditory disorders. In addition, we will examine the genet-
ics of HL within the context of these advances, and their
likely future directions. It is anticipated that genetic testing
will become a standard clinical tool following a patient his-
tory, physical examination, and audiometric profiling for
hearing-impaired patients.
H E R E D I TA RY H E A R I N G L O S S
Hearing loss affects people of all ages and walks of life. It is
estimated that over 275 million people suffer from moder-
ate to profound HL, contributing a substantial global bur-
den[7]‌. In the face of these statistics, deafness still remains
one of the more poorly understood and neglected condi-
tions. Recent years have witnessed an increased incidence of
HL, contributed by a growing aging population and higher
noise-exposure levels[8]. The World Health Organization
(WHO) recognized the scale of this problem, and began
to prioritize HL as a major public health concern, help-
ing to increase awareness. Their key proposal was to invest
in research, with the future projection of clinical appli-
cations[7]. Since 2005, with the implementation of new
genomic technologies[9], the potential for improved clinical
diagnostics and therapeutic treatments are becoming more
of a possibility than ever before.
 Middle Inner
(A) ear ear (C) Tectorial membrane (D)
External ear Stria vascularis
marginal cells
Collagen bundles
Vestibule
Kcne1
Tectorin fibrils K+
Kcnq1
Auditory
Temporal nerve K+
bone
Scala vestibuli Endolymph
perilymph e
bran K+
Cochlea
(B)
r ’s mem
Tympanic sne
Reis Endolymph
membrane K+
with high K+
Ossicles
Ear canal (150mM, 85mV)

(E)
Stereocilia

Ventral Scala typmani

(H)
cochlear perilymph

Suppor
nucleus

cell
ting
Ventral Outer

Prestin
cochlear hair cell
nucleus

kcnq4
(G) Stereocilium K+
Whirlin K+
Myosin
Spiral ganglion K+
in

s XVa K+
Myosin armon

Auditory nerve link

Tip
VIIa b

c d h 15
Sans H

3/p
cdh2 C×26/C×30 gap junction
3)
ks 2
(F) i n cdh
Actin filaments

p dl
Ti 5 an
h 1
cd
(p Stereocilia
b in
Myosin armon

Ankel links
Cuticular plate
H

vlgr1/usherin
VIIa

My I cldn14
os in V nV Outer
osi
I My hair cell

Figure 43.1 
Schematic illustration of the human ear. (A) The ear consists of the outer, middle, and inner ear. (B) Cross-section through one turn of a
cochlea. The sensory epithelium is composed of three rows of outer hair cells, one row of inner hair cells, and supporting cells. Proteins encoding
deafness genes CDH23, CLDN14, GJA1, KCNQ4, MYH9, MYH14, MYO3A, MYO6, MYO7A, MYO15A, PCDH15, POU4F3, PRES, OTOF,
STRC, TFCP2L3, TMC1, TMPRSS3, USH1C, and WFS1 are expressed in inner and outer hair cells. (C) The tectorial membrane is composed
of thin tectorin fibrils and heavy collagen bundles. Proteins encoding deafness genes COL11A2, OTOA, OTOG, and TECTA are expressed in the
tectorial membrane. (D) Marginal cells of the stria vascularis. Proteins encoding deafness genes ATP6B1, BSND, CLCNKA, CLCNKB, EDN3,
EDNRB, GJB2, GJB6, KCNE1, KCNQ1, MITF, MYH14, TFCP2L3, and TMPRSS3 are expressed in the stria vascularis. (E) Single outer hair
cell, surrounded by two Deiters’ supporting cells. Proteins encoding deafness genes CLDN14, GJA1, GJB2, GJB6, KCNQ4, SLC26A4, TFCP2L3,
and TMPRSS3 are expressed in the supporting cells. (F) Enlargement of the hair cell bundle with actin-based stereocilia. (G) A single stereocilium
containing actin filaments. Proteins encoding deafness genes ESPN, PCDH15, STRC, TMIE, and WHRN are expressed in the stereocilia. (H) The
inner hair cells are innervated by the spiral ganglion cells, which in turn produce action potentials that are transmitted via the auditory nerve to
the brain. Proteins encoding deafness genes GJB1, KCNQ1, MPZ, NDP, NDRG1, OTOF, PCDH15, PMP22, SBF2, SLC26A4, TMPRSS3, and
WFS1 are expressed in the spiral ganglion. Modified from[203].

6 6 4   •   G eno m ics in C linical P ractice

 C L A S S I FI C AT I O N O F H E R E D ITA RY
HEARING LOSS
The hearing apparatus is made up of three parts: the external,
middle, and inner ears, which function as one unit (Figure
43.1). Mechanical sound waves captured by the external
ear move along the auditory canal and strike the tympanic
membrane, leading to the vibration of the middle-ear ossi-
cles. These vibrations agitate the cochlea of the fluid-filled
inner ear, causing stimulation of sensory hair cells that con-
vert these vibrations into electrical neural impulses. These
impulses are transmitted to the hearing center of the brain,
the auditory cortex, where they are processed, and trans-
lated into what we recognize as sound. Normal hearing
occurs between 0 to 20 dB. Individuals with mild (27–40
dB), moderate (41–55 dB), or moderate–severe (56–70
dB) HL are referred to as hard of hearing, as distinct from
individuals with severe (71–90 dB) and profound HL (>90
dB), who are regarded as deaf[10]. HL is classified according
to etiology (genetic, environmental, or idiopathic factors),
onset (congenital, pre- or post-lingual), type (conductive,
sensorineural, or mixed), severity (mild, moderate, severe,
and profound), signal frequencies (low, middle, high),
association with other disorders (syndromic HL, SHL; or
non-syndromic HL, NSHL), and afflicted ear (unilateral or
bilateral)[11].
The etiology of HL is heterogeneous, involving a host
of genetic defects (~60–70%)[1]‌, environmental factors
(~30%), a combination thereof, or of unknown cause
(~10%). Environmental factors, mutual to all ages, include
infections (meningitis, measles, chickenpox, mumps, oti-
tis media), exposure to ototoxic medications, and acoustic
trauma, such as head injuries. In prenatal and childhood
HL, various perinatal complications and maternofetal
infections, such as rubella, cytomegalovirus (CMV), toxo-
plasmosis, herpes, and syphilis, are additional major compli-
cations. For adults, excessive noise exposure is an additional
complication[11]. Improved medical care during pregnancy
and in infants has resulted in a reduced prevalence of HL
caused by environmental factors, especially in developed
countries. Consequently, genetic causes, inherited or spon-
taneous, account for the largest proportion of prelingual
and childhood deafness[1]. Late-onset forms of deafness
result from genetic or environmental factors, or a combina-
tion of both, although the increasing numbers of families
discovered with these forms of HL suggests an underesti-
mated genetic contribution[11].
Approximately 2–3 out of 1,000 newborns are born with
congenital deafness or will develop severe-to-profound HL
prior to speech-acquisition[12]. During the transition from
G eno m ic A pplications in Audiological   Medicine   •   6 6 5
childhood to adulthood, the prevalence increases to 3.5
out of 1,000 children who will become hard-of-hearing[1]‌.
These cases are typically progressive and are less severe. In
the adult population, HL affects 4% of persons younger
than 45 years, and up to 50% of people by age 80.
The different types of HL are classified according to the
affected part of the hearing system. Any interference with
the passage of sound to the cochlea, via obstruction or dis-
ease affecting the external or middle ear, results in conduc-
tive hearing loss. Sensorineural HL (SNHL) is caused by
the destruction or dysfunction of the delicate sensory hair
cells in the inner ear, or of the auditory nerve. Mixed hear-
ing loss, as the name suggests, refers to a combination of
these two types[11].
D ET EC T I O N O F T H E H E A R I N G L O S S
Hearing impairment during infancy and early childhood,
if not treated, might impede suitable auditory stimulation,
resulting in dramatic consequences for language, communi-
cation, and learning, and later cognitive and psychosocial
functioning[11, 13, 14]. In adults, particularly in the elderly, the
progressive loss of hearing has been associated with a poorer
quality of life, social isolation, and negative impact on psy-
chological well-being[15,16].
Newborn Screening for Hearing Loss
Routine newborn screening by auditory brainstem response
(ABR) or oto-acoustic emission (OAE) testing has facili-
tated the early identification of HL[17]. Early detection of
HL, combined with early identification of the underly-
ing genes, may enhance rehabilitation. Characterization
of the proteins encoded by the deafness genes enables a
comprehensive understanding of the biological mecha-
nisms involved in the physiology and pathology of hearing.
Recent examples include: the USH1 genes that are involved
in the function and formation of the hair bundle, the main
actor of the mechanotransduction channel; mouse mutants
for Triobp and stereocilin that revealed the contribution
of the stereocilia rootlets to the hair bundle’s function and
its effect on speech intelligibility; and mutations in the
TECTA and TECTB genes that revealed that hair bundles
are stimulated by the tectorial membrane (reviewed in[18]).
These are only a few examples showing that discovering
deafness genes and the mechanisms by which they lead to
deafness is the key for deciphering the principles underlying
auditory processes, which in turn pave the way for future
genetic therapy.
 Genetic Testing for Hearing Loss
Genetic testing is vital to help identify the genetic basis of
HL for individuals who have already manifested the disease,
and for healthy individuals interested in knowing what dis-
eases they might develop, or in planning for a family. The
benefits for individuals, focusing on reproductive aspects,
include the possibility of detecting carriers, determining
the chance of HL’s reappearance in future generations, or
helping to eliminate any uncertainty or unnecessary testing
in the future. The advantages for affected individuals are
extensive. Accurate identification of the genetic etiology
can aid in improved direct medical management in terms
of prevention, prognosis, and treatment[19]. In cases where
the cause of HL is determined to be associated with a syn-
drome, effective medical monitoring of associated anoma-
lies in organs such as the eyes, heart, and kidneys may be
implemented. Moreover, crucial prognostic information,
and predictions of potential medical complications, can be
derived about future hearing, and management of patients
can be adjusted accordingly. From a psychological and emo-
tional aspect, many families seek genetic testing in order to
know the precise cause of HL[20]. Other associated benefits
for the individual and their families include the provision
of timely information regarding the probability of recur-
rence in future offspring, and improving early diagnosis
for potentially affected siblings and close relatives. Today,
molecular genetic testing and evaluation are becoming stan-
dard protocol in the etiological diagnosis of HL.
MONOGENIC (MENDELIAN)
H E R E D I TA RY H E A R I N G L O S S
Monogenic HL is extremely heterogeneous, with over 130
chromosomal loci mapped to the human genome, more
than 65 causally associated deafness genes, and within them,
more than 1,000 described mutations (Hereditary Hearing
Loss Homepage, hereditaryhearingloss.org; Deafness
Variation Database, http://deaf​nessvariationdatabase.
com/​). HL can be classified according to the presence (syn-
dromic) or absence (non-syndromic) of distinctive clini-
cal features[21]. The vast majority of the cases of hereditary
etiology is non-syndromic (~70%), while syndromic cases
represent ~30% of hereditary HL. Each of these categories
can be further sub-classified on the basis of mode of inheri-
tance and causal mutation. In most cases documented to
date, inherited HL is monogenic[22]. In view of the present
quantity of mapped loci, it is anticipated that many more
genes involved in HL will be identified. Genes involved in
auditory function belong to a variety of different protein
families with diverse functions, which, when disrupted,
result in a range of phenotypes, including transcription fac-
tors (POU4F3, POU3F4, PAX3, TFCP2L3), gap junction
proteins (connexin 26 and  32), ion channels (KCNQ1,
KCNE1, and KCNQ4), extracellular matrix molecules
(α-tectorin, otoancorin, and COL11A2), molecular
motors (myosin IIIA, myosin VI, myosin VIIA, and myosin
XVA), and structural proteins (otoferlin and diaphanous
1) (Figure 43.1). These proteins play a role in the develop-
ment, differentiation, maintenance, and functioning of the
six sensory organs of the inner ear, much like a finely tuned
orchestra.
In this section, we introduce both NSHL and SHL,
describing the most common genetic causes of each, and pro-
vide the most relevant information for other, less common
types. The aim of this section is not to provide an exhaus-
tive description of genes implicated in HL, but to provide
an overview that sufficiently allows the non-specialist to
appreciate the complexity of the genetics of HL. The reader
is encouraged to refer to curated databases that regularly
update the genes as they are identified.
N O N-SY N D RO M I C H E A R I N G L O S S
NSHL is the most common form of HL, involving only
anomalies of the inner ear. The different gene loci are
designated by DFN (for DeaFNess), and are classified
according to their mode of inheritance:  DFNA (autoso-
mal dominant), DFNB (autosomal recessive), and DFNX
(X-linked), followed by a number designating the order in
which the locus was discovered. Among forms of inheri-
tance, autosomal recessive is the most frequent (~77%),
followed by dominant (~22%), X-linked (~1%) or mito-
chondrial (<1%)[23]. For prelingual NSHL, most reported
families demonstrate autosomal recessive inheritance. For
postlingual NSHL, most reported families demonstrate
autosomal dominant inheritance, some maternal inheri-
tance due to mitochondrial mutations, and a few autosomal
recessive forms. They are characteristically progressive in
nature[11]. We will present the major genes associated with
NSHL that play a role in key cell structures that are central
to the process of hearing
GJ B2 (C O N N E X I N 26) A N D GJ B6
(C O N N E X I N 30)
A total of 20 genes encoding for connexin proteins have
been identified in humans[24]. Mutations in three of these
genes, Cx26 (GJB2), Cx30 (GJB6), and Cx31 (GJB3), have

6 6 6   •   G eno m ics in C linical P ractice

also been linked to either non-syndromic or syndromic HL.
Mutations in GJB2 are responsible for DFNB1 autosomal
recessive HL[3]‌. Aberrations in the GJB2 gene constitute up
to 50% of severe to profound prelingual recessive deafness
in several populations worldwide, with over 200 known
mutations[25]. These mutations manifest clinical heteroge-
neity, including mostly recessive mutations for congenital
severe to profound NSHL, but some cause mild to mod-
erate or progressive NSHL[26,27]. Involvement of recessive
mutations in auditory neuropathy (AN) was reported[28],
and dominant mutations for NSHL and for SHL involving
skin disease exist as well[29]. The most frequent GJB2 muta-
tions are 35delG, 167delT, and 235delC, with varied preva-
lence among populations[22].
GJB2 is a relatively small gene, consisting of two exons,
located on the short arm of chromosome 13. The entire
coding region of the gene is contained within a single
protein-coding exon, exon 2.  This helps simplify genetic
testing and clinical genetic diagnosis, and permits routine
testing. Exon 2 of GJB2 encodes connexin 26, a protein
that is a member of a large family of gap junction proteins.
These proteins form intercellular channels that regulate the
direct passage of inorganic ions and small metabolites, thus
mediating electrical and chemical cell-to-cell communica-
tion. Connexin 26 is specifically involved in the recycling of
potassium ions in the endolymph of the cochlea, following
stimulation of the sensory hair cells. It is believed that dis-
ruption in GJB2 disrupts this potassium ion flow, leading
to HL[30].
Geneticists, however, noticed that between 10–42% of
patients with mutations in connexin 26 only carried one
mutant GJB2 allele, and other familial cases showed link-
age to this locus, without any GJB2 mutations. It was there-
fore hypothesized that another gene mapping close to GJB2
might be involved. Indeed, three deletions were reported,
resulting in the truncation of the neighboring connexin
gene, GJB6[31–34]. Given the high prevalence of DFNB1
deafness, clinical genetic diagnosis for GJB2 and GJB6
mutations has become the standard care for the diagnosis of
patients with NSHL.
A D D IT I O NA L N O N-SY N D RO M I C
HEARING LOSS GENES
The most prevalent NSHL genes worldwide include
SLC26A4, MYO15A, OTOF, CDH23, and TMC1.
Dozens of mutations in these genes lead to HL. The num-
ber of mutations in the other genes is lower, and many of
them have been reported in consanguineous families or
single families[22].
G eno m ic A pplications in Audiological   Medicine   •   6 6 7
The SLC26A4 gene is currently considered to be the
second most frequent cause of NSHL, with an estimate of
up to 4% of NSHL worldwide[35]. SLC26A4, located on
the long arm of chromosome 7, encodes the pendrin pro-
tein. Pendrin, a member of the solute carrier 26 family, is a
transmembrane anion exchanger, predicted to have up to 15
transmembrane domains[36,37]. Pendrin is an electroneutral
anion exchanger mainly expressed in the thyroid, kidney,
and inner ear (reviewed in[38]). In the inner ear, pendrin is
expressed in the endolymphatic sac and duct; in the saccule,
utricle, and ampulla; and in different cochlear cell types[39].
Pendrin acts as an exchanger of Cl–/HCO3– and has a role
in secreting bicarbonate into the endolymph, thus control-
ling the pH of the endolymph, which is crucial for fluid
and ion homeostasis, affecting overall hearing function[40].
Over 100 mutations in the SLC26A4 gene are linked with
either Pendred’s syndrome (PS), involving enlargement of
the thyroid gland, or the DFNB4 form of NSHL, with or
without enlarged vestibular aqueduct (EVA) or Mondini,
a cochlea of one-and-one-half turns instead of the normal
two-and-one-half turns[41]. Clinical heterogeneity is usually
explained by the severity and type of mutations, whereas
nonsense and frameshift mutations, truncating most of the
protein or causing nonsense-mediated-decay (NMD), are
thought to be involved in more severe forms of the disease,
or in SHL vs. NSHL.
An X-linked form of deafness, DFNX2, involves partial
hypoplasia of the cochlea or Mondini, EVA, and a character-
istic stapes gusher, a flow of cerebrospinal fluid that occurs
as a consequence of surgery to release stapes fixation, where
the stapes bone of the middle ear is fixed in position[42].
Mutations have been found in POU3F4, a transcription
factor in subclass  III of the POU superfamily, including
point mutations, small or complete deletions of the gene,
and inversions in the coding region[42,43]. Deletions and
inversions upstream of the coding region are also known to
be associated with DFNX2 deafness[42, 44]. The deafness phe-
notype resulting from these mutations affects mostly males,
with phenotypes ranging from conductive, to mixed, to sen-
sorineural deafness, while female carriers have been known
to have late-onset hearing loss[45].
The genes described are only a small part of the com-
plete list of genes known to be involved in HL, but they rep-
resent the crucial impact that each singular gene has in the
complex auditory system and the consequences of muta-
tions that disrupt their normal function. The genes and
the proteins they encode are extremely diverse in function,
expression, size, and structure. Nevertheless, mutations in
different genes may lead to the same phenotype, while iden-
tical mutations may cause different phenotypes, suggesting
common pathways on one hand, and the involvement of
factors such as modifier genes or epigenetics on the other.
Both scenarios emphasize the necessity of the identifica-
tion of all genes involved: their early detection by molecular
screening, and protein characterization for understanding
of the auditory system, will accelerate comprehensive clini-
cal rehabilitation.
H ET E R O G E N E O U S D E A F N E S S
DISORDER S
OTO S C L E RO S I S
Whether otosclerosis is categorized as a monogenic or
complex disorder is still not determined, as the details of
the genetic basis for this disease have not yet been eluci-
dated. What is clear is that the otosclerosis cases that do
have a genetic component are genetically heterogeneous,
since at least nine loci have been identified[46]. Otosclerosis
is a common disorder, with a prevalence of 0.2% to 1%
in the Caucasian population[47]. It is caused by abnormal
bone-remodeling foci in the otic capsule of the human
temporal bone, which at the end of the process becomes a
dense sclerotic mass. The location of these sclerotic lesions
in the stapedio-vestibular joint prevents the motion of the
stapes in the oval window, resulting in clinical otosclerosis;
affecting hearing and frequently accompanied by tinnitus
and/or vertigo[48–51]. The age of onset for clinical otoscle-
rosis is typically 20 to 40 years, but juvenile otosclerosis is
known as well, usually with an onset not before five years
of age, although isolated cases of congenital otosclerosis are
also known[52,53]. Otosclerosis begins with conductive HL,
usually developing into mixed HL. Approximately 10% of
individuals with otosclerosis ultimately develop severe to
profound sensorineural HL[54]. A  large variety of charac-
teristics exists inside and between families, including age of
onset, presence of tinnitus and/or vertigo, progress, severity,
laterality, and other aspects of the disease. While some cases
have a positive family history, showing a mode of autoso-
mal dominant inheritance, with variable expressivity and
25–40% penetrance, most are sporadic cases with no family
history. Linkage analysis and genome scans of large fami-
lies with otosclerosis have revealed (as of this writing) nine
different loci for otosclerosis, OTSC1–OTSC8 (reviewed
in[46]) and OTSC10[55], indicating that at least nine distinct
genes are involved in the familial cases, but not in the spo-
radic ones.
Genetic association studies have pointed to several
genetic variants thought to increase the risk for otosclerosis.
6 6 8   •   G eno m ics in C linical P ractice
These include genes involved in bone remodeling, COL1A1,
TGFB1, BMP2, BMP4; genes in the hormonal pathway,
AGT and ACE; and autoimmune-associated HLA genes
(reviewed in[46,56]), all with functions that could be linked
to the development of otosclerosis. Recently, a new gene,
RELN, was identified by a genome-wide association study
(GWAS) in eight different populations, both European and
non-European, including close to 4,000 samples[57]. The
Reelin gene, encoding a secretory glycoprotein, was known
to be involved in brain development and synaptic plasticity
and shown to express in the stapes footplate and in the inner
ear, though its function remains unclear. Heterogeneity
appears to underlie the basis of the disease, and at least in
some populations, the genetic basis is triggered or modified
by environmental factors.
AU D I TO RY N EU RO PAT H Y
Auditory neuropathy (AN) is defined as a disturbed
transmission of the auditory signals from the inner ear
to the auditory nerve and auditory brainstem, resulting
in disrupted temporal encoding and neural synchrony[58].
It is characterized by normal outer hair cell function and
abnormal auditory nerve function[59], with a clinical pic-
ture of absent ABR but normal OAEs and/or cochlear
microphonics (CMs)[60]. The level of HL varies widely—
from normal pure tone results with difficulty in under-
standing speech, as revealed by discrimination results, to
severe HL that is usually also accompanied by lower dis-
crimination than expected, particularly in a noisy back-
ground. The onset ranges from birth to adulthood, and
the underlying etiology involves environmental factors
including prematurity, hyperbilirubinemia, brain dam-
age, drug exposure, and AIDS, as well as genetic causes,
including dominant, recessive, and mitochondrial genes
involved in syndromic or non-syndromic auditory neu-
ropathy (reviewed in[61]). The lesion can be located at the
inner hair cells, at the synapse between the inner hair cells
and the auditory nerve, or along the auditory nerve itself
up to the auditory cortex[58]. The different locations of
lesions represent the extremely heterogeneous etiology
that results in a wide spectrum of heterogeneous pheno-
types, recently termed “auditory neuropathy spectrum
disorder” (ANSD) (reviewed in[62]). The prevalence of
ANSD ranges from 0.5% to 15% among sensorineural
hearing-impaired populations[58,63].
The genes involved include OTOF (DFNB9), PJVK
(DFNB59), DIAPH3 (AUNA1), GJB2, ANUX1, and
12SrRNA for non-syndromic ANSD; PPM22, MPZ,
NL-F, NDRG1, GJB3, GJB1, OPA1, TMEM126A, FXN,
WFS1, TIMM8A (DPP), and 11778mtDNA for syn-
dromic ANSD (reviewed in[61, 64, 65]).
OTOF encodes the otoferlin protein[66]. Otoferlin is
involved in the hair cell presynaptic mechanism and plays a
role both in synaptic vesicle exocytosis at the hair cell affer-
ent ribbon synapse and in vesicle replenishment[67]. Up to
70 mutations are known in the OTOF gene, and the same
mutations might be involved in HL with or without AN,
most likely depending on modifiers or environmental fac-
tors, which remain to be defined. HL due to OTOF muta-
tions is mostly prelingual, severe or profound, and cochlear
implants yield good results, as expected of the presynaptic
role of the OTOF gene[65].
The PJVK gene encodes pejvakin, which is expressed in
hair cells, supporting cells, spiral ganglion, vestibule, and
cell bodies of neurons in the first three levels of the affer-
ent auditory pathway: the cochlear nuclei, superior olivary
complex, and the inferior colliculus[68]. Eight PJVK reces-
sive mutations are known[68–73], resulting in severe to pro-
found HL, with or without AN.
The DIAPH3 gene is responsible for an autosomal dom-
inant form of AN, AUNA1, with late-onset HL, beginning
in the second decade of life with rapid progression to pro-
found HL[74]. DIAPH3 encodes the diaphanous-3 protein,
a member of the formin protein family associated with cel-
lular activities involving regulation of actin polymerization,
including maintenance of cell and stereocilia shape and
vesicle trafficking. Mutations in DIAPH3 are thought to
disrupt the function of the distal portion of the auditory
nerve fibers, since cochlear implantation in these patients
result in improvement of auditory functions and recovery
of electrically evoked brainstem potentials. DIAPH3 muta-
tions are involved in a gain-of-function mechanism lead-
ing to over-expression of diaphanous 3 protein[74], which is
suggested to cause shape disruption of the dendritic spines
in the distal portions of auditory nerve fibers, interrupting
hair-cell function and progressing to profound deafness
and, finally, to loss of OAEs.
The AUNX1 locus on Xq23–27.3 is associated with
X-linked recessive AN[75]. These cases manifest, at the
beginning, HL with an audiological picture of AN, and
therefore are misdiagnosed as ANSD, but later they
develop into a peripheral sensory neuropathy, suggested
to be due to demyelination and axonal loss of the auditory
nerve[61].
The T1095C mutation in the mitochondrial 12S rRNA
gene has been detected in patients with moderate deafness
and AN[76]. As for the other AN mutations in other genes,
phenotypical variability is seen, including HL with or with-
out AN, and syndromic AN[77].
G eno m ic A pplications in Audiological   Medicine   •   6 6 9
Syndromic AN includes different types of Charcot-
Marie-Tooth (CMT) syndrome caused by mutations in
several genes, including PPM22, MPZ, NL-F, NDRG1,
GJB3, and GJB1; optic atrophy, dominant or recessive
(ADOA and AROA), as a result of mutations in OPA1
(R445H) and TMEM126A; Friedreich’s ataxia with FXN
mutations involved; AUNX1 caused by mutations in a gene
linked to Xq23-q27.3; deafness-dystonia-optic neuronopa-
thy (DDON) syndrome, also known as Mohr-Tranebjærg
syndrome, caused by TIMM8A mutations; Wolfram syn-
drome as a result of WFS1 mutations; and Leber’s heredi-
tary optic neuropathy (LHON), caused by mutations in
MTND4 (11778mtDNA) (reviewed in[64]).
Although it has been shown that successful results
of cochlear implantation might depend on the site of the
lesion, which is a reflection of the gene or factor involved,
it is important to emphasize that the distinction between
presynaptic and postsynaptic functioning of certain genes
is not always straightforward. Diagnosis of AN is the key
for rehabilitation. Reduced speech-perception that is dis-
proportional to the HL and presence of OAEs enable a
clear-cut AN diagnosis, particularly if it is accompanied
by peripheral and/or optic neuropathy. Molecular test-
ing for specific gene mutations, which enables localization
of the lesion, has crucial implications for rehabilitation.
Mutations causing presynaptic (OTOF) and postsynaptic
(OPA1 and DIAPH3) disorders of the distal auditory fibers
are predicted to lead to a good cochlear implantation out-
come, since cochlear implantation improves speech percep-
tion by bypassing the site of the lesion. In contrast, cochlear
implantation is expected to have little benefit when AN
involves the entire auditory nerve. The heterogeneity of
ANSD emphasizes the need for early genetic screening
to allow for genetic counseling for optimal treatment and
rehabilitation. Since there is clear evidence that gene–gene
and/or gene–environment interactions affect the geno-
type–phenotype correlations of most of the genes involved,
high-throughput methods might help researchers decipher
the puzzles.
M ITO C H O N D R I A L H E A R I N G L O S S
In the cochlea, both hair cells and stria vascularis are rich
in mitochondria, the source of adenosine triphosphate
(ATP) required for maintaining ionic gradients involved
in signal transduction. Mitochondrial mutations are asso-
ciated with both syndromic and widely variable NSHL.
The HL can be progressive or stable, across all frequencies,
and can be involved in ANSD as well[78]. The variability
is partly explained by heteroplasmy, but it is also thought
to be influenced by interactions with other genes, both
mitochondrial and nuclear, as well as with environmental
factors[79].
The mtDNA mutations include point mutations, dele-
tions, and duplications. The most common mitochon-
drial NSHL mutations are:  A1555G[79], C1494T[80], and
an insertion or deletion of C at nucleotide 961 in the
12S rRNA gene; as well as the mutations A7445G[81],
7472insC[82], T7510C[83], and T7511C[84] in the tRNASer
(UCN) gene. The most frequent mtDNA mutations,
A1555G followed by T1494C, are involved in both syn-
dromic and NSHL, as well as in aminoglycoside-induced
HL[85]. Other, less-frequent tRNA mutations, such as
tRNAThr 15927G>A and tRNASer(UCN) 7444G>A,
are thought to interact with these primary mutations and to
influence their variable phenotypes[86].
SY N D R O M I C H E A R I N G L O S S
Syndromic hearing loss (SHL) is associated with disorders
that result in congenital malformations and anomalies in var-
ious organs, such as retinitis pigmentosa (Usher syndrome),
thyroid goiter and inner ear defects (Pendred syndrome),
craniofacial dysmorphia (Treacher-Collins  syndrome),
renal abnormalities (Alport syndrome), and a long QT syn-
drome (Lange-Nielsen). Over 738 genetic syndromes have
been described in which HL is one of the symptoms[87]. The
Online Mendelian Inheritance in Man database (http://
www.ncbi.nlm.nih.gov/omim/) contains comprehensive
descriptions of the clinical features and molecular genetics
of these syndromes. Associated clinical manifestations can
often be observed; however, variable onset and expressivity
are common, even among related individuals, and may com-
plicate their diagnosis[88].
US H E R SY N D RO M E
Usher syndrome (USH) is an autosomal recessive disorder
characterized by deafness and visual impairments. USH is
the most common genetic cause of deafness-blindness[89],
with a conservatively estimated prevalence of one in 23,000
worldwide[90]. The vision loss is caused by retinitis pigmen-
tosa (RP), a degenerative retinal disease. The SNHL is usually
congenital, with vestibular dysfunction in a portion of the
cases. USH is both genetically and clinically heterogeneous.
Thus far, 12 USH loci and nine causative genes have been
identified (www.hereditaryhearingloss.org). Clinically,
USH is divided into three main subclasses[91]. These forms
are defined by the onset, severity, and progression of the
6 7 0   •   G eno m ics in C linical P ractice
HL, RP, and the presence of a vestibular phenotype[89,  92].
USH Type 1 (USH1) is the severest form, characterized by
profound congenital HL, prepubertal onset of RP, and ves-
tibular areflexia. USH1 accounts for up to 40% of all USH.
Eight USH1 loci (USH1A–H) have been mapped, and
five causative genes have been identified: MYO7A, encod-
ing myosin VIIA; USH1C, encoding harmonin, CDH23,
encoding cadherin 23; PCDH15, encoding protocadherin
15; and USH1G, encoding SANS, a protein that associates
with the USH1C protein, harmonin.
USH2 is the most frequent form of Usher syndrome,
up to 60%, characterized by congenital onset of moderate
to severe SNHL, RP during or after puberty, and normal
vestibular function. Four USH2 loci have been mapped
(USH2A–D), and three causative genes have been iden-
tified:  USH2A, encoding usherin; VLGR1, encoding a
G-coupled protein receptor; and WHRN, encoding whir-
lin. USH1 and USH2 proteins are involved in the develop-
ment and maintenance of the stereocilia in the inner ear.
USH3 is the most variable of these in onset and presen-
tation. The HL is progressive, and vestibular function may
or may not be affected[93]. This clinical type is the least prev-
alent overall (2–4%), although in some regions it is com-
mon, due to regional founder effects[94,  95]. A  single locus,
USH3, has been mapped, with a causative mutation found
in USH3A encoding Clarin1.
P E N D R E D SY N D RO M E
Pendred syndrome (PS) is an autosomal recessive disorder
accounting for 4–10% of prelingual HL worldwide[88, 96]. PS
is characterized by SNHL and the presence of a thyroid goi-
ter that is clinically evident either during childhood (40%)
or adulthood (60%). The HL is usually congenital, bilateral
or unilateral, severe to profound, and sloping in the higher
frequencies. It may be fluctuating or progressive, and patients
may experience variable vestibular dysfunction. HL might be
associated with Mondini dysplasia or an enlarged vestibular
aqueduct (EVA). SLC26A4 mutations, also responsible for
DFNB4, are the most common, detected in approximately
50% of the cases (reviewed in[97]). The other two genes asso-
ciated with Pendred Syndrome, FOXI1 and KCNJ10, are
associated with fewer than 1% of PS individuals[98].
J E RVE L L A N D L A N G E -N I E L S E N
SY N D RO M E
Jervell and Lange-Nielsen syndrome ( JLNS) is an autosomal
recessive disorder characterized by a markedly prolonged
QT interval, a propensity for ventricular arrhythmias,
congenital sensorineural deafness, and a high incidence of
sudden cardiac death[99]. JNLS is relatively rare, although
the precise prevalence of JLNS remains largely unknown
due to high mortality during infancy. JNLS is estimated to
affect between 1.6–6 per 1  million people worldwide[100],
with a higher prevalence in Sweden, most likely due to a
founder effect (1 in 200,000)[101]. JLNS is caused by homo-
zygous or compound heterozygous mutations that affect the
KCNQ1[99] and/or KCNE1[102,103] genes. KCNQ1 encodes a
676 amino acid α subunit, and KCNE1 encodes a smaller
129 amino acid β subunit. These subunits co-assemble to
form a voltage-gated potassium ion (K+) channel that is
involved in the production of endolymph in the inner ear,
and the slow rectifier current in cardiac muscle. About 90%
of all cases are caused by mutations in KCNQ1[104,105], with
mutations in KCNE1 constituting the remaining cases[105].
Molecular testing involves screening of both genes and dele-
tion/duplication analysis of exon(s) in KCNQ1[105,106].
WA A R D E N BU RG SY N D RO M E
Waardenburg syndrome (WS) is an autosomal dominant
inherited disease displaying clinical and genetic heterogene-
ity. WS is accountable for 1–3% of all congenital deafness
and has an estimated prevalence of one in 42,000. WS is
characterized by SNHL of variable degree, dystopia can-
thorum (appearance of wide-set eyes due to a prominent,
broad nasal root) and abnormal pigmentation caused by
melanocyte deficiency. Pigmentation anomalies can occur
in the eyes (complete or incomplete heterochromia iridis
or bright blue eyes), hair (premature graying of eyelashes,
eyebrows, poliosis), skin (pigmentation patches), and the
cochlear stria vascularis. Hirschsprung disease, character-
ized by the blockage of the large intestine due to improper
muscular bowel movements and neurological defects, is
additionally found in a subset of these patients. The syn-
drome is caused by mutations in six genes involved in the
regulation of melanocyte differentiation:  PAX3, MITF,
EDN3, EDNRB, SOX10, and SNAI2[107]. To help manage
the extreme genetic variation, a WS mutation database has
been created (http://grenada.lumc.nl/LOVD2/WS/)[107].
A L P O RT SY N D RO M E
Alport syndrome (AS) is a hereditary disorder of collagen
type VI, with a prevalence of one in 50,000. Clinically, it
presents as a progressive renal disorder characterized by
hematuria, which often leads to renal dysfunction[108,109],
and is accompanied by extra-renal manifestations such
as progressive high-tone sensorineural HL and ocular
G eno m ic A pplications in Audiological   Medicine   •   6 7 1
abnormalities[109]. Genetically, this disorder is heteroge-
neous, caused by mutations in three different collagen type
VI genes (COL4A3[110], COL4A4[110], and COL4A5[111])
found in the basilar membrane, parts of the spiral liga-
ment, and the stria vascularis[112]. It is inherited in an
X-linked manner in approximately 85% of families, caused
by mutations in COL4A5, and autosomal recessively and
dominantly in the remaining cases, caused by mutations in
COL4A3 and COL4A4. The HL is typically apparent only
during late childhood, with around 90% of those afflicted
exhibiting profound deafness by the age of 40[109].
The described syndromes are the most common HL
syndromes; nevertheless, they encompass only a minority
of the syndromes associated with deafness. Hundreds of
genes are involved in SHL, and for the majority of them
HL is the only mutual symptom, while the other symptoms
affect every system and tissue in the human body. This adds
another layer of complexity to the task of understanding the
already complicated auditory system, and obscures the pro-
cess of molecular diagnosis for HL even more, leading again
to the conclusion that high-throughput strategies such as
massively parallel sequencing (MPS) might be best able
to deal with the challenge that the large number of genes
involved present.
P O LYG E N I C C O M P L E X H E A R I N G L O S S
P R E S BYC US I S
Presbycusis, or age-related hearing loss (ARHL), is the
most common sensory handicap in the elderly, with severe
health and social consequences. It is characterized by slow
deterioration of hearing and speech-understanding in noisy
background, beginning at high frequencies and frequently
progressing to a strial pattern of a flat audiogram including
the entire frequency range[113]. The etiology of presbycusis is
complex, as it is the result of a combination of physiologi-
cal degeneration and environmental factors, and medical
conditions and treatment that are suggested to be interact-
ing with susceptibility genes (reviewed in[114]). The envi-
ronmental factors include noise exposure, ototoxic drugs
or diet, aminoglycoside antibiotics, smoking and elevated
blood pressure, and cholesterol levels. The heritability of
ARHL was estimated as responsible for 35–55% of the vari-
ance of ARHL (reviewed in[115]).
The ATP2B2 V586M variant has also been associated
with HL predisposition in connection with a homozygous
mutation in CDH23. In addition, the ATP2B2 V586M
variant was associated with NIHL predisposition, as it was
shown to modify the severity of HL caused by a mutation
in MYO6[116].
Mitochondrial DNA (mtDNA) is a key player in energy
metabolism. It was reported that mutated mtDNA accu-
mulates with ageing in humans[117] and in mice, where it was
shown to play a role in ageing[118]. In addition to point muta-
tions, mtDNA deletions, known as the “common ageing
deletions,” accumulate with age in different tissues, includ-
ing the cochlea. An increased number of mitochondrial
mutations have been implicated in presbycusis[119,  120]. The
deletion mtDNA4,834 associated with AHL in rodents[121]
is equivalent to the human deletion, mtDNA4,977,
detected in temporal bones from individuals with presby-
cusis[122]. This deletion of 4,977 base pairs (mtDNA4977]
between two 13-bp repeats at nucleotides 8470 and 13447
is most common among presbycusis patients[122,  123], while
it was very rare in a control group without age-related HL
(ARHL)[124]. Other AHL mouse models also support a
correlation between mtDNA mutations and presbycusis.
A mutation in the nucleus-encoded catalytic subunit of the
mtDNA polymerase gamma (POLG) in mice was accom-
panied by accumulation of mtDNA mutations that resulted
in premature ageing, including AHL[118,125,126].
Certain mitochondrial haplogroups have been associ-
ated with high prevalence of ARHL[127] (http://www.mito-
map.org/bin/view.pl/MITOMAP/​HaplogroupMarkers​).
Some of the polymorphisms composing haplogroups are
non-synonymous variants that might affect the protein
function[128], and some are rare variants that are thought
to have more predispositional effects than common vari-
ants[129]. In contrast, in a population study, sequencing
of the entire mitochondrial genome of 400 individuals,
including 200 with ARHL and 200 controls, and analyz-
ing association between single mitochondrial variants,
mutations, haplogroups, and environmental factors and the
phenotype, no association with presbycusis was shown[130].
As these studies were mainly conducted by using methods
for detecting mutations in monogenic diseases rather than
complex ones, there is a clear need for more advanced tech-
niques for identifying both common and rare mitochon-
drial and nuclear variants, and elucidating their complex
interactions, as well as the association between the variants
and the environment.
In humans, susceptibility genes for ARHL were
detected mainly by two approaches:  whole-genome scans
and association studies. Whole-genome linkage scans
revealed suggestive linkage to chromosomal regions 11p,
11q13.5, and 14q[131], and detected a linked region on 3q
that includes the DFNA18 locus[132], suggested to be asso-
ciated with ARHL. Association study results linked the
6 7 2   •   G eno m ics in C linical P ractice
NAT2*6A polymorphism, single-nucleotide polymor-
phisms (SNPs) in KCNQ4 and the TFCP2L3 (GRHL2)
gene to presbycusis[133–135].
N O I S E -I N D U C E D H E A R I N G L O S S ( N I H L)
NIHL, or acoustic trauma, is a most common SNHL,
second only to presbycusis. Four- to five-hundred million
people in the United States and Europe are estimated to be
developing NIHL[136]. NIHL is a result of continuous expo-
sure to noise over years, affecting mostly high frequencies,
with a typical notch at 4–6 kHz[137]. As an identical expo-
sure to noise can result in large differences in HL among
different individuals, susceptibility to the damaging effects
is inevitable, and a complex mode of inheritance is assumed.
The wide variability might be due to gene–gene, as well as
gene–environment interactions.
The environmental factors known to affect NIHL
include co-exposure to noise and chemicals; ototoxic
drugs such as aminoglycosides; heat, diet, smoking, high
blood pressure, and cholesterol levels; gender; and age
(reviewed in[138]). As for ARHI, NIHL is probably medi-
ated by reactive oxygen species (ROS). Increased levels
of ROS production in the cochlea are known to play a
role in noise-induced hair cell death[139], making antioxi-
dant defense-associated genes possible candidate genes for
NIHL. Among the cochlear antioxidant enzymes are the
GSH enzymes involved in glutathione metabolism (gluta-
thione S-transferase, GST; glutathione peroxidase, GPX1;
glutathione reductase, GSR)[140]. GST includes GSTM1
and GSTT1 that show human genetic variability, including
up to 50% of Europeans lacking the GSTM1 gene[141] and
25–40% having null genotypes for GSTT1[142]. Individuals
with these null genotypes are suggested to be more suscep-
tible to damage caused by oxidative stress, and therefore,
more prone to the effects of noise[143]. Other antioxidant
genes, Gpx1 and Sod1, were shown in knock-out mice to be
involved in NIHL as well as in ARHL[144,145]. Another asso-
ciation was identified between variations in CAT (catalase)
and NIHL susceptibility[146].
D RU G -I N D U C E D H E A R I N G L O S S
Over 130 drugs and chemicals are known to be potentially
ototoxic to the inner ear, including anti-inflammatory
agents (e.g., salicylates), aminoglycoside antibiotics (strep-
tomycin, kanamycin, neomycin, gentamicin, and more),
antineoplastic agents such as cisplatin and carboplatin,
and loop diuretics, including furosemide and ethacrynic
acid[147]. These include cancer and infectious disease
drugs (cisplatin and aminoglycosides). Drug-induced
HL (DIHL) starts at high frequencies, and with con-
tinuous use of the drug, progresses to finally include the
whole frequency range[148]. Usually, the severity of HL
depends on the dose and the period of time it is taken.
In general, high-dose treatments (loop diuretics and
anti-inflammatory agents) are linked to acute, temporary
HL, while long-term administration of ototoxic drugs
(antineoplastics and aminoglycosides) results in irrevers-
ible HL. A combination of two or more ototoxic drugs has
a larger effect on hearing[149], even though all these effects
differ widely among patients receiving the same treatment,
suggesting involvement of susceptibility genes.
DIHL, like ARHL and NIHL, was shown to be associ-
ated with ROS production[150], suggesting that antioxidant
therapy protects the hearing from ototoxic substances[151–153].
The genes thought to be involved include M40403, pro-
tecting from gentamicin ototoxicity, but not from cisplatin
otoxicity[154]; SOD1, protecting against kanamycin-induced
HL[155] and catalase; and SOD2, protecting hair cells against
aminoglycoside ototoxicity by reducing inner ear oxidative
stress[156]. Even though DIHL can be partly prevented by
adjustment of the optimal doses and the dosing intervals,
the genetic background will remain a major preventing fac-
tor, such as in the case of the mitochondrial A1555G muta-
tion in MTRNR1[157].
M O L E C U L A R D I AG N O S I S
OF HEARING LOSS
Genetic evaluation and diagnostic testing are carried out
only on the most prevalent genes that have been implicated
in HL. There are still many unsolved mapped chromosomal
loci involved in HL, and various forms of HL for which a
genetic basis has been postulated, but not yet proven. In
addition, there is a group of genes that have been identi-
fied, but for which clinical testing is not available, or not
common. For some genes, either the HL gene is extremely
rare, or technically too long and cumbersome to analyze.
Recent developments and the implementation of genomic
technologies contribute significantly to the improvement of
diagnostic genetic testing in the field. Currently, there are
new commercial platforms available that directly sequence
many deafness genes simultaneously, and others that are
being developed for diagnostics that help in the identifica-
tion of genes not previously connected to HL. Molecular
genetic testing has become a dynamic arena, with a continu-
ally growing list of available genetic tests. In the following
section, we introduce genomics, with a brief history, genetic
G eno m ic A pplications in Audiological   Medicine   •   6 7 3
techniques, genomic technologies, and its current and
future diagnostic applications.
G E N O M I C T EC H N I Q U E S
The genomic era (or modern genomics) was launched
with the development of the first rapid DNA-sequencing
method (Sanger sequencing)[158], and the later discovery of
the polymerase chain reaction (PCR)[159]. These methods
revolutionized the way geneticists approached the study
of DNA and made the molecular analysis of genes and
mutations possible. The automation of DNA-sequencing
helped dramatically increase the pace of genome sequenc-
ing, and together with PCR, and the novelty of computer-
ized sequencing data-collection[160], the stage was set for the
Human Genome Project (HGP). In 2001, after a 13-year
effort of more than 3,000 scientists, and at a cost of an esti-
mated $3 billion, the initial sequencing and mapping drafts
of the first human genome sequence were published[4,  5].
Since then, genomics has been driven by technological
advances. Enormous public databases with easily accessible
sequence data were developed following the advent of the
World-Wide Web and powerful computational algorithms,
to help manage large-scale genomic research informa-
tion[161–163]. Further advances in bioinformatics provided
researchers with the ability to mine meaningful biologi-
cal knowledge from volumes of accumulated data through
sophisticated data toolkits. The development of new
high-throughput biotechnologies, such as microarrays[164],
enabled the study of gene-expression and function. Novel
“omics” branches of science were born, such as compara-
tive genomics, proteomics, transcriptomics, metabalomics,
pharmacogenomics, functional genomics, and metagenom-
ics. The post-genomic era has provided professionals from
diverse fields with vast amounts of data and a repertoire of
tools to enable the research into, analysis of, and compre-
hension of the functioning of organisms in health and dis-
ease, at an unprecedented level of molecular detail.
Before the HGP, researchers relied on labor- and
time-intensive techniques for the mapping and identifica-
tion of disorders displaying a Mendelian, or single-gene,
pattern of inheritance. The traditional method for mapping
the disease genes centered on linkage mapping or positional
cloning. Linkage mapping is the process of systematically
scanning the genomes of various members of families
affected by the disorder, using regularly spaced, highly poly-
morphic DNA segments whose exact position is known[165].
Using these genetic markers, investigators can identify
genetic regions associated, or “in linkage,” with the disease
by observing that affected family members share certain
marker variants or alleles located in those regions signifi-
cantly more frequently than would be expected by chance.
These regions can then be isolated, or cloned, for further
analysis and characterization of the responsible genes. In
some instances, these markers fail to map the disease locus
(due to the lack of an easily identifiable marker), and addi-
tional markers are needed, a time-consuming and costly pro-
cess. The development of SNP arrays helped to overcome
this problem in part. “SNPs” refer to the non-pathogenic
changes of single nucleotides in the DNA genome, with a
frequency of more than 1% in the human population. These
polymorphisms are present, on average, in every 300 bases,
and help further narrow the regions of interest for directed
sequencing, significantly hastening the process. The primary
advantage of linkage mapping is that investigators need no
prior knowledge of the physiology or biology underlying
the disorder being studied, which is important for heteroge-
neous disorders, such as HL. Positional cloning has resulted
in the identification of many genes for deafness[23], and in
the identification of regions that may contain such genes
(see the Hereditary Hearing Loss homepage).
The above methods relied on the use of capillary-based,
semi-automated Sanger sequencing, which can achieve
reads of between 800–1,000 bp at a cost of about $500 per
megabase (Mb)[166]. Though highly accurate, this method is
chronophagous and expensive, making the large numbers of
genes required for sequencing a challenge. This was really
the start of the demand for rapid and low-cost sequenc-
ing technologies, which were eventually met in 2005 by
the development of massively parallel sequencing (MPS)
(also known as next-generation sequencing, NGS). As
sequencing technologies are continuously being improved,
the seemingly elusive dream of acquiring a $1,000 genome
in less than a day is close to realization. In contrast to the
traditional Sanger sequencing method, MPS platforms
do large-scale sequencing and can generate hundreds of
gigabases of sequences in a single run. (For more detailed
discussions, readers are referred to comprehensive reviews
in[166–168].)
MPS results are aligned to the latest human-reference
genome sequence using conventional bioinformatics align-
ment tools[169]. This alignment enables any differences
between the sequencing reads and the reference genome to
be distinguished. Bioinformatic tools are used to detect vari-
ants. Further annotation involves determining whether a
variant is unique by searching the SNP and OMIM databases
(http://www.ncbi.nlm.nih.gov/proj​ects/SNP/​; http://www.
ncbi.nlm.nih.gov/omim), and by determining allele fre-
quencies using data from other large sequencing proj-
ects such as the National Heart Lung Blood Institute
6 7 4   •   G eno m ics in C linical P ractice
Grand Opportunity (NHLBI GO) Exome Sequencing
Project (http://evs.​gs.washington.edu/EVS/​) and the
1000 Genomes Project (http://www.1000genomes.org/).
Potential damage to protein structure and function is pre-
dicted using algorithms such as Polyphen[170], SIFT[171],
and ConSurf[172], while taking sequence-conservation into
consideration. The mode-of-inheritance pattern is a crucial
factor for variant filtering and prioritizing, as the candidate
variants are being chosen to match the mode of inheritance,
and each potential candidate left after filtration is first
screened for segregation with the HL in the family.
This interpretation of data, and the subsequent trans-
lation of findings into something of a practical or clinical
value, is easy if a known deafness mutation is detected, but it
is challenging in cases where all genes known to be involved
in deafness are excluded and novel candidates are revealed.
Even though MPS has advanced our ability to sequence
genomes rapidly, sequencing and computational expendi-
tures still remain high, and are cumbersome. Techniques
have been developed that allow genomic regions of inter-
est to be selectively captured from a patient’s DNA before
sequencing. This intermediate resolution has the advantages
of higher coverage of regions of diagnostic interest at a higher
throughput, faster, cheaper, and with fewer ensuing data to
analyze[6,166]. If no mutation is detected in a targeted cap-
ture, the next inevitable step should be exome-sequencing
that includes all coding regions of all genes in the genome.
Presently, genomic-enrichment approaches are the method
of choice, providing the most fruitful clinical platforms[6]‌.
Technically, any subset of the genome can be targeted, but
in the field of audiology specifically, only the target capture
of a small subset of genes of interest[169], and the enrichment
of the whole exome of a patient[173], have been explored
successfully, helping to solve several unidentified mapped
NSHL loci[173–175], novel NSHL mutations in known deaf-
ness genes[176], and elucidating both novel[177–179] and previ-
ously identified[180] causative mutations of SHL. Many of
these discoveries were made with a limited proband num-
ber, but even more impressive are the numerous experi-
ments wherein only a single case was sequenced[173, 174, 176–179].
This method has most certainly transformed approaches to
studying Mendelian HL disorders[176], and there are high
expectations that it will facilitate the unraveling of volumi-
nous causative variants, genes, and genotype–phenotype
correlations. Indeed, an analysis of the success rate of whole
exome sequencing (WES) evaluates a 60–80% success rate
for Mendelian disorders[181].
Apart from being utilized for mutation discovery, WES
is currently being harnessed for routine diagnostic appli-
cations in the clinic. This option is being progressively
explored in syndromes with associated HL[182,183], and
in other genetic ailments[184,185]. Diagnosis of HL in the
near future is likely to make use of a combination of these
approaches, leaning towards WES with falling sequencing
costs.
G E N ET I C D I AG N O S I S A N D
EVA LUAT I O N: P RO P O S E D WO R K FL OW
Despite the significant advances in genomic applications
and a deeper understanding of the molecular genetics of
HL, identifying the precise genetic cause in a particular
individual still remains difficult. The first step should always
be a detailed clinical investigation to exclude syndromic
causes of deafness, followed by a thorough examination of
all available phenotypical clues for diagnosis. Here, we aim
to provide a manageable clinical approach to the genetic
diagnosis of HL.
Once a genetic etiology is suspected, a comprehensive
and detailed patient history must be collected. The prena-
tal, perinatal, and postnatal history of the patient should
be collected, and the developmental history of the patient
should also be explored in order to disclose any congenital
malformations, motor or perception dysfunctions that may
indicate vestibular defects or indicate SHL. A detailed fam-
ily history provides the principal indications of a genetic
cause, and must be astutely and cautiously approached.
A full understanding of the general health and hearing of
siblings, parents, grandparents, and other close relatives
must be gained in order to help define the inheritance pat-
tern within the family. Any additional information regard-
ing the onset, severity, and type of HL of affected members
must be recorded, preferably on the constructed pedigree.
It is also relevant to check for consanguinity and family eth-
nic background (especially if shared and/or from regions
known for founder effects). The genetic evaluation of fam-
ily members is of paramount importance in the diagnostic
process, and in assisting informed decisions regarding the
choice of genes to be molecularly tested for. These tests will
ultimately help validate their specific diagnosis.
Caution should be exercised if the patient’s HL is sus-
pected to have a genetic origin, but indicates no family
history. The phenotypical expression of affected family
members may be masked by reduced penetrance or variable
expression of a HL gene mutation, the family may have an
X-linked inherited disorder in which the parents shown no
phenotype, or the patient may possess a sporadic mutation.
Audiograms are important components of the assessment
process[19], and where feasible should be compared to those
of other affected family members.
G eno m ic A pplications in Audiological   Medicine   •   6 7 5
Subsequently, a physical examination is carried out and
should ideally exclude all syndromic causes of HL through
the identification of signature phenotypical features, in con-
junction with typical ancillary testing (to name several of
them: ophthalmology for Usher syndrome; TSH and per-
chlorate discharge test for Pendred syndrome; urinalysis
and renal ultrasound for Alport syndrome; EKG for Jervells
and Lange-Nielsens syndrome; CT scans for Mondini
malformation).
Genetic testing forms part of the assessment to help
confirm or rule out a specific genetic diagnosis. Once
syndromic HL has been ruled out, the clinician needs to
define the mode of inheritance according to the family
history. If it is recessive, the first step in mutational gene
analysis should be connexin gene testing, since GJB2 is the
most common cause of NSHL, with mutations in this gene
accounting for 50% of patients with autosomal recessive
HL[186], and 20% of all congenital HL[187]. A further argu-
ment for the routine screening of this gene is the unclassifi-
able clinical presentation of the HL. Since screening of this
gene is both fast and relatively inexpensive, incorporation
of its testing as part of a routine work-up in the diagnosis of
NSHL of unknown cause is warranted. If excluded, other
common mutations in the patient’s specific ethnic popu-
lation should be screened for. Mutations in specific genes
can be examined based on associated mutations within
the same ethnic groups, and one can make use of current
genomic tests. We suggest MPS at this point, especially
considering the falling prices in MPS. The take-home mes-
sage is that the medical workup is key in narrowing down
possible genetic causes, and guiding the choice of the most
appropriate genomic technique that is both cost-efficient
and results in a timely diagnosis.
G E N O M I C A P P R OAC H
TO   M A N AG E M E N T O F T H E
HEARING LOSS
Today, clinicians are being faced with an overload of data
regarding a vast number and variety of genes involved
in NSHL and SHL. This may, rightfully so, appear over-
whelming and impenetrable; however, learning to access
and manage this complex information is critical in the
application of the most up-to-date information to patient
care. Many reliable resources are available online that can
aid in the identification of genetic HL disorders and provide
information on the most current genetic tests[188], includ-
ing the Hereditary Hearing Loss Homepage and Deafness
Variation Database.
 N EWB O R N A N D A D U LT AU D I O L O G I C A L
A N D G E N ET I C S C R E E N I N G
Universal newborn hearing screening is recommended by
the U.S. Preventive Services Task Force (USPSTF; screen-
ing, http://www.uspreventiveser​vicestaskforce.org/uspstf/
uspsnbhr.htm​). In many countries worldwide,
newborn-screening is performed a day or two after birth,
before releasing the infant from the hospital[189,  190].
According to the criteria of the USPSTF, infants at
increased risk for HL, with negative results at the first
screening, require repeated screening by age 24–30 months.
The category of elevated risk for HL includes newborns
who spent more than two days in a neonatal intensive care
unit (NICU); infants suffering from congenital syndromes,
family history of SNHL, or syndromes including HL; and
newborns who were exposed to intrauterine infections.
Those with positive results on the first screening require
repeated hearing screening within three months[191,192], fol-
lowed by a full audiological evaluation if positive results
are repeated. The prevalence of congenital HL of 35 dB or
more, which requires intervention, is reported in Western
countries to be 2–3 per 1,000 newborns[193]. In developing
countries, the risk is higher due to environmental factors,
most of which have been eliminated in advanced countries,
such as congenital virus infections such as measles, rubella,
and cytomegalovirus, or social aspects such as increased
rates of consanguinity, which has crucial genetic conse-
quences. A congenital HL can be detected at birth by tran-
sitory evoked oto-acoustic emissions (TEOAE), and by
automated auditory brainstem response (ABR). TEOAE
evaluates cochlear performance and measures outer hair
cell function. ABR assesses retro-cochlear function of the
eighth nerve and the auditory brainstem. These tests offer
the possibility to detect HL early, during the first two to
three years of life, which are crucial for therapy and rehabili-
tation as they are considered the sensitive period of matura-
tion of the auditory pathway, which is almost completed by
two years of age[190]. Thus, an early diagnosis, both audiolog-
ical and genetic, of neonatal HL allows early intervention
and rehabilitation in the most effective time of auditory
development, leading in most cases to better language and
cognitive development[194]. In many cases, the effectiveness
of rehabilitation increases if it is based on genetic diagno-
sis, which allows maximal fitness to the patient’s specific
condition.
To conclude, audiological and genetic screenings com-
plement each other. In some cases, the type of audiogram
or HL, combined with the family history, provide the
direction for the molecular evaluation, as different genes
6 7 6   •   G eno m ics in C linical P ractice
will be screened for congenital profound HL compared
to late-onset, progressive, high-tone HL, or for recessive
SNHL compared to conductive dominant HL. In other
cases, learning about the mutation may predict the type of
HL, the severity and stability versus progression of the HL,
and enable optimal rehabilitation and adaptation to the
upcoming situation. One example is Usher syndrome, in
which some children are misdiagnosed as having congenital
NSHL, as the onset of retinitis pigmentosa is only in the
second decade of life. The earlier the molecular diagnosis
is made, the more intensive the auditory rehabilitation may
be, providing the best result to enable oral communication
after the onset of blindness.
At older ages, any child suspected of possible HL, because
of learning disabilities or speech delay, or those with recur-
rent otitis media, or because of parents’ concern, should be
tested for hearing loss[190]. At ages two to four, children are
evaluated by “Play audiometry,” a behavioral test to obtain
auditory thresholds for speech and frequency-specific stim-
uli. In this test, the child is reacting to each stimulus he hears
by putting a block into a box. Children over four are usually
tested by conventional behavioral auditory evaluation, in
which they are instructed to raise their hand in response to
stimuli, the same technique as used for adults[195].
For adults, the problem arises mostly for elderly people,
as at younger ages people are aware of any deterioration in
hearing and seek audiological evaluation privately, or they
are screened through programs of occupational medicine.
For the elderly, in most cases, in order to diagnose HL, it
is enough to ask them during their physical examination
if they have a hearing problem[196]. A  full otolaryngology
consult and audiological assessment should follow positive
results on hearing screen or when a patient expresses con-
cern about HL, to differentiate between types of HL; for
example, conductive versus SNHL, degree of HL, type of
audiogram, and symmetry of HL. Following audiological
evaluation, molecular testing should be performed. Up to
now, screening of GJB2 was conducted first, and if excluded,
followed by screening for other relevant genes, determined
according to type of HL, age of onset, patient ethnicity, and
family history. However, since hereditary HL is considered
the most heterogeneous Mendelian condition[197], it needs
better DNA diagnostic tools based on the analysis of all
genes, or at least, of all genes known to be associated with
deafness. Recently, high-throughput advanced techniques,
including deep sequencing, were implemented to screen
known deafness genes in a single test[169,198–200], thus increas-
ing the chance of detecting the causative mutation even in
a single proband; but this is not able to solve all inherited
cases, leading to the conclusion that the next step should
include all genes in the genome. Revealing the genetic cause
leads to a more accurate prognosis of the disease, might con-
tribute to better treatment and rehabilitation by helping the
clinician choose the best hearing aid or other intervention
needed (e.g., an operation in case of otosclerosis), it was
proven to play a role in parental guilt-relief[201], and it can
lead to improved quality of life. In addition, knowing the
etiology can help the patient prevent further deterioration
of the hearing by avoiding environmental factors known
to affect HL, as described above for the complex forms of
HL, including mitochondrial deafness, NIHL, ARHL,
and DIHL. Moreover, knowing the causative mutation has
important implications for the entire family regarding pre-
vention and genetic counseling.
F U T U R E P RO S P EC T S
Advances in the evaluation of patients with hereditary HL
have been remarkable in recent years. The field has gone
from the discovery of mutations in one major gene for con-
genital HL, to mutations in over 100 genes associated with
rare forms of HL. The diagnostic value has been enhanced
by the research defining the mechanisms of the mutations.
For example, finding a mutation in the SYNE4 gene led to
diagnostics in two families, as well as to a study of a mouse
model for Syne4-deafness, which determined that a change
in the position of the nucleus is associated with the HL[202].
The future prospects of hereditary HL are bright, since the
number of genes and mutations will continue to increase as
the technology advances at a breathtaking rate. Eventually,
the availability of reliable diagnostics, coupled with an
understanding of the mechanisms causing HL in the
patients, will help lead us toward productive therapeutics.
REFERENCES
1. Morton, C.C. & Nance, W.E. Newborn hearing screening--a silent
revolution. N Engl J Med 354, 2151–2164 (2006).
2. Allen, G.E. Mendel and modern genetics:  the legacy for today.
Endeavour 27, 63–68 (2003).
3. Kelsell, D.P., Dunlop, J., Stevens, H.P., Lench, N.J., Liang, J.N.,
Parry, G., et al. Connexin 26 mutations in hereditary non-syndromic
sensorineural deafness. Nature 387, 80–83 (1997).
4. Venter, J.C., Adams, M.D., Myers, E.W., Li, P.W., Mural, R.J.,
Sutton, G.G., et al. The sequence of the human genome. Science 291,
1304–1351 (2001).
5. Lander, E.S., Linton, L.M., Birren, B., Nusbaum, C., Zody, M.C.,
Baldwin, J., et  al. Initial sequencing and analysis of the human
genome. Nature 409, 860–921 (2001).
6. Shearer, A.E., Hildebrand, M.S., Sloan, C.M., & Smith, R.J.
Deafness in the genomics era. Hear Res 282, 1–9 (2011).
7. World Health Organization. Deafness and Hearing Impairment,
(2012). Available at http://www.who.int/​mediacentre/factsheets/​.
G eno m ic A pplications in Audiological   Medicine   •   6 7 7
8. Agrawal, Y., Platz, E.A., & Niparko, J.K. Prevalence of hear-
ing loss and differences by demographic characteristics among
US adults:  data from the National Health and Nutrition
Examination Survey, 1999–2004. Arch Intern Med 168,
1522–1530 (2008).
9. Margulies, M., Egholm, M., Altman, W.E., Attiya, S., Bader, J.S.,
Bemben, L.A., et  al. Genome sequencing in microfabricated
high-density picolitre reactors. Nature 437, 376–380 (2005).
10. ASHA Guidelines for screening for hearing impairment and middle
ear disorders. ASHA 31, 71–77 (1989).
11. Kalatzis, V. & Petit, C. The fundamental and medical impacts of
recent progress in research on hereditary hearing loss. Hum Mol
Genet 7, 1589–1597 (1998).
12. Mehra, S., Eavey, R.D., & Keamy, D.G., Jr. The epidemiology of
hearing impairment in the United States: newborns, children, and
adolescents. Otolaryngol Head Neck Surg 140, 461–472 (2009).
13. Ruben, R.J. Effectiveness and efficacy of early detection of hearing
impairment in children. Acta Otolaryngol Suppl 482, 127–131; dis-
cussion 132–125 (1991).
14. Moeller, M.P., Osberger, M.J., & Eccarius, M. Language and learn-
ing skills of hearing-impaired students. Receptive language skills.
ASHA Monogr 41–53 (1986).
15. Ciorba, A., Bianchini, C., Pelucchi, S., & Pastore, A. The impact of
hearing loss on the quality of life of elderly adults. Clin Interv Aging
7, 159–163 (2012).
16. Dalton, D.S., Cruickshanks, K.J., Klein, B.E., Klein, R., Wiley, T.L.,
& Nondahl, D.M. The impact of hearing loss on quality of life in
older adults. Gerontologist 43, 661–668 (2003).
17. Sininger, Y.S., Martinez, A., Eisenberg, L., Christensen, E., Grimes,
A., & Hu, J. Newborn hearing screening speeds diagnosis and access
to intervention by 20–25  months. J Am Acad Audiol 20, 49–57
(2009).
18. Richardson, G.P., de Monvel, J.B., & Petit, C. How the genetics
of deafness illuminates auditory physiology. Annu Rev Physiol 73,
311–334 (2011).
19. Arnos, K.S. The implications of genetic testing for deafness. Ear
Hear 24, 324–331 (2003).
20. Withrow, K.A., Tracy, K.A., Burton, S.K., Norris, V.W., Maes, H.H.,
Arnos, K.S., et al. Impact of genetic advances and testing for hearing
loss:  results from a national consumer survey. Am J Med Genet A
149A, 1159–1168 (2009).
21. Kochhar, A., Hildebrand, M.S., & Smith, R.J. Clinical aspects of
hereditary hearing loss. Genet Med 9, 393–408 (2007).
22. Hilgert, N., Smith, R.J., & Van Camp, G. Forty-six genes causing
nonsyndromic hearing impairment: which ones should be analyzed
in DNA diagnostics? Mutat Res 681, 189–196 (2009).
23. Morton, C.C. Genetics, genomics and gene discovery in the audi-
tory system. Hum Mol Genet 11, 1229–1240 (2002).
24. Willecke, K., Eiberger, J., Degen, J., Eckardt, D., Romualdi, A.,
Guldenagel, M., et al. Structural and functional diversity of connexin
genes in the mouse and human genome. Biol Chem 383, 725–737
(2002).
25. Denoyelle, F., Marlin, S., Weil, D., Moatti, L., Chauvin, P.,
Garabedian, E.N., et  al. Clinical features of the prevalent
form of childhood deafness, DFNB1, due to a connexin-26
gene defect:  implications for genetic counselling. Lancet 353,
1298–1303. (1999).
26. Chan, D.K., Schrijver, I., & Chang, K.W. Connexin-26-associated
deafness:  phenotypic variability and progression of hearing loss.
Genet Med 12, 174–181 (2010).
27. Snoeckx, R.L., Huygen, P.L., Feldmann, D., Marlin, S., Denoyelle,
F., Waligora, J., et al. GJB2 mutations and degree of hearing loss: a
multicenter study. Am J Hum Genet 77, 945–957 (2005).
28. Cheng, X., Li, L., Brashears, S., Morlet, T., Ng, S.S., Berlin, C., et al.
Connexin 26 variants and auditory neuropathy/dys-synchrony
among children in schools for the deaf. Am J Med Genet A 139,
13–18 (2005).
 29. Iossa, S., Marciano, E., & Franze, A. GJB2 Gene Mutations in
Syndromic Skin Diseases with Sensorineural Hearing Loss. Curr
Genomics 12, 475–785 (2011).
30. Martinez, A.D., Acuna, R., Figueroa, V., Maripillan, J., & Nicholson,
B. Gap-junction channels dysfunction in deafness and hearing loss.
Antioxid Redox Signal 11, 309–322 (2009).
31. Feldmann, D., Le Marechal, C., Jonard, L., Thierry, P., Czajka, C.,
Couderc, R., et al. A new large deletion in the DFNB1 locus causes
nonsyndromic hearing loss. Eur J Med Genet 52, 195–200 (2009).
32. del Castillo, I., Villamar, M., Moreno-Pelayo, M.A., del Castillo,
F.J., Alvarez, A., Telleria, D., et al. A deletion involving the connexin
30 gene in nonsyndromic hearing impairment. N Engl J Med 346,
243–249 (2002).
33. del Castillo, F.J., Rodriguez-Ballesteros, M., Alvarez, A., Hutchin,
T., Leonardi, E., de Oliveira, C.A., et al. A novel deletion involving
the connexin-30 gene, del(GJB6-d13s1854), found in trans with
mutations in the GJB2 gene (connexin-26) in subjects with DFNB1
non-syndromic hearing impairment. J Med Genet 42, 588–594
(2005).
34. Lerer, I., Sagi, M., Ben-Neriah, Z., Wang, T., Levi, H., & Abeliovich,
D. A deletion mutation in GJB6 cooperating with a GJB2 mutation
in trans in non-syndromic deafness: A novel founder mutation in
Ashkenazi Jews. Hum Mutat 18, 460. (2001).
35. Albert, S., Blons, H., Jonard, L., Feldmann, D., Chauvin, P.,
Loundon, N., et al. SLC26A4 gene is frequently involved in non-
syndromic hearing impairment with enlarged vestibular aqueduct in
Caucasian populations. Eur J Hum Genet 14, 773–779 (2006).
36. Royaux, I.E., Suzuki, K., Mori, A., Katoh, R., Everett, L.A., Kohn,
L.D., et al. Pendrin, the protein encoded by the Pendred syndrome
gene (PDS), is an apical porter of iodide in the thyroid and is
regulated by thyroglobulin in FRTL-5 cells. Endocrinology. 141,
839–845. (2000).
37. Dossena, S., Rodighiero, S., Vezzoli, V., Nofziger, C., Salvioni, E.,
Boccazzi, M., et  al. Functional characterization of wild-type and
mutated pendrin (SLC26A4), the anion transporter involved in
Pendred syndrome. J Mol Endocrinol 43, 93–103 (2009).
38. Dror, A.A., Brownstein, Z., & Avraham, K.B. Integration of human
and mouse genetics reveals pendrin function in hearing and deaf-
ness. Cell Physiol Biochem 28, 535–544 (2011).
39. Everett, L.A., Morsli, H., Wu, D.K., & Green, E.D. Expression pat-
tern of the mouse ortholog of the Pendred's syndrome gene (Pds)
suggests a key role for pendrin in the inner ear. Proc Natl Acad Sci U
S A 96, 9727–9732 (1999).
40. Kim, H.M. & Wangemann, P. Failure of fluid absorption in the endo-
lymphatic sac initiates cochlear enlargement that leads to deafness in
mice lacking pendrin expression. PLoS One 5, e14041 (2010).
41. Pera, A., Dossena, S., Rodighiero, S., Gandia, M., Botta, G., Meyer,
G., et al. Functional assessment of allelic variants in the SLC26A4
gene involved in Pendred syndrome and nonsyndromic EVA. Proc
Natl Acad Sci U S A 105, 18608–18613 (2008).
42. de Kok, Y.J., van der Maarel, S.M., Bitner-Glindzicz, M., Huber,
I., Monaco, A.P., Malcolm, S., et al. Association between X-linked
mixed deafness and mutations in the POU domain gene POU3F4.
Science 267, 685–688 (1995).
43. Bitner-Glindzicz, M., Turnpenny, P., Hoglund, P., Kaariainen, H.,
Sankila, E.M., van der Maarel, S.M., et  al. Further mutations in
Brain 4 (POU3F4) clarify the phenotype in the X-linked deafness,
DFN3. Hum Mol Genet 4, 1467–1469 (1995).
44. de Kok, Y.J., Vossenaar, E.R., Cremers, C.W., Dahl, N., Laporte,
J., Hu, L.J., et al. Identification of a hot spot for microdeletions in
patients with X-linked deafness type 3 (DFN3) 900 kb proximal to
the DFN3 gene POU3F4. Hum Mol Genet 5, 1229–1235 (1996).
45. Marlin, S., Moizard, M.P., David, A., Chaissang, N., Raynaud, M.,
Jonard, L., et al. Phenotype and genotype in females with POU3F4
mutations. Clin Genet 76, 558–563 (2009).
46. Ealy, M. & Smith, R.J. Otosclerosis. Adv Otorhinolaryngol 70,
122–129 (2011).
6 7 8   •   G eno m ics in C linical P ractice
47. Gordon, M.A. The genetics of otosclerosis: a review. Am J Otol 10,
426–438 (1989).
48. Menger, D.J. & Tange, R.A. The aetiology of otosclerosis: a review
of the literature. Clin Otolaryngol Allied Sci 28, 112–120 (2003).
49. Zhao, F., Wada, H., Koike, T., Ohyama, K., Kawase, T., & Stephens,
D. Middle ear dynamic characteristics in patients with otosclerosis.
Ear Hear 23, 150–158 (2002).
50. Gros, A., Vatovec, J., & Sereg-Bahar, M. Histologic changes on sta-
pedial footplate in otosclerosis. Correlations between histologic
activity and clinical findings. Otol Neurotol 24, 43–47 (2003).
51. Chole, R.A. & McKenna, M. Pathophysiology of otosclerosis. Otol
Neurotol 22, 249–257 (2001).
52. Robinson, M. Juvenile otosclerosis. A  20-year study. Ann Otol
Rhinol Laryngol 92, 561–565 (1983).
53. Lescanne, E., Bakhos, D., Metais, J.P., Robier, A., & Moriniere, S.
Otosclerosis in children and adolescents: a clinical and CT-scan sur-
vey with review of the literature. Int J Pediatr Otorhinolaryngol 72,
147–152 (2008).
54. Ramsay, H.A. & Linthicum, F.H., Jr. Mixed hearing loss in otoscle-
rosis: indication for long-term follow-up. Am J Otol 15, 536–539
(1994).
55. Schrauwen, I., Weegerink, N.J., Fransen, E., Claes, C., Pennings,
R.J., Cremers, C.W., et al. A new locus for otosclerosis, OTSC10,
maps to chromosome 1q41–44. Clin Genet 79, 495–497 (2011).
56. Schrauwen, I. & Van Camp, G. The etiology of otosclerosis: a com-
bination of genes and environment. Laryngoscope 120, 1195–1202
(2010).
57. Schrauwen, I., Ealy, M., Huentelman, M.J., Thys, M., Homer, N.,
Vanderstraeten, K., et al. A genome-wide analysis identifies genetic
variants in the RELN gene associated with otosclerosis. Am J Hum
Genet 84, 328–338 (2009).
58. El-Badry, M.M. & McFadden, S.L. Evaluation of inner hair cell and
nerve fiber loss as sufficient pathologies underlying auditory neu-
ropathy. Hear Res 255, 84–90 (2009).
59. Zeng, F.G., Kong, Y.Y., Michalewski, H.J., & Starr, A. Perceptual
consequences of disrupted auditory nerve activity. J Neurophysiol
93, 3050–3063 (2005).
60. Sanyelbhaa Talaat, H., Kabel, A.H., Samy, H., & Elbadry, M.
Prevalence of auditory neuropathy (AN) among infants and
young children with severe to profound hearing loss. Int J Pediatr
Otorhinolaryngol 73, 937–939 (2009).
61. Manchaiah, V.K., Zhao, F., Danesh, A.A., & Duprey, R. The genetic
basis of auditory neuropathy spectrum disorder (ANSD). Int J
Pediatr Otorhinolaryngol 75, 151–158 (2011).
62. Roush, P., Frymark, T., Venediktov, R., & Wang, B. Audiologic man-
agement of auditory neuropathy spectrum disorder in children:  a
systematic review of the literature. Am J Audiol 20, 159–170 (2011).
63. Madden, C., Rutter, M., Hilbert, L., Greinwald, J.H., Jr., & Choo,
D.I. Clinical and audiological features in auditory neuropathy. Arch
Otolaryngol Head Neck Surg 128, 1026–1030 (2002).
64. Santarelli, R. Information from cochlear potentials and genetic
mutations helps localize the lesion site in auditory neuropathy.
Genome Med 2, 91 (2010).
65. Del Castillo, F.J. & Del Castillo, I. Genetics of isolated auditory neu-
ropathies. Front Biosci (Landmark Ed) 17, 1251–1265 (2012).
66. Yasunaga, S., Grati, M., Cohen-Salmon, M., El-Amraoui, A.,
Mustapha, M., Salem, N., et  al. A mutation in OTOF, encoding
otoferlin, a FER-1-like protein, causes DFNB9, a nonsyndromic
form of deafness. Nat Genet 21, 363–369 (1999).
67. Roux, I., Safieddine, S., Nouvian, R., Grati, M., Simmler, M.C.,
Bahloul, A., et  al. Otoferlin, defective in a human deafness form,
is essential for exocytosis at the auditory ribbon synapse. Cell 127,
277–289 (2006).
68. Delmaghani, S., del Castillo, F.J., Michel, V., Leibovici, M., Aghaie,
A., Ron, U., et al. Mutations in the gene encoding pejvakin, a newly
identified protein of the afferent auditory pathway, cause DFNB59
auditory neuropathy. Nat Genet 38, 770–778 (2006).
 69. Schwander, M., Sczaniecka, A., Grillet, N., Bailey, J.S., Avenarius,
M., Najmabadi, H., et al. A forward genetics screen in mice identi-
fies recessive deafness traits and reveals that pejvakin is essential for
outer hair cell function. J Neurosci 27, 2163–2175 (2007).
70. Ebermann, I., Walger, M., Scholl, H.P., Charbel Issa, P., Luke, C.,
Nurnberg, G., et  al. Truncating mutation of the DFNB59 gene
causes cochlear hearing impairment and central vestibular dysfunc-
tion. Hum Mutat 28, 571–577 (2007).
71. Collin, R.W., Kalay, E., Oostrik, J., Caylan, R., Wollnik, B., Arslan,
S., et  al. Involvement of DFNB59 mutations in autosomal reces-
sive nonsyndromic hearing impairment. Hum Mutat 28, 718–723
(2007).
72. Hashemzadeh Chaleshtori, M., Simpson, M.A., Farrokhi, E., Dolati,
M., Hoghooghi Rad, L., Amani Geshnigani, S., et al. Novel muta-
tions in the pejvakin gene are associated with autosomal recessive
non-syndromic hearing loss in Iranian families. Clin Genet 72,
261–263 (2007).
73. Shahin, H., Walsh, T., Rayyan, A.A., Lee, M.K., Higgins, J.,
Dickel, D., et al. Five novel loci for inherited hearing loss mapped
by SNP-based homozygosity profiles in Palestinian families. Eur J
Hum Genet 18, 407–413 (2010).
74. Schoen, C.J., Emery, S.B., Thorne, M.C., Ammana, H.R., Sliwerska,
E., Arnett, J., et  al. Increased activity of Diaphanous homolog 3
(DIAPH3)/diaphanous causes hearing defects in humans with
auditory neuropathy and in Drosophila. Proc Natl Acad Sci U S A
107, 13396–13401 (2010).
75. Wang, Q.J., Li, Q.Z., Rao, S.Q., Lee, K., Huang, X.S., Yang, W.Y.,
et al. AUNX1, a novel locus responsible for X linked recessive audi-
tory and peripheral neuropathy, maps to Xq23–27.3. j Med Genet
43, e33 (2006).
76. Wang, Q., Li, R., Zhao, H., Peters, J.L., Liu, Q., Yang, L., et  al.
Clinical and molecular characterization of a Chinese patient with
auditory neuropathy associated with mitochondrial 12S rRNA
T1095C mutation. Am J Med Genet A 133A, 27–30 (2005).
77. Thyagarajan, D., Bressman, S., Bruno, C., Przedborski, S., Shanske,
S., Lynch, T., et al. A novel mitochondrial 12SrRNA point muta-
tion in parkinsonism, deafness, and neuropathy. Ann Neurol 48,
730–736 (2000).
78. Chennupati, S.K., Levi, J., Loftus, P., Jornlin, C., Morlet, T., &
O'Reilly, R.C. Hearing loss in children with mitochondrial disor-
ders. Int J Pediatr Otorhinolaryngol 75, 1519–1524 (2011).
79. Fischel-Ghodsian, N. Mitochondrial deafness. Ear Hear 24,
303–313 (2003).
80. Zhao, H., Li, R., Wang, Q., Yan, Q., Deng, J.H., Han, D., et  al.
Maternally inherited aminoglycoside-induced and nonsyndromic
deafness is associated with the novel C1494T mutation in the mito-
chondrial 12S rRNA gene in a large Chinese family. Am J Hum
Genet 74, 139–152 (2004).
81. Tekin, M., Duman, T., Bogoclu, G., Incesulu, A., Comak, E., Fitoz,
S., et  al. Frequency of mtDNA A1555G and A7445G mutations
among children with prelingual deafness in Turkey. Eur J Pediatr
162, 154–158 (2003).
82. Hutchin, T.P., Navarro-Coy, N.C., Van Camp, G., Tiranti, V.,
Zeviani, M., Schuelke, M., et  al. Multiple origins of the mtDNA
7472insC mutation associated with hearing loss and neurological
dysfunction. Eur J Hum Genet 9, 385–387 (2001).
83. Hutchin, T.P., Parker, M.J., Young, I.D., Davis, A.C., Pulleyn,
L.J., Deeble, J., et  al. A novel mutation in the mitochondrial
tRNA(Ser(UCN)) gene in a family with non-syndromic sensori-
neural hearing impairment. j Med Genet 37, 692–694 (2000).
84. Ishikawa, K., Tamagawa, Y., Takahashi, K., Kimura, H., Kusakari, J.,
Hara, A., et al. Nonsyndromic hearing loss caused by a mitochon-
drial T7511C mutation. Laryngoscope 112, 1494–1499 (2002).
85. Guan, M.X. Mitochondrial 12S rRNA mutations associated with
aminoglycoside ototoxicity. Mitochondrion 11, 237–245 (2011).
86. Zheng, J., Ji, Y., & Guan, M.X. Mitochondrial tRNA mutations
associated with deafness. Mitochondrion 12, 406–413 (2012).
G eno m ic A pplications in Audiological   Medicine   •   6 7 9
87. Toriello, H., Reardon, W., & Gorlin, R.J. Hereditary Hearing Loss
and Its Syndromes. Oxford, UK: Oxford University Press, 2004.
88. Friedman, T.B., Schultz, J.M., Ben-Yosef, T., Pryor, S.P., Lagziel, A.,
Fisher, R.A., et al. Recent advances in the understanding of syn-
dromic forms of hearing loss. Ear Hear 24, 289–302 (2003).
89. Millan, J.M., Aller, E., Jaijo, T., Blanco-Kelly, F., Gimenez-Pardo,
A., & Ayuso, C. An update on the genetics of usher syndrome. J
Ophthalmol 2011, 417217 (2011).
90. Boughman, J.A., Vernon, M., & Shaver, K.A. Usher syndrome: def-
inition and estimate of prevalence from two high-risk populations.
J Chronic Dis 36, 595–603 (1983).
91. Cohen, M., Bitner-Glindzicz, M., & Luxon, L. The changing face
of Usher syndrome: clinical implications. Int J Audiol 46, 82–93
(2007).
92. Smith, R.J., Berlin, C.I., Hejtmancik, J.F., Keats, B.J., Kimberling,
W.J., Lewis, R.A., et  al. Clinical diagnosis of the Usher syn-
dromes. Usher Syndrome Consortium. Am J Med Genet 50,
32–38 (1994).
93. Sadeghi, M., Cohn, E.S., Kimberling, W.J., Tranebjaerg, L., &
Moller, C. Audiological and vestibular features in affected subjects
with USH3:  a genotype/phenotype correlation. Int J Audiol 44,
307–316 (2005).
94. Pakarinen, L., Karjalainen, S., Simola, K.O., Laippala, P., &
Kaitalo, H. Usher's syndrome type 3 in Finland. Laryngoscope 105,
613–617 (1995).
95. Ness, S.L., Ben-Yosef, T., Bar-Lev, A., Madeo, A.C., Brewer, C.C.,
Avraham, K.B., et al. Genetic homogeneity and phenotypic vari-
ability among Ashkenazi Jews with Usher syndrome type III. j Med
Genet 40, 767–772 (2003).
96. Schrijver, I. Hereditary non-syndromic sensorineural hearing
loss:  transforming silence to sound. J Mol Diagn 6, 275–284
(2004).
97. Dossena, S., Nofziger, C., Brownstein, Z., Kanaan, M., Avraham,
K.B., & Paulmichl, M. Functional characterization of pendrin
mutations found in the Israeli and Palestinian populations. Cell
Physiol Biochem 28, 477–484 (2011).
98. Alasti, F., Van Camp, G., & Smith, R.J.H. Pendred syndrome/
DFNB4. (1993). In: R.A. Pagon, Adam M.P., Ardinger H.H., Bird
T.D., Dolan C.R., Fong C.T., Smith R.J.H., and Stephens K., edi-
tors. Source GeneReviews, [Internet]. Seattle (WA): University of
Washington, Seattle; 1993–2014.
99. Neyroud, N., Tesson, F., Denjoy, I., Leibovici, M., Donger, C.,
Barhanin, J., et al. A novel mutation in the potassium channel gene
KVLQT1 causes the Jervell and Lange-Nielsen cardioauditory
syndrome. Nat Genet 15, 186–189 (1997).
100. Naik, A. Long QT syndrome revisited. J Assoc Physicians India 55
Suppl, 58–61 (2007).
101. Winbo, A., Stattin, E.L., Diamant, U.B., Persson, J., Jensen, S.M.,
& Rydberg, A. Prevalence, mutation spectrum, and cardiac phe-
notype of the Jervell and Lange-Nielsen syndrome in Sweden.
Europace 14, 1799–1806 (2012).
102. Tyson, J., Tranebjaerg, L., Bellman, S., Wren, C., Taylor, J.F.,
Bathen, J., et al. IsK and KvLQT1: mutation in either of the two
subunits of the slow component of the delayed rectifier potassium
channel can cause Jervell and Lange-Nielsen syndrome. Hum Mol
Genet 6, 2179–2185 (1997).
103. Schulze-Bahr, E., Wang, Q., Wedekind, H., Haverkamp, W.,
Chen, Q., Sun, Y., et  al. KCNE1 mutations cause jervell and
Lange-Nielsen syndrome. Nat Genet 17, 267–268 (1997).
104. Tyson, J., Tranebjaerg, L., McEntagart, M., Larsen, L.A.,
Christiansen, M., Whiteford, M.L., et al. Mutational spectrum in
the cardioauditory syndrome of Jervell and Lange-Nielsen. Hum
Genet 107, 499–503 (2000).
105. Schwartz, P.J., Spazzolini, C., Crotti, L., Bathen, J., Amlie, J.P.,
Timothy, K., et al. The Jervell and Lange-Nielsen syndrome: natu-
ral history, molecular basis, and clinical outcome. Circulation 113,
783–790 (2006).
 106. Eddy, C.A., MacCormick, J.M., Chung, S.K., Crawford, J.R., Love,
D.R., Rees, M.I., et  al. Identification of large gene deletions and
duplications in KCNQ1 and KCNH2 in patients with long QT
syndrome. Heart Rhythm 5, 1275–1281 (2008).
107. Pingault, V., Ente, D., Dastot-Le Moal, F., Goossens, M., Marlin,
S., & Bondurand, N. Review and update of mutations causing
Waardenburg syndrome. Hum Mutat 31, 391–406 (2010).
108. Gleeson, M.J. Alport's syndrome: audiological manifestations and
implications. J Laryngol Otol 98, 449–465 (1984).
109. Hertz, J.M. Alport syndrome. Molecular genetic aspects. Dan Med
Bull 56, 105–152 (2009).
110. Mochizuki, T., Lemmink, H.H., Mariyama, M., Antignac, C.,
Gubler, M.C., Pirson, Y., et al. Identification of mutations in the
alpha 3(IV) and alpha 4(IV) collagen genes in autosomal recessive
Alport syndrome. Nat Genet 8, 77–81 (1994).
111. Barker, D.F., Hostikka, S.L., Zhou, J., Chow, L.T., Oliphant, A.R.,
Gerken, S.C., et al. Identification of mutations in the COL4A5 col-
lagen gene in Alport syndrome. Science 248, 1224–1227 (1990).
112. Cosgrove, D., Samuelson, G., & Pinnt, J. Immunohistochemical
localization of basement membrane collagens and associated pro-
teins in the murine cochlea. Hear Res 97, 54–65 (1996).
113. Phillips, S.L., Gordon-Salant, S., Fitzgibbons, P.J., &
Yeni-Komshian, G. Frequency and temporal resolution in elderly
listeners with good and poor word recognition. J Speech Lang Hear
Res 43, 217–228 (2000).
114. Huang, Q. & Tang, J. Age-related hearing loss or presbycusis. Eur
Arch Otorhinolaryngol 267, 1179–1191 (2010).
115. Liu, X.Z. & Yan, D. Ageing and hearing loss. J Pathol 211, 188–197
(2007).
116. Schultz, J.M., Yang, Y., Caride, A.J., Filoteo, A.G., Penheiter,
A.R., Lagziel, A., et  al. Modification of human hearing loss by
plasma-membrane calcium pump PMCA2. N Engl J Med 352,
1557–1564 (2005).
117. Michikawa, Y., Mazzucchelli, F., Bresolin, N., Scarlato, G., &
Attardi, G. Aging-dependent large accumulation of point muta-
tions in the human mtDNA control region for replication. Science
286, 774–779 (1999).
118. Kujoth, G.C., Hiona, A., Pugh, T.D., Someya, S., Panzer, K.,
Wohlgemuth, S.E., et al. Mitochondrial DNA mutations, oxidative
stress, and apoptosis in mammalian aging. Science 309, 481–484
(2005).
119. Fischel-Ghodsian, N., Bykhovskaya, Y., Taylor, K., Kahen, T.,
Cantor, R., Ehrenman, K., et al. Temporal bone analysis of patients
with presbycusis reveals high frequency of mitochondrial muta-
tions. Hear Res 110, 147–154 (1997).
120. Seidman, M.D., Khan, M.J., Bai, U., Shirwany, N., & Quirk,
W.S. Biologic activity of mitochondrial metabolites on aging and
age-related hearing loss. Am J Otol 21, 161–167 (2000).
121. Seidman, M.D., Bai, U., Khan, M.J., & Quirk, W.S. Mitochondrial
DNA deletions associated with aging and presbyacusis. Arch
Otolaryngol Head Neck Surg 123, 1039–1045 (1997).
122. Bai, U., Seidman, M.D., Hinojosa, R., & Quirk, W.S. Mitochondrial
DNA deletions associated with aging and possibly presbycusis: a
human archival temporal bone study. Am J Otol 18, 449–453
(1997).
123. Dai, P., Yang, W., Jiang, S., Gu, R., Yuan, H., Han, D., et  al.
Correlation of cochlear blood supply with mitochondrial DNA
common deletion in presbyacusis. Acta Otolaryngol 124, 130–136
(2004).
124. Seidman, M.D., Bai, U., Khan, M.J., Murphy, M.J., Quirk, W.S.,
Castora, F.L., et al. Association of mitochondrial DNA deletions
and cochlear pathology:  a molecular biologic tool. Laryngoscope
106, 777–783 (1996).
125. Trifunovic, A., Wredenberg, A., Falkenberg, M., Spelbrink, J.N.,
Rovio, A.T., Bruder, C.E., et al. Premature ageing in mice express-
ing defective mitochondrial DNA polymerase. Nature 429,
417–423 (2004).
6 8 0   •   G eno m ics in C linical P ractice
126. Crawley, B.K. & Keithley, E.M. Effects of mitochondrial muta-
tions on hearing and cochlear pathology with age. Hear Res 280,
201–208 (2011).
127. Manwaring, N., Jones, M.M., Wang, J.J., Rochtchina, E., Howard,
C., Newall, P., et  al. Mitochondrial DNA haplogroups and
age-related hearing loss. Arch Otolaryngol Head Neck Surg 133,
929–933 (2007).
128. Kivisild, T., Shen, P., Wall, D.P., Do, B., Sung, R., Davis, K., et al.
The role of selection in the evolution of human mitochondrial
genomes. Genetics 172, 373–387 (2006).
129. Gorlov, I.P., Kimmel, M., & Amos, C.I. Strength of the purify-
ing selection against different categories of the point mutations
in the coding regions of the human genome. Hum Mol Genet 15,
1143–1150 (2006).
130. Bonneux, S., Fransen, E., Van Eyken, E., Van Laer, L., Huyghe, J.,
Van de Heyning, P., et al. Inherited mitochondrial variants are not
a major cause of age-related hearing impairment in the European
population. Mitochondrion 11, 729–734 (2011).
131. DeStefano, A.L., Gates, G.A., Heard-Costa, N., Myers, R.H., &
Baldwin, C.T. Genomewide linkage analysis to presbycusis in the
Framingham Heart Study. Arch Otolaryngol Head Neck Surg 129,
285–289 (2003).
132. Garringer, H.J., Pankratz, N.D., Nichols, W.C., & Reed, T. Hearing
impairment susceptibility in elderly men and the DFNA18 locus.
Arch Otolaryngol Head Neck Surg 132, 506–510 (2006).
133. Unal, M., Tamer, L., Dogruer, Z.N., Yildirim, H., Vayisoglu, Y.,
& Camdeviren, H. N-acetyltransferase 2 gene polymorphism and
presbycusis. Laryngoscope 115, 2238–2241 (2005).
134. Van Eyken, E., Van Laer, L., Fransen, E., Topsakal, V., Lemkens, N.,
Laureys, W., et al. KCNQ4: a gene for age-related hearing impair-
ment? Hum Mutat 27, 1007–1016 (2006).
135. Van Laer, L., Van Eyken, E., Fransen, E., Huyghe, J.R., Topsakal,
V., Hendrickx, J.J., et  al. The grainyhead like 2 gene (GRHL2),
alias TFCP2L3, is associated with age-related hearing impairment.
Hum Mol Genet 17, 159–169 (2008).
136. Alberti, P.W. Noise, the most ubiquitous pollutant. Noise Health 1,
3–5 (1998).
137. Melnick, W., Industrial hearing conservation. In Handbook of
Clinical Audiology. J. Katz, editor. Baltimore, MD:  Williams &
Wilkins; 1985.
138. Abreu-Silva, R.S., Rincon, D., Horimoto, A.R., Sguillar, A.P.,
Ricardo, L.A., Kimura, L., et al. The search of a genetic basis for
noise-induced hearing loss (NIHL). Ann Hum Biol 38, 210–218
(2011).
139. Henderson, D., Bielefeld, E.C., Harris, K.C., & Hu, B.H. The role
of oxidative stress in noise-induced hearing loss. Ear Hear 27, 1–19
(2006).
140. Jacono, A.A., Hu, B., Kopke, R.D., Henderson, D., Van De Water,
T.R., & Steinman, H.M. Changes in cochlear antioxidant enzyme
activity after sound conditioning and noise exposure in the chin-
chilla. Hear Res 117, 31–38 (1998).
141. Board, P.G. Biochemical genetics of glutathione-S-transferase in
man. Am J Hum Genet 33, 36–43 (1981).
142. Pemble, S., Schroeder, K.R., Spencer, S.R., Meyer, D.J., Hallier,
E., Bolt, H.M., et  al. Human glutathione S-transferase theta
(GSTT1):  cDNA cloning and the characterization of a genetic
polymorphism. Biochem J 300 ( Pt 1), 271–276 (1994).
143. Rabinowitz, P.M., Pierce Wise, J., Sr., Hur Mobo, B., Antonucci,
P.G., Powell, C., & Slade, M. Antioxidant status and hearing func-
tion in noise-exposed workers. Hear Res 173, 164–171 (2002).
144. Ohlemiller, K.K., McFadden, S.L., Ding, D.L., Lear, P.M., & Ho,
Y.S. Targeted mutation of the gene for cellular glutathione peroxi-
dase (Gpx1) increases noise-induced hearing loss in mice. J Assoc
Res Otolaryngol 1, 243–254 (2000).
145. Ohlemiller, K.K., McFadden, S.L., Ding, D.L., Flood, D.G.,
Reaume, A.G., Hoffman, E.K., et  al. Targeted deletion of the
cytosolic Cu/Zn-superoxide dismutase gene (Sod1) increases
 susceptibility to noise-induced hearing loss. Audiol Neurootol 4,
237–246 (1999).
146. Konings, A., Van Laer, L., Pawelczyk, M., Carlsson, P.I., Bondeson,
M.L., Rajkowska, E., et al. Association between variations in CAT
and noise-induced hearing loss in two independent noise-exposed
populations. Hum Mol Genet 16, 1872–1883 (2007).
147. Seligmann, H., Podoshin, L., Ben-David, J., Fradis, M., & Goldsher,
M. Drug-induced tinnitus and other hearing disorders. Drug Saf
14, 198–212 (1996).
148. Forge, A. & Schacht, J. Aminoglycoside antibiotics. Audiol
Neurootol 5, 3–22 (2000).
149. Van Laer, L. & Van Camp, G., Applications in audiological
medicine. In Genomics and Clinical Medicine. D. Kumar, editor.
Oxsford, UK: Oxford University Press; 2008.
150. Henderson, D., McFadden, S.L., Liu, C.C., Hight, N., & Zheng,
X.Y. The role of antioxidants in protection from impulse noise.
Ann N Y Acad Sci 884, 368–380 (1999).
151. Minami, S.B., Sha, S.H., & Schacht, J. Antioxidant protection in a
new animal model of cisplatin-induced ototoxicity. Hear Res 198,
137–143 (2004).
152. Schacht, J. Antioxidant therapy attenuates aminoglycoside-induced
hearing loss. Ann N Y Acad Sci 884, 125–130 (1999).
153. Song, B.B., Anderson, D.J., & Schacht, J. Protection from gentami-
cin ototoxicity by iron chelators in guinea pig in vivo. J Pharmacol
Exp Ther 282, 369–377 (1997).
154. McFadden, S.L., Ding, D., Salvemini, D., & Salvi, R.J. M40403,
a superoxide dismutase mimetic, protects cochlear hair cells from
gentamicin, but not cisplatin toxicity. Toxicol Appl Pharmacol 186,
46–54 (2003).
155. Sha, S.H., Zajic, G., Epstein, C.J., & Schacht, J. Overexpression
of copper/zinc-superoxide dismutase protects from kanamycin-
induced hearing loss. Audiol Neurootol 6, 117–123 (2001).
156. Kawamoto, K., Sha, S.H., Minoda, R., Izumikawa, M., Kuriyama,
H., Schacht, J., et al. Antioxidant gene therapy can protect hearing
and hair cells from ototoxicity. Mol Ther 9, 173–181 (2004).
157. Palomar Garcia, V., Abdulghani Martinez, F., Bodet Agusti, E.,
Andreu Mencia, L., & Palomar Asenjo, V. Drug-induced otoxic-
ity: current status. Acta Otolaryngol 121, 569–572 (2001).
158. Sanger, F., Nicklen, S., & Coulson, A.R. DNA sequencing with
chain-terminating inhibitors. Proc Natl Acad Sci U S A 74,
5463–5467 (1977).
159. Bartlett, J.M. & Stirling, D. A short history of the polymerase
chain reaction. Methods Mol Biol 226, 3–6 (2003).
160. Smith, L.M., Sanders, J.Z., Kaiser, R.J., Hughes, P., Dodd, C.,
Connell, C.R., et  al. Fluorescence detection in automated DNA
sequence analysis. Nature 321, 674–679 (1986).
1 61. Benson, D.A., Karsch-Mizrachi, I., Lipman, D.J., Ostell, J.,
& Wheeler, D.L. GenBank. Nucleic Acids Res 34, D16–20
(2006).
162. Kent, W.J., Sugnet, C.W., Furey, T.S., Roskin, K.M., Pringle, T.H.,
Zahler, A.M., et al. The human genome browser at UCSC. Genome
Res 12, 996–1006 (2002).
163. Birney, E., Andrews, T.D., Bevan, P., Caccamo, M., Chen, Y.,
Clarke, L., et al. An overview of Ensembl. Genome Res 14, 925–928
(2004).
164. Shalon, D., Smith, S.J., & Brown, P.O. A DNA microarray system
for analyzing complex DNA samples using two-color fluorescent
probe hybridization. Genome Res 6, 639–645 (1996).
165. Weber, J.L. & May, P.E. Abundant class of human DNA polymor-
phisms which can be typed using the polymerase chain reaction.
Am J Hum Genet 44, 388–396 (1989).
166. Shendure, J. & Ji, H. Next-generation DNA sequencing. Nat
Biotechnol 26, 1135–1145 (2008).
167. Mardis, E.R. The impact of next-generation sequencing technol-
ogy on genetics. Trends Genet 24, 133–141 (2008).
168. Metzker, M.L. Sequencing technologies - the next generation. Nat
Rev Genet 11, 31–46 (2010).
G eno m ic A pplications in Audiological   Medicine   •   6 8 1
169. Brownstein, Z., Friedman, L.M., Shahin, H., Oron-Karni, V., Kol,
N., Abu Rayyan, A., et al. Targeted genomic capture and massively
parallel sequencing to identify genes for hereditary hearing loss in
Middle Eastern families. Genome Biol 12, R89 (2011).
170. Adzhubei, I.A., Schmidt, S., Peshkin, L., Ramensky, V.E.,
Gerasimova, A., Bork, P., et al. A method and server for predicting
damaging missense mutations. Nat Methods 7, 248–249 (2010).
171. Ng, P.C. & Henikoff, S. SIFT: Predicting amino acid changes that
affect protein function. Nucleic Acids Res 31, 3812–3814 (2003).
172. Landau, M., Mayrose, I., Rosenberg, Y., Glaser, F., Martz, E.,
Pupko, T., et al. ConSurf 2005: the projection of evolutionary con-
servation scores of residues on protein structures. Nucleic Acids Res
33, W299–302 (2005).
173. Walsh, T., Shahin, H., Elkan-Miller, T., Lee, M.K., Thornton,
A.M., Roeb, W., et al. Whole exome sequencing and homozygos-
ity mapping identify mutation in the cell polarity protein GPSM2
as the cause of nonsyndromic hearing loss DFNB82. Am J Hum
Genet 87, 90–94 (2010).
174. Zheng, J., Miller, K.K., Yang, T., Hildebrand, M.S., Shearer, A.E.,
DeLuca, A.P., et al. Carcinoembryonic antigen-related cell adhe-
sion molecule 16 interacts with alpha-tectorin and is mutated in
autosomal dominant hearing loss (DFNA4). Proc Natl Acad Sci U
S A 108, 4218–4223 (2011).
175. Delmaghani, S., Aghaie, A., Michalski, N., Bonnet, C., Weil,
D., & Petit, C. Defect in the gene encoding the EAR/EPTP
domain-containing protein TSPEAR causes DFNB98 profound
deafness. Hum Mol Genet 21, 3835–3844 (2012).
176. Sirmaci, A., Edwards, Y.J., Akay, H., & Tekin, M. Challenges in
whole exome sequencing:  an example from hereditary deafness.
PLoS One 7, e32000 (2012).
177. Pierce, S.B., Chisholm, K.M., Lynch, E.D., Lee, M.K., Walsh,
T., Opitz, J.M., et al. Mutations in mitochondrial histidyl tRNA
synthetase HARS2 cause ovarian dysgenesis and sensorineural
hearing loss of Perrault syndrome. Proc Natl Acad Sci U S A 108,
6543–6548 (2011).
178. Sirmaci, A., Walsh, T., Akay, H., Spiliopoulos, M., Sakalar,
Y.B., Hasanefendioglu-Bayrak, A., et  al. MASP1 mutations in
patients with facial, umbilical, coccygeal, and auditory findings of
Carnevale, Malpuech, OSA, and Michels syndromes. Am J Hum
Genet 87, 679–686 (2010).
179. Klein, C.J., Botuyan, M.V., Wu, Y., Ward, C.J., Nicholson, G.A.,
Hammans, S., et al. Mutations in DNMT1 cause hereditary sen-
sory neuropathy with dementia and hearing loss. Nat Genet 43,
595–600 (2011).
180. Winkelmann, J., Lin, L., Schormair, B., Kornum, B.R., Faraco, J.,
Plazzi, G., et al. Mutations in DNMT1 cause autosomal dominant
cerebellar ataxia, deafness and narcolepsy. Hum Mol Genet 21,
2205–2210 (2012).
181. Gilissen, C., Hoischen, A., Brunner, H.G., & Veltman, J.A. Disease
gene identification strategies for exome sequencing. Eur J Hum
Genet 20, 490–497 (2012).
182. Montenegro, G., Powell, E., Huang, J., Speziani, F., Edwards, Y.J.,
Beecham, G., et al. Exome sequencing allows for rapid gene identi-
fication in a Charcot-Marie-Tooth family. Ann Neurol 69, 464–470
(2011).
183. Lieber, D.S., Vafai, S.B., Horton, L.C., Slate, N.G., Liu, S.,
Borowsky, M.L., et al. Atypical case of Wolfram syndrome revealed
through targeted exome sequencing in a patient with suspected
mitochondrial disease. BMC Med Genet 13, 3 (2012).
184. Choi, M., Scholl, U.I., Ji, W., Liu, T., Tikhonova, I.R., Zumbo,
P., et  al. Genetic diagnosis by whole exome capture and mas-
sively parallel DNA sequencing. Proc Natl Acad Sci U S A 106,
19096–19101 (2009).
185. Bonnefond, A., Durand, E., Sand, O., De Graeve, F., Gallina, S.,
Busiah, K., et al. Molecular diagnosis of neonatal diabetes mellitus
using next-generation sequencing of the whole exome. PLoS One 5,
e13630 (2010).
 186. Harris, K.C., Erbe, C.B., Firszt, J.B., Flanary, V.A., & Wackym, P.A.
A novel connexin 26 compound heterozygous mutation results in
deafness. Laryngoscope 112, 1159–1162 (2002).
187. Kelley, P.M., Harris, D.J., Comer, B.C., Askew, J.W., Fowler, T.,
Smith, S.D., et al. Novel mutations in the connexin 26 gene (GJB2)
that cause autosomal recessive (DFNB1) hearing loss. Am J Hum
Genet 62, 792–799 (1998).
188. Cohen, M. & Phillips, J.A., 3rd Genetic approach to evaluation of
hearing loss. Otolaryngol Clin North Am 45, 25–39 (2012).
189. Gilbey, P., Kraus, C., Ghanayim, R., Sharabi-Nov, A., & Bretler, S.
Universal newborn hearing screening in Zefat, Israel: the first two
years. Int J Pediatr Otorhinolaryngol 77, 97–100 (2013).
190. Neumann, K. & Indermark, A. Validation of a new TEOAE-AABR
device for newborn hearing screening. Int J Audiol 51, 570–575
(2012).
191. Hall, J.W., 3rd, Smith, S.D., & Popelka, G.R. Newborn hearing
screening with combined otoacoustic emissions and auditory
brainstem responses. J Am Acad Audiol 15, 414–425 (2004).
192. Norton, S.J., Gorga, M.P., Widen, J.E., Folsom, R.C., Sininger, Y.,
Cone-Wesson, B., et al. Identification of neonatal hearing impair-
ment: a multicenter investigation. Ear Hear 21, 348–356 (2000).
193. Neumann, K., Gross, M., Bottcher, P., Euler, H.A.,
Spormann-Lagodzinski, M., & Polzer, M. Effectiveness and effi-
ciency of a universal newborn hearing screening in Germany. Folia
Phoniatr Logop 58, 440–455 (2006).
194. Dettman, S.J., Pinder, D., Briggs, R.J., Dowell, R.C., & Leigh, J.R.
Communication development in children who receive the cochlear
implant younger than 12 months: risks versus benefits. Ear Hear
28, 11S-18S (2007).
6 8 2   •   G eno m ics in C linical P ractice
195. George, P., Farrell, T.W., & Griswold, M.F. Hearing loss: help for
the young and old. J Fam Pract 61, 268–277 (2012).
196. Gates, G.A., Murphy, M., Rees, T.S., & Fraher, A. Screening for
handicapping hearing loss in the elderly. J Fam Pract 52, 56–62
(2003).
197. Raviv, D., Dror, A.A., & Avraham, K.B. Hearing loss: a common
disorder caused by many rare alleles. Ann N Y Acad Sci 1214,
168–179 (2010).
198. Shearer, A.E., DeLuca, A.P., Hildebrand, M.S., Taylor, K.R.,
Gurrola, J., 2nd, Scherer, S., et al. Comprehensive genetic testing
for hereditary hearing loss using massively parallel sequencing. Proc
Natl Acad Sci U S A 107, 21104–21109 (2010).
199. Tang, W., Qian, D., Ahmad, S., Mattox, D., Todd, N.W., Han,
H., et al. A low-cost exon capture method suitable for large-scale
screening of genetic deafness by the massively-parallel sequencing
approach. Genet Test Mol Biomarkers 16, 536–542 (2012).
200. Schrauwen, I., Sommen, M., Corneveaux, J.J., Reiman, R.A.,
Hackett, N.J., Claes, C., et  al. A sensitive and specific diagnostic
test for hearing loss using a microdroplet PCR-based approach and
next generation sequencing. Am J Med Genet A 161A, 145–152
(2013).
201. Chapple, A., May, C., & Campion, P. Lay understanding of genetic
disease: a British study of families attending a genetic counseling
service. J Genet Couns 4, 281–300 (1995).
202. Horn, H.F., Brownstein, Z., Lenz, D.R., Shivatzki, S., Dror, A.A.,
Dagan-Rosenfeld, O., et  al. The LINC complex is essential for
hearing. J Clin Invest 123, 740–750 (2013).
203. Dror, A.A. & Avraham, K.B. Hearing loss: mechanisms revealed by
genetics and cell biology. Annu Rev Genet 43, 411–437 (2009).
 44.
GENETICS AND GENOMICS OF SKIN DISEASES I:
ATOPIC DERMATITIS AND OTHER
SKIN COMPLEX DISEASES
Nilesh Morar
INTRODUCTION
Atopic dermatitis (AD) or atopic eczema is a chronic,
itchy, inflammatory skin condition that predominantly
affects the skin flexures. Approximately 70% of cases start
in children younger than five years of age (Williams, 2000).
The prevalence has increased three fold over the last three
decades in developed countries, with a higher prevalence
in urban regions and in higher social classes (Taylor et al.,
1984). The lifetime prevalence in children is 10–20%,
and in adults, 1–3% (Schultz-Larsen, 2002). Though 60%
of children with AD are free of symptoms in adolescence
(Rystedt, 1985), up to 50% may have recurrences in adult-
hood (Lammintausta, 1991). Asthma develops in 30% of
children with AD, and allergic rhinitis in 35% (Luoma
et al., 1983).
D I AG N O S T I C C R I T E R I A
There have been substantial variations in the disease’s defi-
nitions and diagnoses due to its variable clinical appear-
ance and distribution patterns, and the intermittent nature
of the disease. Based on the original consensus criteria by
Hanifin and Rajka (1980), a set of minimum and validated
discriminators have been determined by the U.K. Working
Party (Williams et  al., 1994). The diagnosis requires evi-
dence of itchy skin plus three or more of the following: his-
tory of involvement of the skin creases, history of asthma
or hay fever, history of generally dry skin, onset in a child
under two years of age, and visible flexural dermatitis. These
robust criteria are crucial for effective and reproducible
genetic studies.
683
ATO P I C D E R M AT I T I S A N D ATO P Y
Atopic dermatitis, asthma, and hay fever are considered to
be part of a common syndrome of atopic diseases. The word
atopy means “strange disease” and was coined in 1923 by
Coca (Coca and Cooke, 1923). The atopic state is recog-
nized by positive skin-prick tests to common allergens, by
the presence of allergen-specific immunoglobulin E (IgE) in
serum, and by elevations of total serum IgE. Eighty percent
of infants with AD have raised total serum IgE levels ( Juhlin
et  al., 1969). Two types of AD have been described:  an
extrinsic type associated with IgE-mediated sensitization,
and an intrinsic type, which is clinically identical, with nor-
mal total serum IgE and no specific IgE responses to com-
mon allergens ( Johansson et al., 2001).
The association between AD and atopy is not clear-cut,
as both forms of AD have the same phenotype clinically.
There is some suggestion from hospital studies that higher
IgE levels are correlated with severe forms of AD, worse
long-term prognosis (Flohr et al., 2004), and an increased
likelihood of developing asthma. There are immunological
differences between these two types of AD that lie in differ-
ent expression levels in skin-related cytokines that determine
T-cell activation patterns ( Jeong et al., 2003). Inflammatory
dendritic epidermal cells bear the high-affinity receptor for
IgE (FcεRI) on their cells’ surface (Novak et al., 2003). This
allows allergens penetrating the impaired skin barrier in AD
to be focused and efficiently presented by FcεRI-bound IgE
to T cells (Novak et al., 2003; Novak and Bieber, 2003). In
intrinsic AD, there is lower surface expression of FcεRI on
epidermal dendritic cells (Oppel et al., 2000). Interleukin
(IL) 13 induces the production of IgE by B cells (Punnonen
et al., 1993). IL13 is decreased in intrinsic AD (Punnonen
et al., 1993; Jeong et al., 2003). Intrinsic and extrinsic forms
of AD may, however, be part of a continuum in the natural
history of AD (Novembre et al., 2001). The fact that phe-
notypically both are identical suggests that IgE sensitization
is not a prerequisite for developing the skin lesions of AD.
Omalizumab, a monoclonal anti-IgE antibody currently
approved for the treatment of asthma, failed to improve
AD in any patients who were treated (Krathen et al., 2005;
Lane et al., 2006). The diagnostic criteria of AD can also be
fulfilled in the absence of elevated IgE (Hanifin and Rajka,
1980; Williams et  al., 1994). Subgroup analysis in future
genetic studies may help dissect these differences better.
Therefore, though atopic mechanisms dominate our under-
standing of the pathogenesis of AD, other mechanisms may
predispose to AD independently of atopy.
G E N ET I C S T U D I E S O F ATO P I C
D E R M AT I T I S
In genetic-epidemiological and population-based twin
studies of AD, the concordance rates for monozygotic
twins were 0.72 to 0.86, and 0.21 to 0.23 for dizygotic twins
(Larsen and Holm, 1986; Schultz Larsen, 1993). Asthma
shows similar concordance rates of 0.65 in monozygotic
twins and 0.25 in dizygotic twins (Duffy et  al., 1990).
Total serum IgE shows heritability of approximately 47.3%
(Palmer et al., 2000). Strong genetic factors therefore under-
lie the development of AD and atopic disorders. However,
inheritance of AD is complex and does not follow Mendel’s
laws of inheritance. Many genes may be contributory. In
rare Mendelian diseases, polymorphisms commonly cause
mutations by altering protein-coding sequences. In com-
mon diseases such as AD, gene functions may be altered by
subtler mechanisms. These may affect the initiation of gene
transcription in exons or factors that affect gene expression
and splicing in introns. The genes causing complex disease
are likely to show high-frequency allelic variation (fre-
quency of >1%), which is the basis of the “common disease,
common variant” hypothesis (Cargill et al., 1999; Halushka
et  al., 1999; Cheung and Spielman, 2002). There is com-
plex interplay between environmental factors and the alleles
of many genes and interactions between genes themselves
(epistasis), which play an important role in determining dis-
ease expression.
Two traditional approaches are used to identify
genes:  case-control comparisons of polymorphisms in
known candidate genes, or the identification of new genes
by positional cloning. Recently, with the sequencing of the
human genome, genome-wide gene-expression profiling
6 8 4   •   G enomics in C l inica l P ractice
with DNA microarrays has allowed global analyses of the
expression of thousands of genes, and is a high-throughput
method for gene identification.
Genomic approaches to understanding AD are new,
and significant inroads have been made only in the last five
years. To date, the following observations have been made
in the genetics of AD that have given us new insights into
the pathogenesis of AD.
POSITIONAL CLONING STUDIES
In positional cloning, families are studied to identify
chromosomal regions that are co-inherited or linked with
disease. A “genome screen” is a systematic search for such
regions using polymorphic microsatellite markers. The
markers or microsatellites typically contain repeat sequences
of DNA, such as CA repeats, that vary between individu-
als. Linkage analyses search regions of the genome for areas
with a higher than expected number of shared alleles among
affected individuals within a family (Carlson et al., 2004).
Genetic linkage is said to exist if the marker and the phe-
notype are co-inherited. Genetic linkage identifies broad
chromosomal regions linked to disease, normally exceeding
10  cM (~10  million bases). This area has to be saturated
with closely spaced markers to reduce the interval for local-
ization to between 5–40 Mb (megabases). However, fortu-
nately there is non-random association between alleles of
individual single nucleotide polymorphisms (SNPs) and
neighboring SNPs, known as linkage disequilibrium, which
extends for shorter distances than genetic linkage. The
remaining chromosomal region of 0.5–1 Mb can then be
systematically dissected. Because no assumptions need be
made about the disease’s etiology, positional cloning can
identify novel genes and mechanisms. It has been highly
successful in identifying genes underlying single-gene disor-
ders such as cystic fibrosis and Huntington disease. Genome
screens, however, are difficult to replicate because they are
sensitive to population stratification, disease definition, the
panels of markers chosen for genotyping, environmental
factors, and the frequency of the phenotype in the popu-
lation. Also, stringent statistical criteria are needed to con-
firm whether linkage is real or not that take into account
the hundreds of markers tested and the many phenotypes
studied (Lander and Kruglyak, 1995).
There have been four genome screens for AD (Cookson
et al., 2001; Lee et al., 2000; Bradley et al., 2002). The first
screen, carried out in German and Scandinavian families
with childhood AD, showed linkage to AD on chromo-
some 3q21 (Lee et  al., 2000). There was also linkage of
total serum IgE to 3q21. The second genome screen was in
British families identified through a proband with child-
hood AD attending a tertiary referral clinic (Cookson et al.,
2001). Linkage to AD or to AD and asthma combined was
found on chromosomes 1q21, 17q25, and 20p. These are
coincident with regions linked to psoriasis (Tomfohrde
et  al., 1994; Capon et  al., 1999; Trembath et  al., 1997).
Linkage to total serum IgE was also found on chromosomes
5q31 and 16qtel. In both studies, evidence of linkage to
total serum IgE was lower than evidence of linkage to AD.
An adult Swedish study found evidence of linkage to AD
to chromosome 3p24-22 (Bradley et al., 2002). For another
phenotype, AD combined with raised allergen-specific
IgE levels, a suggestive linkage was found to chromosome
region 18q21. When a semi-quantitative phenotype look-
ing at severity scores in AD was used, suggestive linkage was
found to chromosomal regions 3q14, 13q14, 15q14-15,
and 17q21. It is possible that the 3q14 and 17q21 loci cor-
respond to the previously identified AD loci in children.
Chromosome 13q14 has been previously linked to chil-
dren with AD (Beyer et al., 2000) and to atopy and asthma
(Anderson et al., 2002). The most recent genome screen was
in Danish families ascertained via affected sib pairs with
both clinical AD and confirmed specific allergy. Linkage
to AD was found for 3p26-24, 4p15-14, and 18q11-12
(Haagerup et  al., 2004). A  study of selected candidate
regions found evidence of linkage of AD to the 14q11 locus
in Swedish families (Soderhall et al., 2001).
C A N D I DAT E - G E N E S T U D I E S
Candidate genes are studied based on their known biologi-
cal functions in relation to disease. The genes are sequenced
and known and newly identified polymorphisms are geno-
typed in cases and controls. A selection of candidate-gene
studies in AD will be discussed further on.
T H E H I G H-A FFI N IT Y R EC E P TO R F O R
I G E ( F CΕR I )
FcεRI is a multimeric cell-surface receptor that binds the
constant region of the IgE molecule and is expressed on
basophils, mast cells, monocytes, platelets, and eosinophils.
Its expression is also upregulated on both Langerhans cells
and inflammatory dendritic epidermal cells ( Jurgens et al.,
1995; Novak et al., 2002). The beta subunit of FcεRI, FcεRIβ,
serves as an amplifier of FcεRI biological function and sta-
bilizes the surface expression of the receptor (Turner and
Kinet, 1999; Lin et al., 1996). Cross-linking of IgE bound
G enetics and G enomics of S k in D iseases I   •   6 8 5
to its receptor on cells by multivalent allergens initiates a
cascade of events resulting in allergic immune responses.
Mast cells and basophils are involved in the early, immediate
response, which is marked by cellular degranulation and the
release of pre-formed proinflammatory mediators, includ-
ing histamine, proteases, leukotrienes, and prostaglandins,
which cause the symptoms of allergic reactions.
The FcεRIβ gene on chromosome 11q 12-13 has been
linked to atopy (Cookson et al., 1989; Young et al., 1992).
Polymorphisms in this gene have been associated with
AD in two panels of British families in which the major-
ity of probands had severe skin disease and 60% had severe
asthma (Cox et al., 1998). Significant association with two
intronic SNPs was seen for AD and asthma. Significant
sharing of maternal alleles has been seen for AD. As none of
these variants were functional, their relevance to AD is not
known. Coding variants have been identified that are asso-
ciated with atopy but not with AD (Traherne et al., 2003).
M A S T C E L L C H Y M A S E ( MC C)
Mast cells (MC) are key effector cells of IgE-dependent
immediate reactions and also to certain IgE-mediated
late-phase reactions (Williams and Galli, 2000). MCC
is a chymotryptic serine protease present in the secretory
granules of MC, and it has various functions, including
promoting endothelial and epithelial permeability and the
recruitment of neutrophils and leukocytes in vivo (Miller
and Pemberton, 2002). Its gene, CMA1, maps close to
the long arm of chromosome 14q11. This region has been
previously linked to AD (Soderhall et al., 2001). MCC is
increased in chronic AD skin lesions (Badertscher et  al.,
2005), and a potential role of chymase in contributing to
the skin barrier defect has been suggested (Groneberg
et  al., 2005). There have been reports of significant asso-
ciations between a CMA1 promotor polymorphism and
AD in Japanese adults and school children (Mao et  al.,
1996; Mao et  al., 1998; Tanaka et  al., 1999). The lat-
ter two studies found that the association with AD was
independent of total serum IgE levels, suggesting that the
mechanism by which MCC influences disease outcome
is IgE-independent. These finding were not confirmed by
another Japanese study (Kawashima et al., 1998) or by an
Italian study (Pascale et al., 2001). Recently, this polymor-
phism was associated with serum total IgE levels in adult
AD in a family-based association study in Caucasians
(Iwanaga et al., 2004), and with the AD phenotype alone in
a German case-control study (Weidinger et al., 2005). This
supports the notion that CMA1 serves as a candidate gene
for AD. The therapeutic potential of chymase inhibitors in
AD has been successfully explored in preliminary animal
studies where symptoms and signs of AD were ameliorated
with MCC inhibitors (Imada et al., 2002; Watanabe et al.,
2002).
C- C C H E MO K I N E S : R A N T E S ( R EGU L AT E D
O N AC T I VAT I O N, N O R M A L T- C E L L
E X P R E S S E D A N D S EC R ET E D) A N D EOTAX I N
Infiltration of inflammatory cells in tissues is regulated by
chemokines (Alam, 1997). RANTES is a CC chemokine,
and enhanced serum levels have been found in AD that
correlated with total serum IgE levels and eosinophil num-
bers (Kaburagi et al., 2001). A functional point mutation
in the proximal promoter region of the RANTES gene,
which results in a new consensus binding site for the GATA
transcription factor family, has been identified on chromo-
some 17q 11.2 (Nickel et al., 2000). The mutant allele was
associated with AD but not with asthma in children in the
German Multicentre Allergy Study (MAS-90). Using tran-
sient transfection of human cell lines, the polymorphism
was also shown to exert higher transcriptional activity.
This polymorphism is associated with enhanced RANTES
production in AD patients; hence, the polymorphism is
involved in AD by upregulating RANTES protein expres-
sion (Bai et al., 2005). These findings were not replicated,
however, in a study of 128 Hungarian children with AD
(Kozma et al., 2002).
Eotaxin is a potent chemoattractant and activator of
eosinophils and TH2 lymphocytes. The eotaxin gene is
located at chromosome 17q21.1-q21.2 adjacent to the chro-
mosome 17 linkage region. Eotaxin is overexpressed in AD
skin (Yawalkar et al., 1999; Taha et al., 2000) and serum,
and plasma levels are also increased (Hossny et al., 2001).
Numerous polymorphisms have been identified, includ-
ing a coding SNP that results in a non-conservative amino
acid change of alanine to threonine (Tsunemi et al., 2002).
This SNP and SNPs in the promoter and exon regions are
not associated with susceptibility to AD, but two of them
in the promoter region are associated with the total serum
IgE levels in AD. Tacrolimus ointment, a macrolide lactone
that is effective in treating AD, suppresses the expression of
eotaxin and RANTES in AD skin (Park et al., 2005).
T H E C Y TO K I N E G E N E C LUS T E R
The cytokine gene cluster is on chromosome 5q31-33.
Genetic linkage of genes in this region to AD (Forrest et al.,
1999; Beyer et al., 2000; Soderhall et al., 2001) and atopy
(Tanaka et al., 1999) and the total serum IgE was found in
6 8 6   •   G enomics in C l inica l P ractice
several studies (Cookson et  al., 2001). It contains several
interleukins, SPINK5 (the gene encoding the serine prote-
ase inhibitor LEKTI), GM-CSF (granulocyte-macrophage
colony-stimulating factor), CD14 antigen, and TIM-1
(T-cell immunoglobulin domain mucin domain protein 1).
Polymorphisms in IL13 (Liu et  al., 2000; Tsunemi et  al.,
2002; He et al., 2003; Hummelshoj et al., 2003), IL4 (He
et al., 2003; Kawashima et al., 1998), CD14 (Lange et al.,
2005), GM-CSF (Rafatpanah et  al., 2003), TIM-1 (Chae
et al., 2003), and SPINK5 (Nishio et al., 2003; Walley et al.,
2001; Kato et al., 2003) have been associated with AD.
SPINK5 has been identified as the gene that is defec-
tive in Netherton syndrome (NS) (Chavanas et al., 2000a;
Bitoun et al., 2002). NS is a rare, severe, autosomal reces-
sive genodermatosis in which atopy is a universal mani-
festation ( Judge et  al., 1994). LEKTI is expressed on
epithelial and mucosal surfaces as well as in the thymus
(Chavanas et al., 2000a; Magert et al., 1999). A number
of mutations have been found in the SPINK5 gene fol-
lowing sequencing of NS patients (Chavanas et al., 2000).
The majority result in premature-termination codons
causing abolition in protein production. Further poly-
morphisms were found upon sequencing the gene in chil-
dren with severe AD and normal controls (Walley et al.,
2001). Genotyping the SNPs in two independent families
showed significant associations between the Glu420Lys
variant and atopy and AD (Walley et  al., 2001). These
findings have been confirmed in two Japanese studies
(Nishio et al., 2003), but not in a population in Northern
Germany (Folster-Holst et al., 2005).
Significant trypsin-inhibiting activity has been dem-
onstrated for three of the 15 LEKTI domains (Magert
et al., 2002). LEKTI is also implicated in the cornification
process (Descargues et  al., 2005). Many allergens are ser-
ine proteinases, and their protease activity may encourage
their allergenicity (Thomas et  al., 2002). Proteinases are
also involved in T cell and B cell maturation. Proteinase
inhibitors are important negative regulators of proteinase
action in vivo (Magert et  al., 2002). Taken together, the
findings in NS are an example of a monogenic disorder
providing insights into one of the plausible mechanisms
in AD where there is a barrier defect, increased skin per-
meability, and predisposition to recurrent infections. The
house dust mite Dermatophagoides sp’s feces (Thomas et al.,
2002)  and Staphylococcus aureus (Miedzobrodzki et  al.,
2002) are potent sources of proteinases in AD skin lesions
that enhance the disruption of the skin barrier. Treatment
of AD with alpha 1–protease inhibitors was effective in a
pilot study in patients with recalcitrant AD (Wachter and
Lezdey, 1992).
 T H E E P I D E R M A L D I FFE R E N T I AT I O N in the integrity, differentiation, and immunity of the skin
C O M P L E X ( E D C) A N D ATO P I C barrier.
D E R M AT IT I S

The peak of linkage of AD on chromosome 1q21 overlies

D E F E C T I VE S K I N B A R R I E R I N ATO P I C
the epidermal differentiation complex (EDC). These genes
D E R M AT I T I S
regulate epidermal differentiation. They are expressed dur-
ing terminal differentiation of the epidermis.
Several lines of evidence point towards a defective skin
The EDC comprises a number of gene families, such
barrier predisposing to AD. There is a global decrease in
as the S100 calcium-binding proteins, small proline-rich
lipids in AD skin and selective deficiency in stratum cor-
proteins, and the late-expressed cornified envelope (CE)
neum ceramide content (Melnik et al., 1988), which is cor-
proteins (Mischke et  al., 1996; Marshall et  al., 2001). It
related with diminished barrier function in AD (Di Nardo
also includes members of the peptidoglycan-recognition
et al., 1998). Ceramide-dominant moisturizers are effective
protein family and a number of single-copy genes such as
therapeutically, as they hydrate the delipidized stratum cor-
loricrin, involucrin, trichohyalin, and filaggrin. The forma-
neum and reduce transepidermal water loss (Chamlin et al.,
tion of the CE is characteristic of terminally differentiated
2002). Studies of the EDC gene products transcribed within
epidermal keratinocytes. Loricrin is the major CE pro-
terminally differentiating keratinocytes also reveal that the
tein and is cross-linked by both disulfide and N epsilon-
skin is not a passive barrier. The S100 calcium-binding pro-
(gamma-glutamyl) lysine isodipeptide bonds to the CE
tein family (S100 proteins, profilaggrin, filaggrin) have a
(Hohl et  al., 1991). Mutations in the loricrin gene have
wide range of immunological and structural functions, as
been seen in Vohwinkel’s syndrome, a mutilating kerato-
described above. Percutaneous sensitization with house
derma (Maestrini et  al., 1996). DNA microarray analy-
dust mite antigen after barrier disruption elicits a TH2
sis has shown decreased expression of loricrin in AD skin
cytokine response in mice (Kondo et  al., 1998), which is
(Sugiura et al., 2005).
relevant to the disrupted barrier in AD. The NS gene that
S100 proteins are low-molecular-weight calcium-binding
encodes the serine protease inhibitor LEKTI is found in the
proteins of the EF-hand superfamily. They are induced
uppermost epidermis and pilosebaceous units and has been
during terminal differentiation of the epidermis and are
associated with AD (Walley et al., 2001). Stratum corneum
involved in the regulation of numerous cellular processes,
chymotryptic enzyme (SCCE), which is a serine protease,
such as cell-cycle progression and differentiation (Schafer
plays a central role in desquamation, and a polymorphism
et  al., 1995). S100A7 (psoriasin) is highly upregulated in
in SCCE has been associated with AD (Vasilopoulos et al.,
psoriatic epidermis as well as in primary human keratino-
2004). Treatment with protease inhibitors also ameliorates
cytes undergoing abnormal differentiation (Hoffmann
symptoms of AD (Imada et al., 2002).
et al., 1994). It is a potent chemotaxin for neutrophils and
CD4 + T lymphocytes ( Jinquan et al., 1996). Profilaggrin
and trichohyalin are expressed later in terminally differ- ATO P I C D E R M AT I T I S A N D A S T H M A
entiating granular cells and are also members of the S100
family. Profilaggrin, by a process of dephosphorylation and Asthma is an inflammatory disease of the airways of the
proteolysis, is processed to filaggrin, which functions as a lung. It is characterized by narrowing of the airways due to
matrix for packing keratin filaments in differentiated cor- inflammation and mucous secretion, and bronchospasm as
nified cells (Dale et al., 1985). Decreased expression levels a result of contraction of bronchial smooth muscles from
of filaggrin were shown in AD skin by DNA microarrays non-specific stimuli. There is strong association between
(Sugiura et al., 2005) and immunohistochemistry (Seguchi AD and asthma, with 60% of children with AD also having
et al., 1996). Recently, loss-of-function genetic variants in asthma (Cox et al., 1998). Though severe AD and asthma
the gene encoding filaggrin (FLG) have been associated often occur together (Williams et  al., 1999), both condi-
with AD in populations of European origin (Palmer et al., tions respond to different determinants in population stud-
2006). These variants in FLG are also the cause of icthyosis ies (Riedler et al., 2001). Children with parental AD have
vulgaris, which is often associated with the atopic diathesis a high risk for AD compared to children with parental
(Smith et al., 2006). Further studies are required to identify asthma, or parental allergic rhinitis (Dold et al., 1992), sug-
variants in non-European populations. The EDC therefore gesting the presence of eczema-specific genes. More than a
contains several candidate genes for AD that are involved dozen genome screens have been reported for asthma and

G enetics and G enomics of S k in D iseases I   •   6 8 7

its related phenotypes (Daniels et al., 1996; Howard et al.,
2001; Ober et  al., 2000; Wjst et  al., 1999; Hizawa et  al.,
1998; Mathias et  al., 2001; Dizier et  al., 2000; Laitinen
et  al., 2001; Hakonarson et  al., 2002; Koppelman et  al.,
2002; Haagerup et  al., 2002; Meyers et  al., 2005; Wang
et  al., 2005; Dizier et  al., 2005; Bouzigon et  al., 2004).
Three genes underlying asthma have been identified by
positional cloning. ADAM33 on chromosome 20p13 is a
membrane-anchored metalloprotease with diverse func-
tions such as the shedding of cell-surface proteins such as
cytokines and cytokine receptors (Van Eerdewegh et  al.,
2002). PHF11 on chromosome 13q12 contains two PHD
(Plant homeo domain) zinc fingers and probably regulates
transcription (Zhang et al., 2003). The final asthma gene is
DPP10, which encodes a homologue of dipeptidyl pepti-
dases (DPPs) that cleaves terminal dipeptides from cyto-
kines and chemokines (Allen et  al., 2003). PHF11 may
play a role in AD, as chromosome 13q12 does show linkage
to AD (Beyer et al., 2000) and polymorphisms in PHF11
are strongly associated with increased IgE levels in families
comprising children with AD (Zhang et  al., 2003). The
protein-binding zinc fingers PHF11 encodes may modify
immunoglobulin production and clonal expansion of B
cells (Zhang et al., 2003).
The results from the asthma genome screens show
loci that are not generally shared with regions of link-
age to AD, with a few exceptions (Cookson et al., 2001),
suggesting that both diseases may result from distinct
mechanisms.
D E F E C T I VE I N N AT E I M MU N I T Y
A N D ATO P I C D E R M AT I T I S
Innate immunity is a primitive and conserved response
in which the host’s primary defenses recognize
pathogen-associated molecular patterns (PAMPs) on micro-
organisms by germline-encoded pathogen-recognition
receptors (PRRs) (Weidinger et  al., 2005). An intact
innate immune response to microbial organisms is impor-
tant in skin, as the skin is the initial barrier to microbes
and the environment. CD14 receptors (Baldini et  al.,
1999; Koppelman et al., 2001), Toll-like receptors (TLRs)
(Arbour et  al., 2000), and CARD15 (NOD2) (Carneiro
et al., 2004) are PRRs that are important in innate immune
responses. They interact with bacterial lipopolysaccharides
that are associated with IFN-γ dominated TH1 (T-helper)
immunological responses. Allergic responses are driven by
TH2 responses (Romagnani S, 2000). Variations in genes
responding to microbial matter can affect the TH1/TH2
6 8 8   •   G enomics in C l inica l P ractice
balance and predispose to allergic disorders (Kabesch et al.,
2003). Recently it was found that T-helper (TH) cell help
and generation of T-dependent antigen-specific antibody
responses require activation of TLRs in B cells (Pasare
and Medzhitov, 2005). Hence, apart from inducing innate
immune responses, TLRs can mediate control of adaptive
immunity.
Keratinocytes are highly active immunologically and
express PRRs such as CD14 and TLR-4 (Song et al., 2002).
They can induce inflammatory responses without preinduc-
tion by other cells. Patients with AD are prone to infections
by, for example, Staphylococcus aureus, Herpes simplex, and
Mollusca contagiosa viruses. Keratinocytes in AD have also
been shown to be deficient in defensins, which are compo-
nents of the innate immune system that normally have anti-
microbial activity against bacteria, viruses, and fungi (Ong
et al., 2002).
Polymorphisms in PRR such as CARD4 (Weidinger
et al., 2005), CARD15 (Kabesch et al., 2003), and CD14
(Lange et al., 2005) have been associated with AD. A TLR2
polymorphism has been associated with Staphylococcus
aureus infection and has been shown to be significantly
less responsive to bacterial peptides (Lorenz et  al., 2000).
Patients carrying this variant were found to have a distinct
phenotype of severe AD, suggesting that the TLR2 poly-
morphism may increase the susceptibility to infections and
chronic colonization of the skin in AD (Ahmad-Nejad
et al., 2004).
G E N ET I C S O F ATO P I C D E R M AT I T I S
C O M PA R E D TO OT H E R
I N F L A M M ATO RY D I S E A S E S
C RO H N ’S D I S E A S E
CARD15 polymorphisms are associated with susceptibil-
ity to Crohn’s disease (Hampe et al., 2001; Ogura et al.,
2001) and with AD and allergic rhinitis (Kabesch et al.,
2003). Recently, associations between CARD15 variants
and ulcerative colitis have also been described (McGovern
et  al., 2003). CARD4 has recently been associated with
inflammatory bowel disease (McGovern et  al., 2005),
AD (Weidinger et al., 2005) and atopy (Weidinger et al.,
2005; Hysi et al., 2005). This shared genetic background
between inflammatory bowel disease and atopy suggests
that defective recognition of microbes at barrier defenses
in skin and mucosa contributes to excessive responses that
could be either TH1- or TH2-dominated (Kabesch et al.,
2003).
 PSORIASIS
The genome screens in the U.K.  subset do not overlap as
commonly with asthma as one would expect, but rather
with psoriasis-susceptibility loci (Cookson et  al., 2001).
The German genome-screen locus on chromosome 3q21
also overlaps with another psoriasis locus (Enlund et  al.,
1999). This suggests that these genes in these chromosomal
regions are polymorphic and have general effects on dermal
inflammation and immunity.
E P I D E R MO DY S P L A S I A
VE R RU C I F O R M I S A N D L E P RO SY
It is notable that the linkage region to AD and psoriasis maps
to the same region on chromosome 17qter that is a susceptibil-
ity locus for epidermodysplasia verruciformis (EV) (Ramoz
et al., 1999). This is a lifelong disease characterized by infec-
tion with oncogenic human papilloma virus type 5 (HPV5).
Patients with psoriasis also harbor a reservoir for HPV5
(Majewski et al., 1998). Genetic susceptibility to leprosy also
shows linkage to chromosome 20p (Tosh et al., 2002), which
is the area of linkage to the distinctive phenotype of AD and
asthma combined in the U.K. genome screen (Cookson et al.,
2001). These shared areas of linkage of AD with infectious
diseases could underscore the importance of infections or
dysregulation of responses to infections as trigger factors for
the sequelae that determine the AD phenotype.
AU TO I M MU N E D I S E A S E S
The 17q25 locus also shows linkage to multiple sclerosis
(Kuokkanen et al., 1997; Sawcer et al., 1996) and rheuma-
toid arthritis ( Jawaheer et al., 2001). The 20p locus has also
been linked to systemic lupus erythematosus (Gaffney et al.,
2000). ADAM33 is localized to this region but does not
seem to be a candidate for linkage (Van Eerdewegh et al.,
2002).
M AT E R N A L T R A N S M I S S I O N
O F   ATO P I C D E R M AT I T I S
Maternal influence of linkage to atopic disease was first
described for chromosome 11q, which harbors the FcεRIβ
locus where maternal linkage was seen for the atopy phe-
notype as well as skin-prick tests, radioallergosorbent
tests (RASTs) and elevated total IgE (Moffatt et al., 1992;
Cookson et al., 1992). Polymorphisms in FcεRIβ that are
associated with AD and asthma described earlier showed
significant transmission only of maternally derived alleles
G enetics and G enomics of S k in D iseases I   •   6 8 9
for these two phenotypes (Cox et al., 1998). The reason for
this maternal influence could be related to maternal-fetal
interactions influencing the infant’s IgE responses via the
placenta or breast milk (Cox et  al., 1998), or to genomic
imprinting, where the allele from one parent is differen-
tially expressed compared to the allele from the other par-
ent (Cookson et  al., 1992; Hall, 1990). Parent-of-origin
effects have also been noted in other immunological dis-
eases, such as type 1 diabetes (Bennett and Todd, 1996;
Warram et  al., 1984), rheumatoid arthritis (Koumantaki
et al., 1997), psoriasis (Burden et al., 1998), inflammatory
bowel disease (Akolkar et al., 1997), and selective IgA defi-
ciency (Vorechovsky et al., 1999). It has been suggested that
parent-of-origin effects are somewhat adaptive and a gen-
eral phenomenon affecting several immune-related loci and
diseases (Cookson, 2002).
A D U LT ATO P I C D E R M AT I T I S
Adult-onset AD is a poorly defined illness. There are lim-
ited data on adults who present for the first time with AD
in adulthood (Bannister and Freeman, 2000; Ingordo et al.,
2003; Kawashima et  al., 1989; Tay et  al., 1999; Ozkaya,
2005). Apart from typical lesions, these patients may pres-
ent with atypical morphologies such as non-flexural involve-
ment, and nummular and prurigo-like lesions (Ozkaya,
2005). In a study of the genetic basis of AD persisting into
adulthood, polymorphisms in candidate genes involved in
asthma, atopy, and childhood AD were tested for associa-
tion in an adult cohort. A polymorphism in FcεRIβ and a
SNP in the Toll-like receptor-9 (TLR9) were found to be
associated with adult AD. Associations were, however,
weaker than those seen in childhood studies, suggesting the
role of environmental influences or different genes involved
in adult AD (Moffatt et al., 2005). Genetic studies in adults
with AD may be confounded by the fact that they may
not be accommodated in certain criteria such as the U.K.
Working Party criteria (Williams et al., 1994), commonly
used in epidemiological studies where age of onset and clin-
ical distribution of lesions such as flexural involvement are
important diagnostic criteria (Ozkaya, 2005).
M I C R OA R R AY S T U D I E S
O F   ATO P I C D E R M AT I T I S
Genomic-expression profiling using DNA microarrays
allows the simultaneous comparisons of thousands of mes-
senger RNAs hybridized to cDNA clones on chips, and
can identify disease-specific tissue responses by comparing
gene-expression levels. There has been a limited number of
gene-expression studies with small sample sizes using DNA
microarrays in AD. RNA extraction from skin biopsies
showed a distinct pattern of gene expression in AD com-
pared to psoriasis (Nomura et al., 2003). AD skin showed
increased expression of CC chemokines (CCL-13/MCP-4,
CCL-18/PARC, and CCL-27/CTACK) known to attract
TH2 cells, and psoriasis skin expressed CXC chemokines,
which are known to elicit TH1 responses. There was also
decreased expression of antimicrobial genes in AD skin
lesions compared to psoriasis skin lesions (Nomura et  al.,
2003). A  limiting factor in these studies is tissue hetero-
geneity in biopsy samples that may limit interpretation
of expression results, since tissue heterogeneity is a major
confounding variable in microarray experiments (Group
TTABPW, 2004). There have been a few studies in which
isolated cell types were arrayed. When mRNA transcripts
in peripheral blood mononuclear cells were compared in
AD patients and normal controls, four transcripts—IFN-γ,
TRAIL (TNF-related apoptosis-inducing ligand), ISGF-3
(STAT1), and defensin-1—were found to be differen-
tially expressed (Heishi et al., 2002). Monocytes from AD
patients were also examined, and various genes were upreg-
ulated, including those involved in MHC class 1 antigen
presentation, recognition of microorganisms, and apopto-
sis (Nagata et al., 2003). Both these studies, however, had
modest sample sizes.
Recently there have been novel studies using microar-
rays to study complex traits. Classical quantitative traits
represent gross clinical measurements that may not be
related directly to the cellular or biological processes giv-
ing rise to that trait. Using transcript abundances as quan-
titative traits (known as expression quantitative trait loci,
or eQTLs) is more directly related to pathophysiological
processes (Schadt et al., 2003). As these traits are directly
related to the effects of genetic polymorphisms, they offer
greater power than clinical phenotypes. By combining
gene-expression data with genetic-inheritance information,
it has been demonstrated that gene-expression data can be
used to refine the disease phenotype and implicate pathways
for disease causation (Schadt et al., 2003). The heritability
of gene expression traits has been demonstrated in segregat-
ing human populations (Monks et al., 2004).
For most complex diseases, there has been only limited
success of replication with linkage analysis, which fails to
explain the genetic epidemiology of the disease (Altmuller
et al., 2001). Success depends on various factors that are dif-
ficult to control. Confounding factors include the low heri-
tability of complex traits, the modest effect of these traits,
6 9 0   •   G enomics in C l inica l P ractice
different polymorphic markers used in different studies,
and variations in disease definition and severity. Association
analysis is more powerful for the detection of common dis-
ease alleles that confer moderate disease risks (Risch and
Merikangas, 1996), though candidate-gene studies also
have their drawbacks when the fundamental cause of the
disease is not known.
With the completion of the human genome sequence,
improvements in SNP genotyping technology and the ini-
tiation of the International HapMap Project (International
HapMap Project, 2003), genome-wide association studies
(reviewed in Hirschhorn and Daly, 2005) are now possible,
in which a dense set of SNPs across the whole genome is
genotyped to survey the most common genetic variations
for a role in disease or to identify the quantitative traits that
are risks factors for disease (Hirschhorn and Daly, 2005).
The genome-wide association approach is unbiased, as no
assumption is made about the causal gene. The genome-wide
approach affords increased power and efficiency over link-
age studies. This approach has been successful in identifying
genes associated with susceptibility to myocardial infarc-
tion (Ozaki et  al., 2002), age-related macular degenera-
tion (Klein et al., 2005), rheumatoid arthritis (Yamamoto
and Yamada, 2005), and Parkinson disease (Maraganore
et al., 2005). In order to study the genetic basis of natural
variation in gene expression, genome-wide linkage analy-
sis and the determinants of ~1,000 expression phenotypes
were mapped (Morley et al., 2004). These determinants of
human gene expression were recently mapped by regional
and genome-wide association (Cheung et  al., 2005). The
genome-wide association analysis confirmed the linkage
results and pointed to the same location as the genome scans
in about 50% of the phenotypes (Cheung et al., 2005). This
suggests that large-scale association studies are feasible to
identify genes for complex traits.
C O N C LU S I O N
The standard model—that atopic diseases are mediated by
IgE responses to common allergens, and that the inflam-
matory milieu is maintained by T cells skewed to enhance
continued IgE production—is now being challenged. The
results from the genome screens and candidate-gene studies
for AD suggest that the predisposition to AD rests within
the skin itself. Genetic studies have identified classes of genes
by candidate-gene studies that are associated with atopic
mechanisms (FcεRIβ, MCC, CC chemokines), mediators
of the TH2-dominated inflammatory milieu (IL4, IL13),
barrier function and protease activity of the skin (FLG,
SPINK5, SCCE), and innate immunity genes (CD14,
TLR2, CARD15, CARD4). Genome screens have identi-
fied few regions containing candidate genes. The advent of
DNA microarrays with eQTL profiling and whole-genome
association studies will identify novel pathways and genes in
this complex disease.
Identifying the alleles that affect the risk of developing
AD will help improve our understanding of disease etiology
and classification, and their genes and proteins may serve as
future therapeutic targets.
AC K N OW L E D G E M E N T S
Dr. Nilesh Morar is supported by the Wellcome Trust and
has also received funding from START (Skin Treatment
and Research Trust, London).
REFERENCES
Ahmad-Nejad P, Mrabet-Dahbi S, Breuer K, Klotz M, Werfel T, Herz U,
et al. (2004). The Toll-like receptor 2 R753Q polymorphism defines
a subgroup of patients with atopic dermatitis having severe pheno-
type. J Allergy Clin Immunol 113:565–567.
Akolkar PN, Gulwani-Akolkar B, Heresbach D, Lin XY, Fisher S, Katz
S, et al. (1997). Differences in risk of Crohn’s disease in offspring
of mothers and fathers with inflammatory bowel disease. Am J
Gastroenterol 92:2241–2244.
Alam R. (1997). Chemokines in allergic inflammation. J Allergy Clin
Immunol 99:273–277.
Allen M, Heinzmann A, Noguchi E, Abecasis G, Broxholme J, Ponting
CP, et  al. (2003). Positional cloning of a novel gene influencing
asthma from chromosome 2q14. Nat Genet 35:258–263.
Altmuller J, Palmer LJ, Fischer G, Scherb H, Wjst M. (2001).
Genomewide scans of complex human diseases: true linkage is hard
to find. Am J Hum Genet 69:936–950.
Anderson GG, Leaves NI, Bhattacharyya S, Zhang Y, Walshe V,
Broxholme J, et  al. (2002). Positive association to IgE levels
and a physical map of the 13q14 atopy locus. Eur J Hum Genet
10:266–270.
Arbour NC, Lorenz E, Schutte BC, Zabner J, Kline JN, Jones M, et al.
(2000). TLR4 mutations are associated with endotoxin hypore-
sponsiveness in humans. Nat Genet 25:187–191.
Badertscher K, Bronnimann M, Karlen S, Braathen LR, Yawalkar N.
(2005). Mast cell chymase is increased in chronic atopic dermatitis
but not in psoriasis. Arch Dermatol Res 296:503–506.
Bai B, Tanaka K, Tazawa T, Yamamoto N, Sugiura H. (2005). Association
between RANTES promoter polymorphism–401A and enhanced
RANTES production in atopic dermatitis patients. J Dermatol Sci
39:189–191.
Baldini M, Lohman IC, Halonen M, Erickson RP, Holt PG, Martinez
FD. (1999). A polymorphism* in the 5′ flanking region of the
CD14 gene is associated with circulating soluble CD14 levels and
with total serum immunoglobulin E. Am J Respir Cell Mol Biol
20:976–983.
Bannister MJ, Freeman S. (2000). Adult-onset atopic dermatitis.
Australas J Dermatol 41:225–228.
Bennett ST, Todd JA. (1996). Human type 1 diabetes and the insulin
gene: principles of mapping polygenes. Annu Rev Genet 30:343–370.
G enetics and G enomics of S k in D iseases I   •   6 9 1
Beyer K, Nickel R, Freidhoff L, Bjorksten B, Huang SK, Barnes KC,
et  al. (2000). Association and linkage of atopic dermatitis with
chromosome 13q12–14 and 5q31–33 markers. J Invest Dermatol
115:906–908.
Bitoun E, Chavanas S, Irvine AD, Lonie L, Bodemer C, Paradisi M, et al.
(2002). Netherton syndrome:  disease expression and spectrum of
SPINK5 mutations in 21 families. J Invest Dermatol 118:352–361.
Bradley M, Soderhall C, Luthman H, Wahlgren CF, Kockum I,
Nordenskjold M. (2002). Susceptibility loci for atopic dermatitis on
chromosomes 3, 13, 15, 17 and 18 in a Swedish population. Hum
Mol Genet 11:1539–1548.
Bouzigon E, Dizier MH, Krahenbuhl C, Lemainque A, Annesi-Maesano
I, Betard C, et  al. (2004). Clustering patterns of LOD scores for
asthma-related phenotypes revealed by a genome-wide screen in 295
French EGEA families. Hum Mol Genet 13:3103–3113.
Burden AD, Javed S, Bailey M, Hodgins M, Connor M, Tillman D.
(1998). Genetics of psoriasis:  paternal inheritance and a locus on
chromosome 6p. J Invest Dermatol 110:958–960.
Cargill M, Altshuler D, Ireland J, Sklar P, Ardlie K, Patil N, et al. (1999).
Characterization of single-nucleotide polymorphisms in coding
regions of human genes. Nat Genet 22:231–238.
Carlson CS, Eberle MA, Kruglyak L, Nickerson DA. (2004). Mapping
complex disease loci in whole-genome association studies. Nature
429:446–452.
Chae SC, Song JH, Lee YC, Kim JW, Chung HT. (2003). The associa-
tion of the exon 4 variations of Tim-1 gene with allergic diseases in
a Korean population. Biochem Biophys Res Commun 312:346–350.
Chamlin SL, Kao J, Frieden IJ, Sheu MY, Fowler AJ, Fluhr JW, et  al.
(2002). Ceramide-dominant barrier repair lipids alleviate child-
hood atopic dermatitis: changes in barrier function provide a sensi-
tive indicator of disease activity. J Am Acad Dermatol 47:198–208.
Chavanas S, Bodemer C, Rochat A, Hamel-Teillac D, Ali M, Irvine AD,
et  al. (2000a). Mutations in SPINK5, encoding a serine protease
inhibitor, cause Netherton syndrome. Nat Genet 25:141–142.
Chavanas S, Garner C, Bodemer C, Ali M, Teillac DH, Wilkinson J,
et  al. (2000b). Localization of the Netherton syndrome gene to
chromosome 5q32, by linkage analysis and homozygosity mapping.
Am J Hum Genet 66:914–921.
Cheung VG, Spielman RS. (2002). The genetics of variation in gene
expression. Nat Genet 32 Suppl:522–525.
Cheung VG, Spielman RS, Ewens KG, Weber TM, Morley M, Burdick
JT. (2005). Mapping determinants of human gene expression by
regional and genome-wide association. Nature 437:1365–1369.
Capon F, Novelli G, Semprini S, Clementi M, Nudo M, Vultaggio P, et al.
(1999). Searching for psoriasis susceptibility genes in Italy: genome
scan and evidence for a new locus on chromosome 1. J Invest
Dermatol 112:32–35.
Carneiro LA, Travassos LH, Philpott DJ. (2004). Innate immune rec-
ognition of microbes through Nod1 and Nod2:  implications for
disease. Microbes Infect 6:609–616.
Coca AF, Cooke RA. (1923). On the phenomenon of hypersensitive-
ness. J Immunol 8:163–182.
Cookson WO, Sharp PA, Faux JA, Hopkin JM. (1989). Linkage between
immunoglobulin E responses underlying asthma and rhinitis and
chromosome 11q. Lancet 1:1292–1295.
Cookson WO, Young RP, Sandford AJ, Moffatt MF, Shirakawa T, Sharp
PA, et al. (1992). Maternal inheritance of atopic IgE responsiveness
on chromosome 11q. Lancet 340:381–384.
Cookson WO, Ubhi B, Lawrence R, Abecasis GR, Walley AJ, Cox HE,
et al. (2001). Genetic linkage of childhood atopic dermatitis to pso-
riasis susceptibility loci. Nat Genet 27:372–373.
Cookson W. (2002). Genetics and genomics of asthma and allergic dis-
eases. Immunol Rev 190:195–206.
Cox HE, Moffatt MF, Faux JA, Walley AJ, Coleman R, Trembath RC,
et  al. (1998). Association of atopic dermatitis to the beta subunit
of the high affinity immunoglobulin E receptor. Br J Dermatol
138:182–187.
Dale BA, Resing KA, Lonsdale-Eccles JD. (1985). Filaggrin:  a keratin
filament associated protein. Ann N Y Acad Sci 455:330–342.
Daniels SE, Bhattacharrya S, James A, Leaves NI, Young A, Hill MR,
et al. (1996). A genome-wide search for quantitative trait loci under-
lying asthma. Nature 383:247–250.
Descargues P, Deraison C, Bonnart C, Kreft M, Kishibe M,
Ishida-Yamamoto A, et  al. (2005). Spink5-deficient mice mimic
Netherton syndrome through degradation of desmoglein 1 by epi-
dermal protease hyperactivity. Nat Genet 37:56–65.
Di Nardo A, Wertz P, Giannetti A, Seidenari S. (1998). Ceramide and
cholesterol composition of the skin of patients with atopic dermati-
tis. Acta Derm Venereol 78:27–30.
Dizier MH, Besse-Schmittler C, Guilloud-Bataille M, Annesi-Maesano
I, Boussaha M, Bousquet J, et al. (2000). Genome screen for asthma
and related phenotypes in the French EGEA study. Am J Respir Crit
Care Med 162:1812–1818.
Dizier MH, Bouzigon E, Guilloud-Bataille M, Betard C, Bousquet
J, Charpin D, et  al. (2005). Genome screen in the French EGEA
study: detection of linked regions shared or not shared by allergic
rhinitis and asthma. Genes Immun 6:95–102.
Dold S, Wjst M, von Mutius E, Reitmeir P, Stiepel E. (1992). Genetic
risk for asthma, allergic rhinitis, and atopic dermatitis. Arch Dis
Child 67:1018–1022.
Duffy DL, Martin NG, Battistutta D, Hopper JL, Mathews JD. (1990).
Genetics of asthma and hay fever in Australian twins. Am Rev Respir
Dis 142:1351–1358.
Enlund F, Samuelsson L, Enerback C, Inerot A, Wahlstrom J, Yhr M,
et  al. (1999). Psoriasis susceptibility locus in chromosome region
3q21 identified in patients from southwest Sweden. Eur J Hum
Genet 7:783–790.
Flohr C, Johansson SG, Wahlgren CF, Williams H. (2004). How atopic
is atopic dermatitis? J Allergy Clin Immunol 114:150–158.
Folster-Holst R, Stoll M, Koch WA, Hampe J, Christophers E, Schreiber
S. (2005). Lack of association of SPINK5 polymorphisms with
nonsyndromic atopic dermatitis in the population of Northern
Germany. Br J Dermatol 152:1365–1367.
Forrest S, Dunn K, Elliott K, Fitzpatrick E, Fullerton J, McCarthy M,
et  al. (1999). Identifying genes predisposing to atopic eczema. J
Allergy Clin Immunol 104:1066–1070.
Gaffney PM, Ortmann WA, Selby SA, Shark KB, Ockenden TC, Rohlf
KE, et al. (2000). Genome screening in human systemic lupus ery-
thematosus: results from a second Minnesota cohort and combined
analyses of 187 sib-pair families. Am J Hum Genet 66:547–556.
Groneberg DA, Bester C, Grutzkau A, Serowka F, Fischer A, Henz BM,
et al. (2005). Mast cells and vasculature in atopic dermatitis—poten-
tial stimulus of neoangiogenesis. Allergy 60:90–97.
Group TTABPW. (2004). Expression profiling—best practices for
data generation and interpretation in clinical trials. Nat Rev Genet
5:229–237.
Haagerup A, Bjerke T, Schiotz PO, Binderup HG, Dahl R, Kruse TA.
(2002). Asthma and atopy—a total genome scan for susceptibility
genes. Allergy 57:680–686.
Haagerup A, Bjerke T, Schiotz PO, Dahl R, Binderup HG, Tan Q, et al.
(2004). Atopic dermatitis—a total genome-scan for susceptibility
genes. Acta Derm Venereol 84:346–352.
Hakonarson H, Bjornsdottir US, Halapi E, Palsson S, Adalsteinsdottir
E, Gislason D, et al. (2002). A major susceptibility gene for asthma
maps to chromosome 14q24. Am J Hum Genet 71:483–1491.
Hall JG. (1990). Genomic imprinting. Arch Dis Child 65:1013–1015.
Halushka MK, Fan JB, Bentley K, Hsie L, Shen N, Weder A, et  al.
(1999). Patterns of single-nucleotide polymorphisms in candidate
genes for blood-pressure homeostasis. Nat Genet 22:239–247.
Hampe J, Cuthbert A, Croucher PJ, Mirza MM, Mascheretti S, Fisher
S, et al. (2001). Association between insertion mutation in NOD2
gene and Crohn’s disease in German and British populations. Lancet
357:1925–1928.
6 9 2   •   G enomics in C l inica l P ractice
Hanifin JM, Rajka G. (1980). Diagnostic features of atopic dermatitis.
Acta Derm Venereol Suppl (Stockh) 92:44–47.
He JQ, Chan-Yeung M, Becker AB, Dimich-Ward H, Ferguson AC,
Manfreda J, et al. (2003). Genetic variants of the IL13 and IL4 genes
and atopic diseases in at-risk children. Genes Immun 4:385–389.
Heishi M, Kagaya S, Katsunuma T, Nakajima T, Yuki K, Akasawa A,
et al. (2002). High-density oligonucleotide array analysis of mRNA
transcripts in peripheral blood cells of severe atopic dermatitis
patients. Int Arch Allergy Immunol 129:57–66.
Hirschhorn JN, Daly MJ. (2005). Genome-wide association studies for
common diseases and complex traits. Nat Rev Genet 6:95–108.
Hizawa N, Freidhoff LR, Chiu YF, Ehrlich E, Luehr CA, Anderson JL,
et  al. (1998). Genetic regulation of Dermatophagoides pteronyssi-
nus–specific IgE responsiveness: a genome-wide multipoint linkage
analysis in families recruited through 2 asthmatic sibs. Collaborative
Study on the Genetics of Asthma (CSGA). J Allergy Clin Immunol
102:436–442.
Hoffmann HJ, Olsen E, Etzerodt M, Madsen P, Thogersen HC, Kruse
T, et al. (1994). Psoriasin binds calcium and is upregulated by cal-
cium to levels that resemble those observed in normal skin. J Invest
Dermatol 103:370–375.
Hohl D, Mehrel T, Lichti U, Turner ML, Roop DR, Steinert PM.
(1991). Characterization of human loricrin. Structure and func-
tion of a new class of epidermal cell envelope proteins. J Biol Chem
266:6626–6636.
Hossny E, Aboul-Magd M, Bakr S. (2001). Increased plasma eotaxin in
atopic dermatitis and acute urticaria in infants and children. Allergy
56:996–1002.
Howard TD, Whittaker PA, Zaiman AL, Koppelman GH, Xu J, Hanley
MT, et al. (2001). Identification and association of polymorphisms
in the interleukin-13 gene with asthma and atopy in a Dutch popu-
lation. Am J Respir Cell Mol Biol 25:377–384.
Hummelshoj T, Bodtger U, Datta P, Malling HJ, Oturai A, Poulsen LK,
et al. (2003). Association between an interleukin-13 promoter poly-
morphism and atopy. Eur J Immunogenet 30:355–359.
Hysi P, Kabesch M, Moffatt MF, Schedel M, Carr D, Zhang Y, et  al.
(2005). NOD1 variation, immunoglobulin E and asthma. Hum Mol
Genet 14:935–941.
Imada T, Komorita N, Kobayashi F, Naito K, Yoshikawa T, Miyazaki M,
et al. (2002). Therapeutic potential of a specific chymase inhibitor in
atopic dermatitis. Jpn J Pharmacol 90:214–217.
Ingordo V, D’Andria G, D’Andria C. (2003). Adult-onset atopic dermati-
tis in a patch test population. Dermatology 206:197–203.
International HapMap Project. (2003). Nature 426:789–796.
Iwanaga T, McEuen A, Walls AF, Clough JB, Keith TP, Rorke S, et al.
(2004). Polymorphism of the mast cell chymase gene (CMA1) pro-
moter region: lack of association with asthma but association with
serum total immunoglobulin E levels in adult atopic dermatitis. Clin
Exp Allergy 34:1037–1042.
Jawaheer D, Seldin MF, Amos CI, Chen WV, Shigeta R, Monteiro J,
et al. (2001). A genomewide screen in multiplex rheumatoid arthri-
tis families suggests genetic overlap with other autoimmune diseases.
Am J Hum Genet 68:927–936.
Jeong CW, Ahn KS, Rho NK, Park YD, Lee DY, Lee JH, et al. (2003).
Differential in vivo cytokine mRNA expression in lesional skin of
intrinsic vs. extrinsic atopic dermatitis patients using semiquantita-
tive RT-PCR. Clin Exp Allergy 33:1717–1724.
Jinquan T, Vorum H, Larsen CG, Madsen P, Rasmussen HH, Gesser
B, et  al. (1996). Psoriasin:  a novel chemotactic protein. J Invest
Dermatol 107:5–10.
Johansson SG, Hourihane JO, Bousquet J, Bruijnzeel-Koomen C,
Dreborg S, Haahtela T, et  al. (2001). A revised nomenclature for
allergy. An EAACI position statement from the EAACI nomencla-
ture task force. Allergy 56:813–824.
Judge MR, Morgan G, Harper JI. (1994). A clinical and immunological
study of Netherton’s syndrome. Br J Dermatol 131:615–621.
Juhlin L, Johansson GO, Bennich H, Hogman C, Thyresson N. (1969).
Immunoglobulin E in dermatoses. Levels in atopic dermatitis and
urticaria. Arch Dermatol 100:12–16.
Jurgens M, Wollenberg A, Hanau D, de la Salle H, Bieber T. (1995).
Activation of human epidermal Langerhans cells by engagement
of the high affinity receptor for IgE, Fc epsilon RI. J Immunol
155:5184–5189.
Kabesch M, Peters W, Carr D, Leupold W, Weiland SK, von Mutius E.
(2003). Association between polymorphisms in caspase recruitment
domain containing protein 15 and allergy in two German popula-
tions. J Allergy Clin Immunol 111:813–817.
Kaburagi Y, Shimada Y, Nagaoka T, Hasegawa M, Takehara K, Sato S.
(2001). Enhanced production of CC-chemokines (RANTES,
MCP-1, MIP-1alpha, MIP-1beta, and eotaxin) in patients with
atopic dermatitis. Arch Dermatol Res 293:350–355.
Kato A, Fukai K, Oiso N, Hosomi N, Murakami T, Ishii M. (2003).
Association of SPINK5 gene polymorphisms with atopic dermatitis
in the Japanese population. Br J Dermatol 148:665–669.
Kawashima T, Kobayashi S, Miyano M, Ohya N, Naruse C, Tokuda
Y. (1989). [Senile type atopic dermatitis]. Nippon Hifuka Gakkai
Zasshi 99:1095–1103.
Kawashima T, Noguchi E, Arinami T, Kobayashi K, Otsuka F,
Hamaguchi H. (1998a). No evidence for an association between a
variant of the mast cell chymase gene and atopic dermatitis based
on case-control and haplotype-relative-risk analyses. Hum Hered
48:271–274.
Kawashima T, Noguchi E, Arinami T, Yamakawa-Kobayashi K,
Nakagawa H, Otsuka F, et  al. (1998b). Linkage and association
of an interleukin 4 gene polymorphism with atopic dermatitis in
Japanese families. J Med Genet 35:502–504.
Klein RJ, Zeiss C, Chew EY, Tsai JY, Sackler RS, Haynes C, et al. (2005).
Complement factor H polymorphism in age-related macular degen-
eration. Science 308:385–389.
Kondo H, Ichikawa Y, Imokawa G. (1998). Percutaneous sensitization
with allergens through barrier-disrupted skin elicits a Th2-dominant
cytokine response. Eur J Immunol 28:769–779.
Koppelman GH, Reijmerink NE, Colin Stine O, Howard TD,
Whittaker PA, Meyers DA, et al. (2001). Association of a promoter
polymorphism of the CD14 gene and atopy. Am J Respir Crit Care
Med 163:965–969.
Koppelman GH, Stine OC, Xu J, Howard TD, Zheng SL, Kauffman
HF, et  al. (2002). Genome-wide search for atopy susceptibil-
ity genes in Dutch families with asthma. J Allergy Clin Immunol
109:498–506.
Koumantaki Y, Giziaki E, Linos A, Kontomerkos A, Kaklamanis
P, Vaiopoulos G, et  al. (1997). Family history as a risk fac-
tor for rheumatoid arthritis:  a case-control study. J Rheumatol
24:1522–1526.
Kozma GT, Falus A, Bojszko A, Krikovszky D, Szabo T, Nagy A, et al.
(2002). Lack of association between atopic eczema/dermatitis syn-
drome and polymorphisms in the promoter region of RANTES and
regulatory region of MCP-1. Allergy 57:160–163.
Krathen RA, Hsu S. (2005). Failure of omalizumab for treatment of
severe adult atopic dermatitis. J Am Acad Dermatol 53:338–340.
Kuokkanen S, Gschwend M, Rioux JD, Daly MJ, Terwilliger JD, Tienari
PJ, et al. (1997). Genomewide scan of multiple sclerosis in Finnish
multiplex families. Am J Hum Genet 61:1379–1387.
Laitinen T, Daly MJ, Rioux JD, Kauppi P, Laprise C, Petays T, et  al.
(2001). A susceptibility locus for asthma-related traits on chromo-
some 7 revealed by genome-wide scan in a founder population. Nat
Genet 28:87–91.
Lander E, Kruglyak L. (1995). Genetic dissection of complex
traits: guidelines for interpreting and reporting linkage results. Nat
Genet 11:241–247.
Larsen FS, Holm NV, Henningsen K. (1986). Atopic dermatitis.
A genetic-epidemiologic study in a population-based twin sample.
J Am Acad Dermatol 15:487–494.
G enetics and G enomics of S k in D iseases I   •   6 9 3
Lammintausta K, Kalimo K, Raitala R, Forsten Y. (1991). Prognosis
of atopic dermatitis. A prospective study in early adulthood. Int J
Dermatol 30:563–568.
Lane JE, Cheyney JM, Lane TN, Kent DE, Cohen DJ. (2006).
Treatment of recalcitrant atopic dermatitis with omalizumab. J Am
Acad Dermatol 54:68–72.
Lange J, Heinzmann A, Zehle C, Kopp M. (2005). CT genotype of pro-
motor polymorphism C159T in the CD14 gene is associated with
lower prevalence of atopic dermatitis and lower IL-13 production.
Pediatr Allergy Immunol 16:456–457.
Lee YA, Wahn U, Kehrt R, Tarani L, Businco L, Gustafsson D, et  al.
(2000). A major susceptibility locus for atopic dermatitis maps to
chromosome 3q21. Nat Genet 26:470–473.
Lin S, Cicala C, Scharenberg AM, Kinet JP. (1996). The Fc (epsi-
lon) RI beta subunit functions as an amplifier of Fc (epsilon) RI
gamma-mediated cell activation signals. Cell 85:985–995.
Liu X, Nickel R, Beyer K, Wahn U, Ehrlich E, Freidhoff LR, et al. (2000).
An IL13 coding region variant is associated with a high total serum
IgE level and atopic dermatitis in the German multicenter atopy
study (MAS-90). J Allergy Clin Immunol 106:167–170.
Lorenz E, Mira JP, Cornish KL, Arbour NC, Schwartz DA. (2000).
A novel polymorphism in the Toll-like receptor 2 gene and its
potential association with staphylococcal infection. Infect Immun
68:6398–6401.
Luoma R, Koivikko A, Viander M. (1983). Development of asthma,
allergic rhinitis and atopic dermatitis by the age of five years. A pro-
spective study of 543 newborns. Allergy 38:339–346.
McGovern DP, Van Heel DA, Negoro K, Ahmad T, Jewell DP. (2003).
Further evidence of IBD5/CARD15 (NOD2) epistasis in the sus-
ceptibility to ulcerative colitis. Am J Hum Genet 73:1465–1466.
McGovern DP, Hysi P, Ahmad T, van Heel DA, Moffatt MF, Carey A,
et  al. (2005). Association between a complex insertion/deletion
polymorphism in NOD1 (CARD4) and susceptibility to inflam-
matory bowel disease. Hum Mol Genet 14:1245–1250.
Maestrini E, Monaco AP, McGrath JA, Ishida-Yamamoto A, Camisa C,
Hovnanian A, et al. (1996). A molecular defect in loricrin, the major
component of the cornified cell envelope, underlies Vohwinkel’s
syndrome. Nat Genet 13:70–77.
Magert HJ, Standker L, Kreutzmann P, Zucht HD, Reinecke M,
Sommerhoff CP, et al. (1999). LEKTI, a novel 15-domain type of
human serine proteinase inhibitor. J Biol Chem 274:21499–21502.
Magert HJ, Kreutzmann P, Standker L, Walden M, Drogemuller K,
Forssmann WG. (2002). LEKTI: a multidomain serine proteinase
inhibitor with pathophysiological relevance. Int J Biochem Cell Biol
34:573–576.
Majewski S, Favre M, Orth G, Jablonska S. (1998). Is human papilloma-
virus type 5 the putative autoantigen involved in psoriasis? J Invest
Dermatol 111:541–542.
Mao XQ, Shirakawa T, Yoshikawa T, Yoshikawa K, Kawai M, Sasaki S,
et al. (1996). Association between genetic variants of mast-cell chy-
mase and eczema. Lancet 348:581–583.
Mao XQ, Shirakawa T, Enomoto T, Shimazu S, Dake Y, Kitano H, et al.
(1998). Association between variants of mast cell chymase gene and
serum IgE levels in eczema. Hum Hered 48:38–41.
Maraganore DM, de Andrade M, Lesnick TG, Strain KJ, Farrer MJ,
Rocca WA, et  al. (2005). High-resolution whole-genome associa-
tion study of Parkinson disease. Am J Hum Genet 77:685–693.
Marshall D, Hardman MJ, Nield KM, Byrne C. (2001). Differentially
expressed late constituents of the epidermal cornified envelope. Proc
Natl Acad Sci U S A 98:13031–13036.
Mathias RA, Freidhoff LR, Blumenthal MN, Meyers DA, Lester L, King
R, et al. (2001). Genome-wide linkage analyses of total serum IgE
using variance components analysis in asthmatic families. Genet
Epidemiol 20:340–355.
Melnik B, Hollmann J, Plewig G. (1988). Decreased stratum corneum
ceramides in atopic individuals—a pathobiochemical factor in xero-
sis? Br J Dermatol 119:547–549.
Meyers DA, Postma DS, Stine OC, Koppelman GH, Ampleford EJ,
Jongepier H, et al. (2005). Genome screen for asthma and bronchial
hyperresponsiveness:  interactions with passive smoke exposure. J
Allergy Clin Immunol 115:1169–1175.
Miedzobrodzki J, Kaszycki P, Bialecka A, Kasprowicz A. (2002).
Proteolytic activity of Staphylococcus aureus strains isolated from the
colonized skin of patients with acute-phase atopic dermatitis. Eur J
Clin Microbiol Infect Dis 21:269–276.
Miller HR, Pemberton AD. (2002). Tissue-specific expression of mast
cell granule serine proteinases and their role in inflammation in the
lung and gut. Immunology 105:375–390.
Mischke D, Korge BP, Marenholz I, Volz A, Ziegler A. (1996). Genes
encoding structural proteins of epidermal cornification and S100
calcium-binding proteins form a gene complex (“epidermal differen-
tiation complex”) on human chromosome 1q21. J Invest Dermatol
106:989–992.
Moffatt MF, Sharp PA, Faux JA, Young RP, Cookson WO, Hopkin JM.
(1992). Factors confounding genetic linkage between atopy and
chromosome 11q. Clin Exp Allergy 22:1046–1051.
Moffatt MF AM, Reynolds NJ, Meggitt ST, Cookson WO, Barker JN.
(2005). A Case-Control Study Investigating Candidate Genes in
Adult Eczema. 35th Annual European Society of Dermatological
Research Meeting. Tubingen, Germany: Blackwell, A22.
Monks SA, Leonardson A, Zhu H, Cundiff P, Pietrusiak P, Edwards S,
et al. (2004). Genetic inheritance of gene expression in human cell
lines. Am J Hum Genet 75:1094–1105.
Morley M, Molony CM, Weber TM, Devlin JL, Ewens KG, Spielman
RS, et  al. (2004). Genetic analysis of genome-wide variation in
human gene expression. Nature 430:743–747.
Nagata N, Oshida T, Yoshida NL, Yuyama N, Sugita Y, Tsujimoto
G, et  al. (2003). Analysis of highly expressed genes in mono-
cytes from atopic dermatitis patients. Int Arch Allergy Immunol
132:156–167.
Nickel RG, Casolaro V, Wahn U, Beyer K, Barnes KC, Plunkett BS,
et al. (2000). Atopic dermatitis is associated with a functional muta-
tion in the promoter of the C-C chemokine RANTES. J Immunol
164:1612–1616.
Nishio Y, Noguchi E, Shibasaki M, Kamioka M, Ichikawa E, Ichikawa K,
et al. (2003). Association between polymorphisms in the SPINK5
gene and atopic dermatitis in the Japanese. Genes Immun 4:515–517.
Nomura I, Gao B, Boguniewicz M, Darst MA, Travers JB, Leung DY.
(2003a). Distinct patterns of gene expression in the skin lesions of
atopic dermatitis and psoriasis: a gene microarray analysis. J Allergy
Clin Immunol 112:1195–1202.
Nomura I, Goleva E, Howell MD, Hamid QA, Ong PY, Hall CF, et al.
(2003b). Cytokine milieu of atopic dermatitis, as compared to pso-
riasis, skin prevents induction of innate immune response genes. J
Immunol 171:3262–3269.
Novak N, Kraft S, Haberstok J, Geiger E, Allam P, Bieber T. (2002). A
reducing microenvironment leads to the generation of Fc epsilon
RI high inflammatory dendritic epidermal cells (IDEC). J Invest
Dermatol 119:842–849.
Novak N, Allam JP, Bieber T. (2003). Allergic hyperreactivity to micro-
bial components: a trigger factor of “intrinsic” atopic dermatitis? J
Allergy Clin Immunol 112:215–216.
Novak N, Bieber T. (2003). Allergic and nonallergic forms of atopic dis-
eases. J Allergy Clin Immunol 112:252–262.
Novembre E, Cianferoni A, Lombardi E, Bernardini R, Pucci N,
Vierucci A. (2001). Natural history of “intrinsic” atopic dermatitis.
Allergy 56:452–453.
Ober C, Tsalenko A, Parry R, Cox NJ. (2000). A second-generation
genomewide screen for asthma-susceptibility alleles in a founder
population. Am J Hum Genet 67:1154–1162.
Ogura Y, Bonen DK, Inohara N, Nicolae DL, Chen FF, Ramos R, et al.
(2001). A frameshift mutation in NOD2 associated with suscepti-
bility to Crohn’s disease. Nature 411:603–606.
6 9 4   •   G enomics in C l inica l P ractice
Ong PY, Ohtake T, Brandt C, Strickland I, Boguniewicz M, Ganz T,
et  al. (2002). Endogenous antimicrobial peptides and skin infec-
tions in atopic dermatitis. N Engl J Med 347:1151–1160.
Oppel T, Schuller E, Gunther S, Moderer M, Haberstok J, Bieber T, et al.
(2000). Phenotyping of epidermal dendritic cells allows the differ-
entiation between extrinsic and intrinsic forms of atopic dermatitis.
Br J Dermatol 143:1193–1198.
Ozaki K, Ohnishi Y, Iida A, Sekine A, Yamada R, Tsunoda T, et  al.
(2002). Functional SNPs in the lymphotoxin-alpha gene that are
associated with susceptibility to myocardial infarction. Nat Genet
32:650–654.
Ozkaya E. (2005). Adult-onset atopic dermatitis. J Am Acad Dermatol
52:579–582.
Palmer LJ, Burton PR, Faux JA, James AL, Musk AW, Cookson WO.
(2000). Independent inheritance of serum immunoglobulin E con-
centrations and airway responsiveness. Am J Respir Crit Care Med
161:1836–1843.
Palmer CN, Irvine AD, Terron-Kwiatkowski A, Zhao Y, Liao H, Lee SP,
et  al. (2006). Common loss-of-function variants of the epidermal
barrier protein filaggrin are a major predisposing factor for atopic
dermatitis. Nat Genet 38(4):441–446.
Park CW, Lee BH, Han HJ, Lee CH, Ahn HK. (2005). Tacrolimus
decreases the expression of eotaxin, CCR3, RANTES and
interleukin-5 in atopic dermatitis. Br J Dermatol 152:1173–1181.
Pasare C, Medzhitov R. (2005). Control of B-cell responses by Toll-like
receptors. Nature 438:364–368.
Pascale E, Tarani L, Meglio P, Businco L, Battiloro E, Cimino-Reale G,
et al. (2001). Absence of association between a variant of the mast
cell chymase gene and atopic dermatitis in an Italian population.
Hum Hered 51:177–179.
Punnonen J, Aversa G, Cocks BG, McKenzie AN, Menon S, Zurawski
G, et al. (1993). Interleukin 13 induces interleukin 4-independent
IgG4 and IgE synthesis and CD23 expression by human B cells. Proc
Natl Acad Sci U S A 90:3730–3734.
Rafatpanah H, Bennett E, Pravica V, McCoy MJ, David TJ, Hutchinson
IV, et  al. (2003). Association between novel GM-CSF gene poly-
morphisms and the frequency and severity of atopic dermatitis. J
Allergy Clin Immunol 112:593–598.
Ramoz N, Rueda LA, Bouadjar B, Favre M, Orth G. (1999). A suscep-
tibility locus for epidermodysplasia verruciformis, an abnormal pre-
disposition to infection with the oncogenic human papillomavirus
type 5, maps to chromosome 17qter in a region containing a psoria-
sis locus. J Invest Dermatol 112:259–263.
Riedler J, Braun-Fahrlander C, Eder W, Schreuer M, Waser M, Maisch S,
et al. (2001). Exposure to farming in early life and development of
asthma and allergy: a cross-sectional survey. Lancet 358:1129–1133.
Risch N, Merikangas K. (1996). The future of genetic studies of complex
human diseases. Science 273:1516–1517.
Romagnani S. (2000). The role of lymphocytes in allergic disease. J
Allergy Clin Immunol 105:399–408.
Rystedt I. (1985). Long term follow-up in atopic dermatitis. Acta Derm
Venereol Suppl (Stockh) 114:117–120.
Sawcer S, Jones HB, Feakes R, Gray J, Smaldon N, Chataway J, et  al.
(1996). A genome screen in multiple sclerosis reveals susceptibility
loci on chromosome 6p21 and 17q22. Nat Genet 13:464–468.
Schafer BW, Wicki R, Engelkamp D, Mattei MG, Heizmann CW.
(1995). Isolation of a YAC clone covering a cluster of nine S100
genes on human chromosome 1q21:  rationale for a new nomen-
clature of the S100 calcium-binding protein family. Genomics
25:638–643.
Schadt EE, Monks SA, Drake TA, Lusis AJ, Che N, Colinayo V, et al.
(2003). Genetics of gene expression surveyed in maize, mouse and
man. Nature 422:297–302.
Schultz Larsen F. (1993). Atopic dermatitis:  a genetic-epidemiologic
study in a population-based twin sample. J Am Acad Dermatol
28:719–723.
Schultz-Larsen FHJ. (2002). Epidemiology of atopic dermatitis.
Immunol Allergy Clin North Am 22:1–24.
Seguchi T, Cui CY, Kusuda S, Takahashi M, Aisu K, Tezuka T. (1996).
Decreased expression of filaggrin in atopic skin. Arch Dermatol Res
288:442–446.
Smith FJ, Irvine AD, Terron-Kwiatkowski A, Sandilands A, Campbell
LE, Zhao Y, et al. (2006). Loss-of-function mutations in the gene
encoding filaggrin cause ichthyosis vulgaris. Nat Genet 38:337–342.
Soderhall C, Bradley M, Kockum I, Wahlgren CF, Luthman H,
Nordenskjold M. (2001). Linkage and association to candidate regions
in Swedish atopic dermatitis families. Hum Genet 109:129–135.
Song PI, Park YM, Abraham T, Harten B, Zivony A, Neparidze N,
et  al. (2002). Human keratinocytes express functional CD14 and
Toll-like receptor 4. J Invest Dermatol 119:424–432.
Sugiura H, Ebise H, Tazawa T, Tanaka K, Sugiura Y, Uehara M, et al.
(2005). Large-scale DNA microarray analysis of atopic skin lesions
shows overexpression of an epidermal differentiation gene cluster in
the alternative pathway and lack of protective gene expression in the
cornified envelope. Br J Dermatol 152:146–149.
Taha RA, Minshall EM, Leung DY, Boguniewicz M, Luster A, Muro
S, et  al. (2000). Evidence for increased expression of eotaxin and
monocyte chemotactic protein-4 in atopic dermatitis. J Allergy Clin
Immunol 105:1002–1007.
Tanaka K, Sugiura H, Uehara M, Sato H, Hashimoto-Tamaoki T,
Furuyama J. (1999). Association between mast cell chymase geno-
type and atopic eczema: comparison between patients with atopic
eczema alone and those with atopic eczema and atopic respiratory
disease. Clin Exp Allergy 29:800–803.
Tay YK, Khoo BP, Goh CL. (1999). The profile of atopic dermatitis in a
tertiary dermatology outpatient clinic in Singapore. Int J Dermatol
38:689–692.
Taylor B, Wadsworth J, Wadsworth M, Peckham C. (1984). Changes in
the reported prevalence of childhood eczema since the 1939–45
war. Lancet 2:1255–1257.
Thomas WR, Smith WA, Hales BJ, Mills KL, O’Brien RM. (2002).
Characterization and immunobiology of house dust mite allergens.
Int Arch Allergy Immunol 129:1–18.
Tomfohrde J, Silverman A, Barnes R, Fernandez-Vina MA, Young
M, Lory D, et  al. (1994). Gene for familial psoriasis susceptibil-
ity mapped to the distal end of human chromosome 17q. Science
264:1141–1145.
Tosh K, Meisner S, Siddiqui MR, Balakrishnan K, Ghei S, Golding M,
et al. (2002). A region of chromosome 20 is linked to leprosy suscep-
tibility in a South Indian population. J Infect Dis 186:1190–1193.
Traherne JA, Hill MR, Hysi P, D’Amato M, Broxholme J, Mott R,
et  al. (2003). LD mapping of maternally and non-maternally
derived alleles and atopy in Fc epsilon RI-beta. Hum Mol Genet
12:2577–2585.
Trembath RC, Clough RL, Rosbotham JL, Jones AB, Camp RD,
Frodsham A, et  al. (1997). Identification of a major susceptibil-
ity locus on chromosome 6p and evidence for further disease loci
revealed by a two stage genome-wide search in psoriasis. Hum Mol
Genet 6:813–820.
Tsunemi Y, Saeki H, Nakamura K, Sekiya T, Hirai K, Fujita H, et  al.
(2002a). Eotaxin gene single nucleotide polymorphisms in the
promoter and exon regions are not associated with susceptibility
to atopic dermatitis, but two of them in the promoter region are
associated with serum IgE levels in patients with atopic dermatitis. J
Dermatol Sci 29:222–228.
Tsunemi Y, Saeki H, Nakamura K, Sekiya T, Hirai K, Kakinuma T,
et  al. (2002b). Interleukin-13 gene polymorphism G4257A is
G enetics and G enomics of S k in D iseases I   •   6 9 5
associated with atopic dermatitis in Japanese patients. J Dermatol
Sci 30:100–107.
Turner H, Kinet JP. (1999). Signalling through the high-affinity IgE
receptor Fc epsilon RI. Nature 402:B24–30.
Van Eerdewegh P, Little RD, Dupuis J, Del Mastro RG, Falls K, Simon
J, et al. (2002). Association of the ADAM33 gene with asthma and
bronchial hyperresponsiveness. Nature 418:426–430.
Vasilopoulos Y, Cork MJ, Murphy R, Williams HC, Robinson DA, Duff
GW, et al. (2004). Genetic association between an AACC insertion
in the 3′UTR of the stratum corneum chymotryptic enzyme gene
and atopic dermatitis. J Invest Dermatol 123:62–66.
Vorechovsky I, Webster AD, Plebani A, Hammarstrom L. (1999).
Genetic linkage of IgA deficiency to the major histocompatibility
complex: evidence for allele segregation distortion, parent-of-origin
penetrance differences, and the role of anti-IgA antibodies in disease
predisposition. Am J Hum Genet 64:1096–1109.
Wachter AM, Lezdey J. (1992). Treatment of atopic dermatitis with
alpha 1-proteinase inhibitor. Ann Allergy 69:407–414.
Walley AJ, Chavanas S, Moffatt MF, Esnouf RM, Ubhi B, Lawrence
R, et  al. (2001). Gene polymorphism in Netherton and common
atopic disease. Nat Genet 29:175–178.
Wang JY, Lin CG, Bey MS, Wang L, Lin FY, Huang L, et  al. (2005).
Discovery of genetic difference between asthmatic children with
high IgE level and normal IgE level by whole genome linkage dis-
equilibrium mapping using 763 autosomal STR markers. J Hum
Genet 50:249–258.
Warram JH, Krolewski AS, Gottlieb MS, Kahn CR. (1984). Differences
in risk of insulin-dependent diabetes in offspring of diabetic moth-
ers and diabetic fathers. N Engl J Med 311:149–152.
Watanabe N, Tomimori Y, Saito K, Miura K, Wada A, Tsudzuki M, et al.
(2002). Chymase inhibitor improves dermatitis in NC/Nga mice.
Int Arch Allergy Immunol 128:229–234.
Weidinger S, Rummler L, Klopp N, Wagenpfeil S, Baurecht HJ, Fischer
G, et al. (2005). Association study of mast cell chymase polymor-
phisms with atopy. Allergy 60:1256–1261.
Williams H, Robertson C, Stewart A, Ait-Khaled N, Anabwani G,
Anderson R, et  al. (1999). Worldwide variations in the preva-
lence of symptoms of atopic eczema in the International Study
of Asthma and Allergies in Childhood. J Allergy Clin Immunol
103:125–138.
Williams HC, Burney PG, Pembroke AC, Hay RJ. (1994). The U.K.
Working Party’s diagnostic criteria for atopic dermatitis. III.
Independent hospital validation. Br J Dermatol 131:406–416.
Williams HC. (2000). Epidemiology of atopic dermatitis. Clin Exp
Dermatol 25:522–529.
Wjst M, Fischer G, Immervoll T, Jung M, Saar K, Rueschendorf F, et al.
(1999). A genome-wide search for linkage to asthma. German
Asthma Genetics Group. Genomics 58:1–8.
Yamamoto K, Yamada R. (2005). Genome-wide single nucleotide
polymorphism analyses of rheumatoid arthritis. J Autoimmun 25
Suppl:12–15.
Yawalkar N, Uguccioni M, Scharer J, Braunwalder J, Karlen S, Dewald B,
et al. (1999). Enhanced expression of eotaxin and CCR3 in atopic
dermatitis. J Invest Dermatol 113:43–48.
Young RP, Sharp PA, Lynch JR, Faux JA, Lathrop GM, Cookson WO,
et al. (1992). Confirmation of genetic linkage between atopic IgE
responses and chromosome 11q13. J Med Genet 29:236–238.
Zhang Y, Leaves NI, Anderson GG, Ponting CP, Broxholme J, Holt
R, et al. (2003). Positional cloning of a quantitative trait locus on
chromosome 13q14 that influences immunoglobulin E levels and
asthma. Nat Genet 34:181–186.
 45.
GENETICS AND GENOMICS OF SKIN DISEASES, II:
GENOMICS OF PIGMENTATION AND SKIN CANCER
Eugene Healy
INTRODUCTION
Amongst the many physical distinguishing features, skin
pigmentation is by far the most obvious and attractive dif-
ferentiating feature across all human populations. Although
“politically incorrect,” terms like black and white peoples
are still in common use in some continents. From the
evolutionary perspective, skin pigmentation offers a good
example of selection by enormous ecological pressures span-
ning thousands of years of human existence. Differences in
responses to sunshine by fair-skinned peoples compared to
dark-skinned peoples is probably a good example of “selec-
tion advantage.” This is a good model of so-called Darwinian
factors’ offering an advantage to cope with selection pres-
sures according to ambient levels of sunshine. These factors
are now accepted to be analogous to hereditary factors best
described as genetic or genomic. These factors are highlighted
in this chapter with particular reference to their clinical sig-
nificance and utilization in understanding variations in skin
and hair pigmentation, the pathogenesis of the skin cancer,
histological and molecular diagnoses, therapeutic choices,
and prognostication.
S K I N A N D H A I R P I G M E N TAT I O N
Skin and hair pigmentation are determined mainly by
the total and relative amounts of two types of mela-
nin:  brown-black eumelanin and red-yellow phaeomela-
nin, in these tissues (Prota and Thomson, 1976; Thody
et  al., 1991). Melanin is synthesized in membrane-bound
organelles called melanosomes by specialized pigment cells
(melanocytes) in the hair follicle and basal layer of human
epidermis, with eumelanin synthesized in ellipsoid melano-
somes and phaeomelanin in spherical melanosomes (Seiji
et  al., 1963; Jimbow et  al., 1983). Melanin-laden mela-
nosomes are transported in vivo along the melanocyte’s
696
dendrites and then transferred to the growing hair and to
epidermal keratinocytes. Melanocytes are derived from the
neural crest, migrating as melanoblasts to the skin during
embryogenesis (Rawles, 1947). In humans, the numbers
of epidermal melanocytes vary according to skin site, with
more on the face and genitals, but they do not differ sig-
nificantly across racial groups (Staricco and Pinkus, 1957;
Glimcher et al., 1973). However, melanocytes in Negroid
skin produce greater numbers of and larger melanosomes
than Caucasian melanocytes, and the skin of people from
darker-skinned populations generally contains more mela-
nin than does Caucasian skin (Szabo et  al., 1969; Hunt
et al., 1995).
Based on the differences in skin cancer incidence and
susceptibility to sunburn between racial groups, melanin (in
particular, eumelanin) is considered photoprotective, pre-
venting ultraviolet radiation (UVR)–induced DNA dam-
age and cancer development in the skin (Crombie, 1979;
Kaidbey et al., 1979; Cress and Holly, 1997). This photo-
protective role is further supported by the fact that melanin
that has been transferred into neighboring keratinocytes in
the basal layer of human epidermis tends to cluster above the
nucleus as a type of “biological sunhat” (Figure 45.1). One
of the most important environmental influences on human
skin pigmentation is UVR from sunshine, which stimulates
melanin synthesis and its redistribution within the epider-
mis, resulting in tanning. However, tanning ability differs
between individuals, and this variation in individual tanning
responses following UVR exposure, coupled with variation
in sunburning responses, has been used as a mechanism for
phenotyping humans based on their “skin type.”
N O R M A L VA R I AT I O N I N S K I N
A N D H A I R C O L O R
Human skin and hair color varies widely. Normal skin color
ranges from very light to very dark, and normal hair color

Figure 45.1 
Masson-Fontana stain showing the accumulation of
melanosomes above keratinocyte nuclei, producing a “biological
sun-hat.”
ranges from blond, to red, to black. In addition, Caucasian
people with similarly light skin coloring differ in their abil-
ity to tan following exposure to UVR, both after single
doses and after multiple repeated exposures. This variation
in response to UVR was originally used to grade patients
with psoriasis according to their “Fitzpatrick skin type”
in order to calculate the UVR treatment doses for these
patients and avoid UVR-induced burning (Fitzpatrick,
1988). However, Fitzpatrick skin type is based on a combi-
nation of erythemal and tanning responses, and many sub-
jects do not fit neatly into its distinct subgroups (Rampen
et al., 1988). Despite this, in the absence of simple alterna-
tives, the Fitzpatrick scale or modifications of it have been
employed in studies investigating the genetics of normal
skin color. Previous studies had suggested that the range in
skin color from fair-skinned Caucasians to dark-skinned
Negroes results from the additive interaction of three to six
genes (Harrison and Owen, 1964; Stern, 1970). It is known
that over 350 loci affect pigmentation in mice (Montoliu
et  al., 2012), but it is thought that significantly fewer of
these loci affect pigmentation in humans, with alterations
in the following genes reported to be associated with varia-
tions in normal human pigmentation:  melanocortin 1
receptor (MC1R), solute carrier family 24 (sodium/potas-
sium/calcium exchanger) member 5 (SLC24A5), solute
carrier family 45 member 2 (SLC45A2, also known as
membrane-associated transporter protein MATP), tyrosi-
nase (TYR), tyrosinase-related protein 1 (TYRP1), OCA2
(also known as the P gene), agouti signaling protein (ASIP),
proopiomelanocortin (POMC), KIT ligand (KITLG), and
interferon regulatory factor 4 (IRF4).
G enetics and G enomics o f S kin D iseases , I I   •   6 9 7
To date, alterations of the MC1R gene on chromosome
16q24.3 have been most comprehensively shown to be
involved in normal variation in human skin and hair color.
MC1R encodes for a 7-pass transmembrane G-protein
coupled receptor, which is expressed by melanocytes. Its
ligand, alpha melanocyte-stimulating hormone (αMSH),
stimulates pigmentation of melanocytes and melanoma
cells in culture and causes skin darkening when injected
into humans (Lerner and McGuire, 1961). Although the
human receptor is only 317 amino acids in length, numer-
ous polymorphisms have been identified in the single 951
base pairs (bp) exon. Alternative splicing of second exons
can lead to a larger receptor with an additional 33 or 65
amino acids (Tan et al., 1999; Rouzaud et al., 2006), and
intergenic splicing of MC1R transcripts with the down-
stream tubulin-beta-III gene has also been reported (Dalziel
et al., 2011). Investigations have indicated that MC1R vari-
ants/polymorphisms in the single-exon-encoded receptor
(the dominant isoform in human skin and hair) are asso-
ciated with red hair and with fair skin type in Caucasians
(Valverde et al., 1995; Box et al., 1997; Smith et al., 1998;
Healy et al., 2000; Flanagan et al., 2000). The variant recep-
tors commonly found in red hair and fair skin bind αMSH
with similar affinity as the wild-type receptor, but exhibit
reduced signaling via cyclic-AMP, (adenosine monophos-
phate) yet retain the ability to signal via the ERK (extracel-
lular signal-regulated kinases) pathway (Schioth et al., 1999;
Robinson and Healy, 2002; Newton et  al., 2005; Herraiz
et al., 2011). In addition, some variants are expressed at a
lower level at the cell surface, accounting for part of their
compromised function (Beaumont et al., 2005). Transgenic
mice have demonstrated that “red-haired” variants fail to
rescue pigmentation to the same extent as wild-type MC1R,
indicating that these variants preferentially promote the
synthesis of phaeomelanin in vivo (Healy et al., 2001). As a
general rule, individuals with two variant MC1R alleles that
compromise function tend to have red hair, whereas those
with a single variant MC1R allele tend to have fair skin.
Perhaps predictably, considering the association between
red hair, fair skin, and freckling, MC1R variants have also
been associated with freckling (Smith et al., 1998; Bastiaens
et al., 2001a).
SLC24A5 on chromosome 15q21.1, whose ortholog is
responsible for the zebra fish “golden” phenotype, is involved
in determining variation in skin color between Caucasian
and Asian or Negroid races (Lamason et al., 2005). Several
reports of an evolutionarily conserved ancestral allele, Ala111
of SLC24A5, in the majority of African and East Asian
populations, but a Thr111 allele in European and American
populations, in addition to the relationship of these alleles
to variation in skin pigmentation levels measured by reflec-
tance in African-American and African-Caribbean admixed
populations, suggest that SLC24A5 is a determinant of
human skin color (Lamason et al., 2005; Cook et al., 2009).
Similarly, variation in this gene has been implicated as a fac-
tor influencing intrinsic skin pigmentation in a South Asian
population (Stokowski et al., 2007). SLC24A5 encodes for
a potassium-dependent sodium calcium exchanger trans-
membrane protein, called NCKX5, which is expressed
in the melanosome, and its knockdown by siRNA or tar-
geted mutation leads to a reduction in melanin production
(Chi et  al., 2006; Ginger et  al., 2008; Vogel et  al., 2008).
Conversely, in zebra fish golden mutants (with lighter pig-
mentation due to an alteration in slc24a5), rescue of darker
pigmentation by wild-type slc24a5 has been reported.
Certain alleles, including promoter polymorphisms,
of the SLC45A2 (MATP) gene on chromosome 5p13.2
have been associated with lighter or darker hair, skin, and
eye color in Caucasians, and found at different frequencies
in Caucasian and non-Caucasian populations (Nakayama
et al., 2002; Yuasa et al., 2004; Graf et al., 2005; Yuasa et al.,
2006; Graf et  al., 2007; Lucotte et  al., 2010). Functional
analyses in melanocytes/melanoma cell lines has shown
that the polymorphisms decrease SLC45A2 transcription,
and pigmentary alterations secondary to SLC45A2 muta-
tions have been reported in medaka fish, horses, chickens,
and Japanese quail (Fukamachi et  al., 2001; Mariat et  al.,
2003; Graf et  al., 2007; Gunnarsson et  al., 2007; Cook
et al., 2009).
In earlier reports investigating variation in normal skin
and hair color, no sequence variation in several pigmenta-
tion genes (including TYR, TYRP1, etc.) was noted in
relation to skin type or hair color (Tripathi et  al., 1991;
Johnston et  al., 1992; Box et  al., 1998). However, more
recently, coding variants in the TYR gene on 11q14.3 have
been associated with tanning ability, eye color, and freckles
(Sulem et al., 2007; Nan et al., 2009), and alterations at the
TYRP1 locus on 9p23 have been associated with eye color
and identified as a possible cause of Melanesian blond hair in
Solomon Islanders (Sulem et al., 2008; Kenny et al., 2012).
Some studies have suggested that the OCA2 gene on 15q12
may also influence variation in normal eye and hair color,
but additional evidence indicates that a single-nucleotide
polymorphism (SNP) in a putative regulatory element
within an intron in the upstream HERC2 locus may be
causative for lighter pigmentation by reducing expression of
the OCA2/P protein in melanocytes (Eiberg et al., 2008;
Han et  al., 2008; Sturm et  al., 2008; Mengel-From et  al.,
2009). Associations between SNPs in the ASIP gene on
20q11.2, which codes for the agouti protein, and lighter
6 9 8   •   G enomics in C l inica l P ractice
skin, freckling, and red/light hair and eye color have been
reported (Kanetsky et al., 2002; Bonilla et al., 2005; Sulem
et  al., 2008). The function of agouti has been extensively
investigated in studies on mice and cultured melanocytes,
which have demonstrated that this protein reduces sig-
naling via the melanocortin 1 receptor, thus indicating
that SNPs in ASIP probably affect this signaling pathway.
Conversely, low-frequency alterations in the POMC gene
on 2p23.3, which codes for proopiomelanocortin/αMSH,
have been documented in subjects deficient in POMC who
have red hair and/or lighter skin pigmentation, but obvious
pigmentary features are not seen in all cases (Krude et al.,
1998; Clément et  al., 2008). Sequence variation within
and upstream of the KITLG gene on 12q22 has also been
reported in association with hair color, and it is thought
(although not yet shown) that this might influence num-
bers of melanocytes in the hair follicle through reduced
KIT ligand/KIT receptor signaling (Sulem et  al., 2007;
Mengel-From et al., 2009). Alterations in the IRF4 gene on
6p25.3 show association with skin color/tanning and eye
color, but the exact mechanism of their effects on pigmenta-
tion is unclear (Sulem et al., 2007; Han et al., 2008).
INHERITED DISORDER S
O F   P I G M E N TAT I O N
Disease processes affecting pigmentation can be congeni-
tal or acquired, and involve many aspects of melanocyte
biology, including migration and survival of melanoblasts,
melanocyte cell viability, melanin synthesis, and melanin
distribution from melanocytes to keratinocytes. Many of
these disorders of pigmentation can affect animals as well as
humans, and much of the knowledge about the genetics of
normal and abnormal human pigmentation is derived origi-
nally from mouse genetics. In some cases where mutations
in the relevant gene cause a pigmentary disorder, SNPs in
the same gene result in variation in normal skin, eye, or hair
pigmentation.
Failures of melanocyte development and migration may
occur, such as in piebaldism, due to mutations in the KIT
proto-oncogene on 4q12 (Fleischman et al., 1991; Giebel
and Spritz, 1991). KIT is a transmembrane tyrosine kinase
receptor whose ligand is stem cell factor (also known as kit
ligand, steel factor, and mast cell growth factor). Signaling
through the receptor is necessary for melanoblast survival
during embryogenesis as they migrate through the meso-
derm, and for subsequent proliferation of melanoblasts/
melanocytes in the epidermis, in addition to its require-
ment for normal hematopoeisis and gametogenesis (Steel
et al., 1992; Yoshida et al., 1996). Mutations in the SNAI2
(SLUG) gene on 8q11 have been reported less frequently
in human piebaldism (Sánchez-Martín et  al., 2003). The
various forms of Waardenberg syndrome (WS types I–
III) are mediated by effects on microphthalmia-associated
transcription factor (MITF), which is critical for neural
crest as well as melanocyte development, either by muta-
tions of MITF on 3p14-p13 or mutations in genes (PAX3
on 2q35 and SOX10 on 22q13) whose products regulate
MITF gene expression (Watanabe et  al., 1998; Pingault
et al., 1998; Bondurand et al., 2000). WS type IV (associ-
ated with aganglionic megacolon) is caused by mutations
in the endothelin-B receptor (EDNRB) gene on 13q22
and endothelin-3 gene (EDN3 on 20q13.32, encoding the
endothelin-B receptor ligand), whose products normally
upregulate MITF, leading to defects in signaling required
for the terminal migration of melanoblasts (and enteric
neuron precursor cells) (Edery et al., 1996; Lee et al., 2003;
Sato-Jin et  al., 2008). Deletions of SOX10 can also cause
WS type IV as well as being responsible for some cases of
WS type II (Bondurand et al., 2007).
Other pigmentary disorders are due to defects in mela-
nin synthesis. In oculocutaneous albinism type 1 (OCA1),
missense, nonsense, and frameshift mutations in the
tyrosinase (TYR) gene have been reported (Tomita et al.,
1989; Giebel et  al., 1990; King et  al., 1991a). Tyrosinase
catalyzes the formation of various melanin intermedi-
ates during the synthesis of melanin (e.g., dihydroxyphe-
nylalanine [DOPA]/DOPAquinone from tyrosine, and
indole-5,6-quinone from dihydroxyindole). Based on the
clustering of mutations in certain areas (including those
encoding copper-binding sites necessary for enzymatic
activity), it was considered that mutations in both copies
of the TYR gene resulted in inactive forms of tyrosinase
in the melanosome (King et  al., 1991a). However, tyrosi-
nase normally traffics from the endoplasmic reticulum via
the trans-Golgi network to the melanosome, and subse-
quent investigation identified that mutant tyrosinase may
be retained by the endoplasmic reticulum and fail to reach
the melanosome (Halaban et al., 2000; Chaki et al., 2011).
A  temperature-sensitive variant of OCA1 has also been
described, with some peripheral pigmentation on colder
areas of skin, including the arms and legs of affected sub-
jects, and is due to the missense mutations in TYR allow-
ing the retention of partial tyrosinase activity at lower
temperatures (31oC) but resulting in almost complete lack
of enzyme activity at 37oC (Giebel et al., 1991; King et al.,
1991b).
Oculocutaneous albinism type II (OCA2, tyrosinase-
positive albinism) can be phenotypically similar in
G enetics and G enomics o f S kin D iseases , I I   •   6 9 9
Caucasians to OCA1, but is more frequent in Negroid pop-
ulations in whom it causes yellow (or yellow-red) hair and
multiple pigmented freckles ranging from millimeters to
centimeters in diameter. In contrast to OCA1, pigmented
nevi may be present in Caucasian individuals with OCA2.
Nystagmus and photophobia may also be present. Missense,
nonsense, and frameshift mutations in the OCA2 (P) gene,
the human homolog of the mouse pink-eyed dilution locus,
and deletions of this locus are responsible (chromosome
15q11.2-q12) (Rinchik et al., 1993; Durham-Pierre et al.,
1994; Lee et  al., 1994). OCA2 encodes for a transmem-
brane protein that has been proposed to function in con-
trolling the pH within the melanosome; thus P gene defects
are thought to impair tyrosinase activity due to their failure
to regulate the pH (Puri et al., 2000; Ancans et al., 2001).
The phenotype in darker subjects suggests that this causes
a problem with the synthesis of eumelanin, but that pha-
eomelanin can still be produced. In addition, MC1R gene
variants can modify the OCA2 phenotype, resulting in red
rather than yellow/blond hair in affected individuals (King
et  al., 2003). In both the Prader-Willi and the Angelman
syndromes, which result from a combination of genomic
imprinting and heterozygous deletion of 15q, the presence
of lighter pigmentation may be due to the deletion of a sin-
gle copy of the OCA2 gene (Gardner et al., 1992).
In other syndromic disorders, pigmentary anomalies are
associated with defects affecting lysosome-related organ-
elles, which include melanosomes. In Hermansky-Pudlak
syndrome, of which there are at least seven genotypes,
tyrosinase-positive albinism is associated with a platelet
defect and pigmented deposits in cells of the reticuloendo-
thelial system (Hermansky and Pudlak, 1959). The pheno-
type results from mutations causing a variety of defects in
the biogenesis of lysosomes and related organelles (reviewed
in Wei, 2006). Chediak-Higashi syndrome is a rare autoso-
mal recessive syndrome with diluted pigmentation of the
skin, hair, and eyes, and markedly increased susceptibility to
bacterial and viral infections, often resulting in death during
the first decade of life. There are abnormal inclusions in a
variety of cells, including giant granules in leukocytes. The
disease is due to mutations in the lysosomal-trafficking regu-
lator (LYST) gene on chromosome 1q42-43 (Barbosa et al.,
1996; Nagle et al., 1996). This allows the normal synthesis of
secretory lysosomes, but these then fuse to form giant secre-
tory lysosomes with defective function (and hence defective
neutrophil, natural killer cell activity, etc.) (Abo et al., 1982;
Stinchcombe et al., 2000). In melanosomes, the lack of abil-
ity to fuse to the cell membrane may prevent their passage
to neighboring keratinocytes, and recent evidence also sug-
gests that there may be a failure to repair plasma-membrane
lesions because of the inability of lysosomes to fuse with the
cell membrane (Huynh et al., 2004).
VI T I L I G O
Vitiligo, a complex disorder in which the death of melano-
cytes in the skin (and also in the hair bulb) results in localized
hypopigmented patches, is considered an acquired disease of
unknown cause, but there is strong evidence of genetic fac-
tors in its pathogenesis. It can occur at any age, but half of all
cases begin within the first two decades. It affects all races, but
is more obvious (and socially debilitating) in darker-skinned
populations. Typically, the patches are white, but lesions in
darker-skinned subjects may exhibit a tanned margin between
the normal brown skin color and the central white area (“tri-
chrome vitiligo”; Figure 45.2); in some cases, the entire patch
may be this intermediate color. The margins of the lesions are
often irregular, appearing to “invade” the normally pigmented
skin. The number and size of the lesions and extent of skin
involvement are very variable, but they commonly involve
the periorificial areas (eyes, mouth, nose), axillae, groin, geni-
tals, and extensor bony surfaces along the elbows, knees, and
digits; the latter sites of predilection may be due to repeated
trauma at these sites, as traumatized skin may develop vitiligo
(known as “the Koebner phenomenon”). The exact mecha-
nism of initial melanocyte-cell death has been debated, but
the disease is considered an autoimmune disease, with factors
such as the generation of reactive oxygen species that render
melanocytes immunogenic also involved in the disease pro-
cess (Schallreuter et al., 2008).
A family history of vitiligo is seen in up to a third of
cases (Lerner 1959; Nath et  al., 1994; Boisseau-Garsaud
Figure 45.2 
Trichrome vitiligo in an Asian subject. Hair pigmentation is
preserved within affected skin.
7 0 0   •   G enomics in C l inica l P ractice
et  al., 2000). Various HLA (human leucocyte antigen)
associations and/or linkage to MHC (major histocom-
patibility complex) class  I  and class  II gene regions have
been reported, but these differ according to the individual
studies (Dunston and Halder, 1990; Lorini et  al., 1992;
al-Fouzan et al., 1995; Buc et al., 1998; Zamani et al., 2001;
Arcos-Burgos et  al., 2002; Tastan et  al., 2004; de Vijlder
et al., 2004; Fain et al., 2006; Jin et al., 2010a). Earlier stud-
ies on susceptibility to vitiligo have shown associations with
polymorphisms in genes involved in autoimmunity and in
other candidate genes, including the gene encoding lym-
phoid protein tyrosine phosphatase (PTPN22) on 1p13.2
(Canton et  al., 2005), the cytotoxic T lymphocyte anti-
gen4 (CTLA4) gene on 2q33 (Blomhoff et al., 2005), the
angiotensin-converting enzyme (ACE) gene on 17q23 ( Jin
et al., 2004; Akhtar et al., 2005), the transporter associated
with antigen processing-1 (TAP1) gene on 6p21.3 (Casp
et al., 2003), and the catalase (CAT) gene on 11p13 (Casp
et al., 2002); however, a later study failed to find associations
with some of these genes (ACE, CAT) and suggested that
the association with CTLA4 was a secondary phenomenon
due to the co-occurrence of vitiligo with other autoimmune
diseases that result from SNPs at that locus (Birlea et  al.,
2011). In a study investigating linkage in subjects with vit-
iligo from families with systemic lupus erythematosus, evi-
dence for linkage was detected on chromosome 17p13, and
subsequent investigations identified associations of SNPs in
NALP1, a gene involved in the regulation of innate immu-
nity, with vitiligo and other autoimmune diseases (Nath
et al., 2001; Jin et al., 2007). An earlier genome-wide scan in
vitiligo identified linkage to chromosome 1p31 (the AIS1
locus), with weaker linkage at other loci on several chromo-
somes (Fain et al., 2003), and later genome-wide association
studies highlighted SNPs in multiple immune-related genes
associated with this hypopigmentary disorder, including
forkhead box P1 (FOXP1) on 3p14.1, C-C chemokine
receptor type 6 (CCR6) on 6q27, CD80 on 3q21, PTPN22
on 1p13, interleukin-2-receptor alpha chain (IL2RA) on
10p15.1, FOXP3 on Xp11.23, thymic stromal lympho-
poietin (TSLP) on 5q22.1, and X-box binding protein 1
(XBP1) on 22q12, among others ( Jin et al., 2010b; Birlea
et al., 2011, Jin et al., 2012). Associations were also noted
with SNPs in the pigmentation genes Tyr, MC1R, and
OCA2-HERC2, which are likely to act as immune targets in
melanocytes, possibly through more efficient presentation
of these altered proteins to the immune system ( Jin et al.,
2010a; Jin et al., 2012).
Although the identification of various genes that increase
susceptibility to vitiligo has led to an improved understand-
ing of this disorder, further studies are necessary to delineate
their individual roles in the immune and non-immune com-
ponents in the pathogenesis of melanocyte cell loss from
the epidermis. Treatment of vitiligo is difficult and gener-
ally unsatisfactory, with some physicians advocating potent
topical steroids, ultraviolet radiation, and minigrafting, and
others adopting a less interventional approach with cos-
metic camouflage and sunscreens. Repigmentation, when
it occurs, often begins at the hair follicles and spreads out-
wards, consistent with melanocyte precursors in the stem
cell niche in the hair follicle not being affected by the same
pathological process (Nishimura et al., 2002).

SKIN CANCER
Figure 45.3 
Genetic and environmental interactions: basal cell carcinoma
Several different types of cancer affect the skin, including occurring on the medial left cheek in a 43-year-old oil-rig worker with
the fair skin, red hair, and freckling phenotype.
basal cell carcinoma (BCC, from keratinocytes), squamous
cell carcinoma (SCC, from keratinocytes), melanoma
(from melanocytes or melanocytic nevi), Merkel cell car- develop de novo in clinical practice. In addition, a rapidly
cinoma (from Merkel cells), cutaneous T-cell lymphoma growing but self-healing keratinizing tumor, the keratoc-
(from lymphocytes), angiosarcoma (from endothelial cells anthoma, may resemble SCC clinically but is less common
lining blood vessels), and fibrosarcoma (from fibroblasts). than SCC.
In addition, the metastatic spread of internal cancers to the UVR is the most important environmental risk factor
skin can happen occasionally. However, the vast majority of for SCCs and BCCs, with a relationship between cumula-
skin cancer comprises SCC, BCC, and melanoma, with the tive UVR and SCC, while intermittent rather than cumu-
combination of SCC and BCC referred to as non-melanoma lative exposure may be more important for BCC (Vitasa
skin cancer (NMSC). Although “NMSC” is widely used, it et al., 1990; Kricker et al., 1995; Rosso et al., 1996). SCCs
is not an ideal term, and “keratinocyte-derived skin cancer” can also arise from prolonged exposure to infrared radia-
is a more appropriate descriptor. Depending on the depth tion, chronic inflammation, and in certain genodermatoses
of invasion of the primary tumor and other clinical and his- (e.g., dystrophic epidermolysis bullosa, keratitis ichthyosis,
tological features, melanomas and SCCs may metastasize, and deafness [KID] syndrome). Exposure to arsenic and
whereas BCCs metastasize infrequently. photochemotherapy (PUVA) increases the risk of BCC
Skin cancer incidence (melanoma and NMSC) has
risen steadily over the past several decades, and is predicted
to increase further in the next two decades. NMSC is the
most common human cancer, with approximately 13 mil-
lion cases annually worldwide, and its increasing incidence
in aging Western populations now constitutes an important
public health problem (Lucas et al., 2009). BCCs (Figure
45.3) arise in hair-bearing epidermis (most areas of skin
except palms and soles) and may be of follicular or epider-
mal origin (Kruger et  al., 1999; Epstein, 2011). Different
clinical/histological subtypes of BCC exist (nodular, scle-
rosing, superficial, and morphoeic types) and are likely to
reflect genetic heterogeneity within the tumors. The inci-
dence of BCCs exceeds that of squamous cell carcinomas
(Figure 45.4) by about 4 to 1. In contrast to BCC, for which
there is no clinically identifiable precursor lesion, premalig-
nant skin lesions (e.g., actinic keratosis and Bowen’s disease) Figure 45.4 
A squamous cell carcinoma on the back of the hand of an
may develop into SCC, but it is thought that most SCCs elderly individual with fair skin.

G enetics and G enomics o f S kin D iseases , I I   •   7 0 1

and SCC, and immunosuppression increases the frequency
of NMSC, with a 50- to 250-fold higher risk of SCC and a
5- to 10-fold greater risk of BCC in organ-transplant recipi-
ents (McGregor and Proby, 1995; Bouwes Bavinck, 1995;
Stern and Lunder, 1998). Male pattern baldness, which is
genetically determined, also increases the risk of actinic ker-
atoses and SCCs on the scalp secondary to loss of the sun
protection provided by scalp hair.
UVR is also the most important environmental risk
factor for the majority of melanomas, with genetic stud-
ies providing unequivocal evidence for the role of UVR
in melanoma pathogenesis; and melanomas are seldom
seen in clinical practice on sun-protected skin such as the
buttocks or the female scalp (Green et  al., 1993; Hodis
et  al., 2012). However, melanomas can occasionally arise
in sun-protected sites such as the vulva or in the menin-
ges/brain in subjects with large congenital melanocytic
nevi who have abnormally high numbers of melanocytes
in the meninges (termed leptomeningeal melanocytosis)
(Kovalyshyn et  al., 2009; Tcheung et  al., 2012). Multiple
melanocytic nevi (Figure 45.5) are a risk factor for mela-
noma development in sporadic cases and in families with
an inherited predisposition to this tumor, but debate exists
about whether most melanomas arise de novo from melano-
cytes or from melanocytic nevi, and the observation of cells
resembling nevus cells within a melanoma has been inter-
preted as nevoid differentiation of melanoma cells by some
pathologists, but signifying a prior existing melanocytic
nevus by others (Clark et al., 1984; Ackerman and Mihara,
1985). Melanoma occasionally occurs in sun-protected
skin such as the sole of the foot (Figure 45.6), but exposure
to UVR, especially short, intense exposure, is considered
the major environmental risk factor for this tumor. Some
Figure 45.5 
Multiple melanocytic nevi, especially if clinically or
pathologically atypical, are associated with increased melanoma risk.
7 0 2   •   G enomics in C l inica l P ractice
Figure 45.6 
A superficial spreading melanoma on the sole of the foot,
showing characteristic irregularity of color and shape.
evidence suggests that UVR exposure during childhood in
amounts sufficient to cause sunburn is particularly harmful
(Whiteman et al., 2001).
G E R M L I N E G E N ET I C
S US C E P T I B I L I T Y TO S K I N C A N C E R
Skin cancer is more frequent in Caucasians, especially those
of Celtic ancestry, than in other racial groups (Crombie,
1979; Cress and Holly, 1997). Indeed, people with fair skin,
freckling, and red or light-colored hair are more at risk of
skin cancer, and susceptibility to skin cancer is associated
with many of the genes involved in lighter pigmentation.
For example, earlier studies revealed that MC1R gene vari-
ants increase the risk of sporadic melanoma and melanoma
in families with atypical mole syndrome secondary to
cyclin-dependent kinase inhibitor 2A (CDKN2A) germline
alterations, and subsequent larger studies have confirmed
these observations (Valverde et al., 1996; Palmer et al., 2000;
Box et al., 2001a; Kennedy et al., 2001; van der Velden et
al., 2001; Gudbjartsson et al., 2008; Demenais et al., 2010;
Ibarrola-Villava et  al., 2012). MC1R variants also predis-
pose to NMSC as well as precursor lesions, and also pro-
mote BCC growth in younger adults less than 40 years old,
and the development of multiple BCCs (Smith et al., 1998;
Bastiaens et  al., 2001b; Box et  al., 2001b; Gudbjartsson
et  al., 2008; Ferrucci et  al., 2012). Recent evidence sug-
gests that the higher phaeomelanin:eumelanin ratio seen in
MC1R variant subjects remains photoprotective, and that
the main pigmentary reason for increased carcinogenesis
when MC1R function is impaired is that a lower amount
of eumelanin reduces photoprotection (Robinson et  al.,
2010). In addition, non-pigmentary mechanisms such as a
reduction in DNA-repair and effects on apoptosis and pro-
liferation constitute part of the reason for MC1R variants’
predisposing to skin cancer (Robinson and Healy, 2002;
Bohm et al., 2005; Kadekaro et al., 2005).
SNPs in other pigmentation genes that confer higher
risks for melanoma and BCC include ASIP, TYR, and
IRF4, with the IRF4 SNP also being involved in SCC
development, a TYRP1 variant involved in eye color asso-
ciated with melanoma development, and SLC45A2 associ-
ated with risk of BCC and SCC (Gudbjartsson et al., 2008;
Stacey et  al., 2009; Han et  al., 2011). A  meta-analysis of
genetic association studies in melanoma indicated MC1R,
TYR, TYRP1, SLC45A2, and ASIP as having strong epide-
miological credibility in this tumor and confirmed MC1R,
TYR, and SLC45A2 as showing a genome-wide statistically
significant association with melanoma (Chatzinasiou et al.,
2011). Interestingly, as was seen with MC1R variants, SNPs
in a number of these pigmentation genes remain associated
with skin cancer after controlling for pigmentation pheno-
type, thus it is possible that these also mediate skin cancer
susceptibility (at least partially) via non-pigmentary mecha-
nisms; but it must be remembered that phenotyping for skin
type can be challenging (Rampen et al., 1988), and further
work is necessary to determine whether non-pigmentary
effects of these gene products permit (or promote) skin
carcinogenesis. The important role of melanin in protect-
ing against skin cancer development can be seen from the
fact that patients with cutaneous albinism, particularly in
countries with high solar exposure, have a high prevalence
of SCC at an early age (Perry and Silverberg, 2001).
A combination of candidate-gene association studies
and genome-wide association studies (GWAS) has iden-
tified a wide variety of SNPs in or close to other genes
likely to be involved in melanoma and non-melanoma skin
cancer. For example, SNPs near exocyst complex com-
ponent 2 (EXOC2) on 6p25 and ubiquitin-associated
domain-containing protein 2 (UBAC2) on 13q32 are
associated with risk of BCC and SCC (Nan et al., 2011).
Variants on 1p36 and 1q42 have also been associated with
BCC, as have variants in keratin 5 (KRT5) on 12q13, near
CDKN2A, and CDKN2B on 9p21, near kruppel-like fac-
tor 14 (KLF14) on 7q32, and at the telomerase reverse
transcriptase and the cleft lip and palate transmembrane
protein 1-like locus (TERT-CLPTM1L) on 5p15.33 (with
the latter variant conversely protecting against melanoma)
and in the TP53 gene on 17p13.1 (Stacey et al., 2008, 2009,
2011). In the inherited disorder, Gorlin syndrome (also
known as nevoid basal cell carcinoma syndrome), germline
inactivating mutations in the Patched 1 (PTCH1) gene on
9q22.3 give rise to multiple basal cell carcinomas (Hahn
G enetics and G enomics o f S kin D iseases , I I   •   7 0 3
et al., 1996; Johnson et al., 1996). PTCH is a member of
the highly conserved hedgehog signaling pathway, and the
use of hedgehog inhibitors for inoperable/metastatic BCCs
has shown some beneficial therapeutic effects (Sekulic et al.,
2012).
For melanoma, genetic susceptibility includes polymor-
phisms in CDKN2A and adjacent MTAP genes on 9p21,
as well as in ATM on 11q22.3 and MX2 on 21q22.3, and
SNPs at 1q21.3, 1q42.12, and adjacent to the CASP8 gene
on chromosome 2 (Bishop et al., 2009; Barrett et al., 2011;
MacGregor et al., 2011). Furthermore, variants in MTAP,
as well as at PLA2G6 on 22q13.1, have been reported to
be associated with the development of increased numbers
of acquired melanocytic nevi, whereas certain MC1R vari-
ants have been associated with development of congenital
melanocytic nevi (Falchi et al., 2009; Kinsler et al., 2012).
Alterations in CDKN2A have also been recognized for
a long time as the cause of susceptibility in families with
an inherited form of melanoma (Hussussian et  al., 1994;
Gruis et al., 1995), as have inherited mutations in CDK4
on 12q14 (Zuo et al., 1996).
Various other genes whose products are involved either
in immune function, xenobiotic detoxification (including
detoxification of carcinogens and mutagens) or DNA repair,
have been reported as associated with skin cancer. Some
HLA associations have been reported in BCC and/or SCC,
including in organ-transplant recipients with SCC, but these
vary between studies, whereas interleukin-10 (IL-10) poly-
morphisms have been identified in association with BCC
and of possible significance for SCC (Bouwes Bavinck et al.,
1997; Bouwes Bavinck et al., 2000; Welsh et al., 2011). The
glutathione S-transferase GSTM1 null genotype increases
the risk of actinic keratoses and predisposes to BCC,
whereas a polymorphism of GSTM3 has also been shown
to increase the risk for BCC (Carless et al., 2002; Lear et al.,
1996; Lear et al., 1997; Yengi et al., 1996). In the rare con-
dition known as genodermatosis xeroderma pigmentosum
(XP), due to mutations in different genes of the DNA repair
complex, impaired capacity to repair UV-induced DNA
damage results in a dramatically increased risk of melanoma
and non-melanoma skin cancer (Cleaver, 1968; reviewed
in DiGiovanna and Kraemer, 2012). Polymorphisms in
some of these XP genes may be associated with sporadic
skin cancer, but the results of different studies vary, and a
meta-analysis suggests that a SNP in ERCC2/XPD may
constitute a low-penetrance melanoma susceptibility gene in
the general population (Mocellin et al., 2009). In addition,
it has been reported that polymorphisms in some of these
genes (e.g., ERCC5 [XPG] and ERCC2 [XPD]) influence
prognosis in patients with sporadic cutaneous melanoma
(Schrama et al., 2011). In the rare recessive disorder epider-
modysplasia verruciformis (EV), due to mutations in the
genes TMC6/EVER1 and TMC8/EVER2 encoding related
calnexin-associated transmembrane proteins, human papil-
loma virus types 5 and 8 are associated with warty lesions
and SCC in the sun-exposed skin of patients (Ramoz et al.,
2002). In the dominant disorder Ferguson-Smith syndrome,
consisting of multiple self-healing squamous-carcinoma-like,
locally invasive skin tumors resembling keratoacanthomas,
the causative gene is TGFBR1, with mutations resulting
in loss of receptor function, consistent with its acting as a
tumor-suppressor gene (Goudie et al., 2011).
S O M AT I C G E N ET I C A LT E R AT I O NS
IN SKIN CANCER
UVR causes multiple effects in skin, including DNA dam-
age (cyclobutane pyrimidine dimers, 6-4 photoproducts,
strand breaks, etc.) in melanocytes and keratinocytes, which,
if not repaired, lead to permanent genetic alterations/gene
mutation (Hemminki et  al., 2001). As in other cancer
types, genetic alterations that activate proto-oncogenes or
inactivate tumor-suppressor genes involved in the control
of the cell cycle lead to clonal expansion of mutated mela-
nocytes and/or keratinocytes. Further exposure to UVR
causes more DNA damage, leading to mutation in other
genes involved in cell–cell interactions, cell–extracellular
matrix interactions, evasion of immune-regulation, and so
forth, enabling the neoplastic cells to invade into the der-
mis and ultimately metastasize. Although tumors contain
many somatic genetic abnormalities, all of these may not
be relevant to the development of the tumor, and some of
these may be secondary genetic alterations due to later-stage
genomic instability. However, it is now recognized that
uncontrolled genomic instability is likely to be detrimen-
tal for the tumor, and that, in order for tumors to persist,
there needs to be maintenance of sufficient genomic stabil-
ity to avoid tumor cell death and allow the primary tumor
to produce distant metastasis (Kauffmann et  al., 2008).
Variations in gene expression, as identified with gene arrays,
and in the expression of non-coding RNAs/microRNAs,
are likely to be secondary phenomena resulting from other
primary genetic alterations, but may constitute a mecha-
nism through which the primary genetic alterations exert
an effect. In addition, genetic manipulation or alterations
may promote, or be involved in, skin cancer development in
mice, but one cannot extrapolate from findings in mice to
human skin cancer in the absence of documentary evidence
that the same gene is mutated or altered in the germline, or
somatically in the cancer, in humans.
7 0 4   •   G enomics in C l inica l P ractice
In the case of tumor-suppressor genes, some effects may
be seen from haploinsufficiency, but frequently inactivation
of both copies of the gene in tumors occurs through a variety
of mechanisms, including mutation, allelic loss, promoter
methylation, and so on. Previously, investigations looking
for allelic loss (also termed “loss of heterozygosity”) were
employed to look for evidence of tumor-suppressor genes
in skin cancers. Somatic genetic alterations detected within
sporadic cutaneous melanomas included losses of chromo-
somes 9p, 10q, and 6q, with a variable but lower frequency
of loss of other chromosome arms (Healy et al., 1996). Loss
of 9p occurs in early-stage melanomas with a depth of inva-
sion of less than 1 mm, whereas losses of 10q and 6q seem to
occur later and are associated with a poorer clinical outcome
(Healy et al., 1995; Healy et al., 1998). In SCC, allelic losses
were seen on chromosomes 13q, 9p, 17p, 17q, and 3p, but
actinic keratoses (which are benign dysplastic lesions with
the potential to develop into SCC) have a similar degree of
loss of heterozygosity (Quinn et al., 1994; Rehman et al.,
1994). However, BCCs display a different allelotype, with
loss of heterozygosity largely limited to 9q, 9p, and 1q (Bare
et  al., 1992; Quinn et  al, 1994). Conversely, comparative
genomic hybridization and fluorescence in situ hybridiza-
tion (FISH) have documented chromosomal gains (which
may simulate oncogene activity) in the different types of
skin cancer. These gains differ somewhat in the same type of
tumor between studies, but include 1q, 6p, 7q, 8q, and 20q
in melanoma; 1p, 1q, 8q, 13q, 14q, 16q, and 20q in SCC;
and 6p, 6q, and 9p in BCC (Barks et al., 1997; Bastian et al.,
1998; Ashton et al., 2001; Clausen et al., 2006; Kabbarah
et al., 2010; Pesz et al., 2012).
At the gene level, mutations in numerous genes have
been reported in human skin cancers, and the following
paragraphs do not attempt to list every gene, but aim to
concentrate more on providing an overview of the sub-
ject area. Although somatic deletions or mutations of the
CDKN2A tumor-suppressor gene on 9p are seen in patients
with inherited melanoma who carry germline alterations in
this gene, mutations in CDKN2A are uncommon in spo-
radic cases despite loss of heterozygosity at this locus, and
inactivation of the second allele seems to occur through
alternative mechanisms, including promoter CpG meth-
ylation and homozygous deletion (Hussussian et al., 1994;
Gruis et al., 1995; Healy et al., 1996; Gonzalgo et al., 1997;
Fujimoto et  al., 1999; Grafstrom et  al., 2005). Somatic
alterations in MITF have been documented in melanoma,
more frequently in metastases than in the primary tumor,
and subsequent reports have indicated that germline altera-
tions in MITF also predispose to sporadic and familial
melanoma (Garraway et al., 2005; Bertolotto et al., 2011;
Yokoyama et  al., 2011). Mutations of BRAF on chromo-
some 7q34 are frequent in melanoma on skin exposed to
ultraviolet radiation, but are also seen at a high frequency
in acquired melanocytic nevi; thus are permissive but insuf-
ficient on their own for the malignant phenotype (Davies
et al., 2002; Pollock et al., 2003; Kumar et al., 2004). The
V600E alteration is the most common BRAF mutation in
melanoma, and BRAF inhibitors (alone or in combina-
tion with MEK inhibition) targeting the enhanced BRAF,
and downstream MAP kinase signaling secondary to this
activating mutation, have recently shown some promise
in the treatment of metastatic melanoma, with a median
progression-free survival of approximately seven months for
the BRAF inhibitor alone, and of almost 10 months for the
BRAF and MAPK kinase inhibitors combined (Flaherty
et  al., 2012; Sosman et  al., 2012). By contrast, codon 61
mutations in the NRAS gene on 1p13.2 occur in mela-
nomas that are BRAF wild-type, but also affect the Ras/
Raf/mitogen-activated protein kinase signaling pathway
(van ‘t Veer et al., 1989; Albino et al., 1989; Omholt et al.,
2003). BRAF mutations are rare in mucosal melanomas (in
which UVR does not play a role), and in this type of mela-
noma, mutation of the KIT oncogene seems more relevant
(Beadling et al., 2008; Carvajal et al., 2011). The apoptotic
protease activating factor 1 (APAF1) mediates p53-induced
apoptosis, and loss of its expression through allelic loss and
promoter methylation of APAF1 on chromosome 12q23.1
in primary and metastatic melanoma is thought to render
the cancer more resistant to chemotherapeutic agents and
correlates with a worse prognosis (Soengas et al., 2001; Dai
et  al., 2004; Fujimoto et  al., 2004). The high prevalence
of alterations in APAF1 and in the CDKN2A pathways
may account for the low frequency of TP53 gene muta-
tions, which are observed in approximately 0–20% of cases
(Castresana et  al., 1993; Lubbe et  al., 1994; Hodis et  al.,
2012). Exome and whole-genome sequencing continue to
identify many other genes that are altered in melanoma, and
the evidence is mounting that melanoma (like other internal
tumors) contains many somatic genetic alterations; how-
ever, comparison with the frequency of mutation in exonic
versus intronic sequences may help differentiate causal
mutations from those that are passenger mutations second-
ary to UVR exposure and genomic instability (Nikolaev
et al., 2011; Stark et al., 2011; Wei et al., 2011; Berger et al.,
2012; Hodis et al., 2012; Krauthammer et al., 2012).
The identification of germline PTCH1 gene mutations
as the cause of Gorlin syndrome and allelic loss on chromo-
some 9q in sporadic BCCs led to investigations of genes in
the hedgehog signaling pathway for somatic alterations in
sporadic BCC. As a result, somatic PTCH1 gene mutations
G enetics and G enomics o f S kin D iseases , I I   •   7 0 5
were noted to be one of the more frequent alterations in
BCC (Gailani et al., 1996; Heitzer et al., 2007). SCCs from
patients with multiple BCCs may contain PTCH1 muta-
tions; however, mutation/loss of heterozygosity (LOH) for
PTCH1 seem less frequent in SCC than in BCC (Eklund
et al., 1998; Ping et al., 2001; Danaee et al., 2006). Genetic
alterations in other members of the hedgehog pathway have
been demonstrated in BCCs; for example, activating sonic
hedgehog (SHH) mutations have been reported in sporadic
BCC and in BCCs from patients with xeroderma pigmen-
tosum, and activating mutations of smoothened (SMOH)
on 7q32.1 have been described in sporadic BCCs lacking
PTCH1 alterations (Oro et al., 1997; Couvé-Privat et al.,
2004; Reifenberger et al, 1998; Xie et al., 1998). In addi-
tion, mutations in the SUFU gene on 10q24.32, which also
forms part of the hedgehog pathway, have occasionally been
identified somatically in BCC, and as a non-PTCH1 cause
of Gorlin syndrome (Reifenberger et  al., 2005; Pastorino
et al., 2009). Hedgehog inhibitors that bind to and inacti-
vate the smoothened protein (which is downstream of sonic
hedgehog and patched in the signaling pathway) have shown
some effect in metastatic and non-resectable BCCs as well
as in other types of tumors, but, as seen with BRAF inhibi-
tors in melanoma, resistance to the therapeutic agent may
develop during treatment (Metcalfe and de Sauvage, 2011;
Sekulic et al., 2012).
The majority of cutaneous SCCs and their precur-
sor lesions, actinic keratoses and Bowens disease, contain
mutations in the TP53 tumor-suppressor gene (Brash
et  al., 1991; Campbell et  al., 1993; Ziegler et  al., 1994).
UVR-induced mutations in the TP53 gene are also found as
expanded clonal microscopic lesions, termed “p53 patches,”
in sun-exposed normal human skin; and, based on similar
Tp53 mutation profiles in skin SCCs and the p53 patches
in mouse models, it is considered that cutaneous SCCs
develop from these p53 patches (Nakazawa et  al., 1994,
Urano et al., 1995; Jonason et al., 1996; Kramata et al., 2005;
Rebel et al., 2005). TP53 mutations have been detected in
about half of all BCCs, and they are also seen in BCCs that
occur secondary to ionizing radiation as well as in those
resulting from UVR exposure (Rady et al., 1992; Mizuno
et  al., 2006). Other genes involved in SCC development
include the CDKN2A gene, which encodes p16(INK4a)
and p14(ARF); and cell-cycle-regulatory proteins in the
p53 and RB pathways, with point mutations or promoter
methylation of p16(INK4a) and p14(ARF) documented
in some skin SCCs (Brown et  al., 2004). Furthermore,
CDNK2A mutations have been associated with shorter
disease-specific survival in skin SCC (Küsters-Vandevelde
et al., 2010). RAS gene mutations have also been reported
 Box 45.1  KEY POINTS
• Skin and hair pigmentation in man are complex
genetic traits.
• Important genes controlling skin color in man
include the MC1R, SLC24A5 (Golden), SLC45A2
(MATP), TYR, TYRP1, OCA2 (P), ASIP, POMC,
KITLG, and IRF4 genes.
• Inherited disorders of pigmentation may affect
melanocyte migration and survival, or synthesis of
melanin and its distribution to keratinocytes.
• The acquired depigmenting disorder vitiligo is
strongly genetically influenced.
Box 45.2  KEY POINTS
• Non-melanoma skin cancer (NMSC), comprising
basal cell carcinoma (BCC) and squamous cell
carcinoma (SCC), is the commonest human
malignancy.
• Ultraviolet radiation (UVR) is the major
environmental factor that causes NMSC and
melanoma, but genetic susceptibility factors,
including fair skin (NMSC and melanoma) and
increased numbers of melanocytic nevi (melanoma),
are important.
• Other genetic susceptibility factors include immune
response genes (mainly for NMSC), detoxifying
enzymes, and DNA-repair mechanisms.
• Inherited disorders predisposing to skin cancer
include xeroderma pigmentosum due to DNA repair
defects (SCCs, BCCs, and melanomas), atypical
mole syndrome due to CDKN2A gene alterations
(melanoma), and Gorlin syndrome due to defects in
the PTCH1 gene (basal cell carcinomas).
• Recent exome-sequencing studies indicate that
sporadic skin tumors can contain multiple somatic
gene defects, but frequent somatic defects in NMSC
include UV-induced mutations of TP53, mutations
of the PTCH1 gene in basal cell carcinoma, and
BRAF and MITF mutations in melanoma.
in NMSC, with activating HRAS mutations observed in a
variable proportion of SCCs and in the precursor lesions,
actinic keratoses (Pierceall et al, 1991; Spencer et al., 1995;
Kreimer-Erlacher et  al, 2001). In addition, HRAS gene
7 0 6   •   G enomics in C l inica l P ractice
mutations in keratinocytes and/or precursor lesions seem
to permit cutaneous SCCs development in patients treated
with BRAF inhibitors for melanoma (Su et al., 2012).
C O N C LU S I O N
The knowledge of the somatic genetic alterations in skin
cancers and the genetic events responsible for the behavior
of the different types of tumors continues to grow, and is
likely to expand in the future. In addition, it is expected
that our understanding of the genotypes of susceptibility
to skin carcinogenesis, and the genotypes of the host and
the tumor and how they influence host–tumor and tumor–
host interactions, will continue to develop, and that this
will lead to improved prevention and treatments for mela-
noma and non-melanoma disorders in future years (Box
45.1 and 45.2).
REFERENCES
Abo T, Roder JC, Abo W, Cooper MD, Balch CM (1982). Natural killer
(HNK-1+) cells in Chediak-Higashi patients are present in nor-
mal numbers but are abnormal in function and morphology. J Clin
Invest 70:193–197.
Ackerman AB, Mihara I (1985). Dysplasia, dysplastic melanocytes, dys-
plastic nevi, the dysplastic nevus syndrome, and the relation between
dysplastic nevi and malignant melanomas. Hum Pathol 16:87–91.
Akhtar S, Gavalas NG, Gawkrodger DJ, et  al. (2005). An insertion/
deletion polymorphism in the gene encoding angiotensin convert-
ing enzyme is not associated with generalised vitiligo in an English
population. Arch Dermatol Res 297:94–98.
Albino AP, Nanus DM, Mentle IR, et al. (1989). Analysis of Ras onco-
genes in malignant melanoma and precursor lesions:  correlation
of point mutations with differentiation phenotype. Oncogene
4:1363–1374.
al-Fouzan A, al-Arbash M, Fouad F, et al. (1995). Study of HLA class I/
IL and T lymphocyte subsets in Kuwaiti vitiligo patients. Eur J
Immunogenet 22:209–213.
Ancans J, Tobin DJ, Hoogduijn MJ, Smit NP, Wakamatsu K, Thody AJ
(2001). Melanosomal pH controls rate of melanogenesis, eumela-
nin/phaeomelanin ratio and melanosome maturation in melano-
cytes and melanoma cells. Exp Cell Res 268:26–35.
Arcos-Burgos M, Parodi E, Salgar M, et al. (2002). Vitiligo: complex seg-
regation and linkage disequilibrium analyses with respect to micro-
satellite loci spanning the HLA. Hum Genet 110:334–342.
Ashton KJ, Weinstein SR, Maguire DJ, Griffiths LR (2001). Molecular
cytogenetic analysis of basal cell carcinoma DNA using comparative
genomic hybridization. J Invest Dermatol 117:683–686.
Barbosa MDFS, Nguyen QA, Tchernev VT, et al. (1996). Identification
of the homologous beige and Chediak-Higashi syndrome genes.
Nature 382:262–265.
Bare JW, Lebo RV, Epstein EH Jr. (1992). Loss of heterozygosity at chro-
mosome 1q22 in basal cell carcinomas and exclusion of the basal cell
nevus syndrome gene from this site. Cancer Res 52:1494–1498.
Barks JH, Thompson FH, Taetle R, et  al. (1997). Increased chromo-
some 20 copy number detected by fluorescence in situ hybridiza-
tion (FISH) in malignant melanoma. Gene Chromosome Canc
19:278–285.
Barrett JH, Iles MM, Harland M, et al. (2011). Genome-wide associa-
tion study identifies three new melanoma susceptibility loci. Nat
Genet 43:1108–1113.
Bastiaens M, ter Huurne J, Gruis N, et al. (2001a). The melanocortin-
1-receptor gene is the major freckle gene. Hum Mol Genet
10:1701–1708.
Bastiaens MT, ter Huurne JA, Kielich C, et al. (2001b). Melanocortin-1
receptor gene variants determine the risk of nonmelanoma skin
cancer independently of fair skin and red hair. Am J Hum Genet
68:884–894.
Bastian BC, LeBoit PE, Hamm H, Brocker EB, Pinkel D (1998).
Chromosomal gains and losses in primary cutaneous melano-
mas detected by comparative genomic hybridization. Cancer Res
58:2170–2175.
Beadling C, Jacobson-Dunlop E, Hodi FS, et al. (2008). KIT gene muta-
tions and copy number in melanoma subtypes. Clin Cancer Res
14:6821–6828.
Beaumont KA, Newton RA, Smit DJ, et al. (2005). Altered cell surface
expression of human MC1R variant receptor alleles associated with
red hair and skin cancer risk. Hum Mol Genet 14:2145–2154.
Berger MF, Hodis E, Heffernan TP, et  al. (2012). Melanoma
genome sequencing reveals frequent PREX2 mutations. Nature
485:502–506.
Bertolotto C, Lesueur F, Giuliano S, et  al. (2011). A SUMOylation-
defective MITF germline mutation predisposes to melanoma and
renal carcinoma. Nature 480:94–98.
Birlea SA, Jin Y, Bennett DC, et al. (2011). Comprehensive association
analysis of candidate genes for generalized vitiligo supports XBP1,
FOXP3, and TSLP. J Invest Dermatol 131:371–381.
Bishop DT, Demenais F, Iles MM, et al. (2009). Genome-wide associa-
tion study identifies three loci associated with melanoma risk. Nat
Genet 41:920–925.
Blomhoff A, Kemp EH, Gawkrodger DJ, et  al. (2005). CTLA4 poly-
morphisms are associated with vitiligo, in patients with concomi-
tant autoimmune diseases. Pigment Cell Res 18:55–58.
Böhm M, Wolff I, Scholzen TE, et al. (2005). Alpha-melanocyte-stimulating
hormone protects from ultraviolet radiation-induced apoptosis and
DNA damage. J Biol Chem 280:5795–5802.
Boisseau-Garsaud AM, Garsaud P, et  al. (2000). Epidemiology of vit-
iligo in the French West Indies (Isle of Martinique). Int J Dermatol
39:18–20.
Bondurand N, Pingault V, Goerich DE, et al. (2000). Interaction among
SOX10, PAX3 and MITF, three genes altered in Waardenburg syn-
drome. Hum Mol Genet 9:1907–1917.
Bondurand N, Dastot-Le Moal F, Stanchina L, et al. (2007). Deletions at
the SOX10 gene locus cause Waardenburg syndrome types 2 and 4.
Am J Hum Genet 81:1169–1185.
Bonilla C, Boxill LA, Donald SA, et  al. (2005). The 8818G allele of
the agouti signaling protein (ASIP) gene is ancestral and is asso-
ciated with darker skin color in African Americans. Hum Genet
116:402–406.
Bouwes Bavinck JN (1995). Epidemiological aspects of immunosuppres-
sion: role of exposure to sunlight and human papillomavirus on the
development of skin cancer. Hum Exp Toxicol 14:98.
Bouwes Bavinck JN, Bastiaens MT, Marugg ME, et al. (2000). Further
evidence for an association of HLA-DR7 with basal cell carcinoma
on the tropi4cal island of Saba. Arch Dermatol 136:1019–1022.
Bouwes Bavinck JN, Claas FH, Hardie DR, Green A, Vermeer BJ,
Hardie IR (1997). Relation between HLA antigens and skin can-
cer in renal transplant recipients in Queensland, Australia. J Invest
Dermatol 108:708–711.
Box NF, Duffy DL, Chen W, et al. (2001a). MC1R genotype modifies
risk of melanoma in families segregating CDKN2A mutations. Am
J Hum Genet 69:765–773.
Box NF, Duffy DL, Irving RE, et al. (2001b). Melanocortin-1 receptor
genotype is a risk factor for basal and squamous cell carcinoma. J
Invest Dermatol 116:224–229.
G enetics and G enomics o f S kin D iseases , I I   •   7 0 7
Box NF, Wyeth JR, Mayne CJ, et  al. (1998). Complete sequence
and polymorphism study of the human TYRP1 gene encoding
tyrosinase-related protein 1. Mamm Genome 9:50–53.
Box NF, Wyeth JR, O’Gorman LE, et  al. (1997). Characterization of
melanocyte stimulating hormone receptor variant alleles in twins
with red hair. Hum Mol Genet 6:1891–1897.
Brash DE, Rudolph JA, Simon JA, et al. (1991). A role for sunlight in
skin cancer:  UV- induced p53 mutations in squamous cell carci-
noma. Proc Natl Acad Sci U S A 88:10124–10128.
Brown VL, Harwood CA, Crook T, et  al. (2004). p16INK4a and
p14ARF tumor suppressor genes are commonly inactivated in cuta-
neous squamous cell carcinoma. J Invest Dermatol 122:1284–1292.
Buc M, Fazekasova H, Cechova E, et  al. (1998). Occurrence rates of
HLA-DRB1, HLA-DQB1, and HLA-DPB1 alleles in patients suf-
fering from vitiligo. Eur J Dermatol 8:13–15.
Canton I, Akhtar S, Gavalas NG, et al. (2005). A single-nucleotide poly-
morphism in the gene encoding lymphoid protein tyrosine phos-
phatase (PTPN22) confers susceptibility to generalised vitiligo.
Genes Immun 6:584–587.
Campbell C, Quinn AG, Ro YS, Angus B, Rees JL (1993). P53 muta-
tions are common and early events that precede tumor invasion in
squamous cell neoplasia of the skin. J Invest Dermatol 100:746–748.
Carless MA, Lea RA, Curran JE, et  al. (2002). The GSTM1 null
genotype confers an increased risk for solar keratosis develop-
ment in an Australian Caucasian population. J Invest Dermatol
119:1373–1378.
Carvajal RD, Antonescu CR, Wolchok JD, et al. (2011). KIT as a thera-
peutic target in metastatic melanoma. JAMA 305:2327–2334.
Casp CB, She JX, McCormack WT (2002). Genetic association of the
catalase gene (CAT) with vitiligo susceptibility. Pigment Cell Res
15:62–66.
Casp CB, She JX, McCormack WT (2003). Genes of the LMP/TAP
cluster are associated with the human autoimmune disease vitiligo.
Genes Immun 4:492–499.
Castresana JS, Rubio MP, Vazquez JJ, et al. (1993). Lack of allelic dele-
tion and point mutation as mechanisms of p53 activation in human
malignant melanoma. Int J Cancer 55:562–565.
Chaki M, Sengupta M, Mondal M, et al. (2011). Molecular and func-
tional studies of tyrosinase variants among Indian oculocutaneous
albinism type 1 patients. J Invest Dermatol 131:260–262.
Chatzinasiou F, Lill CM, Kypreou K, et al. (2011). Comprehensive field
synopsis and systematic meta-analyses of genetic association studies
in cutaneous melanoma. J Natl Cancer Inst 103:1227–1235.
Chi A, Valencia JC, Hu ZZ, et al. (2006). Proteomic and bioinformatic
characterization of the biogenesis and function of melanosomes. J
Proteome Res 5:3135–3144.
Clark WH Jr, Elder DE, Guerry D, et al. (1984). A study of tumor pro-
gression: the precursor lesions of superficial spreading and nodular
melanoma. Hum Pathol 15:1147–1165.
Clausen OP, Aass HC, Beigi M, et  al. (2006). Are keratoacanthomas
variants of squamous cell carcinomas? A  comparison of chromo-
somal aberrations by comparative genomic hybridization. J Invest
Dermatol 126:2308–2315.
Cleaver JE (1968). Defective repair replication of DNA in xeroderma
pigmentosum. Nature 218:652–656.
Clément K, Dubern B, Mencarelli M, et al. (2008). Unexpected endo-
crine features and normal pigmentation in a young adult patient
carrying a novel homozygous mutation in the POMC gene. J Clin
Endocrinol Metab 93:4955–4962.
Cook AL, Chen W, Thurber AE, et  al. (2009). Analysis of cul-
tured human melanocytes based on polymorphisms within the
SLC45A2/MATP, SLC24A5/NCKX5, and OCA2/P loci. J Invest
Dermatol 129:392–405.
Couvé-Privat S, Le Bret M, Traiffort E, et al. (2004). Functional analy-
sis of novel sonic hedgehog gene mutations identified in basal cell
carcinomas from xeroderma pigmentosum patients. Cancer Res
64:3559–3565.
Cress RD, Holly EA (1997). Incidence of cutaneous melanoma among
non-Hispanic whites, Hispanics, Asians, and blacks:  an analysis
of California Cancer Registry data, 1988–93. Canc Causes Contr
8:246–252.
Crombie IK (1979a). Racial differences in melanoma incidence. Br J
Cancer 40:185–193.
Dai DL, Martinka M, Bush JA, Li G (2004). Reduced Apaf-1 expression
in human cutaneous melanomas. Br J Cancer 91:1089–1095.
Dalziel M, Kolesnichenko M, das Neves RP, et al. (2011). Alpha-MSH
regulates intergenic splicing of MC1R and TUBB3 in human mela-
nocytes. Nucleic Acids Res 39:2378–2392.
Danaee H, Karagas MR, Kelsey KT, Perry AE, Nelson HH (2006).
Allelic loss at Drosophila patched gene is highly prevalent in
basal and squamous cell carcinomas of the skin. J Invest Dermatol
126:1152–1158.
Davies H, Bignell GR, Cox C, et al. (2002). Mutations of the BRAF gene
in human cancer. Nature 417:949–954.
de Vijlder HC, Westerhof W, Schreuder GM, de Lange P, Claas FH
(2004). Difference in pathogenesis between vitiligo vulgaris
and halo nevi associated with vitiligo is supported by an HLA
Association study. Pigment Cell Res 17:270–274.
Demenais F, Mohamdi H, Chaudru V, et  al. (2010). Association of
MC1R Variants and Host Phenotypes with Melanoma Risk in
CDKN2A Mutation Carriers:  A  GenoMEL Study. J Natl Cancer
Inst 102(20):1568–1583.
DiGiovanna JJ, Kraemer KH (2012). Shining a light on xeroderma pig-
mentosum. J Invest Dermatol 132:785–796.
Dunston GM, Halder RM (1990). Vitiligo is associated with HLA-DR4
in black patients. A preliminary report. Arch Dermatol 126:56–60.
Durham-Pierre D, Gardner JM, Nakatsu Y, et al. (1994). African origin
of an intragenic deletion of the human P gene in tyrosinase positive
oculocutaneous albinism. Nat Genet 7:176–179.
Edery P, Attie T, Amiel J, et  al. (1996). Mutation of the endothelin-3
gene in the Waardenburg-Hirschsprung disease (Shah-Waardenburg
syndrome). Nat Genet 12:442–444.
Eiberg H, Troelsen J, Nielsen M, et al. (2008). Blue eye color in humans
may be caused by a perfectly associated founder mutation in a regu-
latory element located within the HERC2 gene inhibiting OCA2
expression. Hum Genet 123:177–187.
Eklund LK, Lindström E, Undén AB, et al. (1998). Mutation analysis
of the human homologue of Drosophila patched and the xeroderma
pigmentosum complementation group A  genes in squamous cell
carcinomas of the skin. Mol Carcinog 21:87–92.
Epstein EH Jr. (2011). Mommy—where do tumors come from? J Clin
Invest 121:1681–1683.
Fain PR, Babu SR, Bennett DC, Spritz RA (2006). HLA class  II
haplotype DRB1*04-DQB1*0301 contributes to risk of familial
generalized vitiligo and early disease onset. Pigment Cell Res 19:
51–57.
Fain PR, Gowan K, LaBerge GS, et  al. (2003). A genomewide screen
for generalized vitiligo: confirmation of AIS1 on chromosome 1p31
and evidence for additional susceptibility loci. Am J Hum Genet
72:1560–1564.
Falchi M, Bataille V, Hayward NK, et  al. (2009). Genome-wide asso-
ciation study identifies variants at 9p21 and 22q13 associated with
development of cutaneous nevi. Nat Genet 41:915–919.
Ferrucci LM, Cartmel B, Molinaro AM, et al. (2012). Host phenotype
characteristics and MC1R in relation to early-onset basal cell carci-
noma. J Invest Dermatol 132:1272–1279.
Fitzpatrick TB (1988). The validity and practicality of sun-reactive skin
types I through VI. Arch Dermatol 124:869–871.
Flaherty KT, Infante JR, Daud A, et al. (2012). Combined BRAF and
MEK inhibition in melanoma with BRAF V600 mutations. N Engl
J Med 367:1694–1703.
Flanagan N, Healy E, Ray A, et al. (2000). Pleiotropic effects of the mela-
nocortin 1 receptor (MC1R) gene on human pigmentation. Hum
Mol Genet 9:2531–2537.
7 0 8   •   G enomics in C l inica l P ractice
Fleischman RA, Saltman DL, Stastny V, Zneimer S (1991). Deletion of
the c-kit protooncogene in the human developmental defect piebald
trait. Proc Natl Acad Sci U S A 88:10885–10889.
Fujimoto A, Morita R, Hatta N, Takehara K, Takata M (1999).
P16INK4a inactivation is not frequent in uncultured sporadic pri-
mary cutaneous melanoma. Oncogene 18:2527–2532.
Fujimoto A, Takeuchi H, Taback B, et al. (2004). Allelic imbalance of
12q22–23 associated with APAF-1 locus correlates with poor dis-
ease outcome in cutaneous melanoma. Cancer Res 64:2245–2250.
Fukamachi S, Shimada A, Shima A (2001). Mutations in the gene
encoding B, a novel transporter protein, reduce melanin content in
medaka. Nat Genet 28:381–385.
Gailani MR, Stahle-Backdahl M, Leffell DJ, et  al. (1996). The role of
the human homologue of Drosophila patched in sporadic basal cell
carcinomas. Nat Genet 14:78–81.
Gardner JM, Nakatsu Y, Gondo Y, et al. (1992). The mouse pink-eyed
dilution gene: association with human Prader-Willi and Angelman
syndromes. Science 257:1121–1124.
Garraway LA, Widlund HR, Rubin MA, et  al. (2005). Integrative
genomic analyses identify MITF as a lineage survival oncogene
amplified in malignant melanoma. Nature 436:117–122.
Giebel LB, Spritz RA (1991). Mutation of the KIT (mast/stem cell
growth factor receptor) protooncogene in human piebaldism. Proc
Natl Acad Sci U S A 88:8696–8699.
Giebel LB, Strunk KM, King RA, Hanifin JM, Spritz RA (1990). A
frequent tyrosinase gene mutation in classic, tyrosinase-negative
(type IA) oculocutaneous albinism. Proc Natl Acad Sci U S A
87:3255–3258.
Giebel LB, Tripathi RK, King RA, Spritz RA (1991). A tyrosinase
gene missense mutation in temperature-sensitive type I oculocuta-
neous albinism. A  human homologue to the Siamese cat and the
Himalayan mouse. J Clin Invest 87:1119–1122.
Ginger RS, Askew SE, Ogborne RM, et al. (2008). SLC24A5 encodes
a trans-Golgi network protein with potassium-dependent
sodium-calcium exchange activity that regulates human epidermal
melanogenesis. J Biol Chem 283:5486–5495.
Glimcher ME, Kostick RM, Szabo G (1973). The epidermal melanocyte
system in newborn human skin. A  quantitative histologic study. J
Invest Dermatol 61:344–347.
Gonzalgo ML, Bender CM, You EH, et  al. (1997). Low frequency of
p16/CDKN2A methylation in sporadic melanoma:  comparative
approaches for methylation analysis of primary tumors. Cancer Res
57:5336–5347.
Goudie DR, D’Alessandro M, Merriman B, et  al. (2011). Multiple
self-healing squamous epithelioma is caused by a disease-specific
spectrum of mutations in TGFBR1. Nat Genet 43:365–369.
Graf J, Hodgson R, van Daal A (2005). Single nucleotide polymorphisms
in the MATP gene are associated with normal human pigmentation
variation. Hum Mutat 25:278–284.
Graf J, Voisey J, Hughes I, van Daal A (2007). Promoter polymorphisms
in the MATP (SLC45A2) gene are associated with normal human
skin color variation. Hum Mutat 28:710–717.
Grafstrom E, Egyhazi S, Ringborg U, Hansson J, Platz A (2005).
Biallelic deletions in INK4 in cutaneous melanoma are com-
mon and associated with decreased survival. Clin Cancer Res
11:2991–2997.
Green A, MacLennan R, Youl P, Martin N (1993). Site distribution of
cutaneous melanoma in Queensland. Int J Cancer 53:232–236.
Gruis NA, van der Velden PA, Sandkuijl LA, et al. (1995). Homozygotes
for CDKN2 (p16) germline mutation in Dutch familial melanoma
kindreds. Nat Genet 10:351–353.
Gudbjartsson DF, Sulem P, Stacey SN, et al. (2008). ASIP and TYR pig-
mentation variants associate with cutaneous melanoma and basal
cell carcinoma. Nat Genet 40:886–891.
Gunnarsson U, Hellstrom AR, Tixier-Boichard M, et  al. (2007).
Mutations in SLC45A2 cause plumage color variation in chicken
and Japanese quail. Genetics 175:867–877.
Hahn H, Wicking C, Gailani MR, et al. (1996). Mutations of the human
homolog of Drosophila patched in the nevoid basal cell carcinoma
syndrome. Cell 85:841–851.
Halaban R, Svedine S, Cheng E, et al. (2000). Endoplasmic reticulum
retention is a common defect associated with tyrosinase-negative
albinism. Proc Natl Acad Sci U S A 97:5889–5894.
Han J, Kraft P, Nan H, et al. (2008). A genome-wide association study
identifies novel alleles associated with hair color and skin pigmenta-
tion. PLoS Genet 4: e1000074.
Han J, Qureshi AA, Nan H, et  al. (2011). A germline variant in the
interferon regulatory factor 4 gene as a novel skin cancer risk locus.
Cancer Res 71:1533–1539.
Harrison GA, Owen JJ (1964). Studies on the inheritance of human skin
colour. Ann Hum Genet 28:27–37.
Healy E, Belgaid CE, Takata M, et  al. (1996). Allelotypes of primary
cutaneous melanoma and benign melanocytic nevi. Cancer Res
56:589–593.
Healy E, Belgaid C, Takata M, et al. (1998). Prognostic significance of
allelic losses in primary melanoma. Oncogene 16:2213–2218.
Healy E, Flanagan N, Ray A, et  al. (2000). Melanocortin-1-receptor
gene and sun sensitivity in individuals without red hair. Lancet
355:1072–1073.
Healy E, Jordan SA, Budd PS, et  al. (2001). Functional variation
of MC1R alleles from red-haired individuals. Hum Mol Genet
10:2397–2402.
Healy E, Rehman I, Angus B, Rees JL (1995). Loss of heterozygosity in
sporadic primary cutaneous melanoma. Gene Chromosome Canc
12:152–156.
Healy E, Sikkink S, Rees JL (1996). Infrequent mutation of p16INK4 in
sporadic melanoma. J Invest Dermatol 107:318–321.
Heitzer E, Lassacher A, Quehenberger F, Kerl H, Wolf P (2007). UV fin-
gerprints predominate in the PTCH mutation spectra of basal cell
carcinomas independent of clinical phenotype. J Invest Dermatol
127:2872–2881.
Hemminki K, Xu G, Le Curieux F (2001). Ultraviolet radiation-induced
photoproducts in human skin DNA as biomarkers of damage and
its repair. IARC Sci Publ 154:69–79.
Hermansky F, Pudlak P (1959). Albinism associated with hemor-
rhagic diathesis and unusual pigmented reticular cells in the bone
marrow:  report of two cases with histochemical studies. Blood
14:162–169.
Herraiz C, Journé F, Abdel-Malek Z, Ghanem G, Jiménez-Cervantes
C, García-Borrón JC (2011). Signaling from the human mela-
nocortin 1 receptor to ERK1 and ERK2 mitogen-activated pro-
tein kinases involves transactivation of cKIT. Mol Endocrinol
25:138–156.
Hodis E, Watson IR, Kryukov GV, et al. (2012). A landscape of driver
mutations in melanoma. Cell 150:251–263.
Hunt G, Kyne S, Ito S, Wakamatsu K, Todd C, Thody A (1995).
Eumelanin and phaeomelanin contents of human epidermis and
cultured melanocytes. Pigment Cell Res 8:202–208.
Hussussian CJ, Struewing JP, Goldstein AM, et al. (1994). Germline p16
mutations in familial melanoma. Nat Genet 8:15–21.
Huynh C, Roth D, Ward DM, Kaplan J, Andrews NW (2004).
Defective lysosomal exocytosis and plasma membrane repair
in Chediak-Higashi/beige cells. Proc Natl Acad Sci U S A
101:16795–16800.
Ibarrola-Villava M, Hu HH, Guedj M, et al. (2012). MC1R, SLC45A2
and TYR genetic variants involved in melanoma susceptibility in
southern European populations:  results from a meta-analysis. Eur
J Cancer 48:2183–2191.
Jimbow K, Ishida O, Ito S, et al. (1983). Combined chemical and elec-
tron microscopic studies of pheomelanosomes in human red hair. J
Invest Dermatol 81:506–511.
Jin Y, Birlea SA, Fain PR, et  al. (2010a). Variant of TYR and auto-
immunity susceptibility loci in generalized vitiligo. Nat Genet
42:576–578.
G enetics and G enomics o f S kin D iseases , I I   •   7 0 9
Jin Y, Birlea SA, Fain PR, et al. (2010b). Common variants in FOXP1 are
associated with generalized vitiligo. Nat Genet 42:576–578.
Jin Y, Birlea SA, Fain PR, et al. (2012). Genome-wide association analy-
ses identify 13 new susceptibility loci for generalized vitiligo. Nat
Genet 44:676–680.
Jin Y, Mailloux CM, Gowan K, et al. (2007). NALP1 in vitiligo-associated
multiple autoimmune disease. N Engl J Med 356:1216–1225.
Jin SY, Park HH, Li GZ, Lee HJ, et al. (2004). Association of angioten-
sin converting enzyme gene I/D polymorphism of vitiligo in Korean
population. Pigment Cell Res 17:84–86.
Johnson RL, Rothman AL, Xie J, et  al. (1996). Human homolog of
Patched, a candidate gene for the basal cell nevus syndrome. Science
272:1668–1671.
Johnston JD, Winder AF, Breimer LH (1992). An MboI polymorphism
at codon 192 of the human tyrosinase gene is present in Asians and
Afrocaribbeans. Nucleic Acids Res 20:1433.
Jonason AS, Kunala S, Price GJ, et  al. (1996). Frequent clones of
p53-mutated keratinocytes in normal human skin. Proc Natl Acad
Sci U S A 93:14025–14029.
Kabbarah O, Nogueira C, Feng B, et al. (2010). Integrative genome com-
parison of primary and metastatic melanomas. PLoS One 5:e10770.
Kadekaro AL, Kavanagh R, Kanto H, et al. (2005). alpha-Melanocortin
and endothelin-1 activate antiapoptotic pathways and reduce DNA
damage in human melanocytes. Cancer Res 65:4292–4299.
Kaidbey KH, Agin PP, Sayre RM, Kligman AM (1979). Photoprotection
by melanin—a comparison of black and Caucasian skin. J Am Acad
Dermatol 1:249–260.
Kanetsky PA, Swoyer J, Panossian S, et al. (2002). A polymorphism in
the agouti signaling protein gene is associated with human pigmen-
tation. Am J Hum Genet 70:770–775.
Kauffmann A, Rosselli F, Lazar V, et al. (2008). High expression of DNA
repair pathways is associated with metastasis in melanoma patients.
Oncogene 27:565–573.
Kennedy C, ter Huurne J, Berkhout M, et  al. (2001). Melanocortin 1
receptor (MC1R) gene variants are associated with an increased risk
for cutaneous melanoma which is largely independent of skin type
and hair color. J Invest Dermatol 117:294–300.
Kenny EE, Timpson NJ, Sikora M, et al. (2012). Melanesian blond hair is
caused by an amino acid change in TYRP1. Science 336:554.
King RA, Mentink MM, Oetting WS (1991a). Non-random distribu-
tion of missense mutations within the human tyrosinase gene in
type I (tyrosinase-related) oculocutaneous albinism. Mol Biol Med
8:19–29.
King RA, Townsend D, Oetting W, et al. (1991b). Temperature-sensitive
tyrosinase associated with peripheral pigmentation in oculocutane-
ous albinism. J Clin Invest 87:1046–1053.
King RA, Willaert RK, Schmidt RM, et al. (2003). MC1R mutations
modify the classic phenotype of oculocutaneous albinism type 2
(OCA2). Am J Hum Genet 73:638–645.
Kinsler VA, Abu-Amero S, Budd P, et al. (2012). Germline melanocortin-
1-receptor genotype is associated with severity of cutaneous phe-
notype in congenital melanocytic nevi: a role for MC1R in human
fetal development. J Invest Dermatol 132:2026–2032.
Kovalyshyn I, Braun R, Marghoob A (2009). Congenital melanocytic
naevi. Australas J Dermatol 50:231–240.
Kramata P, Lu YP, Lou YR, Singh RN, Kwon SM, Conney AH (2005).
Patches of mutant p53-immunoreactive epidermal cells induced by
chronic UVB irradiation harbor the same p53 mutations as squa-
mous cell carcinomas in the skin of hairless SKH-1 mice. Cancer Res
65:3577–3585.
Krauthammer M, Kong Y, Ha BH, et al. (2012). Exome sequencing iden-
tifies recurrent somatic RAC1 mutations in melanoma. Nat Genet
44:1006–1014.
Kreimer-Erlacher H, Seidl H, Back B, Kerl H, Wolf P (2001). High
mutation frequency at Ha-ras exons 1–4 in squamous cell carcino-
mas from PUVA-treated psoriasis patients. Photochem Photobiol
74:323–330.
Kricker A, Armstrong BK, English DR, Heenan PJ (1995). A
dose-response curve for sun exposure and basal cell carcinoma. Int
J Cancer 60:482–488.
Krude H, Biebermann H, Luck W, Horn R, Brabant G, Grüters A
(1998). Severe early-onset obesity, adrenal insufficiency and red hair
pigmentation caused by POMC mutations in humans. Nat Genet
19:155–157.
Kruger K, Blume-Peytavi U, Orfanos CE (1999). Basal cell carcinoma
possibly originates from the outer root sheath and/or the bulge
region of the vellus hair follicle. Arch Dermatol Res 291:253–259.
Kumar R, Angelini S, Snellman E, Hemminki K (2004). BRAF muta-
tions are common somatic events in melanocytic nevi. J Invest
Dermatol 122:342–348.
Küsters-Vandevelde HV, Van Leeuwen A, Verdijk MA, et  al. (2010).
CDKN2A but not TP53 mutations nor HPV presence predict
poor outcome in metastatic squamous cell carcinoma of the skin. Int
J Cancer 126:2123–2132.
Lamason RL, Mohideen MA, Mest JR, et al. (2005). SLC24A5, a puta-
tive cation exchanger, affects pigmentation in zebrafish and humans.
Science 310:1782–1786.
Lear JT, Heagerty AH, Smith A, et al. (1996). Multiple cutaneous basal
cell carcinomas: glutathione S-transferase (GSTM1, GSTT1) and
cytochrome P450 (CYP2D6, CYP1A1) polymorphisms influence
tumour numbers and accrual. Carcinogenesis 17:1891–1896.
Lear JT, Smith AG, Heagerty AH, et  al. (1997). Truncal site and
detoxifying enzyme polymorphisms significantly reduce time to
presentation of further primary cutaneous basal cell carcinoma.
Carcinogenesis 18:1499–1503.
Lee HO, Levorse JM, Shin MK (2003). The endothelin receptor-B is
required for the migration of neural crest-derived melanocyte and
enteric neuron precursors. Dev Biol 259:162–175.
Lee ST, Nicholls RD, Bundey S, et al. (1994). Mutations of the P gene
in oculocutaneous albinism, ocular albinism, and Prader-Willi syn-
drome plus albinism. N Engl J Med 330:529–534.
Lerner AB, McGuire JS (1961). Effect of alpha- and beta-melanocyte
stimulating hormones on the skin colour of man. Nature
189:176–179.
Lerner AB (1959). Vitiligo. J Invest Dermatol 32:285–310.
Lorini R, Orecchia G, Martinetti M, Dugoujon JM, Cuccia M (1992).
Autoimmunity in vitiligo:  relationship with HLA, Gm and Km
polymorphisms. Autoimmunity 11:255–260.
Lubbe J, Reichel M, Burg G, Kleihues P (1994). Absence of p53
gene mutations in cutaneous melanoma. J Invest Dermatol 102:
819–821.
Lucas R, McMichael T, Smith W, et al. (2009). Solar ultraviolet radia-
tion:  global burden of disease from solar ultraviolet radiation.
In:  Environmental Burden of Disease Series No. 13. (Prüss-Üstün
A, Zeeb H, Mathers C, Repacholi M, eds.) Geneva: World Health
Organisation; 1–250.
Lucotte G, Mercier G, Diéterlen F, Yuasa I (2010). A decreasing gradient
of 374F allele frequencies in the skin pigmentation gene SLC45A2,
from the north of West Europe to North Africa. Biochem Genet
48:26–33.
Macgregor S, Montgomery GW, Liu JZ, et  al. (2011). Genome-wide
association study identifies a new melanoma susceptibility locus at
1q21.3. Nat Genet 43:1114–1118.
Mariat D, Taourit S, Guérin G (2003). A mutation in the MATP
gene causes the cream coat colour in the horse. Genet Sel Evol
35:119–133.
McGregor JM, Proby CM (1995). Skin cancer in transplant recipients.
Lancet 346:964–965.
Mengel-From J, Wong TH, Morling N, Rees JL, Jackson IJ (2009).
Genetic determinants of hair and eye colours in the Scottish and
Danish populations. BMC Genet 10:88.
Metcalfe C, de Sauvage FJ (2011). Hedgehog fights back: mechanisms
of acquired resistance against Smoothened antagonists. Cancer Res
71:5057–5061.
7 1 0   •   G enomics in C l inica l P ractice
Mizuno T, Tokuoka S, Kishikawa M, Nakashima E, Mabuchi K,
Iwamoto KS (2006). Molecular basis of basal cell carcinogenesis in
the atomic-bomb survivor population: p53 and PTCH gene altera-
tions. Carcinogenesis 27:2286–2294.
Montoliu L, Oetting WS, Bennett DC. Color genes. (2012). European​
Society for Pigment Cell Research. Available at: http://www.​espcr.
org/micemut​.
Mocellin S, Verdi D, Nitti D (2009). DNA repair gene polymor-
phisms and risk of cutaneous melanoma:  a systematic review and
meta-analysis. Carcinogenesis 30:1735–1743.
Nagle DL, Karim MA, Woolf EA, et al. (1996). Identification and muta-
tion analysis of the complete gene for Chediak-Higashi syndrome.
Nat Genet 14:307–311.
Nakayama K, Fukamachi S, Kimura H, Koda Y, Soemantri A, Ishida T
(2002). Distinctive distribution of AIM1 polymorphism among
major human populations with different skin color. J Hum Genet
47:92–94.
Nakazawa H, English D, Randell PL, et  al. (1994). UV and skin can-
cer: specific p53 gene mutation in normal skin as a biologically rel-
evant exposure measurement. Proc Natl Acad Sci U S A 91:360–364.
Nan H, Kraft P, Qureshi AA, et  al. (2009). Genome-wide association
study of tanning phenotype in a population of European ancestry. J
Invest Dermatol 129:2250–2257.
Nan H, Xu M, Kraft P, et  al. (2011). Genome-wide association study
identifies novel alleles associated with risk of cutaneous basal
cell carcinoma and squamous cell carcinoma. Hum Mol Genet
20:3718–3724.
Nath SK, Kelly JA, Namjou B, et al. (2001). Evidence for a susceptibility
gene, SLEV1, on chromosome 17p13 in families with vitiligo-related
systemic lupus erythematosus. Am J Hum Genet 69:1401–1406.
Nath SK, Majumder PP, Nordlund JJ (1994). Genetic epidemiology of
vitiligo:  multilocus recessivity cross-validated. Am J Hum Genet
55:981–990.
Newton RA, Smit SE, Barnes CC, Pedley J, Parsons PG, Sturm RA
(2005). Activation of the cAMP pathway by variant human
MC1R alleles expressed in HEK and in melanoma cells. Peptides
26:1818–1824.
Nikolaev SI, Rimoldi D, Iseli C, et al. (2011). Exome sequencing identi-
fies recurrent somatic MAP2K1 and MAP2K2 mutations in mela-
noma. Nat Genet 44:133–139.
Nishimura EK, Jordan SA, Oshima H, et  al. (2002). Dominant role
of the niche in melanocyte stem-cell fate determination. Nature
416:854–860.
Omholt K, Platz A, Kanter L, Ringborg U, Hansson J (2003). NRAS
and BRAF mutations arise early during melanoma pathogenesis
and are preserved throughout tumor progression. Clin Cancer Res
9:6483–6488.
Oro AE, Higgins KM, Hu Z, et al. (1997). Basal cell carcinomas in mice
overexpressing Sonic hedgehog. Science 276:817–821.
Palmer JS, Duffy DL, Box NF, et  al. (2000). Melanocortin-1 recep-
tor polymorphisms and risk of melanoma:  is the association
explained solely by pigmentation phenotype? Am J Hum Genet
66:176–186.
Pastorino L, Ghiorzo P, Nasti S, et al. Identification of a SUFU germ-
line mutation in a family with Gorlin syndrome. Am J Med Genet
149A:1539–1543.
Perry PK, Silverberg NB. Cutaneous malignancy in albinism. (2001).
Cutis 67:427–430.
Pesz KA, Bieniek A, Makowska I, Sąsiadek MM (2013). Basal cell carci-
noma of the skin: whole genome screening by comparative genome
hybridization revisited. J Cutan Pathol 40(1):25–29; doi:10.1111/
cup.12033.
Pierceall WE, Goldberg LH, Tainsky MA, Mukhopadhyay T,
Ananthaswamy HN (1991). Ras gene mutation and amplification
in human nonmelanoma skin cancers. Mol Carcinog 4:196–202.
Ping XL, Ratner D, Zhang H, et al. (2001). PTCH mutations in squa-
mous cell carcinoma of the skin. J Invest Dermatol 116:614–616.
Pingault V, Bondurand N, Kuhlbrodt K, et al. (1998). SOX10 mutations
in patients with Waardenburg-Hirschsprung disease. Nat Genet
18:171–173.
Pollock PM, Harper UL, Hansen KS, et al. (2003). High frequency of
BRAF mutations in nevi. Nat Genet 33:19–20.
Prota G, Thomson RH (1976). Melanin pigmentation in mammals.
Endeavour 35:32–38.
Puri N, Gardner JM, Brilliant MH (2000). Aberrant pH of melanosomes
in pink-eyed dilution (p)  mutant melanocytes. J Invest Dermatol
115:607–613.
Quinn AG, Sikkink S, Rees JL (1994). Basal cell carcinomas and squa-
mous cell carcinomas of human skin show distinct patterns of chro-
mosome loss. Cancer Res 54:4756–4759.
Rady P, Scinicariello F, Wagner RF Jr, Tyring SK (1992). P53 mutations
in basal cell carcinomas. Cancer Res 52:3804–3806.
Ramoz N, Rueda LA, Bouadjar B, et al. (2002). Mutations in two adja-
cent novel genes are associated with epidermodysplasia verrucifor-
mis. Nat Genet 32:579–581.
Rampen FH, Fleuren BA, de Boo TM, Lemmens WA (1988).
Unreliability of self-reported burning tendency and tanning ability.
Arch Dermatol 124:885–888.
Rawles ME (1947). Origin of pigment cells from the neural crest in the
mouse embryo. Physiol Zool 20:248–266.
Rebel H, Kram N, Westerman A, Banus S, van Kranen HJ, de Gruijl FR
(2005). Relationship between UV-induced mutant p53 patches
and skin tumours, analysed by mutation spectra and by induc-
tion kinetics in various DNA-repair-deficient mice. Carcinogenesis
26:2123–2130.
Rehman I, Quinn AG, Healy E, Rees JL (1994). High frequency of
loss of heterozygosity in actinic keratoses, a usually benign disease.
Lancet 344:788–789.
Reifenberger J, Wolter M, Weber RG, et al. (1998). Missense mutations
in SMOH in sporadic basal cell carcinomas of the skin and primitive
neuroectodermal tumors of the central nervous system. Cancer Res
58:1798–1803.
Reifenberger J, Wolter M, Knobbe CB, et al. (2005). Somatic mutations
in the PTCH, SMOH, SUFUH and TP53 genes in sporadic basal
cell carcinomas. Br J Dermatol 152:43–51.
Rinchik EM, Bultman SJ, Horsthemke B, et al. (1993). A gene for the
mouse pink-eyed dilution locus and for human type II oculocutane-
ous albinism. Nature 361:72–76.
Robinson SJ, Healy E (2002). Human melanocortin 1 receptor (MC1R)
gene variants alter melanoma cell growth and adhesion to extracel-
lular matrix. Oncogene 21:8037–8046.
Robinson S, Dixon S, August S, et al. (2010). Protection against UVR
involves MC1R-mediated non-pigmentary and pigmentary mecha-
nisms in vivo. J Invest Dermatol 130:1904–1913.
Rosso S, Zanetti R, Martinez C, et  al. (1996). The multicentre south
European study “Helios” II: different sun exposure patterns in the
aetiology of basal cell and squamous cell carcinomas of the skin. Br
J Cancer 73:1447–1454.
Rouzaud F, Costin GE, Yamaguchi Y, et al. (2006). Regulation of con-
stitutive and UVR-induced skin pigmentation by melanocortin 1
receptor isoforms. FASEB J 20:1927–1929.
Sánchez-Martín M, Pérez-Losada J, Rodríguez-García A, et al. (2003).
Deletion of the SLUG (SNAI2) gene results in human piebaldism.
Am J Med Genet 122A:125–132.
Sato-Jin K, Nishimura EK, Akasaka E, et al. (2008). Epistatic connec-
tions between microphthalmia-associated transcription factor and
endothelin signaling in Waardenburg syndrome and other pigmen-
tary disorders. FASEB J 22:1155–1168.
Schallreuter KU, Bahadoran P, Picardo M, et al. (2008). Vitiligo patho-
genesis:  autoimmune disease, genetic defect, excessive reactive
oxygen species, calcium imbalance, or what else? Exp Dermatol
17:139–140; discussion, 141–160.
Schioth HB, Phillips SR, Rudzish R, Birch-Machin MA, Wikberg
JES, Rees JL (1999). Loss of function mutations of the human
G enetics and G enomics o f S kin D iseases , I I   •   7 1 1
melanocortin 1 receptor are common and are associated with red
hair. Biochem Biophys Res Commun 260:488–491.
Schrama D, Scherer D, Schneider M, et  al. (2011). ERCC5  p.
Asp1104His and ERCC2  p.Lys751Gln polymorphisms are inde-
pendent prognostic factors for the clinical course of melanoma. J
Invest Dermatol 131:1280–1290.
Sekulic A, Migden MR, Oro AE, et  al. (2012). Efficacy and safety
of vismodegib in advanced basal-cell carcinoma. N Engl J Med
366:2171–2179.
Seiji M, Fitzpatrick TB, Simpson RT, Birbeck MS (1963). Chemical
composition and terminology of specialized organelles (melano-
somes and melanin granules) in mammalian melanocytes. Nature
197:1082–1084.
Smith R, Healy E, Siddiqui S, et al. (1998). Melanocortin 1 receptor vari-
ants in an Irish population. J Invest Dermatol 111:119–122.
Soengas MS, Capodieci P, Polsky D, et al. (2001). Inactivation of the apop-
tosis effector Apaf-1 in malignant melanoma. Nature 409:207–211.
Sosman JA, Kim KB, Schuchter L, et  al. (2012). Survival in BRAF
V600-mutant advanced melanoma treated with vemurafenib. N
Engl J Med 366:707–714.
Spencer JM, Kahn SM, Jiang W, DeLeo VA, Weinstein IB (1995).
Activated ras genes occur in human actinic keratoses, prema-
lignant precursors to squamous cell carcinomas. Arch Dermatol
131:796–800.
Stacey SN, Gudbjartsson DF, Sulem P, et al. (2008). Common variants
on 1p36 and 1q42 are associated with cutaneous basal cell carci-
noma but not with melanoma or pigmentation traits. Nat Genet
40:1313–1318.
Stacey SN, Sulem P, Masson G, et al. (2009). New common variants affect-
ing susceptibility to basal cell carcinoma. Nat Genet 41:909–914.
Stacey SN, Sulem P, Jonasdottir A, et al. (2011). A germline variant in
the TP53 polyadenylation signal confers cancer susceptibility. Nat
Genet 43:1098–1103.
Stark MS, Woods SL, Gartside MG, et  al. (2011). Frequent somatic
mutations in MAP3K5 and MAP3K9 in metastatic melanoma
identified by exome sequencing. Nat Genet 44:165–169.
Staricco RJ, Pinkus H (1957). Quantitative and qualitative data on
the pigment cells of adult human epidermis. J Invest Dermatol
28:33–45.
Steel KP, Davidson DR, Jackson IJ (1992). TRP-2/DT, a new early mela-
noblast marker, shows that steel growth factor (c-kit ligand) is a sur-
vival factor. Development 115:1111–1119.
Stern C (1970). Model estimates of the number of gene pairs involved
in pigmentation variability of the Negro American. Hum Hered
20:165–168.
Stern RS, Lunder EJ (1998). Risk of squamous cell carcinoma and methox-
salen (psoralen) and UV-A radiation (PUVA): a meta-analysis. Arch
Dermatol 134:1582–1585.
Stinchcombe JC, Page LJ, Griffiths GM (2000). Secretory lysosome
biogenesis in cytotoxic T lymphocytes from normal and Chediak
Higashi syndrome patients. Traffic 1:435–444.
Stokowski RP, Krishna Pant PV, Tony Dadd, et al. (2007). A Genomewide
Association Study of Skin Pigmentation in a South Asian Population.
Am J Hum Genet 81(6):1119–1132; doi:10.1086/522235.
Sturm RA, Duffy DL, Zhao ZZ, et  al. (2008). A single SNP in an
evolutionary conserved region within intron 86 of the HERC2
gene determines human blue-brown eye color. Am J Hum Genet
82:424–431.
Su F, Viros A, Milagre C, et  al. (2012). RAS mutations in cutaneous
squamous-cell carcinomas in patients treated with BRAF inhibitors.
N Engl J Med 366:207–215.
Sulem P, Gudbjartsson DF, Stacey SN, et  al. (2007). Genetic determi-
nants of hair, eye and skin pigmentation in Europeans. Nat Genet
39:1443–1452.
Sulem P, Gudbjartsson DF, Stacey SN, et al. (2008). Two newly identi-
fied genetic determinants of pigmentation in Europeans. Nat Genet
40:835–837.
Szabo G, Gerald AB, Pathak MA, Fitzpatrick TB (1969). Racial dif-
ferences in the fate of melanosomes in human epidermis. Nature
222:1081–1082.
Tan CP, McKee KK, Weinberg DH, et al. (1999). Molecular analysis of
a new splice variant of the human melanocortin-1 receptor. FEBS
Lett 451:137–141.
Tastan HB, Akar A, Orkunoglu FE, Arca E, Inal A (2004). Association
of HLA class I antigens and HLA class II alleles with vitiligo in a
Turkish population. Pigment Cell Res 17:181–184.
Tcheung WJ, Selim MA, Herndon JE 2nd, Abernethy AP, Nelson KC
(2012). Clinicopathologic study of 85 cases of melanoma of the
female genitalia. J Am Acad Dermatol 67:598–605.
Thody AJ, Higgins EM, Wakamatsu K, Ito S, Burchill SA, Marks JM
(1991). Pheomelanin as well as eumelanin is present in human epi-
dermis. J Invest Dermatol 97:340–344.
Tomita Y, Takeda A, Okinaga S, Tagami H, Shibahara S (1989). Human
oculocutaneous albinism caused by single base insertion in the
tyrosinase gene. Biochem Biophys Res Commun 164:990–996.
Tripathi RK, Giebel LB, Strunk KM, Spritz RA (1991). A polymorphism
of the human tyrosinase gene is associated with temperature-sensitive
enzymatic activity. Gene Expr 1:103–110.
Urano Y, Asano T, Yoshimoto K, et al. (1995). Frequent p53 accumula-
tion in the chronically sun-exposed epidermis and clonal expansion
of p53 mutant cells in the epidermis adjacent to basal cell carcinoma.
J Invest Dermatol 104:928–932.
Valverde P, Healy E, Jackson I, Rees JL, Thody AJ (1995). Variants of the
melanocyte-stimulating hormone receptor gene are associated with
red hair and fair skin in humans. Nat Genet 11:328–330.
Valverde P, Healy E, Sikkink S, et al. (1996). The Asp84Glu variant of
the melanocortin 1 receptor (MC1R) is associated with melanoma.
Hum Mol Genet 5:1663–1666.
van der Velden PA, Sandkuijl LA, Bergman W, et  al. (2001).
Melanocortin-1 receptor variant R151C modifies melanoma risk
in Dutch families with melanoma. Am J Hum Genet 69:774–779.
van ‘t Veer LJ, Burgering BM, Versteeg R, et al. (1989). N-ras mutations
in human cutaneous melanoma from sun-exposed body sites. Mol
Cell Biol 9:3114–3116.
Vitasa BC, Taylor HR, Strickland PT, et al. (1990). Association of non-
melanoma skin cancer and actinic keratosis with cumulative solar
ultraviolet exposure in Maryland watermen. Cancer 65:2811–2817.
Vogel P, Read RW, Vance RB, Platt KA, Troughton K, Rice DS (2008).
Ocular albinism and hypopigmentation defects in Slc24a5–/– mice.
Vet Pathol 45:264–279.
7 1 2   •   G enomics in C l inica l P ractice
Watanabe A, Takeda K, Ploplis B, Tachibana M (1998). Epistatic rela-
tionship between Waardenburg syndrome genes MITF and PAX3.
Nat Genet 18:283–286.
Wei ML (2006). Hermansky-Pudlak syndrome: a disease of protein traf-
ficking and organelle function. Pigment Cell Res 19:19–42.
Wei X, Walia V, Lin JC, et  al. (2011). Exome sequencing identi-
fies GRIN2A as frequently mutated in melanoma. Nat Genet
43:442–446.
Welsh MM, Karagas MR, Kuriger JK, et  al. (2011). Genetic determi-
nants of UV-susceptibility in non-melanoma skin cancer. PLoS One
6:e20019.
Whiteman DC, Whiteman CA, Green AC (2001). Childhood sun
exposure as a risk factor for melanoma: a systematic review of epide-
miologic studies. Cancer Causes Control 12:69–82.
Xie J, Murone M, Luoh SM, et al. (1998). Activating Smoothened muta-
tions in sporadic basal-cell carcinoma Nature 391:90–92.
Yuasa I, Umetsu K, Harihara S, et al. (2006). Distribution of the F374
allele of the SLC45A2 (MATP) gene and founder-haplotype analy-
sis. Ann Hum Genet 70:802–811.
Yuasa I, Umetsu K, Watanabe G, Nakamura H, Endoh M, Irizawa Y
(2004). MATP polymorphisms in Germans and Japanese:  the
L374F mutation as a population marker for Caucasoids. Int J Legal
Med 118:364–366.
Yengi L, Inskip A, Gilford J, et  al. (1996). Polymorphism at the glu-
tathione S-transferase locus GSTM3:  interactions with cyto-
chrome P450 and glutathione S-transferase genotypes as risk
factors for multiple cutaneous basal cell carcinoma. Cancer Res
56:1974–1977.
Yokoyama S, Woods SL, Boyle GM, et  al. (2011). A novel recurrent
mutation in MITF predisposes to familial and sporadic melanoma.
Nature 480:99–103.
Yoshida H, Kunisada T, Kusakabe M, Nishikawa S, Nishikawa SI
(1996). Distinct stages of melanocyte differentiation revealed
by analysis of non-uniform pigmentation patterns. Development
122:1207–1214.
Zamani M, Spaepen M, Sghar SS, et  al. (2001). Linkage and associa-
tion of HLA class II genes with vitiligo in a Dutch population. Br J
Dermatol 145:90–94.
Ziegler A, Jonason AS, Leffell DJ, et al. (1994). Sunburn and p53 in the
onset of skin cancer. Nature 372:773–776.
Zuo L, Weger J, Yang Q, et  al. (1996). Germline mutations in the
p16INK4a binding domain of CDK4 in familial melanoma. Nat
Genet 12:97–99.
 46.
THE GENETIC AND GENOMIC PRACTICE
OF REPRODUCTIVE MEDICINE
Dhavendra Kumar
INTRODUCTION
The fields of reproductive medicine and obstetrics cover
a very broad range of disease conditions, as well as natu-
ral physiological processes such as pregnancy, parturition,
and the menstrual cycle. Unique in this field of study is the
need to consider the functioning of more than one organ-
ism: mother and fetus (in pregnancy), or mother and father
(in fertilization and infertility). This chapter reviews the evi-
dence and clinical applications of the available genetic and
genomic data and information relevant to the practice of
reproductive medicine. Aspects of the gynecological prac-
tice are beyond the scope of this chapter and are excluded.
However, a section on gynecological cancers is included in
the chapter on cancer genomics (see Chapter 36).
Genomics concerns the study of these genes in context
within our genetic system, rather than as individuals. Large
numbers of genes, or even all of the known genes in the
genome, can have either their level of activity, or their struc-
ture, examined simultaneously. The total number of genes
in the human genome is considerably less (around 23,000;
see Chapter 2) than was previously estimated (32,000 and
39,000)[1]‌. The very existence of genomics as a field of study
owes its success to two major developments. The first of
these was the completion of the Human Genome Project,
providing us with a complete template of the human
genome and setting the scene for the promising new field
of genomic medicine[2]. The second has been the develop-
ment of tools enabling the interrogation of differences in
structure or function of large numbers of genes in the same
experiment. The most commonly used of these tools have
been microarrays[3].
Microarrays consist of a pattern (array) of spots on
a “micro” scale designed to examine thousands of com-
ponents of the genome simultaneously. The two aspects
of the genome most commonly studied are variations in
genetic sequence or structure (such as single-nucleotide
713
polymorphisms [SNPs]) and changes in gene expression or
function (via measurement in differences in mRNA abun-
dance). The predominant use of microarrays in the study
of obstetrics, gynecology, and reproductive medicine, as in
other medical fields, has been to analyze differences in gene
expression[4]‌. A detailed description of the gene-expression
microarray laboratory technique is beyond the scope of this
chapter.
In this chapter, the evidence from the existing literature
dealing with the applications of genomics to obstetrical and
reproductive medicine is used. Gene-expression microar-
rays will be the main genomic studies dealt with, as they
constitute the vast majority of genomic studies published in
the reproductive field.
F E M A L E I N F E RT I L I T Y
C O N G E N I TA L A N O M A L I E S O F T H E
R E P RO D U C T I VE T R AC T
Investigations for infertility should include a thorough clin-
ical assessment supported by relevant internal exploration
and imaging to exclude congenital anatomical anomalies.
This would apply to both male and female partners. In most
cases, this could also include laboratory investigations to
assess adequate sperm quantity and quality, endocrine pro-
file for satisfactory ovulation, chromosome analysis with
new molecular techniques (fluorescence in situ hybridiza-
tion [FISH] and array comparative genomic hybridization
[array CGH]), and, if applicable, relevant molecular genetic
analysis (e.g., cystic fibrosis). Congenital anomalies of the
male reproductive tract are discussed elsewhere in this chap-
ter in the section on male factor infertility.
Developmental anomalies of the female reproductive
tract per se may diminish the natural potential of achiev-
ing a successful and viable pregnancy. The majority of
the anomalies are isolated and reflect unequal and abnor-
mal embryological differentiation of the Müllerian tract.
However, it is likely that abnormal Müllerian canal differ-
entiation could be associated with other malformations,
predominantly of mesodermal origin. There are several rec-
ognizable multiple malformation syndromes that include
Müllerian and as well as cloacal-developmental anomalies
(Table 46.1). Several genes with regulatory sequences and
polymorphisms are now assigned to some of these dysmor-
phic conditions. Diagnostic applications of the molecular
pathology in some of these conditions offer reliable tools
for confirmation of the diagnosis and reproductive coun-
selling. The scope of this chapter, however, limits the space
for a detailed description of the key dysmorphic conditions
with major female reproductive tract anomalies.
In addition to the above-listed malformation syndromes,
anomalous development of the female reproductive tract is
likely indicate a chromosomal disorder, most notably Turner
syndrome. It is essential that detailed cytogenetic analysis
be carried out in any such case. New molecular cytogenetic
techniques (FISH and array CGH) and new genomic tools
(exome and genome sequencing) are invaluable for eluci-
dating the causative underlying structural genomic abnor-
mality. It is important that any such case be investigated at
a dedicated tertiary pediatric gynecology or infertility unit
equipped with the necessary molecular and imaging diag-
nostic facilities, supported by a skilled multidisciplinary
team (pediatric gynecologist, developmental pediatrician,
pediatric surgeon, clinical geneticist, reproductive medicine
clinician, and clinical psychologist).
D I S O R D E R S O F T H E E N D O M ET R IUM
The endometrium is a dynamic tissue that responds to mul-
tiple stimuli, depending on physiological and environmen-
tal conditions, including steroid hormones, an implanting
conceptus, withdrawal of steroid hormones, contraceptive
steroids, selective steroid hormone-receptor modulators,
infection, transient cell populations, and metaplastic and
neoplastic agents[5]‌. Microarray gene-expression profiling

Table 46.1 SELECTION OF SYNDROMES WITH ABNORMAL FEMALE REPRODUCTIVE TRACT ANATOMY

SYNDROME GYNECOLOGICAL ANOMALY OMIM GENE (LOCUS) GENE PRODUCT

Adams-Oliver Vaginal atresia 100300 3q13;3q21 ARHGAP31;
RBP2
Antley-Bixler Hypoplastic labia majora 207410 FGFR2
FGFR3
POR
Vaginal atresia
Apert Vaginal atresia 101200 10q25-q26 FGFR2
Bardet-Biedl Vaginal atresia 209900 Several BBS1-12
Fraser Bicornate uterus 219000 4q21;13q13 FRAS1
FREM2
Vaginal atresia
Fryns Vaginal atresia 229850 1q41 ?
Bicornate uterus
Gorlin Bicornate uterus 109400 9q22-q31 PTCH1
SUFU
Hand-Foot-Genital Absent/small uterus 140000 7p15 HOXA13
Johanson-Blizzard Bicornate uterus 243800 15q14-q21 UBR1
McKusick-Kaufmann Vaginal/cervical atresia 236700 20p12 MKKS
Hydrometrocolpos
MURCS association Vaginal atresia 601076 12q14; 14q31 ??
Roberts-SC phocomelia Bicornate uterus 268300 8p12-8p21 ESCO2
Schinzel-Giedion Bicornate uterus 269150 18q21 SETBP1
Sienomelia sequence Vaginal atresia 182940 ? ?
Urorectal septum Vaginal atresia ? 10q25-10q26?

7 1 4   •   G enomic s in C linical P ractice

allows large numbers of genes to be investigated simultane-
ously, and the resulting gene-expression patterns can then
be correlated with different structural, functional, or clini-
cal parameters[6].
Endometrial gene-expression profiling is vastly differ-
ent when comparing data obtained using endometrial cell
cultures from profiling derived from endometrial biopsies,
as cells may have different responses in vitro than in vivo.
These differences may be due to the processing of the cells,
the culture medium used, and the loss of paracrine interac-
tions with other cells in situ[7]‌. Especially important when
dealing with endometrial tissue is careful clinical character-
ization of the subjects donating the tissue, their hormonal
stage, their age, any history of endometrial pathologies, and
other medical conditions; it is also important to minimize
inter-subject variability and to maximize data reliability.
Gene-Expression Profiling of the Whole
Endometrial Tissue
There are only few studies reporting on endometrial
gene-expression profiling using the affymetrix HG-U95A
chip (12,000 transcripts) or HG-U95 B-E chip (60,000
transcripts)[8, 9].
Kao et al.[8]‌ looked into differential expression between
the average values of four late proliferative (LP) phase
samples (cycle days 8–10) and samples from seven other
individuals in the mid-secretory (MS) leutinising hor-
mone (LH) phase (LH + 8–LH + 10). They found 114
upregulated and 218 downregulated genes in the MS stage.
Borthwick et  al.[9] compared global gene-expression pro-
files of pooled samples of five women in the proliferative
phase (days 9–11) with pooled samples of five women in
the secretory phase (LH + 6–8) and found 89 upregulated
and 57 downregulated genes in the secretory phase. The
experiments were identical, except that Borthwick et  al.
pooled before hybridizing onto the microarray. There is
about 25–30% agreement between the studies for genes
that were upregulated and 10% agreement for genes that
were downregulated.
Carson and colleagues[10] compared three pooled sam-
ples from the early-secretory (ES) phase (LH days 2–4)
with three pooled samples from the MS phase (LH day
7–9) and found 98 upregulated and 119 downregulated
genes in the MS phase. Riesewijk et  al[11] compared gene
expression between five pairs of ES (LH + 2) and MS (LH
+ 7) samples from five individual women. They found 153
upregulated and 58 downregulated genes in the MS phase.
These two groups have only around 10% agree-
ment between their upregulated gene sets and even less
T he G enetic and G enomic P ractice of   R eproductive M edicine   •   7 1 5
agreement between their downregulated genes. Mirkin and
colleagues[12] investigated the gene-expression profiling of
three samples from the ES stage (LH + 3) versus five samples
from MS stage, and found 49 upregulated and 58 down-
regulated genes in the MS stage compared to ES stage of the
menstrual cycle. A detailed comparison of some of the find-
ings of these five studies is summarized in Table 46.2. The
lack of agreement between the different studies highlights
the difficulties faced with a highly variable, dynamic tissue
such as the endometrium.
In an ambitious project, the transcriptional profiling of
human endometrium during the menstrual cycle through a
seven-phase time-course analysis was attempted[13]. In addi-
tion to identifying 425 genes that showed at least two-fold
upregulation in at least one stage of the menstrual cycle, the
study also demonstrated prediction of the menstrual cycle
stages based on the transcriptional profile of the samples,
and identified different patterns of expression for groups of
the genes associated with different endometrial biological
processes such as implantation and menstruation. It is also
suggested that discordant patterns of gene expression may
help identify endometrial samples with subtle abnormali-
ties not readily apparent in routine histopathology[13].
The length and the subdivision of the dendrogram
branches show the relatedness of the samples, with samples
joined by short branches being most similar in expression
profiles. The 43 samples were sorted into nine clusters based
on the similarity of the gene-expression profiles of the sam-
ples:  mid-proliferatory (MP)—early-proliferatory (EP)
(n = 2); EP–MP (n = 5); MP (n = 5); LP–ES (n = 5); ES–
MS (n = 6); MS–LS (n = 7); LS (n = 3); LS–M (n = 2);
M (n  =  4). Samples in these nine groupings agreed with
their histopathology by either being in the same group, or
in an adjacent group, apart from four samples (E355 [LP],
EA30 [MS], EA85 [LP], and EA2 [EP]). These four outli-
ers were removed from further analysis (Figure 46.1). Three
out of the four outliers (355, A30, and A2) found by hier-
archical clustering showed significant inter-observer vari-
ability between the two histopathology evaluations, while
there was agreement between the two pathologists’ reports
for all other samples. Visual inspection of the fourth out-
lier’s (A85) microarray image showed that it had very poor
hybridization. Following the removal of outliers, group M–
EP was excluded from further analysis due to lack of repli-
cates, and the MS and LS groups were merged to produce
the final molecularly defined classification of the menstrual
cycle. The final groupings (Figure 46.1) were:  EP–MP
(n = 5); MP (n = 5); LP–ES (n = 5); ES–MS (n = 6); MS–
LS (n = 7); LS–M (n = 5); and M (n = 4)[13]. It is likely that
the data on the endometrial gene-expression profiling will
Table 46.2  ENDOMETRIAL GENES RELATED TO IMPLANTATION

GENE NAME ACCESSION NUMBER [8]‌ [10] [11] [9] [12]

Upregulated genes
Osteopontin AF052124 + + + + +
Apolipoprotein D J02611 + + + +
Dickkopf/DKK1 AB020315 + + + +
Placental protein-14/ glycodelin J04129 + + +
Decay accelerating factor for M31516 + + + +
­complement (CD55, Cromer
blood group system)
Adipsin/complement factor D M84526 + + +
Guanylate binding protein 2, M55543 + + +
interferon-inducible
Claudin 4/CEP-R AB000712 + + +
Monoamine oxidase A (MAOA) AA420624/M68840 + + + +
Growth arrest and M60974 + + + +
DNA-damage-inducible protein
(GADD45)
NIP2 AB002365 + + +
Serine (or cysteine) proteinase inhibitor, X54486 + + +
clade G (SERPINGI)
Interleukin 15 (IL15) AF031167 + + + +
Annexin IV (ANXA4) M82809 + + + +
Mitogen-activated protein kinase kinase U67156 + + +
kinase 5 (MAP3K5)
Inhibitor of DNA-binding 4, dominant AL022726 + + +
negative helix-loop-helix protein (ID4)
Complement component 1, r subcom- M14058 + + +
ponent (C1R)
Downregulated genes
Olfactomedin-related ER localized U79299 + + + +
protein
Msh homeobox homolog 1 (MSX1) M97676 + + +
GATA-binding protein 2 (GATA2) M68891 + + +

contribute to the understanding of the endometrial biol-
ogy and also improve the accuracy and reproducibility of
endometrial-dating procedures.
Endometrial Stromal Cell Differentiation
Placental trophoblasts invade into the maternal endome-
trial stromal compartment during implantation in humans.
Decidualization of stromal cells is essential for successful
implantation and represents differentiation to an excep-
tional morphological, biosynthetic, and secretory pheno-
type. It is initiated independently of conception during
the late secretory phase of the normal menstrual cycle. The
molecular pathways that are involved in the process of decid-
ualization are very poorly understood. Decidualization in
vitro can be induced by number of factors, including pro-
gesterone primed with estrogen and cyclic adenosine mono
phosphate (cAMP). Few transcriptional profiling studies
have been undertaken to try to shed light on the decidual-
ization process.
In another study[14], human endometrial stromal cells
were decidualized in vitro in response to progesterone
or cAMP, and the gene-expression profile was compared
to non-decidualized stromal cells. Of the 588 genes ana-
lyzed, significant upregulation of cytokines, growth factors,
nuclear transcription factors, cyclin family members, and
mediators of the cAMP signal transduction pathways were
found. Many of these genes have been implicated in decidu-
alization in vivo. The group identified new members of the
transforming growth factor (TGF) β family upregulated

7 1 6   •   G enomic s in C linical P ractice

 Sample Path stage Molecular stage
E353 M
EA81 LS
EA9 LS LS-M
EA28 LS
EA22 LS
E354 LS
EA4
EA5
LS
LS
LS
EA35 MS
EA67 MS
E355
EA34
LP
MS
MS
EA17 MS
EA14 M
EA20 M
EA16 M M
EA54 M
EA66 M
EA13 EP M-EP
E356 ES
EA41 LP
EA58
EA8
ES
ES
LP-ES
EA21 ES
EA15 MS
EA51 MS
EA82
EA84
ES
ES ES-MS
EA40 ES
EA18 MS
EA85 LP
EA10 MP
EA50 EP
E357
EA52
EP
EP EP-MP
E358 MP
EA44 MP
EA62 MP
EA30 MS
EA63 MP MP
EA42 MP
EA77 MP
EA2 EP

Figure 46.1  Differential endometrial samples based on the global gene-expression profile.

during decidualization and expression of neuromodulators
and neurotransmitter receptors, many of which had not
been previously documented in the differentiation process.
Temporal expression and regulation of nearly 7,000
genes during the decidualization process was investigated
by the gene-expression profile of human termed decid-
ual fibroblasts in response to estrogen + progesterone +
8Br-cAMP at 0, 2, 4, 6, 9, 12, and 15 days[15]. Of 6,918 gene
analyzed, 121 genes expressed by more than two-fold, 110
were downregulated, and 50 showed biphasic behavior.
Dynamically regulated genes clustered into nine patterns of
gene expression as a function of time. These genes were bio-
logically classified into five categories: cell and tissue func-
tion, cell and tissue structure, regulation of gene expression,
expressed sequence tags, and functions unknown. Many
genes were reprogrammed (i.e., changed expression) within
specific functional groups.
Endometriosis
Endometriosis is defined as the presence of endometrial
tissue in ectopic locations outside the uterine cavity, typi-
cally on the pelvic peritoneal surfaces of the uterus and its
ligaments, on the ovaries, and in the rectovaginal septum[16].
Between 5% and 10% of women have endometriosis, which
causes a range of symptoms, from pain to infertility.
The uterine (eutopic) endometrium from women with
endometriosis may have endogenous abnormalities that
promote the establishment of the disease. Matrix metal-
loproteinases (MMPs) MMP-7 and MMP-9 are normally

T he G enetic and G enomic P ractice of   R eproductive M edicine   •   7 1 7

expressed in endometrium during menstruation and are
inhibited by progesterone during the secretory phase of
the cycle. However, in women with endometriosis, these
enzymes are constantly expressed during the secretory
phase. This may allow retrograde endometrial tissue to
occupy the peritoneal surface. Other studies also found
that eutopic endometrium from women with endometrio-
sis expressed higher levels of MMP-2, membranous type 1
MMP, MMP-3, and uPA, and lower levels of TIMP-2 than
endometrium from normal women[17].
Up to 50% of women with infertility have endometrio-
sis. Endometriosis-related infertility may be due to toxicity
to sperm and embryos by peritoneal fluid in women with
endometriosis, and implantation failure due to abnormal
expression of several genes (that either regulate or inhibit
implantation) in women with endometriosis compared to
normal women. Eutopic endometrium from women with
endometriosis may be histologically normal but biochemi-
cally abnormal[18]. During the window of implantation,
several genes are aberrantly expressed in women with endo-
metriosis compared to controls, including alpha-v, beta-3
integrin, MMPs (MMP2, MMP3, MMP7, and MMP11),
transcription factors such as hepatocyte nuclear factor and
endometrial bleeding factor, enzymes involved in steroid
hormone metabolism (aromatase, 17b-hydroxysteroid
dehydrogenase), leukemia inhibitory factor (LIF), HOX
genes, and progesterone receptor isoforms[17].
In another study, Affymetrix microarrays were used to
compare gene profiling in endometrial biopsies obtained
during the window of implantation (6 to 8 days after LH
search) between women with and without endometrio-
sis[19]. Of the 12,686 genes analyzed, 91 were significantly
increased (more than two-fold) in their expression, and 115
were decreased more than two-fold. The data associate some
genes with disease pathogenesis, such as aromatase and pro-
gesterone receptor, and revealed the selective dysregulation
of other genes involved in embryonic attachment and toxic-
ity, immune function, and apoptotic responses in a reduced
likelihood of implantation. The group also identified two
groups of genes that differ with respect to the implantation
window and the presence of disease.
Ectopic endometrial tissue differs in many ways from
eutopic endometrium, including clonality of origin, enzymic
activity, and histological and morphological characteristics[20].
Endometrial lesions show evidence of differential expres-
sion of steroid hormone receptors, adhesion molecules, and
their receptors; lower apoptosis; and loss of heterozygosity.
Few groups have used microarrays to compare the transcrip-
tional profile of ectopic and eutopic endometrium. The gene
expression analysis of 4,133 genes in ovarian endometriomas
7 1 8   •   G enomic s in C linical P ractice
versus eutopic endometrium revealed upregulation of β-actin,
α2-actin, vimentin, and several immune genes in ovarian
endometriomas compared to eutopic endometrium[21]. Using
a similar approach, another study reported 15 upregulated
and 337 downregulated genes in ovarian endometriomas
compared to eutopic endometrium during both prolifera-
tive and secretory phases of the menstrual cycle[22]. However,
whether genes upregulated in ovarian endometriomas are the
same as for pelvic peritoneal endometriosis (the most com-
mon type of endometriosis) remains to be determined.
Two major problems in sampling ectopic endometriotic
tissue from the pelvic peritoneum for microarray analysis
are, first, that the tissue samples may be surrounded with
contaminating “normal” tissue from the surrounding ovary
or peritoneal tissue; and second, that the amounts of tissue
collected will often be small. In addition, since electrodia-
thermy is frequently used in laparoscopic “keyhole” surgery
for endometriosis, the RNA may be damaged by the exci-
sional surgery. A  solution to some of these problems may
be the use of laser-capture microdissection to obtain a pure
population of endometriosis cells, as has been used in one
study[23]. It is essential to have a good relationship between
scientist and surgeon in order to obtain optimal samples for
RNA extraction (e.g., by avoiding use of diathermy until
after sample has been removed).
Knowledge of gene expression in normal endometrium
and how it is changing during the menstrual cycle is a pre-
requisite to an understanding of gene-expression data in
endometrial disorders. Studying gene expression of normal
endometrium via the high-throughput approach of DNA
microarrays also allows the identification of new signature
genes or biomarkers of normal physiological functions of
endometrium such as implantation and menstruation. It is
likely that some biomarker genes will be useful in molecular
or immunohistochemical diagnostics, or as molecular tar-
gets for drug discovery and therapeutic intervention.
One of the most promising and clinically relevant appli-
cations of global transcript analysis with microarray is the
gene-expression profiling of endometrial disorders to estab-
lish a genome-based diagnosis. That promise has not been
fulfilled yet, although remarkable progress has been made.
The data generated with microarrays certainly contribute to
our understanding of endometrial disorders.
ENDOCRINE AND HORMONAL
FAC TO R S
The practice of clinical endocrinology and reproduc-
tive medicine requires a great deal of understanding and
knowledge of interactive approaches in clinical diagnosis
and management. Premature ovarian failure and polycystic
ovary syndrome are two major categories that are relevant
to this review on the genetic and genomic perspectives of
reproductive medicine.
P R E M AT U R E OVA R I A N FA I LU R E
Premature ovarian failure (POF) or premature menopause
refers to the development of amenorrhea due to the ces-
sation of ovarian function before the age of 40 years. The
diagnosis is based on elevated follicle-stimulating hor-
mone (FSH) levels, in the menopausal range (usually above
40 IU/l) detected on at least two occasions a few weeks
apart[24]. Women with POF suffer from anovulation and
hypoestrogenism and present with primary or secondary
amenorrhea, infertility, sex steroid deficiency, and elevated
gonadotropins[25]. The condition affects approximately
1% of women, occurring in 10–28% of women with pri-
mary amenorrhea and in 4–18% of those with secondary
amenorrhea[26,27]. Early loss of ovarian function has signifi-
cant psychosocial sequelae and major health implications,
with nearly two-fold age-specific increase in mortal-
ity[28,29]. A  wide spectrum of pathogenic mechanisms may
lead to the development of POF, including chromosomal,
genetic, autoimmune, metabolic (galactosemia), infectious
(mumps), and iatrogenic (anticancer treatments) causes.
However, in a large proportion of cases no cause is found,
and they are classified as idiopathic or karyotypically nor-
mal spontaneous ovarian failure. In one series, up to 30% of
cases were ascertained with an autoimmune cause[30].
Genetics of Premature Ovarian Failure
Most cases of POF are sporadic, with a small number of
familial cases indicating the role of genetic factors in its
pathogenesis[31]. Genetic mechanisms include specific X and
autosomal chromosomal abnormalities and few uncommon
single-gene disorders (Table 46.3). There is insufficient evi-
dence for SNPs and genome-wide copy number variations
(CNVs) to be causally linked to POF. The evidence for
genetic and genomic factors’ being implicated in the patho-
genesis of POF is derived from several studies using conven-
tional investigative tools, including transgenic “knockout”
animals, mutation screening of candidate genes in affected
women, analyzing pedigree data in linkage analysis, and
lately, population genetics[32]. Various genetic mechanisms
implicated in the pathogenesis of POF include reduced
haploinsufficiency (gene dosage effect) and non-specific
chromosome effects that impair meiosis. These can lead to
T he G enetic and G enomic P ractice of   R eproductive M edicine   •   7 1 9
ovarian failure by causing decrease in the pool of primor-
dial follicles, increased atresia of the ovarian follicles due to
apoptosis, or failure of follicle maturation.
Familial Premature Ovarian Failure
The large majority of POF cases are sporadic, with a small
proportion having a reliable family history. The overall
incidence of familial cases is around 4%. However, there
are conflicting data from various studies indicating an up
to one-third of familial or inherited POF[32]. The variation
between reported incidences might be explained by differ-
ences in the definition of POF and the idiopathic form,
by differences in population recruitment, and by selection
and recall bias. Pedigree studies on affected families show
autosomal dominant sex-limited transmission or X-linked
inheritance with incomplete penetrance[33]. An adequate
family history can distinguish between familial and spo-
radic POF. The risk of female relatives’ developing POF
may be higher in familial POF than in sporadic cases. Early
diagnosis of familial predisposition permits prediction
of impending menopause, and susceptible women can be
guided to achieve their reproductive goals by timely plan-
ning of pregnancy[34].
X Chromosome Abnormalities
Familial as well as non-familial X chromosome abnor-
malities have been described in women with POF. These
abnormalities may include an aneuploidy (45X, Turner
syndrome, and 47,XXX), structural changes (deletions,
isochromosomes, inversion) and balanced X-autosome
translocations[35]. Complete or nearly complete absence of
one X chromosome, as seen in Turner syndrome, leads to
ovarian dysgenesis characterized by primary amenorrhea,
short stature, and characteristic phenotypical features.
According to the Lyonization theory, the post-zygotic
inactivation of one X chromosome in all cells in the female
offspring allows dosage compensation of X-linked genes
between males and females[36]. However, several X-linked
genes escape inactivation and are vital for the normal
functioning of the X chromosome, specifically for normal
ovarian functioning[37]. In the presence of only one X chro-
mosome, ovarian follicles degenerate by birth in Turner
syndrome. Histological data indicate that oogenesis pro-
ceeds normally in these individuals until diplotene oocytes
begin to be incorporated into nascent follicles. There is a
subsequent block to the production of complete follicles,
manifesting as fetal follicular atresia. In 80% of cases, the
paternally derived X is lost[38]. Cytogenetic data indicate
 Table 46.3  GENETIC ASPECTS OF PREMATURE OVARIAN FAILURE

CATEGORIES CHROMOSOME GENE GENE LOCUS

Mutations identified: X chromosome genes FMR1 Xq27.3
FMR2 Xq28
BMP15 Xp11.2
Autosomal genes FOXL 3q22-q23
FSHR 2p21-p16
LH receptor 2p21
FSH beta variant 11p13
LH beta 19q13.32
Inhibin A 2q33-q36
GALT 9p13
AIRE 21q22.3
EIF2B2, -4, and -5 14q24.3, 2p23.3, 3q27
NOGGIN 17q22
POLG 15q25
Candidate genes: X chromosome genes DIAPH2 Xq22
DFFRX Xp11.4
XPNPEP2 Xq25
ZFX Xp22.3-p21.3
FSHPRH1 Xq22
XIST Xq13.2
Autosomal genes WTI 11p13
ATM 11q22.3
Mutations not identified: X chromosome genes AT2 Xq22-q23
c-kit 4q12
SOX3 Xq26-q27
Autosomal genes MIS 19p13.3-13.2
SOURCE: Adapted from Goswamy et al., 2005[32].

that most Turner syndrome physical features map to Xp,
resulting from reduced dosage of genes on the short arm
of the X chromosome. Further investigations narrowed
down the search for the affected chromosomal segment
to the 2.6Mb Xp–Yp pseudoautosomal region. Statistical
analyses of genotype–phenotype correlations mapped the
Turner syndrome phenotype, including POF, to a critical
region in Xp11.2–p22.1.X., identical to the correspond-
ing Yp region. Eighteen such candidate genes have been
reported[39], and more are likely to exist.
It is commonly believed that X trisomy (47,XXX),
which affects one in 900 women in the general popula-
tion, has no significant effect on fertility; however, associa-
tion with hypergonadotrophic POF has been reported[40].
Further documentation of associated ovarian failure in this
rare sex-chromosome aneuploidy is in the form of occa-
sional case reports[41]. Its relative prevalence among women
with POF is not known[42]. In one reported series, two of
52 (3.8%) patients with POF had the triple X syndrome[43].
POF has also been reported in a girl with 48XXXX[44].
The underlying mechanism could be analogous to that
observed among patients with Klinefelter’s syndrome.
X-chromosome mosaic individuals (45X/46XX and
45X/47XXX) carry mixed germlines and manifest pheno-
typical abnormalities and POF similar to monosomy X, but
12% are reported to menstruate[45].
Structural X Chromosome Abnormalities
X chromosome deletions associated with POF are more
common than translocations. Deleted X chromosomes
necessarily leave a portion of the normal X unpaired,
and isodicentrics probably interfere with pairing, result-
ing in atresia of the oocytes. While deletions commonly

7 2 0   •   G enomic s in C linical P ractice

involve the short arm of the X chromosome (Xp), the The Fragile X Syndrome A and FMR1
fraction of deletions that show POF is far higher in the
The fragile X syndrome type A (FRAXA) is one of the com-
Xq13–25 region[46]. Deletions at Xp11 result in 50%
mon inherited mental retardation conditions, occurring
primary amenorrhea and 50% secondary amenorrhea.
predominantly in the male. The inheritance pattern does
Deletions at Xq13 usually produce primary amenorrhea.
not follow the classic Mendelian pattern, and the condition
A  phenotypical difference has also been noted between
is often described as an X-linked semi-dominant condition.
distal deletions, associated with preserved ovarian func-
The critical gene is called FMR1, located on Xq27.3, out-
tion, and proximal deletions, associated with ovarian fail-
side the Xq POF-critical region. Pathological expansion
ure. However, the relationship is imperfect, since large
of the trinucleotide repeats located at its 50 UTR region
deletions that remove the whole critical region for POF,
result in abnormal expression of FMR1. Four types of
in Xq21, were not found to be associated with ovarian
alleles are identified, based on the number of repeats: nor-
failure[47].
mal (6–40), gray-zone (41–60), premutated (61–200),
In contrast to generally neutral effects of balanced
and fully mutated (>200). The full mutation is associated
autosomal translocations, balanced X/autosomal trans-
with the phenotype of FRAXA. The FMR1 gene product is
locations very often lead to POF, with more than 100
expressed in oocytes and encodes an RNA-binding protein
cases of post-pubertal women with X/autosomal bal-
involved in translation. Immunohistochemical studies in a
anced translocations reported[48]; the deleterious effect
“full mutated” female fetus, using the fragile X mental retar-
on ovarian functioning results from X breakpoints that
dation protein (FMRP) antibody, demonstrated FMRP
fall on the long arm between Xq13 and Xq26, indicat-
expression in all germ cells surrounded by FMRP-negative
ing a possible “critical region” for normal ovarian func-
paragranulosa and interstitial cells[55]. The Müllerian epithe-
tion between Xq13–q26. Within this region, the most
lium of the fetal fallopian tube was continuously positive in
frequent breakpoints involve two specific regions defined
the control case, whereas the full mutation carrier showed a
as POF loci—POF1 Xq26–qter and POF2 Xq13.3–
discontinuous patchy pattern[55]. In a study of murine Fmr1
Xq21.1[49]. These are separated by a short region in Xq22.
expression pattern, enhanced FMRP levels were seen in the
It has been suggested that chromosome dynamics in the
fetal ovary, while in the mature ovary, no specific FMR1
region could be sensitive to structural changes and that
expression signal was found. As proliferation of oogonia
the resulting unpaired chromosome provokes a pachytene
takes place in fetal ovary, it was suggested that FMRP serves
checkpoint during meiosis, leading to oocyte apopto-
a special function during germ cell proliferation[56].
sis[50]. Distal deletions involving the POF1 locus are asso-
POF was first noted as an unexpected phenotype
ciated with POF at the ages of 24–39  years[51]. Various
among heterozygous carriers of the fragile X premutation in
molecular techniques and bioinformatics have been used
the early 1990s. Subsequent studies showed that FMR1 tri-
to map the POF1 locus and identify the putative genes
nucleotide expansions in the premutation range of 50–200
for POF[52]. The translocations involving the POF2 locus
repeat units, but not full mutations, are associated with
cause POF at an earlier age of 16–21 years[49]. An analysis
POF[57]. The underlying mechanism for this association is
of the X autosome translocations in 11 women with POF
unclear. Presently, there exists firm evidence for a significant
to POF2 locus involving a 15Mb YAC (Yeast artificial
association between fragile X premutation carrier status
chromosome) contig mapped the critical Xq21 region
and premature menopause as shown by both the analysis of
between loci DXS233 and DXS1171[53]. A  single gene
women carrying the premutated allele and the screening of
is unlikely to be responsible for ovary development and/
women affected by POF[58]. The results of an international
or oogenesis; rather, several genes may be present along
collaborative study examining premature menopause in 760
the critical region, and they may be interrupted by the
women from fragile X families showed that 16% of the 395
balanced translocations. However, it must be noted that
premutation carriers had experienced menopause prior to
many breakpoints on the X chromosome are not associ-
the age of 40, compared with none of the 238 full mutation
ated with POF[48].
carriers, and one (0.4%) of the 237 controls[59].
Molecular investigations in women with POF and
The incidence of FRAXA premutations has been shown
experiments on transgenic animal models have led to the
to vary among women with POF, depending on the pro-
identification of a number of candidate genes for POF;
portion of sporadic and familial cases. Thirteen percent of
mutations in these genes have been identified in 10% of
pedigrees with the familial POF and 3% of women with the
POF cases; however, functions of many of these genes are
sporadic form of POF have been found to carry FRAXA
not known[54].

T he G enetic and G enomic P ractice of   R eproductive M edicine   •   7 2 1

premutations, compared with an expected prevalence of
1:590. It is likely that carriers who received the premutation
from their fathers were at a higher risk of POF (28%) than
those who received the premutation from their mothers
(4%), suggesting that POF may be limited to permutations
that are paternally inherited[60]; however, this is not fully
supported in subsequent studies[61]. Nevertheless, from the
practical point of view, FRAXA premutations are certainly
worth seeking in those with familial POF in order to enable
genetic counselling and, hopefully, limit the transmission of
fragile X syndrome to future generations. Some units might
also consider screening for FRAXA premutations in spo-
radic cases.
Fragile Site, Folic Acid Type, Rare (FRAXE), and
Fragile Site Mental Retardation 2 Gene (FMR2)
In patients who have the cytogenetic changes of fragile X
syndrome but who are FMR1-mutation negative, a second
site of fragility (FRAXE) is seen[62], lying approximately
150–600 kb distal to the FRAXA site at Xq28, and simi-
larly folate-sensitive. An excess of small alleles with fewer
than 11 repeats at the FRAXE locus were found in women
with POF[63]. It is believed that microdeletions within
FMR2 may be a significant cause of POF, being found in
1.5% of women with the condition, and in only 0.04% of
the general female population[63].
Bone Morphogenetic Protein
Mutations in the bone morphogenetic protein 15 gene
(BMP15) are associated with POF. Bone morphoge-
netic proteins (BMPs) are extracellular signaling proteins
belonging to the transforming growth factor-b superfamily,
which also includes growth/differentiation factors (GDFs).
BMP15 is an oocyte-specific GDF that stimulates folliculo-
genesis and granulosa cell growth and is expressed in oocytes
during early folliculogenesis. It shares a coincident expres-
sion pattern with the closely related mouse Gdf-9 gene,
which is essential for normal folliculogenesis in mice[64]. The
BMP15 gene maps to Xp11.2 within the Xp POF-critical
region[65]. Two sisters with POF presenting with primary
amenorrhea were reported to have heterozygous BMP15
mutation (A to G transition at base pair 704) that resulted
in a tyr235-to-cys (Y235C) amino acid substitution[66].
The father was a hemizygous carrier, whereas the mother
had a wild-type BMP15 coding sequence. This condition
represents an unusual example of X-linked human disease
exclusively affecting heterozygous females who inherited
the genetic alteration from the unaffected father.
7 2 2   •   G enomic s in C linical P ractice
Autosomal Abnormalities in Premature
Ovarian Failure
Autosomal translocations are uncommon in women with
POF. Most reports of translocations document X/auto-
some balanced translocations, with no common autosomal
breakpoint. Reciprocal translocation between chromo-
somes 2 and 15 (46; XX,t[2;15] [q32.3;q13.3]) was recorded
in a woman with POF[67]. In addition, it is well known that
women with trisomy 21 (Down’s syndrome) are twice as
likely to develop POF as are their corresponding age-related,
chromosomally normal cohort, implying possible etiologi-
cal associations of the autosomal genes on chromosome 21.
Several autosomal loci and genes are now known to
exist that are causally associated with POF (Table 46.4).
Notable examples include the blepharophimosis epicanthus
inversus syndrome (BPES, Online Mendelian Inheritance
in Man database [OMIM] 110100)  and ectodermal dys-
trophy associated with polyendocrinopathy and candidiasis
(APECED; OMIM 240300). BPES, an autosomal domi-
nant genetic condition, is characterized by complex eye-
lid malformation and inverted epicanthus. Two forms are
described—type I with POF in affected females, and type II
not associated with POF[68]. POF-related infertility is inher-
ited as an autosomal dominant sex-limited trait. The BPES-I
associated with POF maps to 3q22–q23, as does type II[69];
the putative gene is FOXL2 (winged helix/forkhead tran-
scription factor gene), and is mutated in both BPES types
I and II[70]. The FOXL2 protein appears predominantly in
the ovary in adult humans, and its corresponding gene is the
first human autosomal gene in which dominant mutations
have been implicated in ovarian maintenance and differen-
tiation. More than 100 intragenic mutations and variants
of FOXL2 are listed[71], including uncommonly pathogenic
association with non-syndromic POF[54]. However, other
studies involving phenotypically normal women affected by
POF did not reveal any mutations in the FOXL2 coding
region[72].
Other major autosomal genes pathologically linked
with POF include FSH receptor (FSHR; 2p21–p16)[73–75];
the LH receptor gene (LHR; 2p21)[76,77]; the FSH
beta-subunit variant[78,79]; the LH beta-subunit variant[80,81];
Inhibins (encodes a glycoprotein)[82,83,84]; GALT (9p13;
gene for autosomal recessive galactosemia); AIRE (gene
for APECED syndrome)[85,86]; eukaryotic initiation factor
2B (EIF2B)[87–90]; Noggin mutation (17q22); the gene for
autosomal dominant proximal symphalangism, SYM1[91–93];
mitochondrial DNA polymerase g mutations (POLG, the
autosomal gene for progressive external ophthalmople-
gia)[94,94]; and some other genes for syndromal associations,
 Table 46.4  CANDIDATE GENES ASSOCIATED WITH PREMATURE OVARIAN FAILURE

GENE MAP PROTEIN PRODUCT REFERENCE(S)

Diaphanous 2 Drosophila homologue (DIAPH2) DIA [98]

Drosophila fat facets related X-linked gene (DFFRX/USP9X) Xp11.4 [99]

X-propyl aminopeptidase 2 (XPNPEP2) Xp25 [100]

X zinc finger gene (ZFX) Xp22.1-22.2 [101]

FSH primary response homologue 1 (FSHPRH1)

Leucine-rich primary response gene-1 (LRPR1), Xq22 [102]

X inactivation-specific transcript (XIST) Xq13.2 [103]

Wilms tumor 1 gene (WT1 /inhibin-alpha gene) 11p13 [104]

Ataxia telangiectasia- ATM 11q22.3 [105]

Angiotensin II type 2 (AT2) receptor genes Xq22–q23) [106]

c-Kit gene (KIT) 4q12. [107]

SRY related HMG-box (SOX) 3 gene Xq26–q27 [108]

Müllerian-inhibiting substance (MIS) gene 19p13.3–p13.2 [109]

including bilateral corneal anesthesia associated with multi-
ple systemic abnormalities[95, 96] and progeria or accelerated
aging[97]. Other candidate genes linked to POF are listed in
Table 46.4.
P O LYC YS T I C OVA R I A N SY N D RO M E
(PCOS)
Polycystic ovarian syndrome (PCOS) is the most com-
mon endocrine condition afflicting women, occurring in
5% of the population[98]. It is characterized by enlarged
ovaries with multiple small follicles, anovulation, and an
androgen-secreting stroma, presumably from the theca cells
of the follicles. Recent genomic studies have improved our
understanding of the molecular mechanisms underlying
PCOS[99,100]. Analysis of the isolated and cultured theca
cells showed an increase in gene transcription for alde-
hyde dehydrogenase 6 and retinol dehydrogenase 2[101].
The gene-expression profiling between normal and PCOS
whole ovarian tissue involved Wnt signaling, extracellular
matrix components, and immunological factors[102]. Using
cultured smooth muscle cells from fibroids and myome-
trium, the autocrine and paracrine action of TGF-beta on
gene expression has been analyzed recently by microar-
ray[103,104]. Fibroid and myometrial smooth muscle cells
were found to respond differently to TGF-beta treatment,
possibly explaining the altered, excessive, extracellular
matrix produced in fibroids compared with myometrium
(Figure 46.2). Estrogen responses in cultured fibroid
smooth-muscle cells have been studied using microarrays,
supporting a role for IGF-1 in mediating estrogen-induced
fibroid growth[105]. It is likely that many other genomic
studies on hormone and growth factor responses in fibroid
smooth muscle cells will be performed in the near future.
If genetic pathways can be elucidated that differ between
fibroids and myometrium, it may be possible to create medi-
cal treatments that selectively inhibit fibroid growth with-
out interfering in myometrial function. This is particularly
important in premenopausal women, whose families have
not yet been completed.
M A L E FAC TO R I N F E RT I L I T Y
Male factor infertility is implicated in about half of all
infertile couples. This includes wide-ranging potential
causes that require thorough evaluation for accurate detec-
tion and management. Couples with a component of male
factor infertility need a systematic evaluation directed
at the apparently normal-looking healthy male partner
to maximize their reproductive potential. A  systematic
approach with a complete medical history in conjunction
with targeted investigations is necessary in all cases of sus-
pected male factor infertility. The physical examination
should be focused on signs of an associated endocrine dis-
order; the presence of gynecomastia; examination of the
penis, scrotum, and testis; and per rectal examination for
prostatic enlargement. The semen analysis is mandatory
in all cases, but it is by no means sufficient to determine
the cause or decide on management plan therapy. Limited
imaging investigations may also assist in evaluating male
factor infertility.
Genetic factors implicated in male factor infertil-
ity include chromosomal abnormalities (47,XXY and

T he G enetic and G enomic P ractice of   R eproductive M edicine   •   7 2 3

 CYR61 mRNA CYR61 mRNA
*
5

6
4

mRNA relative units

5

mRNA relative units

4 3

3 2
2
1
1
0 0
Fibroid Myometrium Fibroid Myometrium

CTGF mRNA CTGF mRNA

3
†
2.5
mRNA relative units

mRNA relative units

2 2
1.5

1 1
0.5

0 0
Fibroid Myometrium
Fibroid Myometrium

COL4A2 mRNA COL4A2 mRNA

*
0.5

0.5 0.4
mRNA relative units

mRNA relative units

0.4
0.3
0.3
0.2
0.2

0.1 0.1

0 0.0
Fibroid Myometrium Fibroid Myometrium

Levels of mRNA expression in 18 paired fibroid and myometrical specimens as determined by quantitative RT-PCR for the
three genes CYR 61, CTGF, and COL4A2. In all three, the microarray result was confimed. mRNA levels in arbitary units
shown on the Y-axis of the graphs, with all results corrected against expression of the housekeeping gene β actin.(† p<0.05;
* p<0.01). Graphs on the left are line graphs linking paired specimens, while on the right are box and whiskers plots.

Figure 46.2  Gene-expression profiling in normal and fibroid myometrium showing skewed variation in selected genes. Source: Adapted from Swartz et al., with
permission[115].

structural Y chromosome abnormalities), single-gene
disorders interfering in the sperm’s motility or propaga-
tion (cystic fibrosis), spermatogenesis (mutations in SRY
and 5-α reductase genes) and mutations in the DAZ gene
cluster[106]. The prognosis for any given couple depends, in
large part, on the cause of the infertility. Without a firm
understanding of the genetics, anatomy, and physiology,
and their interactions necessary to permit full functioning
of the male reproductive system, the evaluation becomes an
inefficient exercise that often fails to elucidate the precise
cause of infertility. Treatment success relies not just on the
clinical diagnosis, but on determining the cause of male fac-
tor infertility as well.
C O N G E N I TA L A N O M A L I E S O F T H E
M A L E R E P RO D U C T I VE T R AC T
Exclusion of a congenital anatomical anomaly of the
male reproductive tract is the essential first step in man-
aging male factor infertility. Most congenital anoma-
lies of the male reproductive tract are sporadic, without
any appreciable recurrent genetic basis (Table 46.5).

7 2 4   •   G enomic s in C linical P ractice

Table 46.5  CHROMOSOMAL ABNORMALITIES ASSOCIATED WITH MALE FACTOR INFERTILITY

PREVALENCE AND PHENOTYPES OF COMMON CHROMOSOMAL ABNORMALITIES ASSOCIATED WITH MALE INFERTILITY

GENETIC ABNORMALITY PHENOTYPE PREVALENCE, %

Chromosomal abnormalities
Klinefelter syndrome
Robertsonian translocation
Y chromosome microdeletions
AZFa deletion
AZFb deletion
AZFc deletion
Partial AZFc deletions
Azoospermia to normozoospermia 5 (total infertile population); 15 (azoospermic)
Azoospermia to severe oligozoospermia 5 (severe oligozoospermia); 10 (azoospermic)
Azoospermia to normozoospermia 0.8 (total infertile population); 1.6 (oligozoospermic);
0.09 (azoospermic)
Azoospermia to oligozoospermia 10–15 (azoospermic); 5–10 (oligozoospermic)
Azoospermia, Sertoli cell–only syndrome 1.5–1.0 (2)
Azoospermia, spermatogenic arrest 1.5–1.0 (2)
Severe oligozoospermia to non-obstructive 6–12
azoospermia
From azoospermia to normozoospermia 3–5 (2)

NOTE: Prevalence listed refers to listed phenotype, unless noted otherwise.

SOURCE: Adapted with permission from O’Flynn et al., 2010[116].

However, genetic factors play a significant role in the cau- Klinefelter’s Syndrome and Partial X Disomy
sation of congenital anomalies; for example, cystic fibrosis
The mandatory chromosome analysis in the male often
transmembrane-conductance-regulator (CFTR)–related
reveals an additional X chromosome (47,XXY) consistent
congenital bilateral absence of the vas deferens (CBAVD).
with the phenotype of Klinefelter’s syndrome. Amongst
Genetic factors underlying a congenital anomaly of the
most infertile couple, the physical features (for example
male reproductive tract include both chromosomal and
subtle male secondary sexual characteristics and soft tes-
Mendelian conditions (Tables 46.5 and 46.6). There are
tes) were known to the male partner but often overlooked.
limited clinically relevant genomic data or information
A  basic clinical examination would confirm the clini-
available on these. It is nevertheless likely that low-risk
cal diagnosis, if supported by X chromosome aneuploidy.
alleles, gene polymorphisms, and genome-wide CNVs
Azoospermia or oligospermia is present in almost all cases.
could have some causal and/or functional roles in male fac-
It is possible to aspirate immature sperms or spermatogonia
tor infertility. Such studies are important, but they might
by direct testicular aspiration or to obtain them at the testic-
be difficult to conduct and interpret due to the pheno-
ular biopsy. Prospects for natural conception are extremely
typical heterogeneity of congenital anomalies of the male
limited; however, it has been reported in rare cases. Most
reproductive tract.
couples would opt for assisted reproduction (in vitro

Table 46.6  THE Y CHROMOSOME GENES INVOLVED IN MALE FACTOR INFERTILITY

GENES ON Y CHROMOSOME WITH SUSPECTED INVOLVEMENT IN MALE FACTOR INFERTILITY

GENE LOCATION REASONS FOR INVESTIGATION
USP9Y AZFa Involved in efficiency of spermatogenesis; deletion or shortening may cause azoospermia, oligozoospermia,
or oligoasthenozoospermia
DBY AZFa Involved in premeiotic germ cell development
RBMY AZFb RNA-binding protein/testis-specific splicing factor; reduced expression in azoospermic men
PRY AZFb Regulation of apoptosis
DAZ AZFc Regulation of translation, meiosis, and germ cell population; codes for RNA-binding proteins; reduced
expression in azoospermic men; partial deletions related to oligozoospermia
CDY Yq Involved in histone replacement
TSPY Yq Regulates timing of spermatogenesis; greater copy number in infertile patients
SOURCE: Adapted with permission from O’Flynn et al., 2010[116].

T he G enetic and G enomic P ractice of   R eproductive M edicine   •   7 2 5

fertilization; IVF) using intra-cytoplasmic sperm inocula-
tion (ICSI). Successful pregnancies are now recorded using
this approach. Genetic counselling is extremely important
as part of the preparation for IVF/ICSI. It is recommended
that pre-implantation genetic diagnosis (PGD) be per-
formed before assisted reproduction therapy (ART) (ICSI/
IVF) to ensure that the offspring is not aneuploid.
Molecular Genetics of Male Factor Infertility
Mutations in several Y-linked genes and interstitial dele-
tions of the Y chromosome may cause impaired spermato-
genesis and anomalous development of male gonads[107].
Y chromosome microdeletions, predominantly of the long
arm, are a frequent cause of infertility in males[108]. A partic-
ular area of interest on Yq is the azoospermia factor region
(AZF region), which contains genes involved in the growth
and development of sperm (AZFa, AZFb, and AZFc), more
prevalent in men who are azoospermic and severely oligo-
zoospermic[109–110]. It is essential to look for the three AZF
deletions when planning ART, because microdeletions are
always passed on to the male offspring, and fertilization and
pregnancy rates are not affected by microdeletions on the
AZFc region when using ICSI[111]. Multiple gene deletions
in the AZFb and AZFc areas can produce a wide range of
infertile phenotypes[112]. Researchers are attempting to
Image of Y chromosome displaying AZF regions and associated genes. Enlarged portion of AZFc region highlights discussed
microdeletions (A) Normal AZFc region; (B) gr/gr deletion; (C) b1/b3 deletion; (D) g1/g3 deletion; (E) gr/gr duplication.
Yp
DBY (DDX3Y)
USP9Y
TPSY
CDY
AZFa
Figure 46.3.  chromosome abnormality/deleted in azoospermia (DAZ) spectrum. Source: Adapted with permission from O’Flynn et al., 2010
7 2 6   •   G enomic s in C linical P ractice
characterize deletions in the AZF region (Figure 46.3)
so they can be used to determine treatment for infertile
males[113].
Deletions in the AZFa region containing USP9Y and
DBY cause Sertoli cell–only syndrome, a condition char-
acterized by the presence of complete Sertoli cells in the
testes but a lack of spermatozoa in the ejaculate[114,115,116].
Shortening or deletion of the USP9Y gene causes azo-
ospermia, oligozoospermia, or oligoasthenozoospermia[117].
Deletions of the AZFb region containing the RBMY and
PRY genes may cause spermatogenesis-arrest at the primary
spermatocyte stage[118–121]. Deletions in the AZFc region
produce a wide range of phenotypes, many of which are
associated with low sperm concentration due to reduced
spermatogenesis[114]. AZFc deletions cause approximately
12% of non-obstructive azoospermia and 6% of severe oli-
gozoospermia[119]. In many cases, men can still achieve fer-
tilization with the assistance of ART. Studies demonstrate
that only the AZFa and AZFb regions are needed to initiate
spermatogenesis, but that, without the AZFc region, sper-
matogenesis will not be completely normal[122]. Complete
deletions of the AZFc region may coexist with partial
Y-chromosome deletions[123]. A deletion of the AZFc region
may lead to sexual reversal. Several studies have found this
deletion to be a premutation for 45,X0 and for the mosaic
phenotype 45,X/46,XY[124].
Yq
RBMY
DAZ
PRY
AZFb AZFc
b1 b2 g1 r1r2 b3 g2 r3r4 g3 b4
A
B gr/gr deletion
C b1/b3 deletion
D g1/g3 deletion
gr/gr duplication
E
.
[116]
 The AZFc region is prone to many smaller subdele-
tions caused by intrachromosomal recombinations[125],
probably due to genomic variation evolved over successive
in response to environmental pressures[126]. The complex
interaction of genes and the environment makes it difficult
to replicate the results of studies and definitively associate
subdeletions with infertile phenotypes[125]. The three most
frequent subdeletions on the AZFc region of the Y chro-
mosome are gr/gr, b1/b3, and g1/g3 (b2/b3)[116]. The gr/gr
subdeletion, which removes half of the content in the
AZFc region, is probably more important in determining
a man’s fertility status[127]. The AZFc region also contains
genes involved in spermatogenesis. The DAZ gene has four
copies on the Y chromosome[128]. DAZ genes are thought
to serve a variety of roles throughout the spermatogenic
process because they are expressed in all stages of germ cell
development[107]. They regulate translation, code for germ
cell–specific RNA-binding proteins, and are involved in the
control of meiosis and the maintenance of the primordial
germ cell population[129]. Deletions of the DAZ genes can
cause a spectrum of phenotypes ranging from oligozoosper-
mia to azoospermia[130]. Additionally, DAZ gene expression
was reduced in azoospermic patients, and partial deletions
of DAZ genes seem to be related to oligozoospermia[131].
It is critical that azoospermic and severely oligozoosper-
mic men be tested for microdeletions, both for accurate
diagnosis and for genetic counseling, before performing
ART[132]. Initial screen for Y chromosome microdeletions
is crucial before employing more expensive and techni-
cally challenging testing methods[133]. Other Y-linked genes
involved spermatogenesis include CDY (Yq), the chromo-
domain protein Y-linked gene and TSPY (Yp)[134,135].
Several autosomal and X-linked Mendelian disorders
are known to complicate male factor infertility. It is impor-
tant to investigate men with the clinical suspicion of these
disorders (Table 46.7). Diagnostic molecular genetic meth-
ods are available for confirmation of the diagnosis. A refer-
ral to the clinical genetic service for investigations and
genetic counselling is important and should be included in
the management plan.
MENDELIAN DISORDER S AND MALE
FAC TO R I N F E RT I L I T Y
E P I G E N ET I C S A N D E P I G E N O M I C S I N M A L E
FAC TO R I N FE RT I L IT Y
Essentially, “epigenetic phenomenon” refers to an alteration
in gene functioning without any DNA sequence changes in
T he G enetic and G enomic P ractice of   R eproductive M edicine   •   7 2 7
the coding regions of the gene. Epigenetic and epigenomic
factors are now increasingly discussed and implicated in a
number of biological processes (see also Chapter 4). Since
spermatogenesis involves complex series of events, it is
vulnerable to an accumulation of genetic and epigenetic
errors that can severely affect the spermatogenic process.
There is evidence to suggest that epigenetics and epigenom-
ics are among the several genetic factors in male factor
infertility[136].
The epigenetic control of gene function and expression
is regulated by several complex processes (see Chapter 4).
Several epigenetic factors that are involved in spermato-
genesis could adversely affect sperm development, fertiliza-
tion, and the development of the early embryo. Similarly,
two sets of protamine genes, P1 and P2, are important
for histone function and chromatin structure. Mutations
in both P1 and P2 are shown to be associated with sperm
DNA damage and abnormally packaged chromatin struc-
ture. Mutations and/or haploinsufficiency of either the P1
or the P2 gene are related to infertility[137,  138]. Abnormal
protamine ratios may also be associated with problems in
epigenetic reprogramming and gene imprinting, and have
been correlated with IVF success and embryo quality[139].
A SNP, G197 T, for P1 may predispose men to DNA frag-
mentation and teratozoospermia or to a high prevalence of
morphologically abnormal sperm in the ejaculate.
Histone markers signify DNA-imprinting control
regions during the formation of spermatozoa[140]. The tran-
scriptional control of gene expression is regulated by the
addition of acetyl, methyl, ubiquitin, and phosphate groups
to histones. Imprinting, the methylation of DNA, deter-
mines which genes from the parental and maternal genomes
are expressed in the embryo and is critical for normal devel-
opment (see Chapter 5). The imprinted regions of DNA are
reset every reproductive cycle, which allows novel parental
imprints to be established on the germ cells. Imprinting is
achieved by the differential marking of DNA regions with
histone modifications, methylation, or possibly both, to
allow only one copy of a gene to remain active.
In fertile men with a normal ejaculate, paternal differen-
tially methylated regions (DMRs) of the DNA should be
methylated, and the maternal DMRs should be unmethyl-
ated[141]. Approximately 14% of infertile patients had abnor-
malities in the DMRs of the paternal imprint, and 21% of
the infertile patients had abnormalities in the maternal
imprint. Most patients with abnormalities in both imprint
regions were oligozoospermic. Additionally, men with
abnormally imprinted DMRs had low success rates with
ART. It was also discovered that oligozoospermic men may
have a higher risk of transmitting imprinting errors to their
Table 46.7  MENDELIAN GENES AND POLYMORPHISMS ASSOCIATED WITH MALE FACTOR INFERTILITY

Autosomal
GENE LOCATION DISORDER PATHOLOGY OF MALE INFERTILITY
CFTR 7q Cystic fibrosis Congenital bilateral absent vas deferens (CBAVD)
SHBG 17 - Sex hormone–binding globulin; mutations result in
low level of spermatogenesis and androgen response
ESR1 6 - Estrogen receptor gene 1—abnormal spermatogenesis
ESR2 14 - Estrogen receptor gene 1—abnormal spermatogenesis
FSHR 2 - Follicle stimulating hormone receptor—inefficient spermatogenesis
DAZL 3 - Autosomal homologue of DAZ—azoospermia
MTHFR 1 - Methylene tetrahydrofolate reductase—abnormal DNA
methylation and inefficient spermatogenesis
INSL3 19 Testicular Insulin-like 3—associated with cryptorchidism;
dysgenesis
LGR8 13 Cryptorchidism Receptor for INSL3—associated with testicular
dysgenesis syndrome
Sex Chromosomal
AR Xq Kennedy disease Androgen receptor gene—regulates conversion of
spermatocytes to round spermatids
USP26 Xq Expressed in testes throughout early spermatogenesis
Involved in histone removal and protein breakdown
Reformation during spermatogenesis
TAF7L Xq Related to autosomal TAF7 gene; regulates spermatogenesis
KS1 Xp Kalman syndrome Idiopathic hypogonadotropic
hypogonadism (IHH) combined with anosmia or hyposmia; male infer-
tility is encountered in some cases; autosomal form linked with FGFR1

children[142]. ART could have negative consequences on the
imprinting of sperm because it may use sperm that are not
yet fully mature, and, consequently, whose epigenetic code
is not established. If the sperm are too immature or abnor-
mal, it is more likely that the offspring could be born with
an imprinting disorder.
The correlation between the incidence of imprinting
disorders and ART in men with abnormal sperm is a con-
troversial topic. A  study found that spermatogonia from
infertile men did not have increased imprinting errors in
comparison with that of normal men[143]. Several researchers
have asserted that ART, such as ICSI, causes imprinting dis-
orders like Angelman syndrome and Beckwith-Wiedemann
syndrome[139]. Angelman syndrome is a rare neurologi-
cal disorder characterized by cognitive defects, seizures,
uncontrolled limb and body movements, spontaneous
laughter, and difficulties with speech development[144]. Two
independent groups reported an increased incidence of
Angelman syndrome in offspring from ICSI procedures[145].
Beckwith-Wiedemann syndrome, also an uncommon
disorder, is characterized by large fetal and organ size, hypo-
glycemia, midline abdominal defects, facial moles, and
enlarged tongues. Children with Beckwith-Wiedemann
syndrome are also at risk for developing tumors. One study
reports the incidence of Beckwith-Wiedemann syndrome
was almost 5% in children conceived by ART in com-
parison with an incidence of less than 1% in the general
population[145].
M I TO C H O N D R I A L G E N E S I N M A L E
FAC TO R I N FE RT I L IT Y
Mitochondrial DNA (mtDNA) inheritance may also have
an impact on male factor infertility (see also Chapter  9).
Abnormal mitochondria are known to interfere with
sperm motility because of aberrations in the mitochondrial
sheath. There is concern about passing mutated mtDNA
to offspring using ICSI in ART, because an entire sperm is
injected into the oocyte and the mtDNA is conserved. In
normal fertilization, the sperm mitochondria are lost along

7 2 8   •   G enomic s in C linical P ractice

with the loss of the sperm tail. In addition, it is also known
that oligospermic males have a relatively higher prevalence
of mtDNA mutations, and various studies have demon-
strated a correlation between abnormal mtDNA and dys-
functional sperm. However, data are conflicting as to the
functional impact of mtDNA injected during ICSI. It is
widely believed that the paternal mtDNA is degraded after
fertilization, and therefore, the functional clinical impact of
mitochondrial abnormality in children conceived through
ICSI/ART is not confirmed.
N EW G E N O M I C A P P L I C AT I O NS
I N M A L E FAC TO R I N FE RT I L IT Y
Incorporating techniques such as genomics, proteomics,
and metabolomics into infertility research could assist us
in creating a complete portrait of the genes involved in
infertility, and would enable improvements in ART for
the development of more-targeted solutions. Microarrays
are valuable tools for the identification of gene-expression
profiles of infertile phenotypes[146]. Examining the simulta-
neous expression of genes allows geneticists to determine
molecular signatures related to infertile phenotypes[147].
Microarray technology is also useful in the examination
of spermatogenesis. An analysis of gene expression over
time could be performed to determine the genes that are
involved in each stage of the process. Genomic analysis can
also be used to determine differentially transcribed genes.
An enhanced understanding of transcription regulation
could help geneticists discover how different expression
patterns affect a patient’s fertility[148]. Additionally, micro-
arrays can be used to study the effect of hormones or growth
factors on gene-expression profiles. Some advantages of
using a microarray are that it is a noninvasive test, and it
is very effective in studying germ cells because they express
4% of the genome. Disadvantages of genomics are that gene
expression can vary between two different samples, and
infertile patients might have pockets of gene expression that
are difficult to detect using microarrays[149].
Proteomics enables the determination of protein-
expression profiles of fertile and infertile men. Proteins are
identified with two-dimensional electrophoresis and mass
spectrometry techniques, and the results are used to cre-
ate maps of the proteome[150]. Spermatozoa are ideal for
the study of protein expression because they do not have
active transcription or translation. Further research in
these fields can continue after the sperm proteome is fully
defined and the components of seminal plasma are identi-
fied[151]. Advances have been made in both of these areas,
and many new proteins have been identified as a result.
T he G enetic and G enomic P ractice of   R eproductive M edicine   •   7 2 9
The identification of protein biomarkers for male factor
infertility will allow unbiased comparison between fertile
and infertile males and will clarify the pathophysiology of
the disease[152]. However, most of the causes of astheno-
zoospermia, or abnormally low sperm-motility levels, are
unknown, even though it is a common infertile phenotype.
An advantageous characteristic of genomic and proteomic
technology is that the results can be confirmed through rep-
lication using other techniques such as Western blots, flow
cytometry, and polymerase chain reaction (PCR). Another
important feature of these tests is that the results provide
a definitive characterization of infertile phenotypes. The
use of technologies like genomics and proteomics is a step
toward creating personalized medical diagnoses by deter-
mining individual causes of infertility[153].
Metabolomics is another emerging area of research in
the evaluation of the role of genetic factors in male factor
infertility. Metabolomics involves measuring the expression
of metabolites, small biomarkers that indicate the function-
ality of a cell, and characterizing them for certain diseases
or physiological states[152]. The identification of the human
metabolome will reveal the functional phenotype of the
system being studied, whether it is a single cell or an entire
organism. Mass spectroscopy, nuclear magnetic resonance
spectroscopy, and other chromatography methods can be
used to create profiles of metabolites. Pathway or cluster
analysis is used to determine subsets of metabolites that can
be used to quantitatively characterize patients for diagno-
sis[152]. By identifying differences in the expression of metab-
olites in infertile phenotypes, new methods of diagnosis
and treatment of male factor infertility can be developed
that are inexpensive and noninvasive[154].
G E N ET I C A S P EC TS O F A S S I S T E D
R E P RO D U C T I O N
Fertility is affected by a wide range of conditions dealt with
elsewhere in this chapter, such as endometriosis and poly-
cystic ovarian syndrome. Because fertility depends upon
function of three major components—the mother, the
father, and the embryo—the genomic/microarray studies
in this area of interest have adopted a wide range of experi-
mental approaches to different aspects of the problem.
A key problem limiting success during in vitro fertiliza-
tion (IVF) is a diminished reserve of ovarian follicles, with
increasing maternal age being the greatest factor influenc-
ing success. Unfortunately, there are no reliable predic-
tors of ovarian reserve to inform patients of the chances
of success with treatment—information that could help
individuals make decisions regarding participation in
often-expensive treatment. Mechanisms of oocyte ageing
have been investigated in a mouse model by one group,
by using oligonucleotide microarrays to compare gene
expression in oocytes collected from the ovaries of “young”
mice (5–6 weeks old) to that of oocytes from “old” mice
(42–45 weeks old)[154]. They found numerous genes with
altered expression in ageing oocytes, including genes
involved in genome stability, oxidative stress, and mito-
chondrial function. Luteinized granulosa cells, collected
from human subjects during oocyte retrieval for IVF, have
been screened by gene-expression microarrays for potential
markers of poor ovarian reserve[155]. By comparing patterns
of gene-expression from women with normal and with
reduced ovarian reserves, the investigators found prelimi-
nary evidence of a molecular expression pattern that may
predict poor ovarian reserve.
It has been shown that up to 61% of first-trimester
miscarriages after IVF have chromosomal abnormalities
detectable by array CGH, a laboratory technique allowing
scanning of whole genomes for structural changes, such as
either missing or extra pieces of chromosomes present in a
cell or tissue. Genomic screening of single cells taken from
embryos prior to transfer, with tools such as array CGH, is
a possible way of reducing miscarriage rates[156]. Successful
use of array CGH in women with recurrent implantation
failure is reported[157]. The feasibility of using array CGH
on single cells taken from embryos has been proven in prin-
ciple, with reported detection of trisomy 13 and 15 with
amplified DNA from a single cell[158]. The cumulus cells
surrounding oocytes in culture have been examined for
gene-expression changes between oocytes that successfully
fertilize and those that do not[159]. Altered expression of 160
genes in cumulus cells of oocytes that failed to fertilize is
reported, pointing to possible markers of embryo quality
that could be tested objectively by sampling of cumulus cells
without harm to developing embryos[160].
Some investigators are employing microarrays to study
male factors in infertility. Gene-expression profiles from
ejaculated spermatozoa of normal fertile men have been
characterized[161]. A  “normal” gene-expression profile in
normal fertile men is important as a baseline to compare
with that of infertile or subfertile men. However, this
might not clearly identify or predict specific individual
genes with dysregulated activity in infertile males. In an
attempt to identify genes with a role in spermatogenesis,
gene-expression profiles of human adult and fetal testis
have been compared, as well as spermatozoal mRNA with
adult testis mRNA[161]. Presumably, genes upregulated in
adult testis and spermatozoa in these experiments, respec-
tively, are increasingly likely to have an important role in
7 3 0   •   G enomic s in C linical P ractice
the spermatogenic process. Testicular biopsies have been
investigated by microarrays, and have been shown to be
able to successfully distinguish between normal testicular
tissue, and testes with absent germ cells (Sertoli cell–only
syndrome)[162]. Whether molecular profiling is capable
of detecting subtler testicular and spermatogenic defects
remain to be seen.
P R E I M P L A N TAT I O N G E N ET I C
TESTING AND SCREENING
Developments in molecular cytogenetic and genomic tech-
niques have offered prospects of selecting an embryo prior to
implantation using IVF-based assisted-reproduction meth-
ods. This approach allows successful reproduction outcomes
in managing recurrent reproductive failure and allowing the
couple in selective reproduction when faced with increased
genetic risk in a future offspring. This includes two distinct
approaches—preimplantation genetic screening and preim-
plantation genetic diagnosis. The preimplantation genetic
screening (PGS) involves the selection of embryos before
transfer into the uterus, to increase the success of assisted
reproduction (see next section). Genetic analysis is carried
out on one or two blastomeres that are microsurgically
removed from the embryo on day 3 of the culture. Embryos
with abnormal chromosomal complements are discarded,
and the selected embryos can be transferred on day 5 or fro-
zen for future transfer. FISH is the main technique, involv-
ing the use of fluorescently labeled DNA probes to paint
fetal DNA in interphase nuclei. Recent reports involve the
use of array CGH, but this is not yet available for clinical
use. Indications for PGS include advanced maternal age,
repeated implantation failure, and idiopathic recurrent
pregnancy loss, and in order to improve pregnancy rates in
single-embryo transfers.
In most cases, the results are helpful in making an appro-
priate embryo selection. However, due to the high level of
chromosomal mosaicism in the cleavage stages of embryonic
development, the interpretation of findings might be dif-
ficult, leading to reanalysis. Secondly, because the contem-
porary FISH methods do not capture the full complement
of chromosome material, the extent to which PGS is use-
ful in improving pregnancy rates and outcomes is debated.
Consequently, there is a move towards employing other,
new methods in PGS. The current FISH technology is not
recommended for the indications noted above. Analysis of
polar bodies may yield improved pregnancy outcomes by
detecting maternal genetic abnormalities in eggs, including
meiotic errors that result in aneuploidy. Newer array-based
methods, including 24-chromosome SNP arrays or virtual
karyotyping, will probably replace FISH because they pro-
vide more genetic information[163].

Preimplantation Genetic Diagnosis

Preimplantation genetic diagnosis (PGD) allows the selec-
tion of disease-free embryos for transfer into the uterus.
Genetic analysis is usually carried out as described for pre-
implantation genetic screening. FISH is used to detect sex
chromosomes and specific chromosomal abnormalities, or
PCR, is used to amplify DNA for molecular diagnosis. The
first births after PGD of structural chromosomal abnormal-
ities with the use of comparative genomic hybridization and
microarray analyses were recently reported[176]. Detection of
mitochondrial DNA mutations is also possible, providing
that they are prevalent in the mitochondrial pool[164].
The first and second polar bodies can be analyzed to
determine the presence of maternal genetic contributions
(i.e., X-linked diseases and autosomal dominant diseases),
including carrier states for Duchenne muscular dystrophy,
incontinentia pigmenti, and neurofibromatosis type 2. PGD
has been successfully used in a number of monogenic domi-
nant, recessive, and sex-linked diseases (www.hfea.gov.
uk/preimplantation-genetic-diagnosis.html). With cur-
rent molecular genetic methods and the rapid inclusion
of the next-generation genomic methods, the diagnosis of
Mendelian disorders is highly accurate, with a false-positive
rate of less than 1%[165]. Misdiagnosis has been attributed
to laboratory error, including transfer of the wrong embryo,
contamination by extra-embryonic material, allele dropout
(when one of the alleles is not amplified on PCR), use of the
wrong probes or primer sets, and chromosomal mosaicism.
Preimplantation genetic diagnosis is increasingly
available in the United States and Europe. However, in
most countries, its practice is relatively unregulated. In
the United Kingdom, legal requirements and guide-
lines are provided by the statutory Human Embryology
and Fertilisation Authority (HEFA; www.hfea.gov.
uk/preimplantation-genetic-diagnosis.html). In addi-
tion, professional societies (e.g., the American Society
for Reproductive Medicine and the European Society of
Human Reproduction and Embryology) have issued guide-
lines and recommended the accreditation of laboratories
performing such genetic diagnoses[165].
R EC E N T A DVA N C E S I N P R E NATA L
D I AG N O S I S
Laboratory techniques that are routinely employed in ana-
lyzing the aborted material, fetal tissue or cells, and parental
Figure 46.4  Standard multi-color FISH karyotype of a normal female.
(http://en.wikipedia.org/wiki/Karyotype)
bloods include tissue culturing, conventional karyotyping,
FISH (Figure 46.4), quantitative fluorescence polymerase
chain reaction (QF-PCR; Figure 46.5); and increasingly,
the new technique of array CGH. Newer techniques
are being rapidly introduced with the increasing use of
new-generation genomic methods (exome sequencing,
Sanger sequencing, whole-genome sequencing and deep
sequencing) in clinical practice; however, they are limited
by technical and cost factors[166].
Aneuploidy (Trisomy 21) Pregnancy
Trisomy 21 (Down syndrome) is the most common cause
of genetic mental retardation, occurring in one in 700 live
births. Testing for Down syndrome is offered as a part
of routine antenatal care for women in most developed
nations, and has been increasingly introduced in develop-
ing countries. Unfortunately, non-invasive screening tests of
maternal serum screening markers are insufficient to detect a
Down syndrome pregnancy. The panel of serum tests used in
antenatal screening for Down’s syndrome includes free-beta
human chorionic gonadoptropin (βhCG), unconjugated
estriol (uE3), plasma protein A  (PAPP-A), and dimeric
inhibin A[167]. Recently, ADAM12, a novel serum marker
with biological properties similar to those of PAPP-A, has
been shown to improve detection rates[168]. Several studies
have advocated improved performance of serum screening
when combined with first-trimester ultrasonographic mea-
surement of nuchal translucency[169]. However, it is difficult
to implement this as a population-based screening program
due to the shortage of adequately trained sonographers and
problems with quality assurance. The objective of antenatal

T he G enetic and G enomic P ractice of   R eproductive M edicine   •   7 3 1

A

Figure: 46.5a 
QF-PCR detection of common aneuploidies. QF-PCR electropherogram “trace” showing a result consistent with trisomy for
chromosome 21: Markers D21S1435 and D21S1437 show bi-allelic trisomy, while markers D21S11 and D21S1411 show tri-allelic trisomic results.
Courtesy of Mr. Chris Anderson and Mr. Peter Thompson, Principal Scientists, The Cytogenetic Laboratory, All Wales Genetic Laboratory, University Hospital of Wales, Cardiff, U.K.

Down’s syndrome screening is detecting high-risk pregnan-
cies among younger women (under 35 years of age), who are
then offered one of the two invasive tests for fetal chromo-
some analysis. The invasive tests required (amniocentesis or
chorionic villous sampling of amniotic liquor) entail a mis-
carriage risk of 0.5–1.0%.
Traditionally, fetal chromosome analysis involves cell
culture and full karyotyping. This is slow and expensive, but
assures a comprehensive analysis, minimizing the risk of fail-
ure to detect any other chromosomal abnormality. However,
the rapid introduction of molecular cytogenetic methodol-
ogy led to the increased use of commercially available fluo-
rescent hybridization kits for selective cytogenetic analysis
for chromosomal aneuploidies of 13, 18, and 21 chromo-
somes. The technique has improved to employ QF-PCR.
It is claimed that a selective, but fast and less expensive,
cytogenetic analysis using QF-PCR is more cost-effective
when used in a carefully selected cohort of 17,500 women,
using a combined screening strategy of serum testing with
increased nuchal translucency (>4  mm)[169]. Microarray
studies of trisomy 21 (T21) pregnancies include an attempt
to identify new maternal serum biochemical markers for
Down syndrome pregnancy by cDNA microarrays. Other
approaches for faster detection of chromosomal aneuploi-
dies include the use of cell-free fetal DNA from the amni-
otic fluid to form fluorescent probes for hybridization to
CGH arrays[170].
Cell-Free Fetal DNA
Most prenatal genetic diagnosis (PND) is based on fetal
tissue obtained by invasive methods involving chorionic
villus biopsy (>11 weeks), amniocentesis (>15 weeks),
and fetoscopy (>18 weeks). All invasive PND methods
involve additional risk of miscarriage (1–3%). However,
with high-resolution ultrasound imaging, precise localiza-
tion has considerably reduced fetal risks. Efforts to develop
non-invasive PND (NIPND) led to the consideration of
testing circulating fetal cells in the maternal circulation. It
has long been recognized that nucleated fetal cells reach the

7 3 2   •   G enomic s in C linical P ractice

B

Figure 46.5b 
QF-PCE showing trisomy 18 (Edwards syndrome). QF-PCR electropherogram “trace” showing a result consistent with trisomy for
chromosome 18, Edwards syndrome: Markers D18S535, D18S390, and D19D391 show bi-allelic trisomy; marker D18S386 shows a tri-allelic
trisomic result, and D18S978 is uninformative. Courtesy of Mr. Chris Anderson and Mr. Peter Thompson, Principal Scientists, The Cytogenetic Laboratory, All Wales Genetic Laboratory, University
Hospital of Wales, Cardiff, U.K.

maternal circulatory system, but attempts to isolate these
rare cells from maternal blood (which typically number
1–6 cells per milliliter of maternal blood) and use them
for genetic testing have been disappointing because of low
sensitivity. Cell-free fetal DNA is currently the material
of choice for NIPND. It represents 3–6% of circulating
cell-free DNA in maternal plasma, and it can be detected in
the first trimester of pregnancy, increasing in abundance as
the placenta grows. Cell-free fetal DNA fragments are much
smaller than cell-free maternal DNA, which facilitates
DNA-sequence analysis. Although fetal DNA is detectable
at 5 weeks of gestation, current methods of analysis are unre-
liable before 7 weeks of gestation[171]. Cell-free fetal RNA
and DNA are not released from the fetus but from apoptotic
placental trophoblast cells. These hold greater promise for
genetic testing as a result of advances in DNA-sequencing
methods and informatics[172;173]. In 2007, Down’s syndrome
was detected by the quantitative assay of maternal blood
cell-free RNA for PLAC4, a trophoblast-specific gene
located in the Down’s syndrome region of chromosome 21.
The PLAC4 coding sequence has a SNP that allows deter-
mination of allelic ratios when the fetus is heterozygous for
the SNP[174]. Euploid embryos have an allelic ratio of 1:1.
A ratio of 2:1 indicates a strong likelihood of trisomy 21.
The analysis of mRNAs encoded by different genes on chro-
mosome 21 could improve the sensitivity of this method,
but it has not been widely pursued.
Since the presence of the Y chromosome defines the
male sex, its detection or lack thereof in maternal blood
can be used to determine the fetal sex. A recent review and
meta-analysis of fetal sex determination[175] with the use of
maternal cell-free fetal DNA reported very good sensitiv-
ity, but the greatest sensitivity and specificity in the use of
Y-chromosome sequences to determine sex are obtained
after 20 weeks of gestation, at which time ultrasonography
is the preferred (and inexpensive) method.

T he G enetic and G enomic P ractice of   R eproductive M edicine   •   7 3 3

 In addition to sex determination, the detection of pater-
nal genomic contributions to cell-free fetal DNA can be
used to determine fetal RhD status with high accuracy in
the pregnancy of an RhD-negative woman. This approach
can also be used to detect paternally transmitted, domi-
nant single-gene disorders, including Huntington’s disease,
achondroplasia, and myotonic dystrophy[176]. Carrier status
for cystic fibrosis, hemoglobinopathies, and 21-hydroxylase
deficiency has also been determined.
Recent reports on sequencing of the cell-free
fetal DNA indicate that it has allowed the detec-
tion of under-representation or over-representation of
chromosome-specific sequences, leading to applications in
diagnosing trisomies of chromosomes 13, 18, and 21[176].
Sequencing-based measurements of the proportion of
small DNA fragments derived from chromosome 21 that
exceed a threshold value relative to sequences from euploid
reference samples have been reported to have a positive
predictive value of 96.6% and a negative predictive value
of 100%[176]. This approach is based on “shotgun sequenc-
ing” of the small cell-free fetal DNA fragments and has
the potential of identifying less common, more complex
aneuploid states resulting from unbalanced chromosomal
rearrangements or partial chromosome duplication. Other
potential applications of genetic analysis of cell-free fetal
DNA, including whole-genome sequencing, include detec-
tion of a fetal microdeletion syndrome and copy number
genomic variation[177;178]. If these approaches become tech-
nically feasible, it is not clear whether they would be used
as a screening method or as a diagnostic test. They would
need to be cost-effective, with sufficiently rapid reporting
of results in order to have a meaningful effect on parental
decision-making.
O B S T ET R I C A L M E D I C I N E
Examples of uses of genetic and genomic techniques in the
complex field of obstetrical medicine include recurrent early
pregnancy loss, aneuploidy screening (trisomy 21/Down
syndrome), and managing pre-eclampsia (hypertension of
pregnancy). Other applications, such as genomic factors
associated with pre-term births and unexplained perina-
tal demise, are being intensively investigated and might be
included in coming years[179].
R ECU R R E N T E A R LY P R EG NA N C Y L O S S
Recurrent early pregnancy loss is not uncommon, with
a universally agreed estimate of around one in seven
7 3 4   •   G enomic s in C linical P ractice
pregnancies not progressing beyond the first trimester. It is
widely acknowledged that fetal chromosome abnormalities
account for about 50% of first-trimester pregnancy losses.
Most of these abnormalities are numerical (aneuploid)
abnormalities (86%), and a low percentage is caused by struc-
tural abnormalities (deletions, insertions, ring chromosome
and isochromosome) (6%) or other genetic mechanisms,
including chromosome mosaicism (8%)[180]. The recurrence
risk of numerical abnormalities is consistently low (<1 %),
so karyotyping of fetal material in case of a miscarriage does
not seem worthwhile in daily practice. The majority occur
de novo, and parental karyotypes are inadvertently normal
and are not routinely examined. A direct correlation with
mother’s age at conception (>35  years) is consistently
observed, supporting the argument to introduce preventive
measures including embryo selection by pre-implantation
genetic screening (PGS), fetal chromosome analysis by
chorionic villus biopsy or amniocentesis, and the sophisti-
cated technique of analyzing circulating cell-free fetal DNA
in the maternal circulatory system. Conversely, half of the
structural abnormalities may be inherited from a parent car-
rying a balanced chromosome translocation or inversion.
Parental carriership is found in 4–6% of the couples with
early recurrent miscarriages. In case of parental carriership
of a balanced structural chromosome abnormality, a next
pregnancy may result in a child with an unbalanced struc-
tural chromosome abnormality. This child can have mul-
tiple congenital malformations and/or a mental handicap.
Prenatal diagnosis is therefore recommended.
In addition to demonstrable chromosome abnormali-
ties, other genetic causes implicated in early recurrent mis-
carriages include metabolic and multisystem Mendelian
(single-gene) abnormalities, uniparental disomy, genomic
imprinting, skewed X-linked dominant conditions with
male lethality, and possibly multifactorial/polygenic
causes[180;181].
Immunological and Inflammatory Factors in
Early Pregnancy Loss
Complex biological mechanisms governing fertilization,
embryo development, and fetal growth create a state of
maternal immunological paradox. Physiologically, the
maternal immune system confronts the embryo/fetus with
a host-defense reaction, based on the recognition of pater-
nally derived fetal and placental antigens[180;181]. The mater-
nal immune system is also selectively programmed, however,
to safeguard and prevent rejection of the semi-allogenic
embryo/fetus. Two major types (TH-1 and TH-2) of
maternal immunity are relevant for pregnancy. TH-2 type
immunity is believed to contribute to successful pregnancy,
whilst TH-1 type immunity has been shown to be associ-
ated with idiopathic recurrent early miscarriages[181]. Murine
studies indicate that dominance of TH-1 type dependent
cytokines (interleukin 1, IL-1; interleukin 2, IL-2; tumor
necrosis factor, TNF-a; and interferon, IFN-g) are incom-
patible with a successful pregnancy, whereas a dominance
of TH-2 cytokines (IL-4, IL-6, and IL-10) prevents fetal
wastage[182]. Reports of elevated concentrations of TH-1
cytokines and reduced concentrations of TH-2 cytokines,
including IL-6, among women with early miscarriage are in
accordance with these animal models[183].
IL-6 is a multifunctional cytokine, produced by many dif-
ferent cell types, including immune cells, fibroblasts, endo-
thelial cells, adipocytes, and myocytes. Secretion of IL-6
leads to a stimulation of the hypothalamic-pituitary-adrenal
(HPA) axis during inflammatory processes, promotes
osteoclastogenesis, and participates in the development of
osteoporosis associated with estrogen withdrawal[184]. IL-6
is not constitutively expressed, but is highly inducible and
produced in response to a number of inflammatory stimuli.
IL-6 is generally considered to be a proinflammatory cyto-
kine. IL-6 also has anti-inflammatory properties, as demon-
strated in IL-6 gene knock-out mice[185]. The role of IL-6
expression during pregnancy, as well as its predictive value
for pregnancy outcome, is unclear. Animal studies found
that an increase in lL-6 concentrations precedes uterine
contractions, suggesting that lL-6 plays a role in the physio-
logical mechanisms involved in the propagation of labor. In
human studies, IL-6 production has been described in the
decidua during early pregnancy. Also, IL-6 has been shown
to induce the release of human chorionic gonadotrophin
(hCG) from trophoblasts, leading to a subsequent cascade
of progesterone production, the release of TH2 cytokines
(e.g., IL-6, IL-4), and the suppression of TH1 cytokines[186].
The gene encoding IL-6 has been mapped to chromo-
somal region 7p21[199]. Mutations and polymorphisms with
significantly lower in IL-6 levels are noted in healthy per-
sons[187], associated with Kaposi sarcoma and atherosclero-
sis. These clinical data suggest that carriage of the mutated
IL-6 allele C confers protection against the development
and course of inflammatory diseases. During pregnancy,
IL-6 serum levels are detectable and increased significantly
at the time of delivery[188] and in pregnancies complicated
with intra-amniotic infection. It is now believed that
in normal pregnancy, IL-6 may be regarded as part of an
anti-inflammatory mechanism aimed at maintaining the
pregnancy. There is no firm evidence for either mutation or
polymorphisms in the gene encoding IL-6 to be correlated
with early recurrent miscarriages. However, an indirect,
T he G enetic and G enomic P ractice of   R eproductive M edicine   •   7 3 5
negative contributory effect of IL-6 is likely, since at the
time of initiation of acute inflammatory responses, such as
intra-amniotic infection, IL-6 seems to act as an inducer of
acute phase reactions and as an important player in the elici-
tation of cellular immune responses.
Pre-eclampsia
Pre-eclampsia is a pregnancy-specific syndrome complicat-
ing about 5% of pregnancies and characterized by hyper-
tension and proteinuria[189]. It is a significant obstetrical
problem, accounting for 15–20% of maternal mortality in
developed countries. The etiology is unknown, although
placentas from women with preeclampsia display poor pla-
centation and reduced perfusion. One of the two main the-
ories of the etiology of pre-eclampsia is that ischemia due to
a poorly perfused placenta leads to the release of circulating
toxins to the vascular endothelium, which in turn leads to
the clinical manifestations of the syndrome and secondly,
pre-eclampsia is probably a maternal immune reaction to
the placenta.
There have been a number of microarray studies seeking
to define gene-expression differences in pre-eclamptic pla-
centas compared with those from normal pregnancies. The
HuGeneFL microarray pooled analysis of the RNA from six
normal and six pre-eclamptic placentas revealed 44 upregu-
lated and 15 downregulated genes in pre-eclamptic placen-
tas[190]. In particular, the study found increased expression
of the placental obese gene, as well as elevated plasma levels
of its protein product, leptin, in pre-eclamptic patients. In
another study employing a cDNA microarray specific for
cytokines and interleukins, it was found that 162 of the
221 genes on the array had different expression (defined as
two-fold change) in pre-eclamptic compared with normal
placental tissue[191]. The molecular analysis of pre-eclamptic
placental specimens, using GeneFilter cDNA microar-
rays, showed elevated placental glycogen phosphorylase-M
mRNA, probably due to coexisting hypoxia. Reduced
mRNA levels of HLA-G, which has immunomodulatory
functions, have also been detected in pre-eclamptic placen-
tas by using gene-expression microarray[192].
Direct genomic evidence for hypoxia, presumably
due to inadequate placental vascular development after
implantation, was obtained in a recent microarray study.
Gene-expression profiles by microarray analysis were
obtained from placentas from high-altitude pregnancies
(a relatively hypoxic environment), as well as from pla-
cental tissue explants cultured with 20% (normal) and
3% (hypoxic) oxygen. These expression profiles were then
compared with those from pre-eclamptic placentas. Similar
expression profiles were obtained for high-altitude and
pre-eclamptic placentas as for hypoxic cultured placenta,
supporting the view that the pre-eclamptic placenta func-
tions in a relatively hypoxic environment. In an attempt to
define the systemic endothelial dysfunction hypothesized
to be caused by circulating toxic factors released from
the pre-eclamptic placenta, microarrays were performed
on endothelial cell cultures incubated with serum from
patients with severe pre-eclampsia[193]. It is likely that endo-
thelial dysfunction seen in pre-eclamptic patients could
10. Carson, Daniel.D., Errin Laglow., et  al. Changes in gene expres-
possibly be due to genetic mechanisms other than soluble
toxic factors.
S U M M A RY
Recent additions to our scientific knowledge and tech-
nological developments have considerably revolution-
ized the contemporary practice of reproductive medicine.
Utilization of medical genetic expertise and the application
of conventional genetic laboratory methods are now rou-
tinely employed in managing major referrals to a modern
reproductive medicine facility in a tertiary unit. Major areas
of concern that have benefitted include premature ovar-
ian failure, gynecological conditions like endometriosis,
male factor infertility, pre-implantation genetic screening,
assisted reproduction techniques, early recurrent pregnancy
loss, prenatal diagnosis, pre-eclampsia, and pre-term birth.
The use of microarrays has permeated most areas of study
within the reproductive, obstetrics, and gynecology fields.
The number of diseases and conditions within this broad
field that are currently undergoing genomic studies is too
large to be summarized in this chapter. The rate at which
microarray studies and next-generation genomic methods
are being published makes it certain that this chapter will
always be short of information. Nonetheless, this chapter
provides an overview of the way that genetic and genomic
input and studies are important in a number of areas in the
current practice of reproductive medicine.
REFERENCES
1. Bork, P., R. Copley., Filling in the gaps. Nature, 2001. 409:
p. 818–820.
2. Guttmacher, A.E., and F.S. Collins, Genomic medicine:  a primer.
New England Journal of Medicine, 2002. 347(19): p. 1512–1520.
3. Ishkanian, A.S., et  al., A tiling resolution DNA microarray with
complete coverage of the human genome. Nature Genetics, 2004.
36(3): p. 299–303.
4. Fragouli, E., et  al., Comparative genomic hybridization analysis
of human oocytes and polar bodies. Human Reproduction, 2006.
21(9): p. 2319–2328.
7 3 6   •   G enomic s in C linical P ractice
5. Giudice, L., Elucidating endometrial function in the post-genomic era.
Human Reproduction Update, 2003; 9(3): p. 223–235.
6. Lance Liotta & Emanuel Petricoin. Molecular profiling of human
cancer. Nature Reviews Genetics, 2000. 1(1): p. 48–56.
7. Giudice, L.C. Clinical practice. Endometriosis. New Eng J Medicine,
2010. 362(25): p. 2389–2398.
8. Kao, L.C., S.Tulac et  al. Global gene profiling in human endome-
trium during the window of implantation. Endocrinology, 2002.
143: p. 2119–2138.
9. Borthwick, J.M., D.S. Charnock.Jones., B.D. Tom, M.L. Hull, R.
Teirney, S.C. Phillips, and S.K. Smith. Determination of the transcript
profile of human endometrium. Molecular Human Reproduction,
2003. 9: p. 19–33.
sion during the early to mid-luteal (receptive phase) transition in
human endometrium detected by high-density microarray screening.
Molecular Human Reproduction, 2002. 8: p. 871–879.
11. Riesewijk, A., Julio Martin, et al. Gene expression profiling of human
endometrial receptivity on days LH+2 versus LH+7 by microarray
technology. Molecular Human Reproduction, 2003. 9: p. 253–264.
12. Mirkin, S., M. Arslan, et  al. In search of candidate genes critically
expressed in the human endometrium during the window of implanta-
tion. Human Reproduction, 2005. 20(8): p. 2104–2117.
13. Ponnampalam, A.P., Gareth C.  Weston, A.C. Trajstman, B. Susil,
and Peters A.  Rogers. Molecular classification of human endome-
trial cycle stages by transcriptional profiling. Molecular Human
Reproduction, 2004. 10(12): p. 879–893.
14. Popovici, R.M., L.C. Kao., and L.C. Giudice. Discovery of new
inducible genes in in vitro decidualized human endometrial stro-
mal cells using microarray technology. Endocrinology, 2000.
141(9): p. 3510–3513.
15. Brar, A.K., S. Handwerger, C.A. Kessler, and B.J. Aronow. Gene
induction and categorical reprogramming during in vitro human
endometrial fibroblast decidualization. Physiological Genomics,
2001. 7(2): p. 135–148.
16. Giudice, L.C., L.C. Kao, Endometriosis. Lancet, 2004. 364:
p. 1789–1799.
17. D’Hooghe, T.M., C.K. Kyama, S. Debrock, C. Meuleman, and J.M.
Mwenda, Future directions in endometriosis research. Annals of the
New York Academy of Science, 2004. 1034: p. 316–325.
18. Lessey, B.A., W. Sawin, C.A. Buck, R. Schinnar, W. Bilker, and
B.L. Strom, Aberrant integrin expression in the endometrium of
women with endometriosis. Journal of Clinical and Endocrinological
Metabolism, 1994. 79(2): p. 643–649.
19. Kao, L.C., A. Germeyer, et  al., Expression profiling of endome-
trium from women with endometriosis reveals candidate genes for
disease-based implantation failure and infertility. Endocrinology,
2003. 144(7): p. 2870–2881.
20. Redwine, D., Was Sampson wrong? Fertility and Sterility, 2002.
78(4): p. 686–693.
21. Eyster, K.M., A.L.Boles, J.D. Brannian, and K.A. Hansen, 2002.
DNA microarray analysis of gene expression markers of endometriosis.
Fertility and Sterility, 77(1): p. 38–42.
22. Arimoto, T., et  al., Genome-wide cDNA microarray analy-
sis of gene-expression profiles involved in ovarian endometriosis.
International Journal of Oncology, 2003. 22(3): p. 551–560.
23. Matsuzaki, S., et  al., DNA microarray analysis of gene expression
profiles in deep endometriosis using laser capture microdissection.
Molecular Human Reproduction, 2004. 10: p. 719–728.
24. Conway, G., Premature ovarian failure. British Medical Bulletin,
2000. 56: p. 643–649.
25. Kalantaridou, S., S.R. Davis, and L.M. Nelson, Premature ovarian
failure. Endocrinology and Metabolism Clinics of North America,
1998. 27: p. 989–1006.
26. Coulam, C., S.C. Adamson, and J.F. Annegers. Incidence of pre-
mature ovarian failure. Obstetrics and Gynecology, 1986. 67:
p. 604–606.
 27. Anasti, J., Premature ovarian failure:  an update. Fertility and
Sterility, 1998. 70: p. 1–15.
28. Snowdon, D.A., R.L. Kane, W.L. Beeson, G.L. Burke, J.M. Sprafka,
et  al. Is early natural menopause a biologic marker of health and
aging? Am J Public Health, 1989. 79:  p.  709–714; doi:10.2105/
ajph.79.6.709.
29. Taylor, A., Systemic adversities of ovarian failure. Journal of the
Society of Gynecological Investigation, 2001. 8: p. S7–S9.
30. Conway, G.S., D. Goswami, A. Patel, M.C. Davies, and H.S. Jacobs,
Characterization of idiopathic premature ovarian failure. Fertility
and Sterility, 1996. 65: p. 337–341.
31. Conway, G., Premature ovarian failure. Current Opinion in
Obstetrics and Gynecology, 1997. 9: p. 202–206.
32. Conway, G.S, D. Goswami, Premature ovarian failure. Human
Reproduction Update, 2005. 11(4): p. 391–410.
33. Farkhondeh Pouresmaeili and Zahra Fazeli. Premature Ovarian
Failure:  A  Critical Condition in the Reproductive Potential
with Various Genetic Causes. Int J Fertil Steril, 2014 Apr–Jun.
8(1): p. 1–12.
34. Davison, R., C.R. Quilter, J. Webb, A. Murray et al. A familial case of
X chromosome deletion ascertained by cytogenetic screening of women
with premature ovarian failure. Human Reproduction Update,
1998. 13: p. 3039–3041.
35. Zinn, A., The X chromosome and the ovary. Journal of the Society of
Gynecological Investigation, 2001. 8: p. S34–S36.
36. Lyon, M., The X inactivation centre and X chromosome imprinting.
European Journal of Human Genetics, 1994. 2: p. 255–261.
37. Zinn, A., and J.L. Ross, Turner syndrome and haploinsuffi-
ciency. Current Opinion in Genetics and Development, 1998.
8: p. 322–327.
38. Loughlin, S.A., J. McIver, E. Boyd, A. Carothers, and J.M. Connor,
Analysis of the origin of Turner’s syndrome using polymorphic DNA
probes. Journal of Medical Genetics, 1991. 28: p. 156–158.
39. Lahn, B.P.D., Functional coherence of the human Y chromosome.
Science, 1997. 278: p. 675–680.
40. Jacobs, P.A., and A.G. Baikie, W.M. Court Brown, T.N. MacGregor,
N. Maclean, and D.G. Harnden. Evidence for the existence of the
human “super female”. Lancet. 1959 Sep 26;2(7100):423–425.
41. Holland, C.M., 47,XXX in an adolescent with premature ovarian
failure and autoimmune disease. Journal of Pediatric and Adolescent
Gynecology, 2001. 14(2): p. 77–80.
42. Ko, M.-S., Cytogenetic survey of primary amenorrhea. Japanese
Journal of Human Genetics, 1982. 27(1): p. 35–42.
43. Goswami, R., et al., Prevalence of the triple X syndrome in phenotypi-
cally normal women with premature ovarian failure and its associa-
tion with autoimmune thyroid disorders. Fertility and Sterility, 2003.
80(4): p. 1052–1054.
44. Kaur, A., M. Pandhi, and J.R. Singh, 48,XXXX, a rare aneuploidy.
Balkan Journal of Medical Genetics, 2009. 12(1): p. 65–68.
45. Simpson, J.L., Gonadal dysgenesis and abnormalities of the human
sex chromosomes: current status of phenotypic-karyotypic correlations.
Birth Defects Original Article Series, 1975. 11(4): p. 23.
46. Simpson, J.L., and A. Rajkovic, Ovarian differentiation and
gonadal failure. American Journal of Medical Genetics, 1999.
89(4): p. 186–200.
47. Merry, D.E., et al., Choroideremia and deafness with stapes fixation: a
contiguous gene deletion syndrome in Xq21. American Journal of
Human Genetics, 1989. 45(4): p. 530.
48. Therman, E., R. Laxova, and B. Susman, The critical region on the
human Xq. Human Genetics, 1990. 85(5): p. 455–461.
49. Powell, C.M., et al., Molecular and cytogenetic studies of an X; auto-
some translocation in a patient with premature ovarian failure and
review of the literature. American Journal of Medical Genetics, 1994.
52(1): p. 19–26.
50. Burgoyne, P.S., and T.G. Baker. Meiotic pairing and gametogenic
failure. In Symposia of the Society for Experimental Biology. Symp Soc
Exp Biol, 1984. 38: p. 349–362.
T he G enetic and G enomic P ractice of   R eproductive M edicine   •   7 3 7
51. Tharapel, A.T., et al., Deletion (X)(q26. 1--> q28) in a proband and
her mother:  molecular characterization and phenotypic-karyotypic
deductions. American Journal of Human Genetics, 1993.
52(3): p. 463.
52. Davison, R.M., et al., A familial case of X chromosome deletion ascer-
tained by cytogenetic screening of women with premature ovarian fail-
ure. Human Reproduction, 1998. 13(11): p. 3039–3041.
53. Sala, C., et al., Eleven X chromosome breakpoints associated with pre-
mature ovarian failure (POF) map to a 15-Mb YAC contig spanning
Xq21. Genomics, 1997. 40(1): p. 123–131.
54. Harris, S.E., et  al., Identification of novel mutations in FOXL2
associated with premature ovarian failure. Molecular Human
Reproduction, 2002. 8(8): p. 729–733.
55. Rife, M., et al., Immunohistochemical FMRP studies in a full mutated
female fetus. American Journal of Medical Genetics Part A, 2004.
124(2): p. 129–132.
56. BÃchner, D., et  al., Enhanced expression of the murine Fmr1 gene
during germ cell proliferation suggests a special function in both the
male and the female gonad. Human Molecular Genetics, 1993.
2(12): p. 2043–2050.
57. Partington, M.W., D. Moore, and G.M. Turner, Confirmation of
early menopause in fragile X carriers. American Journal of Medical
Genetics, 1996. 64(2): p. 370–372.
58. Murray, A., et  al., Studies of FRAXA and FRAXE in women with
premature ovarian failure. Journal of Medical Genetics, 1998.
35(8): p. 637–640.
59. Allingham-Hawkins, D.J., Vieri, F., et al., (1999). Fragile X premu-
tation is a significant risk factor for premature ovarian failure: the
International Collaborative POF in Fragile X Study—preliminary
data. American Journal of Medical Genetics. 83: p. 322–325.
60. Hundscheid, R.D.L., et  al., Imprinting effect in premature ovar-
ian failure confined to paternally inherited fragile X premutations.
American Journal of Human Genetics, 2000. 66(2): p. 413–418.
61. Murray, A., S. Ennis, and N. Morton, No evidence for parent of
origin influencing premature ovarian failure in fragile X premu-
tation carriers. American Journal of Human Genetics, 2000.
67(1): p. 253.
62. Sutherland, G.R., and E. Baker, Characterisation of a new rare frag-
ile site easily confused with the fragile X. Human Molecular Genetics,
1992. 1(2): p. 111–113.
63. Murray, A., et al., Microdeletions in FMR2 may be a significant cause
of premature ovarian failure. Journal of Medical Genetics, 1999.
36(10): p. 767–770.
64. Dong, J., et  al., Growth differentiation factor-9 is required during
early ovarian folliculogenesis. Nature. 1996. 10;383(6600):531-5.
65. Aaltonen, J., et al., Human growth differentiation factor 9 (GDF-9)
and its novel homolog GDF-9B are expressed in oocytes during early
folliculogenesis. Journal of Clinical Endocrinology & Metabolism,
1999. 84(8): p. 2744–2750.
66. Di Pasquale, E., P. Beck-Peccoz, and L. Persani, Hypergonadotropic
ovarian failure associated with an inherited mutation of human
bone morphogenetic protein-15 (BMP15) gene. American Journal of
Human Genetics, 2004. 75(1): p. 106–111.
67. Kawano, Y., et  al., Premature ovarian failure associated with a
Robertsonian translocation. Acta Obstetrica et Gynecologica
Scandinavica, 1998. 77(4): p. 467–469.
68. Zlotogora, J., M. Sagi, and T. Cohen, The blepharophimosis, ptosis,
and epicanthus inversus syndrome: delineation of two types. American
Journal of Human Genetics, 1983. 35(5): p. 1020.
69. Amati, P., et  al., A gene for premature ovarian failure associated
with eyelid malformation maps to chromosome 3q22-q23. American
Journal of Human Genetics, 1996. 58(5): p. 1089.
70. Crisponi, L., et al., The putative forkhead transcription factor FOXL2
is mutated in blepharophimosis/ptosis/epicanthus inversus syndrome.
Nature Genetics, 2001. 27(2): p. 159–166.
71. Beysen, D., et  al., The human FOXL2 mutation database. Human
Mutation, 2004. 24(3): p. 189–193.
 72. Bodega, B., et  al., Mutations in the coding region of the FOXL2
gene are not a major cause of idiopathic premature ovarian failure.
Molecular Human Reproduction, 2004. 10(8): p. 555–557.
73. Meduri, G., et  al., Delayed puberty and primary amenorrhea asso-
ciated with a novel mutation of the human follicle-stimulating hor-
mone receptor: clinical, histological, and molecular studies. Journal of
Clinical Endocrinology & Metabolism, 2003. 88(8): p. 3491–3498.
74. Sundblad, V., et al., Controversial role of inhibin -subunit gene in the
aetiology of premature ovarian failure. Human Reproduction, 2006.
21(5): p. 1154–1160.
75. Conway, G.S., E. Conway, and C. Walker, Mutation screening and iso-
form prevalence of the follicle stimulating hormone receptor gene in women
with premature ovarian failure, resistant ovary syndrome and polycystic
ovary syndrome. Clinical Endocrinology, 1999. 51(1): p. 97–99.
76. Arnhold, I.J., A. Lofrano-Porto, and A.C. Latronico, Inactivating
mutations of luteinizing hormone β-subunit or luteinizing hormone
receptor cause oligo-amenorrhea and infertility in women. Hormone
Research in Paediatrics, 2009. 71(2): p. 75–82.
77. Fogli, A., et  al., Ovarian failure related to eukaryotic initiation fac-
tor 2B mutations. American Journal of Human Genetics, 2003.
72(6): p. 1544–1550.
78. Matthews, C.H., et al., Primary amenorrhoea and infertility due to a
mutation in the beta-subunit of follicle stimulating hormone. Nature
Genetics, 1993. 5(1): p. 83–86.
79. Layman, L.C., et  al., Follicle-stimulating hormone beta gene struc-
ture in premature ovarian failure. Fertility and Sterility, 1993.
60(5): p. 852.
80. Takahashi, K., et  al., Increased prevalence of luteinizing hormone
beta-subunit variant in patients with premature ovarian failure.
Fertility and Sterility, 1999. 71(1): p. 96–101.
81. Takahashi, K., et al., Influence of missense mutation and silent muta-
tion of LHβ-subunit gene in Japanese patients with ovulatory disorders.
European Journal of Human Genetics, 2003. 11(5): p. 402–408.
82. Soules, M.R., D.E. Battaglia, and N.A. Klein, Inhibin and reproduc-
tive aging in women. Maturitas, 1998. 30(2): p. 193–204.
83. MacNaughton, J., et  al., Age related changes in follicle stimulat-
ing hormone, luteinizing hormone, oestradiol and immunoreactive
inhibin in women of reproductive age. Clinical Endocrinology, 1992.
36(4): p. 339–345.
84. Petraglia, F., et al., Low levels of serum inhibin A and inhibin B in
women with hypergonadotropic amenorrhea and evidence of high lev-
els of activin A  in women with hypothalamic amenorrhea. Fertility
and Sterility, 1998. 70(5): p. 907–912.
85. Laml, T., et  al., Genetic disorders in premature ovarian failure.
Human Reproduction Update, 2002. 8(5): p. 483–491.
86. Ahonen, P., et  al., Clinical variation of autoimmune
polyendocrinopathy- candidiasis-ectodermal dystrophy (APECED)
in a series of 68 patients. New England Journal of Medicine, 1990.
322(26): p. 1829–1836.
87. Schiffmann, R., et al., Leukodystrophy in patients with ovarian dys-
genesis. Annals of Neurology, 1997. 41(5): p. 654–661.
88. Mathis, S.P., et al., The ovarioleukodystrophy. Clinical Neurology and
Neurosurgery, 2008. 110(10): p. 1035–1037.
89. van der Knaap, M.S., et al., eIF2B-related disorders: antenatal onset
and involvement of multiple organs. American Journal of Human
Genetics, 2003. 73(5): p. 1199–1207.
90. Fogli, A., et al., Screening for known mutations in EIF2B genes in a
large panel of patients with premature ovarian failure. BMC Women’s
Health, 2004. 4(1): p. 8.
91. Gong, Y., et  al., Heterozygous mutations in the gene encoding nog-
gin affect human joint morphogenesis. Nature Genetics, 1999.
21(3): p. 302–304.
92. Groppe, J., et al., Structural basis of BMP signalling inhibition by the
cystine knot protein Noggin. Nature, 2002. 420(6916): p. 636–642.
93. Kosaki, K., et al., Premature ovarian failure in a female with proximal
symphalangism and Noggin mutation. Fertility and Sterility, 2004.
81(4): p. 1137–1139.
7 3 8   •   G enomic s in C linical P ractice
94. Luoma, P., et  al., Parkinsonism, premature menopause, and mito-
chondrial DNA polymerase γ mutations:  clinical and molecular
genetic study. Lancet, 2004. 364(9437): p. 875–882.
95. Cheng, C.J., and S. Stenson, Bilateral corneal anesthesia associated
with diaphragmatic paralysis, ovarian failure, and developmental
delay. Eye & Contact Lens, 2003. 29(4): p. 262–265.
96. Smith, J.A., et  al., Dry eye signs and symptoms in women with
premature ovarian failure. Archives of Ophthalmology, 2004.
122(2): p. 151.
97. Schupf, N., et  al., Early menopause in women with Down’s
syndrome. Journal of Intellectual Disability Research, 1997.
41(3): p. 264–267.
98. Cahill, D.J., and P.G. Wardle, Management of infertility.
BMJ: British Medical Journal, 2002. 325(7354): p. 28.
99. Jansen, E., et al., Abnormal gene expression profiles in human ovaries
from polycystic ovary syndrome patients. Molecular Endocrinology,
2004. 18(12): p. 3050–3063.
100. Wood, J.R., et al., The molecular phenotype of polycystic ovary syn-
drome (PCOS) theca cells and new candidate PCOS genes defined
by microarray analysis. Journal of Biological Chemistry, 2003.
278(29): p. 26380–26390.
101. Levens, E., et al., Fibromodulin is expressed in leiomyoma and myo-
metrium and regulated by gonadotropin-releasing hormone analogue
therapy and TGF-β through Smad and MAPK-mediated signalling.
Molecular Human Reproduction, 2005. 11(7): p. 489–494.
102. Luo, X., et  al., The expression of Abl interactor 2 in leiomy-
oma and myometrium and regulation by GnRH analogue and
transforming growth factor-β. Human Reproduction, 2006.
21(6): p. 1380–1386.
103. Swartz, C.D., et al., Estrogen-induced changes in IGF-I, Myb fam-
ily and MAP kinase pathway genes in human uterine leiomyoma
and normal uterine smooth muscle cell lines. Molecular Human
Reproduction, 2005. 11(6): p. 441–450.
104. O’Flynn O’Brien, K.L., A.C. Varghese, and A. Agarwal, The genetic
causes of male factor infertility: a review. Fertility and Sterility. 2010.
93(1): p. 1–12.
105. Reynolds, N., and H.J. Cooke, Role of the DAZ genes in male fertil-
ity. Reproductive Biomedicine Online, 2005. 10(1): p. 72–80.
106. Krausz, C., G. Forti, and K. McElreavey, The Y chromosome and
male fertility and infertility1. International Journal of Andrology,
2003. 26(2): p. 70–75.
107. Katagiri, Y., et  al., Y chromosome assessment and its implications
for the development of ICSI children. Reproductive Biomedicine
Online, 2004. 8(3): p. 307–318.
108. Foresta, C., E. Moro, and A. Ferlin, Y chromosome microdeletions
and alterations of spermatogenesis. Endocrine Reviews, 2001.
22(2): p. 226–239.
109. Akgul, M., et  al., Cytogenetic abnormalities in 179 cases
with male infertility in Western Region of Turkey:  report and
review. Journal of Assisted Reproduction and Genetics, 2009.
26(2–3): p. 119–122.
110. Simoni, M., E. Bakker, and C. Krausz, EAA/EMQN best practice
guidelines for molecular diagnosis of chromosomal microdeletions.
State of the art 2004. International Journal of Andrology, 2004.
27(4): p. 240–249.
111. Ferlin, A., et al., Molecular and clinical characterization of Y chro-
mosome microdeletions in infertile men:  a 10-year experience in
Italy. Journal of Clinical Endocrinology & Metabolism, 2007.
92(3): p. 762–770.
112. Vogt, P.H., Azoospermia factor (AZF) in Yq11: towards a molecular
understanding of its function for human male fertility and spermato-
genesis. Reproductive Biomedicine Online, 2005. 10(1): p. 81–93.
113. Nuti, F., and C. Krausz, Gene polymorphisms/mutations relevant
to abnormal spermatogenesis. Reproductive Biomedicine Online,
2008. 16(4): p. 504–513.
114. Lardone, M.C., et  al., Quantification of DDX3Y, RBMY1, DAZ
and TSPY mRNAs in testes of patients with severe impairment
 of spermatogenesis. Molecular Human Reproduction, 2007.
13(10): p. 705–712.
115. Jobling, M.A., and C. Tyler-Smith, The human Y chromosome: an
evolutionary marker comes of age. Nature Reviews Genetics, 2003.
4(8): p. 598–612.
116. Krausz, C., et al., Natural transmission of USP9Y gene mutations: a
new perspective on the role of AZFa genes in male fertility. Human
Molecular Genetics, 2006. 15(18): p. 2673–2681.
117. Skaletsky, H., et al., The male-specific region of the human Y chro-
mosome is a mosaic of discrete sequence classes. Nature, 2003.
423(6942): p. 825–837.
118. Ferlin, A., et  al., The human Y chromosome’s azoospermia factor b
(AZFb) region: sequence, structure, and deletion analysis in infertile
men. Journal of Medical Genetics, 2003. 40(1): p. 18–24.
119. Vogt, P.H., Human chromosome deletions in Yq11, AZF candidate
genes and male infertility:  history and update. Molecular Human
Reproduction, 1998. 4(8): p. 739–744.
120. Georgiou, I., et  al., Genetic and epigenetic risks of intracytoplas-
mic sperm injection method. Asian Journal of Andrology, 2006.
8(6): p. 643–673.
121. Zhang, F., et  al., Partial deletions are associated with an increased
risk of complete deletion in AZFc: a new insight into the role of par-
tial AZFc deletions in male infertility. Journal of Medical Genetics,
2007. 44(7): p. 437–444.
122. Patsalis, P.C., et al., Effects of transmission of Y chromosome AZFc
deletions. Lancet, 2002. 360(9341): p. 1222–1224.
123. Ferlin, A., et  al., Male infertility:  role of genetic background.
Reproductive Biomedicine Online, 2007. 14(6): p. 734–745.
124. Bateson, P., et  al., Developmental plasticity and human health.
Nature, 2004. 430(6998): p. 419–421.
125. Lin, Y.W., et al., Partial duplication at AZFc on the Y chromosome is
a risk factor for impaired spermatogenesis in Han Chinese in Taiwan.
Human Mutation, 2007. 28(5): p. 486–494.
126. Ferlin, A., et  al., Human male infertility and Y chromosome dele-
tions:  role of the AZF-candidate genes DAZ, RBM and DFFRY.
Human Reproduction, 1999. 14(7): p. 1710–1716.
127. Yen, H.W., et al., Selective alterations in insulin receptor substrates-1,-2
and-4 in theca but not granulosa cells from polycystic ovaries.
Molecular Human Reproduction, 2004. 10(7): p. 473–479.
128. Reijo, R., et al., Diverse spermatogenic defects in humans caused by
Y chromosome deletions encompassing a novel RNA-binding protein
gene. Nature Genetics, 1995. 10(4): p. 383–393.
129 Writzl, K., B. Zorn, and B. Peterlin, Copy number of DAZ genes in
infertile men. Fertility and Sterility, 2005. 84(5): p. 1522–1525.
130. Sadeghi-Nejad, H., and F. Farrokhi, Genetics of azoospermia: cur-
rent knowledge, clinical implications, and future directions.
Part II:  Y chromosome microdeletions. Urology Journal, 2009.
4(4): p. 192–206.
131. Mitra, A., et  al., Y chromosome microdeletions in azoospermic
patients with Klinefelter’s syndrome. Asian Journal of Andrology,
2006. 8(1): p. 81–88.
132. Schnieders, F., et  al., Testis-specific protein, Y-encoded (TSPY)
expression in testicular tissues. Human Molecular Genetics, 1996.
5(11): p. 1801–1807.
133. Vodicka, R., et al., TSPY gene copy number as a potential new risk
factor for male infertility. Reproductive Biomedicine Online, 2007.
14(5): p. 579–587.
134. Emery, B.R., and D.T. Carrell, The effect of epigenetic sperm abnor-
malities on early embryogenesis. Asian Journal of Andrology, 2006.
8(2): p. 131–142.
135. Lee, K., et al., Premature translation of protamine 1 mRNA causes
precocious nuclear condensation and arrests spermatid differentiation
in mice. Proceedings of the National Academy of Sciences, 1995.
92(26): p. 12451–12455.
136. Cho, C., et al., Haploinsufficiency of protamine-1 or -2 causes infer-
tility in mice. Nature Genetics, 2001. 28(1): p. 82–86.
T he G enetic and G enomic P ractice of   R eproductive M edicine   •   7 3 9
137. Lawrence, L.T., and K.H. Moley. Epigenetics and assisted repro-
ductive technologies:  human imprinting syndromes. In Seminars
in Reproductive Medicine, 2008. Thieme Medical Publishers.
138. Delaval, K., et  al., Differential histone modifications mark mouse
imprinting control regions during spermatogenesis. EMBO Journal,
2007. 26(3): p. 720–729.
139. Kobayashi, H., et al., Bisulfite sequencing and dinucleotide content
analysis of 15 imprinted mouse differentially methylated regions
(DMRs):  paternally methylated DMRs contain less CpGs than
maternally methylated DMRs. Cytogenetic and Genome Research,
2006. 113(1–4): p. 130–137.
140. Kobayashi, H., et al., Aberrant DNA methylation of imprinted loci
in sperm from oligospermic patients. Human Molecular Genetics,
2007. 16(21): p. 2542–2551.
141. Hartmann, S., et al., Genetic imprinting during impaired spermato-
genesis. Molecular Human Reproduction, 2006. 12(6): p. 407–411.
142. Clayton-Smith, J., and L. Laan, Angelman syndrome:  a review of
the clinical and genetic aspects. Journal of Medical Genetics, 2003.
40(2): p. 87–95.
143. Cox, G.F., et al., Intracytoplasmic sperm injection may increase the
risk of imprinting defects. American Journal of Human Genetics,
2002. 71(1): p. 162–164.
144. Maher, E.R., et  al., Beckwith-Wiedemann syndrome and assisted
reproduction technology (ART). Journal of Medical Genetics, 2003.
40(1): p. 62–64.
145. DeBaun, M.R., E.L. Niemitz, and A.P. Feinberg, Association of in
vitro fertilization with Beckwith-Wiedemann syndrome and epi-
genetic alterations of LIT1 and H19. American Journal of Human
Genetics, 2003. 72(1): p. 156–160.
146. He, Z., W.-Y. Chan, and M. Dym, Microarray technology offers a
novel tool for the diagnosis and identification of therapeutic targets for
male infertility. Reproduction, 2006. 132(1): p. 11–19.
147. Lin, Y.-H., et  al., The expression level of septin12 is criti-
cal for spermiogenesis. American Journal of Pathology, 2009.
174(5): p. 1857–1868.
148. Ellis, P.J.I., et  al., Coordinated transcriptional regulation patterns
associated with infertility phenotypes in men. Journal of Medical
Genetics, 2007. 44(8): p. 498–508.
149. Martínez-Heredia, J., et  al., Identification of proteomic differences
in asthenozoospermic sperm samples. Human Reproduction, 2008.
23(4): p. 783–791.
150. Barratt, C.L.R., The human sperm proteome: the potential for new
biomarkers of male fertility and a transformation in our under-
standing of the spermatozoon as a machine. Commentary on the
article “Identification of proteomic differences in asthenozoospermic
sperm samples” by Martinez et  al. Human Reproduction, 2008.
23(6): p. 1240–1241.
151. Pilch, B., and M. Mann, Large-scale and high-confidence proteomic anal-
ysis of human seminal plasma. Genome Biology, 2006. 7(5): p. R40.
152. Deepinder, F., H.T. Chowdary, and A. Agarwal, Role of metabo-
lomic analysis of biomarkers in the management of male infertility.
Expert Review of Molecular Diagnostics, 2007. 7(4): p. 351–358.
153. Fragouli, E., et  al., Comprehensive chromosome screening of polar
bodies and blastocysts from couples experiencing repeated implanta-
tion failure. Fertility and Sterility. 94(3): p. 875–887.
154. Gutierrez-Mateo, C., et al., Aneuploidy study of human oocytes first
polar body comparative genomic hybridization and metaphase II
fluorescence in situ hybridization analysis. Human Reproduction,
2004. 19(12): p. 2859–2868.
155. Chin, K.-V., et al., DNA microarray analysis of the expression profiles
of luteinized granulosa cells as a function of ovarian reserve. Fertility
and Sterility, 2002. 77(6): p. 1214–1218.
156. Margalioth, E.J., A. Ben-Chetrit, M. Gal, and T. Eldar-Geva.
Investigation and treatment of repeated implantation failure follow-
ing IVF-ET. Hum Reprod, 2006 Dec. 21(12): p. 3036–43. Epub
2006 Aug 12.
 157. Alex Simon and Neri Laufer. Assessment and treatment of
repeated implantation failure (RIF). J Assist Reprod Genet, 2012.
29(11): p. 1227–1239; doi:10.1007/s10815-012-9861-4
158. Cedric Le Caignec, Claudia Spits, Karen Sermon, Martine De
Rycke, Bernard Thienpont, Sophie Debrock, Catherine Staessen,
Yves Moreau, Jean-Pierre Fryns, Andre Van Steirteghem, Inge
Liebaers, and Joris R.  Vermeesch. Single-cell chromosomal
imbalances detection by array CGH. Nucleic Acids Res, 2006.
34(9): p. e68; doi:10.1093/nar/gkl336
159. Darryl, L.R. and L. Rebecca. Robker. Molecular mechanisms of ovu-
lation: co-ordination through the cumulus complex. Hum. Reprod.
Update, 2007. 13(3): p. 289–312; doi:10.1093/humupd/dml062
160. Uyar, A., S. Torrealday, and E. Seli.Cumulus and granulosa cell
markers of oocyte and embryo quality. Fertil Steril, 2013 Mar 15.
99(4): p. 979–997; doi:10.1016/j.fertnstert.2013.01.129.
161. Yangxing Zhao, Qiaoli Li, Chenjiang Yao, et al. Characterization
and quantification of mRNA transcripts in ejaculated spermatozoa of
fertile men by serial analysis of gene expression. Hum. Reprod, 2006.
21(6): p. 1583–1590; doi:10.1093/humrep/del027
162. Heike Cappallo-Obermann, Kathrein von Kopylow, Wolfgang
Schulze, and Andrej-Nikolai Spiess. A biopsy sample reduction
approach to identify significant alterations of the testicular tran-
scriptome in the presence of Y-chromosomal microdeletions that
are independent of germ cell composition. Hum Genet, 2010.
128(4): p. 421–431; doi:10.1007/s00439-010-0865-9.
163. Alfarawati, S., et al., First births after preimplantation genetic diagno-
sis of structural chromosome abnormalities using comparative genomic
hybridization and microarray analysis. Human Reproduction.
26(6): p. 1560–1574.
164. Bredenoord, A., et al., Preimplantation genetic diagnosis for mitochon-
drial DNA disorders: ethical guidance for clinical practice. European
Journal of Human Genetics, 2009. 17(12): p. 1550–1559.
165. Geraedts, J.P.M., and G. De Wert, Preimplantation genetic diagno-
sis. Clinical Genetics, 2009. 76(4): p. 315–325.
166. Feero, W.G., et  al., Genomics and perinatal care. New England
Journal of Medicine. 366(1): p. 64–73.
167. Laigaard, J., et  al., ADAM12:  a novel first-trimester maternal
serum marker for Down syndrome. Prenatal Diagnosis, 2003.
23(13): p. 1086–1091.
168. Spencer, K., and K.H. Nicolaides, A first trimester trisomy 13/tri-
somy 18 risk algorithm combining fetal nuchal translucency thickness,
maternal serum free hCG and PAPP-A. Prenatal Diagnosis, 2002.
22(10): p. 877–879.
169. Chitty, L.S., et  al., Fetal nuchal translucency scan and early pre-
natal diagnosis of chromosomal abnormalities by rapid aneuploidy
screening: observational study. BMJ: British Medical Journal, 2006.
332(7539): p. 452.
170. Larrabee, P.B., et  al., Microarray analysis of cell-free fetal DNA in
amniotic fluid: a prenatal molecular karyotype. American Journal of
Human Genetics, 2004. 75(3): p. 485–491.
171. Ehrich, M., et  al., Noninvasive detection of fetal trisomy 21 by
sequencing of DNA in maternal blood: a study in a clinical setting.
American Journal of Obstetrics and Gynecology. 204(3): p. 205.
e1–205. e11.
172. Go, A.T.J.I., J.M.G. van Vugt, and C.B.M. Oudejans, Non-invasive
aneuploidy detection using free fetal DNA and RNA in mater-
nal plasma:  recent progress and future possibilities. Human
Reproduction Update, 17(3): p. 372–382.
173. Lo, Y.M.D., and R.W.K. Chiu, Genomic analysis of fetal nucleic
acids in maternal blood. Annual Review of Genomics and Human
Genetics, 13: p. 285–306.
174. Tsui, N.B.Y., et al., Synergy of total PLAC4 RNA concentration and
measurement of the RNA single-nucleotide polymorphism allelic
ratio for the noninvasive prenatal detection of trisomy 21. Clinical
Chemistry. 56(1): p. 73–81.
7 4 0   •   G enomic s in C linical P ractice
175. Devaney, S.A., et  al., Noninvasive fetal sex determination using
cell-free fetal DNA. JAMA: The Journal of the American Medical
Association. 306(6): p. 627.
176. Wright, C.F., and H. Burton, The use of cell-free fetal nucleic acids
in maternal blood for non-invasive prenatal diagnosis. Human
Reproduction Update, 2009. 15(1): p. 139–151.
177. Fan, H.C., et al., Noninvasive diagnosis of fetal aneuploidy by shotgun
sequencing DNA from maternal blood. Proceedings of the National
Academy of Sciences, 2008. 105(42): p. 16266–16271.
178. Chiu, R.W.K., et al., Non-invasive prenatal assessment of trisomy 21
by multiplexed maternal plasma DNA sequencing: large scale valid-
ity study. BMJ: British Medical Journal. 2011; p. 342; doi: http://
dx.doi.org/10.1136/bmj.
179. Peters, D., et  al., Noninvasive prenatal diagnosis of a fetal
microdeletion syndrome. New England Journal of Medicine.
365(19): p. 1847–1848.
180. Rull, K., L. Nagirnaja, and M. Laan, Genetics of recurrent mis-
carriage:  challenges, current knowledge, future directions. Front.
Genet., 2012. doi:10.3389/fgene.2012.00034
181. Wegmann, T.G., et  al., Bidirectional cytokine interactions in the
maternal-fetal relationship: is successful pregnancy a TH 2 phenom-
enon? Immunology Today, 1993. 14(7): p. 353–356.
182. Hefler, L.A., et al., Polymorphisms of the angiotensinogen gene, the
endothelial nitric oxide synthase gene, and the interleukin-1β gene
promoter in women with idiopathic recurrent miscarriage. Molecular
Human Reproduction, 2002. 8(1): p. 95–100.
183. Unfried, G., et  al., A polymorphism of the interleukin-6 gene pro-
moter and idiopathic recurrent miscarriage. Human Reproduction,
2003. 18(2): p. 267–270.
184. Manolagas, S.C., T. Bellido, and R.L. Jilka, New insights into the
cellular, biochemical, and molecular basis of postmenopausal and
senile osteoporosis: roles of IL-6 and gp130. International Journal of
Immunopharmacology, 1995. 17(2): p. 109–116.
185. Xing, Z., et al., IL-6 is an antiinflammatory cytokine required for
controlling local or systemic acute inflammatory responses. Journal of
Clinical Investigation, 1998. 101(2): p. 311.
186. Nishino, E., et  al., Trophoblast-derived interleukin-6 (IL-6) regu-
lates human chorionic gonadotropin release through IL-6 recep-
tor on human trophoblasts. Journal of Clinical Endocrinology &
Metabolism, 1990. 71(2): p. 436–441.
187. Fishman, D., et  al., The effect of novel polymorphisms in the
interleukin-6 (IL-6) gene on IL-6 transcription and plasma IL-6 lev-
els, and an association with systemic-onset juvenile chronic arthritis.
Journal of Clinical Investigation, 1998. 102(7): p. 1369.
188. Makhseed, M., et  al., Circulating cytokines and CD30 in normal
human pregnancy and recurrent spontaneous abortions. Human
Reproduction, 2000. 15(9): p. 2011–2017.
189. Sibai, B.M., Diagnosis and management of gestational hyper-
tension and preeclampsia. Obstetrics & Gynecology, 2003.
102(1): p. 181–192.
190. Reimer, T., et  al., Microarray analysis of differentially expressed
genes in placental tissue of pre-eclampsia:  upregulation of
obesity-related genes. Molecular Human Reproduction, 2002.
8(7): p. 674–680.
191. Pang, Z.-J., and F.-Q. Xing, Comparative study on the expression of
cytokine-receptor genes in normal and preeclamptic human placen-
tas using DNA microarrays. Journal of Perinatal Medicine, 2003.
31(2): p. 153–162.
192. Hviid, T.V.F., et  al., HLA-G expression in placenta in relation
to HLA-G genotype and polymorphisms. American Journal of
Reproductive Immunology, 2004. 52(3): p. 212–217.
193. Rusterholz, C., S. Hahn, and W. Holzgreve, Role of placentally pro-
duced inflammatory and regulatory cytokines in pregnancy and the
etiology of preeclampsia. In Seminars in Immunopathology, 2007.
29(2): p. 151–162.
 47.
STEM CELL GENOMICS: DEVELOPMENTAL
COMPETENCE
Kyle M. Loh, Bing Lim, and Lay Teng Ang
INTRODUCTION
Stem cells harbor the ability to reconstitute organs and
tissues de novo; therefore, they are the cornerstones of
life: they drive embryonic development, adult homeosta-
sis, and regeneration after injury. A common denomina-
tor of stem cells is the quintessential ability to give rise
to one or more committed cell types, which is other-
wise known as “developmental competence.” Thus stem
cells are positioned at a developmental crossroads at
which they can commit to one of multiple fates: under-
standing how such “decisions” are made is fundamental
to developmental biology. Nevertheless, myriad unan-
swered questions continue to surround the potentiality
of stem cells and how downstream lineage decisions are
executed. How is the developmental destiny of stem cells
(how they know “what” they can become) molecularly
preconfigured? How do stem cells position themselves in
an uncommitted state in a way that they can still access
multiple prospective developmental fates? Upon lineage
commitment, how is a singular developmental program
executed, and how are other prospective fates rescinded?
The advent of various genomics platforms provides a win-
dow of opportunity to access the molecular mechanisms
underlying these phenomena. Drawing from diverse
examples, we argue that stem cell transcription factors
are lineage specifiers that endow stem cells with the com-
petence to differentiate into various progeny. During lin-
eage commitment, extrinsic signals and transcriptional
regulators act within this predetermined range of devel-
opmental potential to specify various daughter-cell types
in a mutually exclusive fashion.
741
S T E M C E L L S , T H E I R D E VE L O PM E N TA L
C O M P ET E N C E , A N D I N D U C I B I L I T Y
D EVE L O PM E N TA L C O M P ET E N C E
A N D E M B RYO G E N E S I S
Developmental competence is a common denominator of
diverse stem cell populations (Waddington, 1940; Wolff,
1968). Unlike terminally differentiated cells, which have
stably acquired a given developmental identity, stem cells/
progenitors are amenable to changes in their cellular iden-
tity elicited by external instructions, which can instruct
them to generate multiple daughter-cell types. These ideas
were similarly conceptualized by Waddington over sev-
enty years ago, who hypothesized that progenitors have
an intrinsic competence, and that extrinsic signals (induc-
ers) acted within this predetermined window of develop-
mental opportunity to induce particular embryonic fates
(Waddington, 1940). Several decades on, in modern stem
cell colloquial terminology, the collective of extrinsic fac-
tors that act on stem cells has been rebranded as the “niche”
(Scadden, 2006; Schofield, 1978).
The developmental competence of a stem cell popula-
tion is delimited by what cell types it can generate (see
Box 47.1). One modern approach to assessing develop-
mental competence is genetic lineage tracing (Blanpain
and Simons, 2013). Lineage tracing involves the perma-
nent genetic labeling of a precursor cell(s) in vivo: upon
its differentiation into one or more lineages, all its progeny
will inheritably carry the genetic label, therefore defin-
ing the repertoire of fates the original stem cell popula-
tion can assume (Blanpain and Simons, 2013). In historic
 Box 47.1  KEY POINTS
• Developmental competence is the ability to generate
one or more daughter cell types.
• Distinct types of stem cells that exist during
embryonic development and in the adult enable
organ formation, homeostasis and tissue repair after
injury.
• The distinction between stem cells and progenitors is
unclear. In certain systems, stem cells and progenitors
interconvert.
• A ubiquitous “stemness” gene is elusive but
emerging evidence indicates the the architecture of
transcription factor networks confers developmental
competence.
•
Developmental competence is conferred and
regulated by extrinsic signaling, lineage specifiers and
chromatin regulators.
• High-throughput genomic technologies (including
chromatin immunoprecipitation-sequencing and
RNA-sequencing) enabled the identification
of human genomic elements that regulate
developmental competence in stem cells and their
lineage specification.
developmental studies, the developmental competence of
stem cell or progenitor populations was classicially defined
by orthotopic or heterotopic transplantation into a recipi-
ent organism (Buckingham and Meilhac, 2011; Weissman
et al., 1978). Orthotopic grafting of cells back into their
native environment tested the typical array of lineages these
cells were intrinsically competent to generate (Weissman
et al., 1978). By contrast, heterotopic transplantation into
other anatomical domains tested whether such progenitors
could be “re-specified” by their extrinsic signaling environ-
ment to adopt other fates. For example, posterior epiblast
(fated to form mesoderm) may be rediverted to adopt ecto-
dermal fates if heterotopically grafted into the anterior end
of the mouse gastrula (Beddington, 1982). Results from
these heterotopic transplantation studies therefore suggest
that the developmental competence of progenitors may be
broader than initially thought, and that lineage commit-
ment boundaries are rather labile. Therefore, redefining the
full scope of developmental potential harbored by various
7 4 2   •   G enomics in C linical P ractice
progenitors (Kawamoto et al., 2010; Rothenberg, 2011)
and how inductive signals act within (or expand) win-
dows of predetermined competence has attracted renewed
interest.
The interaction between inducers and developmentally
competent cells is a recurrent motif occurring throughout
embryonic development for the progressive specification of
new cell types. For example, prospective liver progenitors
within the ventral foregut endoderm are directed to com-
mit to the hepatic lineage by signaling molecules Fgf2 ( Jung
et al., 1999) and Bmp4 (Rossi et al., 2001), secreted by the
surrounding cardiac mesoderm and septum transversum
mesenchyme. Another such interaction is reflected by how
the surface ectoderm secretes Bmp4 to induce nearby inter-
mediate mesoderm to form the nephric duct (Obara-Ishihara
et al., 1999). At a deeper level of complexity, juxtaposition
of two reciprocally signaling progenitor populations pro-
vides an opportunity for self-organization:  for example,
the developing intestinal endoderm (midgut/hindgut)
provides Shh ligand to the adjacent intestinal mesenchyme;
in turn, the intestinal mesenchyme provides a reciprocal
signal (possibly Bmp4) to properly pattern the intestinal
endoderm (Roberts et al., 1995; Roberts et al., 1998). These
interactions between inducers and developmentally compe-
tent cells lay the groundwork for embryogenesis.
Sequential steps of embryonic development are driven
by stem cell populations that possess differing ranges of
developmental competence. Typically, developmental com-
petence becomes progressively restricted as body parts and
organs become increasingly mature. Embryonic develop-
ment is initially spearheaded by cells with expansive devel-
opmental competence. The dozen or so pluripotent cells of
the peri-implantation epiblast at ~E4.5 in mouse devel-
opment are the ancestors to all the hundreds of cell types
of the future adult (Gardner and Rossant, 1979b). Upon
further developmental progression, the competence of plu-
ripotent cells is successively relinquished as they commit
to a number of restricted developmental paths, forming
multipotent stem cell populations harboring the potential
to form particular organ domains or body parts and all the
cell-types within them. Typically, during organ specifica-
tion, multipotent organ-progenitor pools proliferate copi-
ously to expand organ domains, thus delimiting the size
of the future organ in proportion to its surrounding ana-
tomical confines. Multipotent organ stem cell pools fur-
ther differentiate into so-called unipotent progenitors that
are irreversibly committed to form a particular constituent
cell-type. Pancreatic development serves as an example of
this concept of developmental hierarchy. Genetic lineage
tracing in the pancreas has revealed an early multipotent
pancreatic stem cell population harboring the potential to
form all endocrine, exocrine, and ductal lineages (Kopp
et  al., 2011). Such stem cells generate unipotent progeni-
tors that are irreversibly restricted to form either endocrine,
exocrine, or ductal cell-types (Kopp et  al., 2011). Shortly
thereafter, specification of terminally differentiated cell
types commences, yielding the diverse repertoire of pancre-
atic lineages. In this instance, proliferation of multipotent
pancreatic progenitors is essential in order to preconfigure
the size of the future organ: if such progenitor proliferation
is disrupted, a hypoplastic yet perfectly organized pancreas
will arise (Stanger et al., 2007).
S T E M C E L L S A N D S E L F-R E N EWA L
Any discussion of stem cells and developmental competence
must also necessarily encompass self-renewal. The distinc-
tion between stem cells and progenitors is typically drawn
based on self-renewal (Smith, 2006; Weissman, 2000),
which refers to the ability of stem cells to self-replicate and
sustain daughter-cell production whilst continually main-
taining an uncommitted pool of stem cells competent for
future differentiation (Hall and Watt, 1989). For example,
while bona fide “long-term” blood-forming hematopoietic
stem cells (HSCs) can sustain mature blood cell produc-
tion for years (Spangrude et al., 1988b), they also give rise
to “short-term” multipotent progenitors that can produce
all lymphoid and myeloid cell types, but only for weeks or
months (Morrison and Weissman, 1994). In this system, the
“long-term” cells are typically assigned “stem cells,” whereas
the “short-term” cells are known as “progenitors”; but often
the distinction is not as clear, and whether these cellular
activities (or the cells that underlie them) are interchange-
able remains unclear. In fact, in examples we cite below in
multiple organ systems, stem cell and progenitor identities
are probably convertible states that fluctuate in response to
a number of stimuli, including injury (Buczacki et al., 2013;
van Es et al., 2012). Therefore, a rigid distinction between
the two is likely to be a false dichotomy. Self-renewal has
been historically regarded a definitive attribute of stem cells
(Smith, 2006; Weissman, 2000), but it may be given less
importance in light of recent findings.
The concept of stem cell self-renewal is often conflated
with cellular proliferation, as stem cells must harbor the
capacity to self-replicate in order to sustain long-term tis-
sue production. However, whether stem cells replicate at
a slow or fast rate remains unclear. For example, reports
differ as to whether HSCs replicate relatively quickly or
S tem C ell G enomics : D evelopmental C ompetence   •   7 4 3
slowly (Cheshier et  al., 1999)  or slowly (Wilson et  al.,
2008) (dividing either every 2 months or every 6 months,
alternatively). In the developing embryo, “stem cell” popu-
lations divide with extreme rapidity, dividing once every
24 hours, or even more frequently than that (Snow, 1975).
Certain adult stem cells such as hair follicle stem cells ( Jaks
et al., 2008; Waghmare et al., 2008) and intestinal stem cells
(Barker et al., 2007; Snippert et al., 2010) cycle actively, up
to once a day. Perhaps rapid stem cell cycling in some tis-
sues reflects the homeostatic need to vigorously continue
daughter-cell production in organ systems experiencing
rapid turnover (as in the epithelial linings of the skin or
the intestines). However, certain adult stem cell compart-
ments are thought to be rather quiescent, which has been
postulated to prevent stem cell exhaustion and to protect
rare stem cells from oncogenic stresses during cell division
(Lobo et al., 2007; Park and Gerson, 2005; Wang and Dick,
2005). Hence, whether stem cell proliferation is truly pro-
portional to tissue turnover remains unclear.
T H E D R A M AT I S P E R S O N A E O F
THE STEM CELL ENSEMBLE
Though developmental competence is a broad characteris-
tic of stem cells, there are numerous distinct types of stem
cell, disparate in terms of their multilineage potential and
their tissue of residence. The majority of stem cell popula-
tions are restricted in their developmental competence and
can only generate a restricted number of lineages that are
found within their cognate tissue. Nevertheless, rather than
a one-to-one relationship of tissues to their cognate stem
cells, it has been recently realized that a singular tissue may
harbor multiple distinct kinds of stem cells, which might
simultaneously partake in the maintenance or repair of
their resident tissue.
P LU R I P OT E N T MO US E E M B RYO N I C
STEM CELLS
Pluripotent cells possess the potential to generate all the
cell types in the adult body (Smith, 2001) and are found in
the epiblast of mouse embryos during early embryogenesis
(Diwan and Stevens, 1976; Gardner and Rossant, 1979a;
Levak-Svajger and Svajger, 1974) (Figure 47.1). During the
first 3.5 days of embryogenesis (E3.5), the embryo under-
goes multiple cellular divisions to form a hollow sphere (the
early blastocyst) comprising 64 cells, which are spatially seg-
regated to an outer layer (the trophectoderm, which gives
rise to extraembryonic tissues) and an inner layer (the inner
cell mass, the precursor to both hypoblast and epiblast)
(Pan et  al., 2002). After an additional round of cell divi-
sion, the E4.5 128-cell late blastocyst harbors three distinct
lineages: the trophectoderm, hypoblast, and epiblast. This
developmental stage is described as “peri-implantation”
because the 128-cell blastocyst is on the verge of embed-
ding into the uterine wall. Peri-implantation mouse epiblast
cells can be explanted, cultured, and propagated in vitro to
generate self-renewing, pluripotent stem cell lines known as
“mouse embryonic stem cells” (ESCs) (Evans and Kaufman,
1981; Martin, 1981).
ESCs retain the ability to self-renew for extended peri-
ods in culture, suggesting that these have become isolated
from developmental events that would normally direct
epiblast development. The developmental competence
of mouse ESCs is most rigorously assessed using an assay
known as tetraploid complementation. In tetraploid embryos,
only extraembryonic tissues are viable: the cells of the fetus
proper progressively degenerate due to their doubled chro-
mosomal count and are therefore unable to adequately sus-
tain embryogenesis (Snow, 1975; Snow, 1976; Tarkowski
et  al., 1977). Therefore, tetraploid embryos lacking cells
of the body-proper serve as an ideal system to test whether
cells are truly pluripotent. Indeed, when mouse ESCs are
transplanted into host tetraploid embryos, they are fully
capable of differentiating into all fetal cell types, rescuing
embryogenesis and generating viable and fertile progeny in
the tetraploid complementation assay (Nagy et  al., 1993;
Snow, 1975; Snow, 1976). This ability to form viable prog-
eny indicates that ESCs are formally pluripotent in vivo, as
they can generate all bodily cell types when reintroduced
into an embryonic environment.
To stably capture the ephemeral condition of pluripo-
tency, peri-implantation epiblast cells must be explanted
into specific culture conditions that restrain differentiation
and therefore maintain undifferentiated pluripotent cells.
Multiple such culture regimens exist, and typically include
the cytokine leukemia inhibitory factor (LIF) (Smith et al.,
1988; Williams et al., 1988), which promotes self-renewal
and suppresses endoderm/mesoderm specification through
Jak kinase signaling and the transcription factor Stat3
(Bourillot et al., 2009; Niwa et al., 1998). Withdrawal of
LIF typically induces rapid mouse ESC differentiation
(Burdon et al., 2002). Additionally, activation of Wnt/β-
catenin signaling blocks forward developmental progres-
sion to a post-implantation epiblast stage (ten Berge et al.,
2011) and concomitantly acts to enhance mouse ESC pro-
liferation and self-renewal (Wray et al., 2010; Ying et al.,
2008). Finally, abrogating FGF/MAPK signaling inhibits
endoderm, mesoderm, or extraembryonic differentiation,
7 4 4   •   G enomics in C linical P ractice
therefore entrapping cells in an uncommitted state (Burdon
et al., 1999; Ying et al., 2008). These culture environs that
capture self-renewing mouse ESCs resemble the signal-
ing environment that peri-implantation epiblast cells are
exposed to in the late blastocyst, in which LIF/STAT (Do
et al., 2013; Nichols et al., 2001) and Wnt/β-catenin signal-
ing (ten Berge et al., 2011) are active and FGF/MAPK is
suppressed (Nichols et al., 2009).
P LU R I P OT E N T MO US E E P I B L A S T
STEM CELLS
After implantation of the blastocyst into the uterine
wall, by E5.5, the blastocyst undergoes a dynamic phase
of cell proliferation to form an elongated structure, the
egg cylinder (Skreb et al., 1991). The pluripotent epiblast
cells of the peri-implantation stage expand dramatically,
maturing into the post-implantation epiblast (~E5.5).
Post-implantation epiblast cells are the direct anteced-
ents to the embryonic lineages. Different epiblast cells in
specific spatial domains of the post-implantation epiblast
generate different cell types. For example, the epiblast cells
at the proximal posterior region of the embryo ingress
and form the primitive streak (precursor of embryonic
tissues:  definitive endoderm and mesoderm), while the
remaining uningressed anterior epiblast cells give rise to
definitive ectoderm lineages (Lawson et  al., 1991; Tam
and Beddington, 1987).
In vitro explantation of the ~E5.5 post-implantation
epiblast also generates pluripotent stem cell lines, the epi-
blast stem cells (EpiSCs) (Brons et  al., 2007; Tesar et  al.,
2007)  (Figure  47.1). Similar to mouse ESCs and their in
vivo counterparts, EpiSCs are competent to generate the
three embryonic germ layers. However, EpiSCs appear to
be more developmentally “primed” for lineage commit-
ment than are ESCs:  they appear to be on the verge of
endoderm, mesoderm, or ectoderm differentiation (Brons
et  al., 2007; Tesar et  al., 2007)  (akin to the pre-gastrula
egg cylinder epiblast), whereas mouse ESCs (correspond-
ing to blastocyst-stage epiblast) must transit through a
post-implantation epiblast-like state before germ layer
specification (Hayashi et  al., 2011; Zhang et  al., 2010a).
Orthographic grafting of EpiSCs into their native environ-
ment (the post-implantation epiblast) results in formation
of chimeras. This is akin to how mouse ESCs can also inte-
grate into its native tissue, the peri-implantation epiblast, to
generate viable progeny. However, EpiSCs integrate infre-
quently into the inner cell mass (ICM) of pre-implantation
blastocysts (Brons et al., 2007; Tesar et al., 2007), suggest-
ing that the cellular environment of the post-implantation
epiblast may be more conducive to EpiSC integration
(Huang et  al., 2012). Finally, explantation and propaga-
tion of EpiSCs relies on TGFβ/Activin and FGF/MAPK
signaling (Brons et al., 2007; Tesar et al., 2007), which are
diametrically opposite to conditions used for mouse ESC
propagation (which include FGF/MAPK inhibition, for
instance) (Burdon et al., 1999; Ying et al., 2008).
Pluripotent Human Embryonic Stem Cells
While mouse ESCs and mouse EpiSCs can be isolated from
mouse blastocysts, human ESC lines can also be derived
from human blastocysts (Thomson et al., 1998): often
these blastocysts were produced by in vitro fertilization
for clinical purposes. Much like their mouse pluripotent
counterparts, human ESCs can be extensively propagated
in vitro (over a year; Amit et al., 2000) in an uncommit-
ted state, and they readily display the ability to differen-
tiate into endoderm, mesoderm, or ectoderm derivatives.
However, human ESCs appear to be more akin to mouse
EpiSCs than to mouse ESCs: they lack cardinal markers of
mouse ESC identity and are driven by TGFβ/Activin and
FGF/MAPK signaling (Ludwig et al., 2006), much like
EpiSCs. Human and mouse ESCs share a common origin
from the pre-implantation blastocyst, yet they apparently
display many divergent characteristics: lineage relation-
ships between human ESCs, mouse ESCs, and mouse
EpiSCs are not fully understood but have been recently
reviewed (Hanna et al., 2010).
I N D U C E D P LU R I P OT E N T S T E M
CELLS
Through experimental engineering, it is possible to reverse
the process of differentiation to convert differentiated cells
into pluripotent stem cells. Pluripotent stem cells have
been engineered in vitro: in a salient study, mouse fibro-
blast cells were converted to induced pluripotent stem cells
(iPSCs) through the introduction of pluripotency factors
such as Oct4, Sox2, Klf4, and c-Myc (OSKM) (Takahashi
and Yamanaka, 2006). Thus, stem cells may be recreated
from terminally differentiated adult cell types through
reprogramming. The search for the factors sufficient to
induce pluripotency has yielded a variety of different routes
through which pluripotency can be recreated. Recently, a
combination of small molecule inhibitors and microRNAs
(miR-302/miR-367 and a HDAC inhibitor [Anokye-
Danso et al., 2011] or else miR-302/miR-369/miR-200
[Miyoshi et al., 2011]) have been shown to be sufficient to
induce iPSC formation.
S tem C ell G enomics : D evelopmental C ompetence   •   7 4 5
OVE RVI EW O F A D U LT S T E M C E L L S
Adulthood is no less a dynamic situation than embryonic
development—although adult cell-types are arranged in
their ordained positions, constant environmental insults
(e.g., abrasive environments faced by the skin and intestines,
and natural cellular deterioration) continue to erode bodily
membranes and tissues. Thus, stem cell populations must be
recalled to generate their cognate tissues in response to such
injury, whether chronic or acute. Persistent cellular attri-
tion forces stem cells of the blood, stomach, intestines, and
skin (amongst other organs) to continually regenerate new
cell-types, whereas other, rarer stem cell populations in the
liver and other organs are specifically evoked in response to
rarer damage to those organs.
H E M ATO P O I ET I C S T E M C E L L S
In the 1950s, a striking series of observations was noted.
A lethal dose of irradiation extinguished the blood-forming
ability of animals, leading to their death in short order.
However, such irradiated animals could be rescued by trans-
plantation of bone marrow cells from an unirradiated ani-
mal (Congdon, 1957; Thomas et al., 1959; Thomas et al.,
1957; Urso and Congdon, 1957)—a procedure known
as bone marrow transplantation. The characterization of
cells in these transplants eventually led to the discovery of
blood-forming hematopoietic stem cells (HSCs), sources
of the entire hematopoietic system, which possess the capac-
ity to self-renew and differentiate to lymphoid, myeloid, and
erythoid lineages (Lewis and Trobaugh, 1964; Morrison
et  al., 1995; Spangrude et  al., 1988a). HSCs persistently
sustain blood cell production at an incredible rate (with an
estimated 1011 new blood cells being produced every day in
humans [Beerman et al., 2010]) for a lifetime. When seri-
ally transplanted into a succession of recipient mice, HSCs
can inexhaustibly continue blood cell production (Allsopp
et al., 2001; Iscove and Nawa, 1997) for up to five lifetimes
of an adult animal (Allsopp and Weissman, 2002).
The phenotypical capacity of bone marrow cells to pro-
mulgate long-term hematopoiesis and to rescue lethally
irradiated recipients was the basis of a progressive search
to precisely identify the HSC population within this
diverse group of cells, initially led by the Weissman group
(Weissman, 2002) (Figure 47.1). Using cell-surface mark-
ers, they found that the long-term reconstituting activity
of bone marrow cells was first contained in the Thy1low but
not the Thy1high compartment (Muller-Sieburg et al., 1986),
and that long-term reconstituting activity laid in uncom-
mitted cells that were absent for a number of differentiated
lymphoid and myeloid lineage markers (known collectively
 Adult stem cells

Bone marrow Hair follicle Stomach Intestinal Kidney Muscle

Embryonic-like stem cells

Peri-implantation (E4.5) Post-implantation Peri-or pre-implantation Engineered

Mouse embryo (E5.5) Human embryo iPSCs

Trophectoderm Trophectoderm
Proximal epiblast
Primitive endoderm Primitive endoderm
(Hypoblast) Epiblast (Hypoblast)
Epiblast (Post- Epiblast
(Primitive implantation) (Primitive
ectoderm) Distal epiblast ectoderm)

Figure 47.1 The dramatis personae of the stem cell ensemble, including the niches of the adult stem cells (top) and embryonic stem cells (bottom).

as “lineage”). Whereas the transfer of 40,000 whole bone E P IT H E L I A L S T E M C E L L S

marrow cells was sufficient to rescue half of the irradiated
Stem cell populations residing in diverse epithelial tissues,
recipients, this effect was equally reproduced by 30 puri-
such as the stomach, kidney, colon, and hair follicles, have
fied Thy1lowSca1+Lineage-cells (Spangrude et al., 1988b),
recently been defined by genetic lineage-tracing. Despite
demonstrating that HSCs must be harbored within this
their distinct developmental competencies, these stem cells
rare population (representing a more than thousand-fold
commonly express the Wnt co-receptor Lgr5, facilitating
enrichment of HSC activity), and all long-term reconsti-
their discovery.
tuting activity strictly resided within this compartment
For example, Lgr5+ cells residing at the base of pro-
(Uchida and Weissman, 1992). HSCs were successively
spective corpus and pyloric glands of neonatal stomach are
parsed by the discovery of additional cell-surface mark-
multipotent gastric stem cells, which generate the gastric
ers, including c-Kit (Ikuta and Weissman, 1992), CD34
epithelium (Barker et  al., 2010). Hair follicle stem cells
(Osawa et al., 1996) and CD150/Slamf1 (Kiel et al.,
expressing Lgr5, positioned at the midpoint of the hair
2005). These discoveries have together identified a panel
follicle outside the hair bulb (at the bulge) (Figure  47.1),
of surface markers of HSCs to enable their prospective
actively proliferate, live for over 14  months, and regener-
isolation in mice and in human beings. HSCs are best
ate hair follicles ( Jaks et  al., 2008). In the kidney, Lgr5+
described by a composite of lineage markers—for example,
nephron stem cells/progenitors located in S-shaped bodies
in the mouse, CD150+CD34low/-Flk2-Sca-1+c-Kit+Lineage-
(Figure  47.1) generate the thick ascending loop of Henle
(Yilmaz et  al., 2005). Remarkably enough, the introduc-
and distal convoluted tubule (Barker et al., 2012). However,
tion of a single HSC can save the life of a lethally irradiated
unlike the above stem cells that are normally available dur-
mouse (in ~50% of irradiated mice; Kiel et al., 2005),
ing homeostasis, Lgr5+ hepatic stem cells are not found in
demonstrating that contemporarily defined HSC popula-
normal adult livers but are only induced upon liver injury
tions in mice are exceptionally pure. Definition of HSCs in
(Huch et al., 2013). Lineage tracing shows these Lgr5+ stem
human beings, however, remains equivocal and a matter of
cells generate hepatocytes and cholangiocytes upon injury
active experimental investigation, although CD90+ITGα6+
(Huch et al., 2013). Another type of adult epithelial stem
and other combinations of cell-surface markers have been
cell is located in the intestines. These stem cells also express
claimed to delineate human hematopoietic stem cell popu-
Lgr5 (see the more detailed discourse provided below).
lations (Notta et al., 2011).

7 4 6   •   G enomics in C linical P ractice

 I N T E S T I NA L S T E M C E L L S dividing Bmi1-expressing +4 stem cells were shown to act
as a reserve for rapidly replicating Lgr5+ CBCs, as genetic
Due to the constant source of environmental insults
ablation of Lgr5+ CBCs led to no deficits in intestinal
brought with digestion, the epithelial lining of the intes-
homeostasis:  instead, Bmi1+ stem cells entered cell cycle,
tines must sustain a dramatic rate of regenerative turn-
gave rise to Lgr5+ stem cells, and thus fully reconstituted
over:  the intestinal epithelium is made anew every five
intestinal lineages (Tian et  al., 2011). Thus, it appeared
days (Sato and Clevers, 2013). Based on historical obser-
there was a developmental hierarchy within the crypt in
vations, it was thought that intestinal stem cells were
which quiescent +4 stem cells were atop the hierarchy and
harbored near the bottoms of intestinal crypts and there-
gave rise to CBC stem cells that sustained day-to-day crypt
fore partially secluded from the abrasive intestinal lumen
proliferation and were the immediate source of intestinal
(Clevers, 2013). Such stem cells were thought to generate
progeny. However, this is not a unidirectional relationship,
intensely proliferating transit-amplifying progenitors that
as Lgr5+ CBCs can also give rise to Hopx+ +4 stem cells as
would give rise to differentiated intestinal progeny that
well (Takeda et al., 2011). Thus, instead of a strict hierar-
continually moved upwards in the crypt where they would
chy ultimately dictated by +4 stem cells, the intestinal crypt
meet the incisive environment of the intestinal lumen and
system might be regarded as an adaptive oligarchy in which
be shed within several days, thus forming a cellular “con-
both +4 stem cells and CBCs can interconvert and regener-
veyor belt” as the epithelium was constantly regenerated
ate one another.
by newer progeny emerging from the depths of the crypt
Most recent reports have suggested that there may a
(Clevers, 2013).
false dichotomy in the way we regard +4 stem cells and
Modern genetic lineage-tracing has delineated devel-
CBCs in a mutually exclusive fashion. High-resolution
opmental relationships amongst the various intestinal
mapping of molecular markers by multiplexed fluorescent
crypt lineages. Notably, a population of Lgr5+ stem cells
in situ demonstrated half or more of Lgr5+ cells coex-
at the very base of the crypt, known as crypt base colum-
pressed archetypic markers of +4 cells including Bmi1
nar (CBC) cells, were thought to fuel homeostatic intesti-
and mTert (Itzkovitz et al., 2011; Muñoz et al., 2012), sug-
nal regeneration (Barker et al., 2007). Genetic labeling of
gesting that these do not represent distinct stem cell lin-
individual Lgr5+ stem cells formally shows they can gen-
eages. Unexpectedly, recent work has found that intestinal
erate all crypt cell-types. Lgr5+ stem cells are evenly inter-
crypt cells coexpressing Lgr5 together with +4 markers are
spersed with differentiated Paneth cells at the base of the
short-lived Paneth cell and enteroendocrine precursors in
crypt: such Paneth cells supply essential signals (e.g., EGF
steady-state conditions (Buczacki et  al., 2013). However,
and Wnt3) necessary to sustain adjacent CBCs (Sato et al.,
upon intestinal injury, these short-lived precursors are pro-
2011). Furthermore, Lgr5+ stem cells autonomously gener-
moted into long-term intestinal stem cells capable of form-
ate Paneth cells (Sato et al., 2009): thus they generate their
ing all intestinal lineages (Buczacki et al., 2013). Likewise,
own niche; an example of developmental self-organization.
physiologically short-term Dll1+ secretory progenitors
Unexpectedly, Lgr5+ stem cells divide once a day (Barker
gain an expansion of responsibility upon injury, where-
et al., 2007), suggesting that they are a highly proliferative
upon they ascend to Lgr5+ status and become full-fledged
stem cell population.
multipotent intestinal stem cells (van Es et  al., 2012).
In apparent contrast to these findings, it was found
Therefore, the extrinsic signaling environment can substan-
that a putatively distinct intestinal stem cell population
tially expand the multilineage competence of restricted
also existed at a slightly higher level of the intestinal crypt
progeny: rigid lineage boundaries between various progen-
(the “+4” position). Such +4 stem cells express a cohort
itor and stem cell types have become increasingly unclear.
of molecular markers, including Bmi1 (Sangiorgi and
Capecchi, 2008), Hopx (Takeda et  al., 2011), and mTert
(Montgomery et  al., 2011), and they similarly harbor the
MUS C L E S T E M C E L L S
ability to generate all crypt cell-types at a clonal level. Yet,
in contrast to CBCs, +4 stem cells divide infrequently and Stem cells also reside in the skeletal muscles. Muscle-resident
thus retain dilutable DNA labels, leading to their assign- “satellite cells” are a heterogeneous population of myoblasts
ment as “label-retaining cells.” Coexistence of +4 and that closely associate with muscle fibers and persist in the
CBC stem cells in regenerative intestinal epithelium can be basal lamina of the sarcolemma (Figure 47.1). Regeneration
thought to reflect developmental redundancy designed to of muscle tissue after acute injury is mediated by these
safeguard lifelong intestinal rejuvenation. However, slowly tissue-resident muscle stem cells. They are recruited to sites

S tem C ell G enomics : D evelopmental C ompetence   •   7 4 7

of muscle damage, proliferate, and form multinucleated
myotubes (Chen and Goldhamer, 2003; Mauro, 1961).
Radiolabel tracing experiments show that activated satel-
lite cells generated new myonuclei and satellite cells fol-
lowing injury (Charge and Rudnicki, 2004; Heslop et al.,
2001; Schultz, 1996; Smith and Schofield, 1997; Yao and
Kurachi, 1993). Serial transplantation of these cells showed
persisting ability to regenerate new muscles, demonstrat-
ing self-renewal ability (Gross and Morgan, 1999; Morgan
et al., 1994; Yao and Kurachi, 1993).
A D U LT S T E M C E L L S : T H E G E N E S I S
OF CANCER?
Adult stem cells are essential to life and sustain tis-
sue homeostasis and regeneration throughout decades.
However, during such extended periods, long-lived stem
cell populations may accrue genetic lesions that fore-
shadow cancer. Particularly during episodes of tissue
injury in which stem cell proliferation is extrinsically dic-
tated, one possibility is that chronic injury and repeated
recall of stem cells may entrap stem cells in a constitu-
tively self-renewing state evocative of the uncontrolled
proliferation evinced in cancer (Beachy et al., 2004). An
evolutionary perspective is applicable to the way stem
cells and cancer are viewed together. It stands to reason
that, within a stem cell compartment in which mutations
are being continually inflicted, individual stem cells that
gain mutations that enhance self-renewal and/or block
differentiation will gradually out-compete their peers,
and in doing so, will come to dominate the stem cell pop-
ulation (Rossi et  al., 2008). Thus, the stem cell popula-
tion is continually evolving towards a rapidly proliferative
state in which stem cell clones gaining successive muta-
tions that further enhance self-renewal are being selected
for, and then such clones then dominate the population
(Rossi et al., 2008).
Whether stem cells are the genesis of cancer has been
experimentally tested in recent years. Intestinal adeno-
mas were effectuated when oncogene β-catenin was sta-
bilized in Bmi-1+ intestinal stem cells (Sangiorgi and
Capecchi, 2008); likewise, when tumor suppressor Apc
was deleted in Lgr5+ intestinal stem cells (Barker et  al.,
2009). Tumor cells that express Lgr5 were multipotent
cancer stem cells and were able to generate all other cell
types present in the intestinal adenomas (Schepers et al.,
2012). Of note, the same oncogenic perturbations failed
to consistently yield adenomas when they were induced
in transit-amplifying intestinal progenitor cells (Barker
et  al., 2009). Therefore, considerable evidence suggests
7 4 8   •   G enomics in C linical P ractice
that the ultimate tumor-initiating cells may be stem cells.
Of course, these experiments tested whether cancer can
be ectopically initiated via stem cells, but whether stem
cells are the true cells-of-origin of cancer in situ remains
to be formally tested.
In this chapter, we focus on the mechanisms regulating
developmental competence and lineage commitment and
provide topical perspectives where genomics could shed
light on various aspects of stem cell multilineage potential,
self-renewal, or differentiation.
EXTRINSIC SIGNALING AND
D E VE L O PM E N TA L C O M P ET E N C E
In above sections, we have presented developmental com-
petence as a fixed characteristic intrinsic to specific stem
cell or progenitor populations (Waddington, 1940; Wolff,
1968). However, developmental competence can also be
extrinsically programmed to a certain extent: for example,
upon injury, restricted secretory/enteroendocrine precur-
sors within the intestines gain expanded competency to
become intestinal stem cells (Buczacki et al., 2013; van Es
et  al., 2012), an effect mediated by yet-undiscovered sig-
nals. More generally, extrinsic signals play a pivotal role in
altering developmental competence throughout diverse
contexts.
I N S TA L L AT I O N O F D EVE L O PM E N TA L
C O M P ET E N C E BY E X T R I N S I C S I G NA L S
Developmental competence of multipotent stem cells/
progenitors can be conferred or restricted by extrinsic sig-
nals. For example, during renal development in the chick
embryo, where the intermediate mesoderm is the direct pre-
cursor to the kidney fate, the dorsal neural tube surrounding
the kidney morphogenetic field secretes an extrinsic TGFβ
signal (Activin) to induce renal competence within these
cells (Preger-Ben Noon et al., 2009). This renal competence
is then realized by retinoic acid, which induces the expres-
sion of kidney genes like Pax2 and Lhx1 at the posterior
domain of the intermediate mesoderm in vivo (Preger-Ben
Noon et al., 2009).
Extrinsic signals also negatively restrict the devel-
opmental competence of stem cells/progenitors in
vivo. An example can be found during hepatic develop-
ment in the zebrafish embryo, wherein the endoderm
is anteriorly-posteriorly patterned to yield either ante-
rior endoderm (foregut, which gives rise to the liver)
or posterior endoderm (mid/hindgut, which yields
the intestines). Although posterior endoderm is fated
to form intestines, it appears to retain the competence
to form liver:  ectopic expression of liver-inducing sig-
nals, including Wnt8a, can induce ectopic hepatocytes
in the intestinal bulbs (Shin et  al., 2011a). To ensure
that the hepatic-competent intestinal bulbs outside the
liver-forming region are restrained from hepatic specifi-
cation, the surrounding mesenchyme expresses Fgf10a,
which precludes hepatocytes from forming in prospec-
tive intestines (Shin et al., 2011a).
Clearly, extrinsic signaling can endow or restrict devel-
opmental competence: therefore, controlling the ability of
progenitors to respond to external signals must represent a
second mechanism to control competence.
D EV E L O PM E N TA L R EC E P TO R E X P R E S S I O N
E NA B L E S R E S P O NS I VE N E S S TO E X T R I N S I C
S I G NA L S
Extrinsic signals dictate developmental outcomes. Therefore,
developmental competence may be parsimoniously sustained
or restricted by controlling expression of cognate growth fac-
tor receptors.
Binary separation of the myeloid and lymphoid lineages
is one of the earliest events during hematopoiesis, culmi-
nating in myeloid- and lymphoid-restricted progenitors.
Accordingly, common lymphoid progenitors downregulate
IL2Rβ, a receptor for myeloid-inducing interleukin signals
(Kondo et  al., 1997). Strikingly, reintroduction of IL2Rβ
into lymphoid progenitors can resuscitate myeloid com-
petence, respecifying lymphoid-committed progenitors to
illegitimate monocyte and granulocyte fates (Kondo et al.,
1997). Thus, in this instance, early segregation of mutu-
ally exclusive myeloid and lymphoid lineages is initiated by
downregulation of critical cytokine receptors that would
otherwise induce the opposing lineage. Therefore, at this
point, myeloid developmental competence has not yet been
relinquished at the level of chromatin within lymphoid pro-
genitors. Nevertheless, this binary cell-fate decision is con-
firmed shortly thereafter (potentially by chromatin-based
mechanisms), because reintroduction of IL2Rβ into more
committed pro-B cells cannot initiate myeloid respecifica-
tion (Kondo et al., 1997).
In an earlier developmental context, miR-15 and miR-
16 refine the mesendodermal competence of pluripotent
Xenopus blastula cells by repressing expression of TGFβ
receptor Acvr2a (Martello et al., 2007). Apparently broad
Acvr2a expression throughout the blastula is limited by
miR-15/16 expression at the ventral pole, consequently
decreasing TGFβ responsiveness in the ventral blastula
S tem C ell G enomics : D evelopmental C ompetence   •   7 4 9
(Martello et al., 2007). By desensitizing ventral blastula
cells to dorsalizing TGFβ signals, miR-15/16 restricts dorsal
competence (e.g., organizer potential) to the dorsal blastula
(Martello et al., 2007).
In both examples in distinct developmental contexts,
the availability of an obligatory growth factor receptor (or
the lack thereof ) is the decisive determinant of develop-
mental competence.
L I N E AG E -S P E C I F Y I N G
T R A N S C R I P T I O N FAC TO R S C O N F E R
D E VE L O PM E N TA L C O M P ET E N C E
Extrinsic signals clearly influence developmental compe-
tence, as they are responsible for initiating lineage deci-
sions or regulating the capacity of cells to respond to
inductive signals. However, there must be intrinsic actors
within stem cells and progenitors that are responsible for
installing their developmental competence. Transcription
factors are clear determinants of stem cell status and prob-
ably feature prominently in providing stem cells with
their multilineage potential. For example, Pax6 expres-
sion in surface ectoderm is cell-autonomously required
for intrinsic competence to subsequently form the lens
placode (Ashery-Padan et al., 2000; Fujiwara et al., 1994).
However, in virtually all other contexts, how transcrip-
tion factors endow developmental competence remains
obscure. Paradoxically, contemporary models have sug-
gested that stem cell transcription factors actually impede
downstream differentiation in order to restrain stem
cells in an uncommitted, self-renewing state. Therefore,
what are the transcriptional origins of developmental
competence?
A U N I VE R S A L R EGU L ATO R O F “S T E M N E S S”?
A parsimonious explanation of developmental compe-
tence would be that there is a universal “stemness” gene(s)
expressed across all types of stem cells that globally endows
them with some degree of competence. In search of such a
universal regulator, genome-wide microarray analyses were
used to compare the transcriptomes of three different stem
cell classes—the neural stem cells, ESCs, and HSCs: these
analyses identified a common signature of intrinsic factors
(more than 230 genes) that were apparently shared between
embryonic, neural, and hematopoietic stem cells (Ivanova
et al., 2002; Ramalho-Santos et al., 2002; Terskikh et al.,
2001). However, cross-comparison of these datasets across
different laboratories showed only six overlapping genes
(Evsikov, 2003; Fortunel et al., 2003; Ivanova et al., 2002;
Ramalho-Santos et al., 2002). Therefore, a unified molecu-
lar regulator of “stemness” remains elusive.
On the contrary, increasing evidence suggests that,
rather than the expression of common genes across diverse
stem cells, the developmental competence of stem cells
appears to be generally conferred by the organization of
transcription factor networks within stem cells (as opposed
to being generally endowed by the exact transcription fac-
tors within such networks).
T H E P R EC A R I O US -BA L A N C E MO D E L
O F P LU R I P OT E N C Y
How are transcription factors organized in stem cells/pro-
genitors to enable them to simultaneously access diverse
lineage programs? The pluripotential competence of mouse
and human ESCs provides an instructive entry point to this
discussion.
The pluripotent regime of ESCs is cooperatively
governed by a coalition of transcription factors, includ-
ing Oct4, Sox2, Nanog, Esrrb, Sall4, Zic3, and Tbx3
(Chambers et al., 2003; Ivanova et al., 2006; Masui et al.,
2007; Nichols et al., 1998; Niwa et al., 2000; Vallier et al.,
2009)  (Figure  47.2). Individual removal of any given
factor broadly leads to loss of pluripotency and hence
incipient differentiation, demonstrating the intrinsically
cooperative requirement for all factors. It is thought
that stem cell factors concertedly suppress downstream
differentiation-effectors, preventing precocious differ-
entiation and ensuring stable perpetuation of the undif-
ferentiated state. This is referred to as the “ground state”
of pluripotency—a stable stem cell condition in which
differentiation is constitutively repressed ( Jaenisch and
Young, 2008; Silva and Smith, 2008; Young, 2011).
For example, in human ESCs, pluripotency factor
OCT4 is thought to suppress ectodermal and extraem-
bryonic differentiation; SOX2 represses endoderm and
mesoderm specification, and finally, NANOG precludes
ectodermal differentiation (Vallier et al., 2009; Wang et al.,
2012). Therefore, individual pluripotency factors suppress
subsets of prospective developmental outcomes, and coex-
pression of this “trinity” of pluripotency factors (Silva and
Smith, 2008)  collaboratively restrains ESC entry into any
committed lineage, forcibly entrapping them in an undif-
ferentiated state by deferring all alternatives. Likewise, in
mouse ESCs, a cohort of pluripotency factors is thought to
act similarly in order to unilaterally suppress differentiation
(Jaenisch and Young, 2008; Silva and Smith, 2008; Young,
2011). Much like the situation in human ESCs, pluripotency
7 5 0   •   G enomics in C linical P ractice
factors do not repress all lineages, but rather inhibit specific
differentiation programs in mESCs. For instance, Sox2 and
Nanog, respectively, repress endoderm and ectoderm (Vallier
et al., 2009); Oct4, Sox2, or Sall4 specifically inhibit trophec-
toderm commitment (Ivanova et al., 2006; Masui et al., 2007;
Niwa et al., 2000); whereas Prdm14 or Dax1 directly block
primitive endoderm induction (Ma et al., 2010; Niakan et al.,
2006). Given that pluripotency factors strongly repress differ-
entiation, downregulation of pluripotency factors is thought
to be essential for exit from the pluripotent “ground state”
and entry into lineage commitment (Silva and Smith, 2008).
Although decommissioning of stem cell regulators is
thought to be an early event that permits differentiation
(Silva and Smith, 2008), during in vivo developmental
progression, pluripotency factors Oct4 and Nanog are not
downregulated upon epiblast differentiation, and instead
remain expressed in the gastrula-stage primitive streak (PS)
(Hart et  al., 2004; Hoffman et  al., 2013a) alongside PS
transcription factors (e.g., Mixl1) (Mossman et al., 2005).
Likewise, pluripotency factor Sox2 remains expressed
in prospective ectoderm during epiblast differentiation
(Avilion et al., 2003). Therefore, paradoxically, pluripotency
factors are robustly expressed at early stages of differentia-
tion. Do pluripotency factors truly inhibit differentiation
in all instances?
If pluripotency factors were differentiation repres-
sors, one would expect that pluripotency factor overex-
pression would suppress differentiation. However, when
overexpressed, pluripotency factors sometimes promote
ESC lineage specification into specific fates instead of
reinforcing undifferentiated self-renewal. For example,
over-expression of Oct4 in ESCs does not block differ-
entiation—rather, it immediately directs mesodermal or
primitive endoderm differentiation (Niwa et  al., 2000).
Overexpression of Sall4, Tbx3, or Esrrb elicits endoder-
mal determination (Ivanova et  al., 2006; Lu et  al., 2011;
Zhang et al., 2006), and Dax1 instructs trophectodermal
respecification of mESCs (Figure 47.2) (Sun et al., 2009).
Overexpression of Sox2 promotes neuroectodermal dif-
ferentiation (Kopp et al., 2008). Similar effects have been
observed in hESCs as well—it has recently been shown
that NANOG overexpression promotes primitive streak
differentiation in human ES cells (Teo et  al., 2011; Yu
et al., 2011). Therefore, pluripotency factors are not simply
unilaterally repressing differentiation—they also appear to
specify particular developmental outcomes from ESCs.
We propose that pluripotency transcription factors
behave as “lineage specifiers” (Loh and Lim, 2011)  that
direct ESCs to differentiate into particular fates. In order
to promote a singular committed outcome from ESCs,

Mesoderm
Oct4
Trophectoderm
Figure 47.2  Lineage-specifying effects of transcription factors in embryonic stem cells.
pluripotency factors concurrently block differentiation to
mutually exclusive fates while promoting differentiation
to other alternative lineages. For example NANOG spe-
cifically promotes endoderm germ-layer formation from
hESCs while reciprocally repressing the alternate ectoderm
fate (Teo et al., 2011; Vallier et al., 2009; Wang et al., 2012).
Therefore, the previously emphasized differentiation-block-
ing effects of pluripotency factors ( Jaenisch and Young,
2008; Silva and Smith, 2008; Young, 2011) may be explained
in the context of their lineage-specifying activities: they are
excluding alternate fates to channel differentiation to their
lineage of choice. Likewise, ectoderm-specifying factor
Sox2 elicits ectoderm differentiation by blocking endoderm
or mesoderm formation, whereas Oct4 specifically induces
mesoderm commitment of ESCs by precluding ectoderm
induction (Teo et al., 2011; Wang et al., 2012). In each of
these examples, each of these pluripotency factors probably
confer ESCs with the ability to differentiate into a par-
ticular germ-layer (Loh and Lim, 2011). Coexpression of
diverse lineage-specifying pluripotency factors (Oct4, Sox2
and Nanog) therefore endows ESCs with the capacity to
form all germ layers (endoderm, mesoderm, and ectoderm)
and thus, pluripotency (Loh and Lim, 2011).
If the transcriptional determinants of the pluripotent
regime are each advocating a different path of differentia-
tion, this initially appears to be inconsistent with the fact
that ESCs may be stably maintained in an uncommitted
state in culture. We propose a model of transcriptional
competition in which pluripotency factors inducing dif-
ferent developmental outcomes compete with one another,
suppressing one anothers’ lineage-specifying activities and
S tem C ell G enomics : D evelopmental C ompetence   •   7 5 1
Esrrb
Nanog
Endoderm
Tbx3
Sall4
Sox2
Ectoderm
Zic3
therefore generating a temporary (and precarious) state of
developmental indecision. For example, though Sox2 pro-
motes ectoderm formation from ESCs, it is counteracted by
the inhibitory effects of Oct4 and Nanog. Likewise, Nanog
specifies endoderm from ESCs, yet is held in check by
endoderm-suppressing Sox2 (Teo et al., 2011; Wang et al.,
2012).
In conclusion, coexpression of multiple competing
and cross-antagonizing lineage-specifiers in ESCs enables
them to simultaneously access diverse differentiation pro-
grams while maintaining a temporary state of undifferen-
tiated self-renewal (see Box 47.2). Furthermore, stem cell
transcription factors may generally act as differentiation-
promoting lineage-specifiers throughout diverse stem cell
populations, generally directing stem cell differentiation to
a particular daughter lineage at the expense of other cell
types. Concomitant expression of multiple such transcrip-
tion factors leads to a state of undifferentiated indecision
yet endows individual stem cells with the ability to differ-
entiate into any of the lineages comprising the stem cell’s
cognate tissue. This model therefore provides an underlying
molecular explanation for stem cell multilineage potential.
S T E M C E L L G E N O M I C S : C H R O M AT I N
AC C E S S I B I L I T Y E N D OW S T H E
D E VE L O PM E N TA L C O M P ET E N C E O F
STEM CELLS
Developmental competence is the sine qua non of stem cells,
yet since the original introduction of the term over seventy
 Box 47.2  TWO CASE EX AMPLES OF COMPETING LINEAGE SPECIFIERS IN PROGENITORS

• Competing lineage-specifying multipotency factors in pancreatic progenitors

Similar to how Nanog, Oct4 and Sox2 respectively endow ESCs with the competence to differentiate towards multiple
lineage outcomes, a parallel example regards embryonic multipotent pancreatic progenitors (arising from ~E8.5 during
mouse embryogenesis and persisting for several days thereafter). Such pancreatic progenitors are defined by coexpression
of transcription factors Nkx6.1, Sox9 and Ptf1a and they harbor the developmental potency to differentiate into all
pancreatic cell-types (endocrine, exocrine and ductal lineages) (Kopp et al., 2011; Pan and Wright, 2011). Strikingly,
terminal pancreatic differentiation is not accompanied by complete loss of these multipotent pancreatic progenitor
factors. Rather, pancreatic differentiation leads to mutually exclusive resolution of these transcription factors in
distinct spatial domains, yielding Nkx6.1+ endocrine cells (e.g., β-cells), Ptf1a+ exocrine cells and Sox9+ ductal cells,
each of which expresses only one of the three original progenitor factors. For example, though Nkx6.1 and Ptf1a are
concomitantly expressed by multipotent pancreatic progenitors, they functionally promote opposing developmental
agendas—Nkx6.1/Nkx6.2 specify endocrine differentiation, counteracting exocrine-promoting Ptf1a (Schaffer et al.,
2010). Finally, Nkx6.1 and Ptf1a reciprocally inhibit one anothers’ expression(Schaffer et al., 2010). Thus it appears
undifferentiated multipotent pancreatic progenitors are driven by cross-repressive transcription factors, each of which
unilaterally directs progenitor differentiation to a particular differentiated fate while suppressing other prospective fates.
Thus, at the heart of pancreatic progenitors is dynamic transcriptional competition between Nkx6.1, Sox9 and Ptf1a in
which coexpression of these factors endows pancreatic progenitors with multilineage competence.
• Competing lineage-specifying multipotency factors in hepatic progenitors
Lineage-specifying multipotency factors again appear in the earliest multipotent liver progenitors during hepatic
development, which can form either hepatocytes or bile duct cells (cholangiocytes). These hepatic progenitors are defined
by coexpression of multiple transcription factors, including Tbx3, Hnf6a, Hnf1β and Cebpα. Again, each of these pro-
genitor transcription factors induces hepatic progenitors to favor a particular developmental outcome. For example, Sall4
enhances cholangiocyte differentiation while repressing hepatocyte commitment (Oikawa et al., 2009). Conversely, Cebpα
and Tbx3 (Suzuki et al., 2008) repress the cholangiocyte fate choice by repressing biliary regulators (Lüdtke et al., 2009)
Hnf6 (Clotman et al., 2002) and Hnf1β (Coffinier et al., 2002) either directly or indirectly and instead promote hepato-
cyte differentiation formation (Suzuki et al., 2008). We propose that coexpression of these factors endows uncommitted
progenitors with the capacity to differentiate into either of these fates. During subsequent lineage bifurcations, the lineage
specifiers become expressed in a mutually exclusive way, leading to the segregation of the hepatic versus cholangiocyte fates
(similar to the compartmentalization of pancreatic specifiers).

years ago, its molecular origins remain obscure. Of all pos-
sible developmental outcomes, how are a select number
of developmental programs made specifically accessible to
a particular type of stem cell? As aforementioned, extrin-
sic signals and transcription factors endow multilineage
potency. However, ultimately, competence is regulated
at the level of chromatin. Since terminal commitment is
driven by induction of differentiation genes, the compe-
tence to differentiate must lie in the facility of uncommit-
ted stem cells to readily upregulate differentiation effectors
in response to developmental signaling. Hence, a following
proposition is that within stem cells, prospective devel-
opmental genes must be made uniquely accessible within
chromatin such that they may be acutely induced and then
deployed upon extrinsic instruction. Thus, at the crux of
competence is how developmental genes are “prefigured”
for activation in stem cells in the undifferentiated state.
Chromatin is a complex macrobiomolecule that
consists of genomic DNA wrapped in spools around
a protein core, known as histones. The basic unit of
chromatin is the nucleosome core. Each nucleosome
consists of 147 bp of DNA wrapped around a histone
octamer core (Olins and Olins, 2003). The structure
and state of chromatin may affect the accessibility of
DNA to transcriptional machinery. Chromatin state
is associated with post-translational covalent modifi-
cations on the N-terminal tails of histone moieties of
nucleosomes, including post-translational modifica-
tions such as acetylation, phosphorylation, methylation
and ubiquitinylation at key specific residues (Cosgrove
et  al., 2004; Jenuwein, 2001; Strahl and Allis, 2000).
Chromatin-immunoprecipitation followed by hybrid-
ization to DNA-microarray (ChIP-chip) (Heintzman
et al., 2007) or high-throughput-sequencing (ChIP-seq)

7 5 2   •   G enomics in C linical P ractice

 Transcription factors G L O BA L LY AC C E S S I B L E
Genomic DNA 1 2 3 4 C H RO M AT I N A N D C O M P ET E N C E?
Promoter
How does chromatin regulate developmental compe-
tence? Two explanations have been proposed for ESCs.
Crosslink protein to
DNA and sheat DNA One of which is that the chromatin in ESCs is globally
into fragments permissive or “open” to enable transcriptional access to
multiple fate choices. For example, global decondensation
(Efroni et al., 2008) and hyperdynamic histone exchange
Immunoprecipitation
with antibody (Meshorer et  al., 2006)  on ESC chromatin, combined
with overabundance of the euchromatin-associated
H3K4me1 histone modification (Zhu et  al., 2013)  (and
Release and amplify supposed diminution of repression-associated histone
DNA modifications (Efroni et al., 2008)) are thought to gener-
ate a globally permissive transcriptional state throughout
the ESC genome that is essential for broad developmental
Amplify DNA
competence in which many developmental genes are tran-
and tag DNA scriptionally accessible. However, this model of globally
with adoptors
permissive transcription is incompatible with the observa-
dCTP dTTP
dATP dGTP tion that upon lineage induction, only a restricted cohort
of differentiation drivers is invoked. Furthermore, very
Sequence
AGCTACGTGCTACATGAC few lineage outcomes are directly accessible by ESCs (e.g.,
TCGATGCACGATGTACTG
definitive endoderm, mesoderm and definitive ectoderm).
Figure 47.3 
Overview of ChIP-sequencing. ChIP is performed by in situ Therefore, it seems an oversimplification to equate global
cross-linking of a specific protein onto its DNA binding site.
chromatin decondensation in ESCs with access to many
lineage choices. Moreover, whether the globally open
chromatin structure of ESCs contributes to developmen-
( Johnson et al., 2007) are instrumental technologies that
tal competence or whether it is a non-specific byproduct
enable genome-wide profiling of these histone modifica-
remains unclear.
tions (Figure 47.3).
As a developmentally competent cell contains multiple
ChIP is performed by in situ cross-linking of a specific
competing lineage specifiers specifying mutually exclusive
protein onto its DNA binding site. The genomic DNA
lineage outcomes, the regulatory elements controlling the
is sheared and the protein-bound DNA fragment is immu-
expression of these genes to transcription must be amenable
noprecipitated by histone modification-specific antibody.
for activation for these genes to be subsequently expressed.
The immunoprecipitate is enriched for DNA fragments
We propose an explanation in which is that the chromatin
with the histone modification of interest, separated from the
is primed at a restricted number of developmental genes
cross-linked protein, and amplified. Fragments are labeled
associated with prospective lineage outcomes. Chromatin
with adaptors and sequenced. In this way, the DNA sites
priming of developmental genes in uncommitted stem cells
of specific histone marks can be assessed in a genome-wide
may foreshadow accessible developmental outcomes and
manner (Figure 47.3).
precondition such lineage-specifying genes for subsequent
A research consortium, the ENCODE (ENcyclopedia
activation by extrinsic signaling.
of DNA Elements) project was launched in 2007 to profile
the functional regulatory elements encoded in the human
genome using ChIP-seq (Birney et al., 2007). It has since
B I VA L E N T LY M A R K E D P RO M OT E R S
extensively profiled and mapped in many diverse human cell
types, protein-bound regions including the histone modi- As aforementioned (Box 47.3), “bivalent” promoters are
fications of DNA elements (ChIP-seq) (Hoffman et  al., known as “bivalently”-marked because they concomi-
2013b) (Box 47.3). These genomic technologies enabled tantly harbor both activation—and repression-associated
the identification of genomic distribution of histone modi- histone modifications (Figure 47.4). In mouse (Bernstein
fications and their association with gene transcription and et  al., 2006; Xie et  al., 2013)  and human (Pan et  al.,
biological functions. 2007)  ESCs, the promoters of some pattern specification

S tem C ell G enomics : D evelopmental C ompetence   •   7 5 3

 Box 47.3  GENOMIC R EGULATORY ELEMENTS CONTROLLING GENE EXPR ESSION

• Promoters are DNA sequences that enable the binding of RNA polymerase II preinitiation complexes to
active gene transcription (Kim et al., 2005). Promoters can be active or inactive (Heintzman et al., 2007).
Active promoters are bound by RNA polymerase II preinitiation complexes (Kim et al., 2005) or marked by
trimethylation of Lys4 of histone H3 (H3K4me3) (Heintzman et al., 2007). A subset of inactive promoters
marked by both activating and repressive histone modifications are known as bivalently marked promoters (see
below) (Bernstein et al., 2006).
• Enhancers are cis-acting DNA regulatory elements that enhance the expression of its target gene from varying
orientations and positions, including upstream of, downstream of, or within a target gene(Blackwood, 1998; Bulger
and Groudine, 1999).
• Similar to promoters, enhancers can alternate between active or inactive states. Correlation between enhancer
states and the histone modifications have been studied and histone methylation states (Histone H3 on lysine
4—H3K4me1/2/3) and histone acetylation (Histone H3 Lysine 27—H3K27ac) states were found to reflect the
activity of enhancers (Pekowska et al., 2011).
• Active enhancers are marked by H3K27ac and H3K4me1 (Heintzman et al., 2007). Nevertheless, H3K4me1 and
H3K27ac are also highly enriched over gene bodies (Heintzman et al., 2007).
• Inactive or “poised” enhancers can be marked by H3K4me1 only or H3K4me1 together with H3K27me3
(Heintzman et al., 2007) respectively. The H3K4me1 and H3K27me3 labeled enhancers are labelled as “poised”
enhancers since the genes associated with these enhancers are largely inactive (Rada-Iglesias et al., 2010). It is
noteworthy that H3K4me3 is also found on some active enhancers (Pekowska et al., 2011).
• Inactive enhancers can also be devoid of any mark. Latent enhancers for example are regions of the genome that are
unbound by TFs, and that lack enhancer marks but gain them after stimulation (Ostuni et al., 2013).
• Contrary to enhancers, insulators are DNA elements that reduce transcription by obstructing the spread of the
heterochromatin and block enhancers from activating promoters (Mihaly et al., 1998; Mihaly et al., 1997; Udvardy
et al., 1985; Valenzuela and Kamakaka, 2006).

genes (e.g., Brachyury and Nodal) are paradoxically
co-decorated by both activation-associated (H3K4me3)
and repression-associated (H3K27me3) histone modifica-
tions. However, bivalent promoters are thought to repre-
sent some form of transcriptional priming, signaling that
various developmental genes are being poised for activation
while being kept temporarily inactive in ESCs. While such
“bivalently”-marked developmental genes are largely qui-
escent in ESCs, a subset of such bivalent genes is invoked
upon early differentiation. Concomitant with differentia-
tion, the bivalent state at these genes becomes unilaterally
resolved:  developmental genes that become active exclu-
sively gain H3K4me3 and lose H3K27me3 (Bernstein
et al., 2006). By contrast, lineage specifiers associated with
alternate, unrealized fates inherit H3K27me3 and poten-
tially lose H3K4me3 (Bernstein et  al., 2006; Xie et  al.,
2013). Such decommissioning of bivalent domains upon
lineage commitment could track the loss of competence
to access alternate developmental paths, consistent with
increasing lineage restriction.
Although chromatin preaccessibility of developmental
genes in anticipation of differentiation has been phenom-
enologically described in stem cells, is such chromatin prim-
ing is functionally essential for sustaining developmental
competence? For example, in the case of bivalently marked
promoters, loss of either the H3K4me3 methyltransferase
Mll2 (Hu et al., 2013) or the H3K27me3 methyltransfer-
ase component Eed does not significantly alter mouse ESCs
(Hu et  al., 2013). For example, Mll2 depletion (and cor-
responding loss of H3K4me3 at bivalent promoters in the
undifferentiated state) does not seemingly alter the ability
of mESCs to upregulate bivalently marked developmental
genes upon differentiation (Hu et al., 2013). Thus, whether
bivalent marking of promoters causally potentiates devel-
opmental competence, or only correlates with it, remains
unclear.

7 5 4   •   G enomics in C linical P ractice

 Bivalent promoters
Bivalent promoters Active promoters
H3K4me3 H3K4me3
H3K27me3 Differentiation

Pre-enhancer states

Inactive, poised enhancers Active enhancers

H3K4me1/2
Differentiation H3K4me1/2
H3K27ac
H2AZ
H3K4me1/2
H3K27me3 Differentiation H3K4me1/2
H3K27ac
H2AZ
Inactive, latent enhancers
H3K4me1/2
No known H3K27ac Active enhancers also
mark Differentiation typically characterized by
H2AZ H3K4me1/2 and H2AZ as
well
Inactive, H2AZ-primed enhancers H3K4me1/2
H3K27ac
Differentiation
H2AZ H2AZ
H3K4me1/2 H3K4me1/2
Differentiation H3K27ac
H2AZ
H2AZ

Poised RNA polymerase II predeployment

H3K4me3 Elongation
H3K27me3 H3K4me3
Differentiation

Pioneer factors

Pioneer Factor
(e.g. FoxA)

Differentiation

Lineage specifiers compete at chromatin

PRC2 P300/Nanog P300/Nanog

Oct4/Sox2 H3K4me1/2 H3K4me1/2
H3K27me3 H3K27ac
Differentiation

Figure 47.4  Models describing chromatin regulation of stem cell developmental competence and lineage commitment.

Furthermore, it remains unclear whether bivalent
domains truly foreshadow the immediate lineage options
directly available to ESCs. Firstly, many bivalent genes in
ESCs are activated only at far later developmental stages
(long after germ-layer induction or patterning). Secondly, a
subset of bivalent domains may be artifactual. For example,
an apparently large collection of bivalently marked develop-
mental genes reside in Xenopus gastrulae; yet, when gastrulae
were separated into vegetal or animal halves, many appar-
ently “bivalent” genes were exclusively active (H3K4me3+)
or repressed (H3K27me3+) in either the vegetal or animal
domains, such that when whole gastrulae were examined,
these genes superficially appeared bivalent (Akkers et  al.,
2009). Finally, the increasing prevalence of distal enhancer
elements in controlling precise deployment of developmen-
tal genes (whereas proximal promoter elements appear to

S tem C ell G enomics : D evelopmental C ompetence   •   7 5 5

be more cell-type nonspecific in their activity (Heintzman
et al., 2009)) makes it unclear whether bivalent regulation
of promoters has any developmental consequence.
“P R E -E N H A N C E R” S TAT E S
Various developmental enhancer elements harbor increased
chromatin accessibility in uncommitted ESCs (referred to
as enhancer “poising” (Creyghton et al., 2010; Rada-Iglesias
et al., 2010; Rada-Iglesias et al., 2012), perhaps foreshadow-
ing the ability to upregulate cognate developmental genes
upon differentiation. Such enhancer “poising” is thought to
represent another form of how developmental competence
is encoded in chromatin.
Such developmental enhancers often lack activation-
associated mark H3K27ac in the uncommitted state,
becoming activated only upon lineage commitment.
Nevertheless, such developmental enhancers often harbor
euchromatic mark H3K4me1 in ESCs, existing in a state
of “open chromatin” and awaiting activation upon dif-
ferentiation (Creyghton et al., 2010; Rada-Iglesias et al.,
2010; Rada-Iglesias et al., 2012). Some of these “poised”
enhancers also are occupied by repressive mark H3K27me3
as well as histone acetyltransferase p300—the former of
which may restrain poised enhancers from precocious
activation and the latter of which might enable acute his-
tone acetylation upon differentiation (Rada-Iglesias et al.,
2011). Deposition of euchromatic H3K4me1 upon poised
enhancers is thought to enable rapid engagement by tran-
scription factors and chromatin regulators upon differ-
entiation (Calo and Wysocka, 2013), thereby facilitating
transition of “poised” developmental enhancers to a fully
active H3K27ac+ state.
However, developmental enhancers may not need to be
“poised” by H3K4me1 in ESCs to be subsequently upregu-
lated upon differentiation. Of all endodermal enhancers
that become rapidly activated in hESCs (within 1-3 days
of differentiation), only one-third of such endodermal
enhancers are pre-designated by significant H3K4me1
in the undifferentiated state (Loh and Ang et al., 2014).
Likewise, only a minority of mesodermal enhancers are pre-
marked by H3K4me1 in mESCs (Wamstad et al., 2012).
Hence, H3K4me1 captures only a fraction of “poised”
developmental enhancers.
Prospective endodermal enhancers are not solely
pre-marked by H3K4me1:  they exist in an unexpected
diversity of “pre-enhancer” chromatin states within hESCs.
These other pre-enhancer chromatin states include a discrete
class of enhancer totally lack any known histone modifica-
tions (so-called “latent” enhancers (Ostuni et  al., 2013));
7 5 6   •   G enomics in C linical P ractice
other enhancers are decorated with repression-associated
H3K9me3, thought to be a mark of condensed heterochro-
matin (Zhu et al., 2012); and finally one-quarter of future
endoderm enhancers bear histone variant H2AZ in the
apparent absence of other histone modifications (Figure
47.4). Despite these differences in pre-enhancer states, some
associated with euchromatin (e.g., H3K4me1) and others
appearing strongly repressed (e.g., H3K9me3), all of these
“pre-enhancer” states are resolved upon differentiation to
yield an active state characterized by H3K27ac. Thus there
appears to be multiple chromatin preconditioning events by
which enhancers may be prepared for imminent activation.
AC C E S S I N G C L O S E D C H RO M AT I N
BY P I O N E E R FAC TO R S A N D
C H RO M AT I N R E MO D E L E R S
Chromatin accessibility at developmental regulatory ele-
ments (e.g., enhancer “poising” or promoter “bivalency”)
is thought to foreshadow developmental competence: but
how is chromatin made uniquely accessible at these sites
in the first place? What are the functional effectors of
chromatin-level lineage priming?
Chromatin accessibility can be regulated by a special-
ized class of transcription factor known as “gene poten-
tiators” (Zaret, 1999)  or more recently “pioneer factors”
(Zaret et al., 2008), which endow competence for cell type
specification by regulating local chromatin accessibility. To
give some context, chromatin state can markedly alter access
of transcription factors to their cognate binding sites:  for
example, NFI transcription factors are unable to engage
their target sites if DNA is wrapped around a nucleosome
(Piña et  al., 1990). By contrast, ChIP-sequencing and in
vivo foot printing studies show that so-called “pioneer”
transcription factors (e.g., Foxa1 and Foxa2) can bind to
compacted chromatin, initiate chromatin decondensation,
and increase access for subsequent binding of other tran-
scription factors necessary for gene expression (Cirillo and
Zaret, 1999; Cirillo et al., 2002). Therefore, pioneer factors
may be the earliest transcription factors that engage devel-
opmental genes within stem cells and/or progenitors and
precondition such genes for subsequent activation later in
developmental time. For example, in uncommitted foregut
endoderm, Foxa transcription factors already bind to the
liver-specific albumin enhancer, foreshadowing subsequent
liver specification even despite lack of detectable albumin
expression at such stages (Gualdi et  al., 1996). Such pio-
neer factor binding precedes and functionally enables sub-
sequent albumin expression at later developmental phases
(Gualdi et al., 1996). Foxa2 binding is functionally essential
for endoderm specification because it displaces nucleo-
somes from endodermal promoters, therefore enabling
subsequent recruitment of ATP-dependent chromatin
remodeler Brg1/Smarca4 (Li et al., 2012).
Another related instance of developmental “baton-
passing” has been recently revealed for the transcription
factor Foxp3, which is essential for regulatory T cell (Treg)
lineage specification. Interestingly, during the maturation of
naïve Foxp3—T cell precursors to committed Foxp3+ Treg
cells, it seems that Foxp3 is largely unable to infiltrate closed
chromatin:  instead it selectively seeks out already-opened
chromatin domains that have already been exposed in naïve
T cell precursors(Samstein et al., 2012). In Treg precursors,
related transcription factor Foxo1 (which binds the same
consensus motif as Foxp3) prospectively engages distal ele-
ments that are later bound by Foxp3: in this manner, Foxo1
serves as a genomic placeholder and is subsequently suc-
ceeded by Foxp3 upon Treg commitment (Samstein et  al.,
2012).
L I N E AG E -S P EC I F Y I N G T R A NS C R I P T I O N
FAC TO R S C O M P ET E AT T H E L EVE L
O F C H RO M AT I N
How does pioneer transcription factor activity integrate
with the above discussion of lineage-specifying pluripo-
tency factors and chromatin-level “priming” of develop-
mental genes? Of particular interest, pluripotency factors
Oct4 and Sox2 have been proposed to act as pioneer fac-
tors and are apparently able to access closed chromatin in
some contexts (Soufi et  al., 2012). One possibility is that
the pioneering activities of Oct4 and Sox2 in ESCs permit
them to engage dormant germ-layer formation genes and
thereby prepare them for activation: both Oct4 and Sox2
have been clearly documented to engage both the promot-
ers and enhancers of cardinal gastrulation-associated genes
in ESCs ( Jaenisch and Young, 2008). Consistent with
their divergent lineage-specifying activities (above), plu-
ripotency factors OCT4, SOX2 and NANOG compete
over the enhancer elements of developmental genes within
undifferentiated hESCs.
Transcription factor EOMES is an early master regu-
lator of definitive endoderm specification that in inactive
within undifferentiated ESCs(Arnold et al., 2008; Teo
et al., 2011). The EOMES upstream enhancer element is
co-bound by OCT4, SOX2 and NANOG, yet these plu-
ripotency factors are not unilaterally repressing it: pro-
endoderm pluripotency factor NANOG activates the
EOMES enhancer whereas OCT4 and SOX2 are actively
competing to silence it (Teo et al., 2011). Therefore, even
S tem C ell G enomics : D evelopmental C ompetence   •   7 5 7
within undifferentiated hESCs, pluripotency factors act
dichotomously against one another in order, some induc-
ing and others repressing specific developmental genes.
The net outcome is that EOMES is inhibited in hESCs
but positioning of NANOG on its enhancer element
potentially primes EOMES for imminent expression
once SOX2 expression is downregulated in the primitive
streak (see below). Interestingly, it appears that pluripo-
tency factors may bring competing transcriptional coacti-
vator and corepressor complexes to bear in their battle
over EOMES. Both the transcriptional coactivator p300
(H3K27 acetyltransferase) and the transcriptional core-
pressor PRC2 complex (H3K27me3 methyltransferase)
are simultaneously positioned on the EOMES enhancer in
hESCs (Rada-Iglesias et al., 2011): one likely possibility is
that NANOG recruits p300 to activate EOMES whereas
OCT4 or SOX2 counteract this by attracting PRC2. As
a result of such competition, the EOMES enhancer solely
bears repression-associated H3K27me3 and is devoid of
H3K27ac in the undifferentiated state—nevertheless,
p300 prepositioning likely enables the EOMES enhancer
to be rapidly inaugurated, as within 24 hours of endoderm
differentiation, the EOMES enhancer solely gains activa-
tion-associated H3K27ac whereas H3K27me3 is totally
lost (Loh and Ang et al., 2014). Therefore, this provides
an integrated view of how lineage-specifying transcrip-
tion factors and chromatin dynamics are coordinated in
uncommitted stem cell and upon differentiation.
P R E D E P L OY M E N T O F R NA
P O LY M E R A S E AT D EVE L O PM E N TA L
P RO MOT E R S
RNA polymerase is the terminal arbiter of gene expres-
sion and its predeployment to developmental promoters in
ESCs is a form of “preparedness” for gene expression (by
contrast to the significance ascribed to various chroma-
tin states). For example, in Drosophila mesoderm, RNA
polymerase is deployed to the promoters of many inactive
developmental genes but is “stalled” and does not initiate
transcription (Zeitlinger et  al., 2007). RNA polymerase
prepositioning prefigures these developmental genes for
imminent activation, allowing synchronous and precise
execution of developmental programs, as transcription can
immediately ensue upon reception of the appropriate sig-
nals upon lineage commitment. Abrogating such “stalled”
polymerases has deleterious developmental consequences.
For example, disrupting RNA polymerase predeployment
to the Snail gene in prospective Drosophila mesoderm leads
individual cells to stochastically initiate Snail expression at
different times, thus desynchronizing mesoderm invagina-
tion (Lagha et al., 2013).
Likewise, RNA polymerase is predeployed to the pro-
moters of various lineage commitment genes in ESCs
(Guenther et al., 2007; Marks et al., 2012), especially biva-
lently marked promoters carrying both H3K4me3 and
H3K27me3 (Min et  al., 2011). Although its functional
consequences have yet to be revealed in this context, it is
likely RNA polymerase prepositioning on commitment
genes enables differentiation to be rapidly executed.
Figure 47.4 illustrates models describing chromatin reg-
ulation of stem cell developmental competence and lineage
commitment.
E N T RY I N TO L I N E AG E
COMMITMENT
Stem cells simultaneously access multiple prospective devel-
opmental fates, and from this multitude of choices a uni-
lateral decision has to be made. Unilateral commitment
proceeds by the activation of a singular developmental pro-
gram and shut down of mutually exclusive fates (Davidson,
2010; Graf and Enver, 2009).
U N BA L A N C I N G S T E M C E L L
T R A N S C R I P T I O NA L EQ U I L I B R IU M S
I N VO K E E A R LY D I FFE R E N T I AT I O N
At the heart of developmental competence is a confluence of
multiple lineage specifiers cross-antagonising one another
in a precarious equilibrium. Constitution of stem cell tran-
scriptional programs by cross-repressive lineage specifiers
offers a unique strategy for how multipotent and pluripo-
tent stem cell lineage decisions are made—it suggests that
multiple developmental fates are constantly simultaneously
accessible to stem cells and that uncommitted stem cells can
adopt one lineage or another in a mutually exclusive way
due to coexpression of mutually exclusive lineage specifiers.
Thus, the principal question is how in states of such tran-
scriptional indecision or transcriptional conflict what “tips”
such a precarious equilibrium for stem cells to unilaterally
enter one developmental path versus another?
Frequently, extrinsic signals perform the initial “symme-
try breaking” event that unbalances the stem cell transcrip-
tional equilibrium. Pluripotency factors Oct4, Sox2 and
Nanog are broadly coexpressed in the pluripotent epiblast/
ESCs, yet upon gastrulation these factors become segre-
gated such that the anterior epiblast/prospective ectoderm
solely expresses Sox2 whereas posterior epiblast/primitive
7 5 8   •   G enomics in C linical P ractice
streak is marked by Oct4 and Nanog (Hart et  al., 2004;
Hoffman et  al., 2013a). Thus, progressive development
leads to segregation of pluripotency factors in distinct lin-
eages as seen previously for multipotency factors in other
developmental contexts.
How does this segregation occur? The post-implantation
epiblast in vivo and hESCs in vitro are exposed to moderate
levels of TGFβ/Nodal signaling. TGFβ signaling upregu-
lates Oct4 and Nanog while repressing Sox2 (Vallier et al.,
2009; Xu et  al., 2008):  therefore, at intermediate TGFβ
levels, Oct4, Sox2 and Nanog are coexpressed. However,
juxtaposed to the future anterior epiblast is an extraembry-
onic tissue, the anterior visceral endoderm (AVE), which
expresses diffusible TGFβ antagonists including Lefty1 and
Cer1, which generate an anterior-to-posterior gradient of
increasing TGFβ signaling within the epiblast (Figure 47.5)
(Perea-Gomez et al., 2002). Therefore with low TGF in the
anterior epiblast, Oct4 and Nanog are lost and Sox2 is sus-
tained (Mesnard et al., 2006; Shin et al., 2011b), whereas
higher TGFβ signaling in the posterior epiblast leads to
expression of Oct4 and Nanog in the absence of Sox2.
Therefore, either elevating or suppressing TGFβ signaling
destabilizes the pluripotent equilibrium, culminating in
Oct4+Nanog+ primitive streak or Sox2+ definitive ecto-
derm fates. Wnt signaling also serves to untip this equilib-
rium: Wnt signaling follows a similar posterior-to-anterior
gradient due to posterior Wnt3 signaling and opposing
expression of Wnt antagonists by the AVE (e.g., Dkk1).
Likewise, Wnt signaling elevates Nanog whereas it recipro-
cally suppresses Sox2 (Sumi et al., 2013).
Figure 47.5 illustrates the extrinsic signaling events that
spatially control lineage specification in the embryo during
gastrulation.
Asymmetric localization of Sox2 versus Nanog in spa-
tially distinct epiblast compartments is consistent with
their distinct roles in inducing differential germ-layer out-
comes—in hESCs, SOX2 subsequently induces neural
determinant PAX6, thus specifying neuroectoderm (Wang
et al., 2012) whereas NANOG invokes primitive streak reg-
ulator EOMES to affirm primitive streak commitment (Teo
et al., 2011) (see above). How does this occur at the level of
chromatin? As aforementioned, SOX2 and NANOG com-
pete over the EOMES enhancer:  however, loss of SOX2
enables NANOG to activate the EOMES enhancer and
hence induce its expression (Teo et al., 2011). Thus, an ini-
tial and transient unbalancing of competing pluripotency
factors by extrinsic signals segues into germ layer commit-
ment, and subsequently pluripotency factor expression
itself is often lost, marking an irreversible step forward in
developmental progression. Consequently, it could be said

Nodal
BMP
Wnt inhibition
Wnt
BMP inhibition
Nodal inhibition
Posterior primitive
Definitive
Anterior primitive
ectoderm
streak
Definitive
endoderm
High
Nodal
Figure 47.5 
Extrinsic signaling events that spatially control lineage
specification in the embryo during gastrulation.
that upregulation of pluripotency factors (and concomitant
downregulation of others) marks the first step of germ layer
differentiation. Here, the transient upregulation of pluri-
potency factors by extrinsic signals is the necessary deci-
sive step to execute germ layer formation by perturbing the
equilibrium.
R EGU L AT I O N O F MU T UA L LY E XC LUS I VE
L I N E AG E D EC I S I O NS BY C RO S S -R E P R E S S I VE
T R A NS C R I P T I O N FAC TO R S
Stem cells are at a developmental crossroads:  unilateral
commitment to a particular lineage necessarily requires
relinquishing alternate developmental possibilities. After
stem cell transcriptional equilibriums are tipped or way
in uncommitted cells (above), lineage specification must
be consolidated by downstream transcription factors that
“lock in” the prevailing fate. Lineage-specifying transcrip-
tion factors prominently secure mutually exclusive lineage
specification by promoting a specific developmental agenda
and repressing other alternatives (Enver and Greaves, 1998;
Graf and Enver, 2009). Often, competing lineage specifi-
ers not only inhibit one anothers’ lineage options but also
cross-antagonize one another to enhance mutual exclusion.
Cross-repressive transcription factors consolidate one
developmental outcome by blocking off alternatives. For
example, common myeloid progenitors (CMPs) access two
S tem C ell G enomics : D evelopmental C ompetence   •   7 5 9
possible fates:  they can commit either to the erythroid/
megakaryocyte or the granulocyte/macrophage lineage.
This lineage decision is principally driven by opposition
between megakaryocyte/erythroid-inducing Gata1 versus
FGF granulocyte/monocyte-driving PU.1/Sfpi1 (Heyworth
et  al., 2002; Kulessa et  al., 1995; Nerlov and Graf, 1998;
streak
Nerlov et al., 2000). Both transcription factors repress the
alternate fate and one anothers’ expression (Nerlov and Graf,
1998; Nerlov et al., 2000) and such a transcription factor
“cross-antagonism” leads to the developmental separation
of these two fates, culminating in Gata1+PU.1- megakaryo-
cyte/erythroid progenitors (MEPs) versus PU.1+Gata1—
Mesoderm granulocyte/monocyte progenitors (GMPs) (Arinobu
subtypes
et al., 2007).
Similarly, endoderm transcription factor Eomes
(Arnold et  al., 2008; Teo et  al., 2011)  acts at the branch-
point of primitive streak progenitors that can commit
to either definitive endoderm or mesoderm. EOMES is a
key decider in the endoderm-mesoderm bifurcation and
it reciprocally blocks mesoderm formation in order to
generate endoderm:  EOMES directly occupies and sup-
presses a number of mesoderm genes (e.g., BRACHYURY,
KDR and MEOX1) (Teo et al., 2011). Therefore, if Eomes
is ablated, mesoderm is illegitimately generated even in
endoderm-inducing conditions (Teo et al., 2011).
N O N- C O D I N G M I C RO R NA S A N D
T R A N S C R I P T I O N FAC TO R S S I L E N C E
D EVE L O PM E N TA L C O M P ET E N C E
D U R I N G L I N E AG E S P EC I FI C AT I O N
Non-coding RNAs, in particular microRNAs, have been
shown to regulate stem cell lineage specification. The res-
toration of a single microRNA (miR-430) in Dicer mutant
zebrafish that could no longer produce endogenous
microRNA ameliorated deficits in zebrafish neuroectoder-
mal development and neuronal differentiation (Giraldez
et al., 2011). In mammals, specific microRNAs have been
shown to regulate B-cell differentiation (miR-181 (Chen
et  al., 2009)), adipocyte differentiation (miR-143 (Esau
et  al., 2004)), and insulin secretion (miR-375 (Poy et  al.,
2009)).
One way that microRNAs serve to execute differentia-
tion is to reciprocally silence progenitor-state factors, thus
unidirectionally locking developmental progression in the
forward direction. For example, miR-302 represses Oct4
during differentiation to result in its eventual downregu-
lation (Rosa et al., 2009); expression of miR-145 increases
during human ESC differentiation, whereupon it represses
Oct4, Sox2 and Klf4 (Tay et al., 2008a); and finally, miR-296
and miR-470 suppress Nanog in mouse ESCs (Tay et  al.,
2008b). These microRNAs shut down the pluripotency
program, enabling transition into differentiation: therefore,
deletion of microRNA-processing enzymes, such as Dicer
(Bernstein et al., 2003) and Dgcr8 (Wang et al., 2007) leaves
ESCs unable to differentiate.
Suppression of the preceding progenitor state is also a
recurrent motif of many lineage-specifying transcription
factors:  for example, Eomes binds to the Oct4 and Sox2
genes and downregulates their expression during endoderm
commitment (Teo et al., 2011) whereas Pax6 binds to the
Oct4 and Nanog promoters, therefore suppressing them
during ectoderm specification (Zhang et al., 2010b).
C O N C LU S I O N
Historically, developmental lineage tracing and graft-
ing experiments have classically delineated the
progenitor-progeny relationships which form the devel-
opmental hierarchies we know today, with various stem
cell and progenitor lineages positioned at different levels.
These findings laid the foundation of developmental biol-
ogy but were previously confined to phenomenological
descriptions of various progenitor-progeny relationships.
Understanding the causal drivers of each discrete cell-fate
transition within complex developmental hierarchies
remains the overarching challenge of modern developmen-
tal biology:  this is an issue that is optimally addressed by
genomics approaches. Historically, regulatory genes behind
particular lineage transitions were unbiasedly identified by
genome-wide mutagenesis in model organisms (forward
genetics) or genetic ablation of selected candidate regulators
(reverse genetics).
These incredibly informative yet time-consuming
approaches may yet be superseded by studies empowered
by computational inference. For particular developmen-
tal hierarchies, for example the hematopoietic system, we
now have accurate transcriptional profiles of rather pure
populations of major stem cell populations, their progeny
and the intermediates spanning between them (by either
microarrays or RNA-seq) (Seita et  al., 2012). Therefore,
we can make informed guesses of the regulatory genes that
drive various lineage transitions by comparing differentially
expressed genes that differ between proximal developmen-
tal intermediates and can narrow the list of regulatory genes
to functionally test by gain- or loss-of-function analyses.
Even in cell-fate relationships in which knowledge of key
drivers was previously known, historically it has been diffi-
cult to ascribe exact molecular functions to these regulatory
7 6 0   •   G enomics in C linical P ractice
genes. However, ablation of known regulators followed by
genome-wide transcriptional profiling can reveal the exact
genes that are functionally invoked or repressed by them.
ChIP-seq and related technologies can ascertain which
genes are under the direct control of specific transcription
factors or chromatin remodelers.
However, in years to come it will be necessary to employ
descriptive genomics profiling with functional genetic
approaches in equal measure. For example, it was previously
extrapolated that pluripotency factors unilaterally inhibit
differentiation, because pluripotency factors were found
to occupy the promoters of many differentiation genes
in ESCs via ChIP analyses ( Jaenisch and Young, 2008).
Nevertheless, only phenotypic gain—and loss-of-function
approaches later uncovered that pluripotency factors actu-
ally upregulate a subset of these developmental genes and
therefore specify ESC differentiation to particular lineages
(Loh and Lim, 2011; Teo et al., 2011; Wang et al., 2012).
Transcription factor binding identified by computational
approaches cannot be inferred to mean either gene upreg-
ulation or downregulation and therefore we must seek, in
both stem cell biology and developmental biology, to rig-
orously functionally validate all computational predictions.
What challenges remain? This review has emphasized
the historic concept of developmental competence, concep-
tualized seventy years ago and the application of genom-
ics to uncover the underlying mechanisms. However, our
knowledge of its molecular foundation still remains frag-
mentary. We have proposed that lineage-specifying tran-
scription factors within uncommitted stem cells capacitate
specific lineage options, therefore providing a possible
explanation for multilineage competence (Loh and Lim,
2011). Yet, the mechanistic details of how transcription fac-
tors “prime” developmental genes within undifferentiated
stem cells remain unclear, although pioneer factor activity
and corresponding chromatin accessibility at regulatory
elements may be a parsimonious answer. At one end of the
question, we have incomplete knowledge of the mechanistic
workings of competence, and at the other extreme, even the
basic definition of developmental competence has become
increasingly vague, as stem cells and progenitors can appar-
ently traverse lineage boundaries during injury and appre-
ciably expand their lineage potential (Buczacki et al., 2013;
van Es et al., 2012).
AC K N OW L E D G E M E N T S
Lay Teng Ang and Bing Lim are supported by the Singapore
Agency for Science, Technology and Research (A*STAR),
and Kyle M. Loh by the Fannie and John Hertz Foundation,
the U.S. National Science Foundation, and the Davidson
Institute for Talent Development. We thank Xinhong Lim
(Stanford University School of Medicine) for informative
discussions. Jingyao Zhang, Vibhor Kumar, and Shyam
Prabhakar (Genome Institute of Singapore) contributed to
chromatin state analyses that laid the foundation to some
ideas discussed in this article: they also invented the neolo-
gism “pre-enhancer.”
REFERENCES
Akkers, RC, et  al., 2009. A hierarchy of H3K4me3 and H3K27me3
acquisition in spatial gene regulation in Xenopus embryos.
Developmental Cell 17(3):425–434.
Allsopp, RC, S Cheshier, and IL Weissman, 2001. Telomere shortening
accompanies increased cell cycle activity during serial transplan-
tation of hematopoietic stem cells. The Journal of Experimental
Medicine 193(8):917–924.
Allsopp, RC, and IL Weissman, 2002. Replicative senescence of hemato-
poietic stem cells during serial transplantation: does telomere short-
ening play a role? Oncogene 21(21):3270–3273.
Amit, M, et al., 2000. Clonally derived human embryonic stem cell lines
maintain pluripotency and proliferative potential for prolonged
periods of culture. Developmental Biology 227(2):271–278.
Anokye-Danso, F, et al., 2011. Highly efficient miRNA-mediated repro-
gramming of mouse and human somatic cells to pluripotency. Cell
Stem Cell 8(4):376–388.
Arinobu, Y, et  al., 2007. Reciprocal activation of GATA-1 and PU.1
marks initial specification of hematopoietic stem cells into
myeloerythroid and myelolymphoid lineages. Cell Stem Cell
1(4):416–427.
Arnold, SJ, et al., 2008. Pivotal roles for eomesodermin during axis for-
mation, epithelium-to-mesenchyme transition and endoderm speci-
fication in the mouse. Development 135(3):501–511.
Ashery-Padan, R, et  al., 2000. Pax6 activity in the lens primordium is
required for lens formation and for correct placement of a single
retina in the eye. Genes & Development 14(21):2701–2711.
Avilion, AA, et  al., 2003. Multipotent cell lineages in early mouse
development depend on SOX2 function. Genes & Development
17(1):126–140.
Barker, N, et  al., 2010. Lgr5(+ve) stem cells drive self-renewal in the
stomach and build long-lived gastric units in vitro. Cell Stem Cell
6(1):25–36.
Barker, N, et al., 2009. Crypt stem cells as the cells-of-origin of intestinal
cancer. Nature 457(7229):608–612.
Barker, N, et  al., 2012. Lgr5+ve stem/progenitor cells contribute to
nephron formation during kidney development. Cell Reports
2(3):540–552.
Barker, N, et al., 2007. Identification of stem cells in small intestine and
colon by marker gene Lgr5. Nature 449(7165):1003–1007.
Beachy, PA, SS Karhadkar, and DM Berman, 2004. Tissue repair and
stem cell renewal in carcinogenesis. Nature 432(7015):324–331.
Beddington, RS, 1982. An autoradiographic analysis of tissue potency
in different regions of the embryonic ectoderm during gastrulation
in the mouse. Journal of Embryology & Experimental Morphology
69:265–285.
Beerman, I, et al., 2010. Stem cells and the aging hematopoietic system.
Current Opinion in Immunology 22(4):500–506.
Bernstein, BE, et  al., 2006. A bivalent chromatin structure marks key
developmental genes in embryonic stem cells. Cell 125(2):315–326.
S tem C ell G enomics : D evelopmental C ompetence   •   7 6 1
Bernstein, EE, et  al., 2003. Dicer is essential for mouse development.
Nature Genetics 35(3):215–217.
Birney, E, et al., 2007. Identification and analysis of functional elements
in 1% of the human genome by the ENCODE pilot project. Nature
447(7146):799–816.
Blackwood, EM, 1998. Going the distance: a current view of enhancer
action. Science 281(5373):60–63.
Blanpain, C, and BD Simons, 2013. Unravelling stem cell dynamics by
lineage tracing. Molecular Cell Nature Reviews Molecular Cell
Biology14(8):489–502.
Bourillot, P-Y, et al., 2009. Novel STAT3 target genes exert distinct roles
in the inhibition of mesoderm and endoderm differentiation in
cooperation with Nanog. Stem Cells 27(8):1760–1771.
Brons, GM, et  al., 2007. Derivation of pluripotent epiblast stem cells
from mammalian embryos. Nature 448(7150):191–195.
Buckingham, ME, and SM Meilhac, 2011. Tracing cells for tracking cell
lineage and clonal behavior. Developmental Cell 21(3):394–409.
Buczacki, SJA, et al., 2013. Intestinal label-retaining cells are secretory
precursors expressing Lgr5. Nature 495(7439):65–69.
Bulger, M, and M Groudine, 1999. Looping versus linking:  toward a
model for long-distance gene activation. Genes & Development
13(19):2465–2477.
Burdon, T, et al., 1999. Suppression of SHP-2 and ERK signalling pro-
motes self-renewal of mouse embryonic stem cells. Developmental
Biology 210(1):30–43.
Burdon, T, A Smith, and P Savatier, 2002. Signalling, cell cycle and
pluripotency in embryonic stem cells. Trends in Cell Biology
12(9):432–438.
Calo, E, and J Wysocka, 2013. Modification of enhancer chroma-
tin: what, how, and why? Molecular Cell 49(5):825–837.
Chambers, I, et  al., 2003. Functional expression cloning of Nanog,
a pluripotency sustaining factor in embryonic stem cells. Cell
113(5):643–655.
Charge, SBP, and MA Rudnicki, 2004. Cellular and molecu-
lar regulation of muscle regeneration. Physiological Reviews
84(1):209–238.
Chen, CZ, 2009. MicroRNAs modulate haematopoietic lineage differ-
entiation. Science 303(5654):83–86.
Chen, JCJ, and DJ Goldhamer, 2003. Skeletal muscle stem cells.
Reproductive Biology and Endocrinology 1:101.
Cheshier, SH, et  al., 1999. In vivo proliferation and cell cycle kinetics
of long-term self-renewing hematopoietic stem cells. Proceedings of
the National Academy of Science, USA 96(6):3120–3125.
Cirillo, LA, and KS Zaret, 1999. An early developmental transcription
factor complex that is more stable on nucleosome core particles than
on free DNA. Molecular Cell 4(6):961–969.
Cirillo, LA, et  al., 2002. Opening of compacted chromatin by early
developmental transcription factors HNF3 (FoxA) and GATA-4.
Molecular Cell 9(2):279–289.
Clevers, H, 2013. The intestinal crypt, a prototype stem cell compart-
ment. Cell 154(2):274–284.
Clotman, F, et  al., 2002. The onecut transcription factor HNF6 is
required for normal development of the biliary tract. Development
129(8):1819–1828.
Coffinier, C, et  al., 2002. Bile system morphogenesis defects and liver
dysfunction upon targeted deletion of HNF1β. Development
129(8):1829–1838.
Congdon, CC, 1957. Experimental treatment of total-body irradiation
injury: a brief review. Blood 12(8):746–754.
Cosgrove, MS, J D Boeke, and C Wolberger, 2004. Regulated nucleo-
some mobility and the histone code. Nature Structural & Molecular
Biology 11(11):1037–1043.
Creyghton, MP, et  al., 2010. Histone H3K27ac separates active from
poised enhancers and predicts developmental state. Proceedings of
the National Academy of Science, USA 107(50):21931–21936.
Davidson, EH, 2010. Emerging properties of animal gene regulatory net-
works. Nature 468(7326):911–920.
Diwan, SB, and LC Stevens, 1976. Development of teratomas from the
ectoderm of mouse egg cylinders. Journal of the National Cancer
Institute 57(4):937–942.
Do, DV, et al., 2013. A genetic and developmental pathway from STAT3
to the OCT4-NANOG circuit is essential for maintenance of ICM
lineages in vivo. Genes & Development 27(12):1378–1390.
Efroni, S, et  al., 2008. Global transcription in pluripotent embryonic
stem cells. Cell Stem Cell 2(5):437–447.
Enver, T, and M Greaves, 1998. Loops, lineage, and leukemia. Cell
94(1):9–12.
Esau, C, et al., 2004. MicroRNA-143 regulates adipocyte differentiation.
Journal of Biological Chemistry 279(50):52361–52365.
Evans, MJ, and MH Kaufman, 1981. Establishment in culture of pluri-
potential cells from mouse embryos. Nature 292(5819):154–156.
Evsikov, AV, 2003. Comment on “stemness”: transcriptional profiling of
embryonic and adult stem cells” and “a stem cell molecular signa-
ture” (II). Science 302(5644):393c–393.
Fortunel, NO, et al., 2003. Comment on “Stemness”: transcriptional
profiling of embryonic and adult stem cells” and “a stem cell molecu-
lar signature” (I). Science 302(5644):393b–393.
Fujiwara, M, et al., 1994. Uchida rat (rSey): a new mutant rat with cra-
niofacial abnormalities resembling those of the mouse Sey mutant.
Differentiation 57(1):31–38.
Gardner, RL, and J Rossant, 1979a. Investigation of the fate of 4–5 day
post-coitum mouse inner cell mass cells by blastocyst injection.
Journal of Embryology & Experimental Morphology 52:141–152.
Gardner, RL, and J Rossant, 1979b. Investigation of the fate of 4.5 day
post-coitum mouse inner cell mass cells by blastocyst injection.
Journal of Embryology & Experimental Morphology 52:141–152.
Giraldez, AJ, 2011. MicroRNAs regulate brain morphogenesis in zebraf-
ish. Science 308(5723):833–838.
Graf, T, and T Enver, 2009. Forcing cells to change lineages. Nature
462(7273):587–594.
Gross, JG, and JE Morgan, 1999. Muscle precursor cells injected into
irradiated mdx mouse muscle persist after serial injury. Muscle &
Nerve 22(2):174–185.
Gualdi, R, et  al., 1996. Hepatic specification of the gut endoderm
in vitro:  cell signaling and transcriptional control. Genes &
Development 10(13):1670–1682.
Guenther, MG, et  al., 2007. A chromatin landmark and transcription
intiation at most promoters in human cells. Cell 130:77–88.
Hall, PA, and FM Watt, 1989. Stem cells:  the generation and mainte-
nance of cellular diversity. Development 106(4):619–633.
Hanna, JH, K Saha, and R Jaenisch, 2010. Pluripotency and cellu-
lar reprogramming:  facts, hypotheses, unresolved issues. Cell
143(4):508–525.
Hart, AH, et  al., 2004. Identification, cloning and expression analysis
of the pluripotency promoting Nanog genes in mouse and human.
Developmental Dynamics 230(1):187–198.
Hayashi, K, et  al., 2011. Reconstitution of the mouse germ cell
specification pathway in culture by pluripotent stem cells. Cell
146(4):519–532.
Heintzman, ND, et al., 2009. Histone modifications at human enhanc-
ers reflect global cell-type-specific gene expression. Nature
459(7243):108–112.
Heintzman, ND, et  al., 2007. Distinct and predictive chromatin sig-
natures of transcriptional promoters and enhancers in the human
genome. Nature Genetics 39(3):311–318.
Heslop, L, et al., 2001. Transplanted primary neonatal myoblasts can give
rise to functional satellite cells as identified using the Myf5nlacZl
mouse. Gene Therapy 8(10):778–783.
Heyworth, C, et al., 2002. Transcription factor-mediated lineage switch-
ing reveals plasticity in primary committed progenitor cells. EMBO
Journal 21(14):3770–3781.
Hoffman, JA, C-I Wu, and BJ Merrill, 2013a. Tcf7l1 prepares epiblast
cells in the gastrulating mouse embryo for lineage specification.
Development 140(8):1665–1675.
7 6 2   •   G enomics in C linical P ractice
Hoffman, MM, et  al., 2013b. Integrative annotation of chroma-
tin elements from ENCODE data. Nucleic Acids Research
41(2):827–841.
Hu, D, et  al., 2013. The Mll2 branch of the COMPASS family regu-
lates bivalent promoters in mouse embryonic stem cells. Nature
Structural & Molecular Biology. 20:1093–1097; doi:10.1038/
nsmb.2653
Huang, Y, et  al., 2012. In Vivo differentiation potential of epiblast
stem cells revealed by chimeric embryo formation. Cell Reports
2(6):1571–1578.
Huch, M, et al., 2013. In vitro expansion of single Lgr5+ liver stem cells
induced by Wnt-driven regeneration. Nature 494(7436):247–250.
Ikuta, K, and IL Weissman, 1992. Evidence that hematopoietic stem
cells express mouse c-kit but do not depend on steel factor for their
generation. Proceedings of the National Academy of Science, USA
89(4):1502–1506.
Iscove, NN, and K Nawa, 1997. Hematopoietic stem cells expand during
serial transplantation in vivo without apparent exhaustion. Current
Biology: CB 7(10):805–808.
Itzkovitz, S, et al., 2012. Single-molecule transcript counting of stem-cell
markers in the mouse intestine. Nature Cell Biology. 14: 106–114;
doi:10.1038/ncb2384
Ivanova, NB, et al., 2002. A stem cell molecular signature. Science
298(5593):601–604.
Ivanova, N, et al., 2006. Dissecting self-renewal in stem cells with RNA
interference. Nature 442(7102):533–538.
Jaenisch, R, and R Young, 2008. Stem cells, the molecular circuitry of
pluripotency and nuclear reprogramming. Cell 132(4):567–582.
Jaks, V, et al., 2008. Lgr5 marks cycling, yet long-lived, hair follicle stem
cells. Nature Genetics 40(11):1291–1299.
Jenuwein, T, 2001. Translating the histone code. Science
293(5532):1074–1080.
Johnson, DS, et al., 2007. Genome-wide mapping of in vivo protein-DNA
interactions. Science 316:1497.
Jung, J, et al., 1999. Initiation of mammalian liver development from endo-
derm by fibroblast growth factors. Science 284(5422):1998–2003.
Kawamoto, H, et  al., 2010. A map for lineage restriction of progeni-
tors during hematopoiesis: the essence of the myeloid-based model.
Immunological Reviews 238(1):23–36.
Kiel, MJ, et al., 2005. SLAM family receptors distinguish hematopoietic
stem and progenitor cells and reveal endothelial niches for stem
cells. Cell 121(7):1109–1121.
Kim, TH, et al., 2005. A high-resolution map of active promoters in the
human genome. Nature 436(7052):876–880.
Kondo, M, IL Weissman, and K Akashi, 1997. Identification of clono-
genic common lymphoid progenitors in mouse bone marrow. Cell
91(5):661–672.
Kopp, JL, et  al., 2011. Progenitor cell domains in the developing and
adult pancreas. Cell Cycle 10(12):1921–1927.
Kopp, JL, et  al., 2008. Small increases in the level of Sox2 trigger
the differentiation of mouse embryonic stem cells. Stem Cells
26(4):903–911.
Kulessa, H, J Frampton, and T Graf, 1995. GATA-1 reprograms avian
myelomonocytic cell lines into eosinophils, thromboblasts, and
erythroblasts. Genes & Development 9(10):1250–1262.
Lagha, M, et al., 2013. Paused Pol ii coordinates tissue morphogenesis in
the Drosophila embryo. Cell 153(5):976–987.
Lawson, KA, JJ Meneses, and RA Pedersen, 1991. Clonal analysis of
epiblast fate during germ layer formation in the mouse embryo.
Development 113(3):891–911.
Levak-Svajger, B, and A Svajger, 1974. Investigation on the origin of the
definitive endoderm in the rat embryo. Journal of Embryology &
Experimental Morphology 32(2):445–459.
Lewis, JP, and FE Trobaugh, 1964. Haematopoietic stem cells. Nature
204:589–590.
Li, Z, et al., 2012. Foxa2 and H2A.Z mediate nucleosome depletion dur-
ing embryonic stem cell differentiation. Cell 151(7):1608–1616.
Lobo, NA, et al., 2007. The biology of cancer stem cells. Annual Review
of Cell & Developmental Biology 23(1):675–699.
Loh and Ang et al. 2014. Efficient endoderm induction from human plu-
ripotent stem cells by logically directing signals controlling lineage
bifurcations. Cell Stem Cell
Loh, KM, and B Lim, 2011. A precarious balance: pluripotency factors
as lineage specifiers. Cell Stem Cell 8:363–369.
Lu, R, A Yang, and Y Jin, 2011. Dual functions of T-Box 3 (Tbx3) in the
control of self-renewal and extraembryonic endoderm differentia-
tion in mouse embryonic stem cells. Journal of Biological Chemistry
286(10):8425–8436.
Lüdtke, TH-W, et al., 2009. Tbx3 promotes liver bud expansion during
mouse development by suppression of cholangiocyte differentia-
tion. Hepatology 49(3):969–978.
Ludwig, TE, et al., 2006. Derivation of human embryonic stem cells in
defined conditions. Nature Biotechnology 24(2):185–187.
Ma, Z, et  al., 2011. Sequence-specific regulator Prdm14 safeguards
mouse ESCs from entering extraembryonic endoderm fates. Nature
Structural & Molecular Biology. 18(2):120–U175.
Marks, H, et al., 2012. The transcriptional and epigenomic foundations
of ground state pluripotency. Cell 149(3):590–604.
Martello, G, et al., 2007. MicroRNA control of Nodal signalling. Nature
449(7159):183–188.
Martin, GR, 1981. Isolation of a pluripotent cell line from early mouse
embryos cultured in medium conditioned by teratocarcinoma
stem cells. Proceedings of the National Academy of Sciences of the
United States of America 78(12):7634–7638.
Masui, S, et  al.2007 Pluripotency governed by Sox2 via regulation of
Oct3/4 expression in mouse embryonic stem cells. Nature Cell
Biology 9(6):625–635.
Mauro, A, 1961. Satellite cell of skeletal muscle fibers. The Journal of
Biophysical & Biochemical Cytology 9:493–495.
Meshorer, E, et  al., 2006. Hyperdynamic plasticity of chromatin pro-
teins in pluripotent embryonic stem cells. Developmental Cell
10(1):105–116.
Mesnard, D, M Guzman-Ayala, and DB Constam, 2006. Nodal
specifies embryonic visceral endoderm and sustains pluripotent
cells in the epiblast before overt axial patterning. Development
133(13):2497–2505.
Mihaly, J, et  al., 1998. Chromatin domain boundaries in the Bithorax
complex. Cellular and molecular life sciences: Cellular & Molecular
Life Sciences 54(1):60–70.
Mihaly, J, et  al., 1997. In situ dissection of the Fab-7 region of the
bithorax complex into a chromatin domain boundary and a
Polycomb-response element. Development 124(9):1809–1820.
Min, IM, et  al., 2011. Regulating RNA polymerase pausing and tran-
scription elongation in embryonic stem cells. Genes & Development
25(7):742–754.
Miyoshi, N, et al., 2011. Reprogramming of mouse and human cells to plu-
ripotency using mature microRNAs. Cell Stem Cell 8(6):633–638.
Montgomery, RK, et  al., 2011. Mouse telomerase reverse transcrip-
tase (mTert) expression marks slowly cycling intestinal stem cells.
Proceedings of the National Academy of Sciences of the United
States of America 108(1):179–184.
Morgan, JE, et al., 1994. Myogenic cell lines derived from transgenic mice
carrying a thermolabile T antigen: a model system for the derivation
of tissue-specific and mutation-specific cell lines. Developmental
Biology 162(2):486–498.
Morrison, SJ, et al., 1995. The purification and characterization of fetal
liver hematopoietic stem cells. Proceedings of the National Academy
of Sciences, USA 92(22):10302–10306.
Morrison, SJ, and IL Weissman, 1994. The long-term repopulating subset
of hematopoietic stem cells is deterministic and isolatable by pheno-
type. Immunity 1(8):661–673.
Mossman, AK, et al., 2005. Mixl1 and oct4 proteins are transiently coex-
pressed in differentiating mouse and human embryonic stem cells.
Stem Cells & Development 14(6):656–663.
S tem C ell G enomics : D evelopmental C ompetence   •   7 6 3
Muller-Sieburg, CE, CA Whitlock, and IL Weissman, 1986. Isolation
of two early B lymphocyte progenitors from mouse marrow: a com-
mitted pre-pre-B cell and a clonogenic Thy-1-lo hematopoietic stem
cell. Cell 44(4):653–662.
Muñoz, J, et  al., 2012. The Lgr5 intestinal stem cell signature:  robust
expression of proposed quiescent “+4” cell markers. EMBO Journal
31(14):3079–3091.
Nagy, A, et  al., 1993. Derivation of completely cell culture-derived
mice from early-passage embryonic stem cells. Proceedings of the
National Academy of Sciences, USA 90(18):8424–8428.
Nakagawa, M, et  al., 2007. Generation of induced pluripotent stem
cells without Myc from mouse and human fibroblasts. Nature
Biotechnology 26(1):101–106.
Nerlov, C, and T Graf, 1998. PU.1 induces myeloid lineage commitment
in multipotent hematopoietic progenitors. Genes & Development
12(15):2403–2412.
Nerlov, C, et al., 2000. GATA-1 interacts with the myeloid PU.1 tran-
scription factor and represses PU.1-dependent transcription. Blood
95(8):2543–2551.
Niakan, KK, et  al., 2006. Novel role for the orphan nuclear receptor
Dax1 in embryogenesis, different from steroidogenesis. Molecular
Genetics & Metabolism 88(3):261–271.
Nichols, J, et  al., 2001. Physiological rationale for responsiveness of
mouse embryonic stem cells to gp130 cytokines. Development
128(12):2333–2339.
Nichols, J, et al., 1998. Formation of pluripotent stem cells in the mam-
malian embryo depends on the POU transcription factor Oct4.
Cell 95(3):379–391.
Nichols, J, et  al., 2009. Suppression of Erk signalling promotes
ground state pluripotency in the mouse embryo. Development
136(19):3215–3222.
Niwa, H, et  al., 1998. Self-renewal of pluripotent embryonic stem
cells is mediated via activation of STAT3. Genes & Development
12(13):2048–2060.
Niwa, H, J Miyazaki, and AG Smith, 2000. Quantitative expression of
Oct-3/4 defines differentiation, dedifferentiation or self-renewal of
ES cells. Nature Genetics 24(4):372–376.
Notta, F, et  al., 2011. Isolation of single human hematopoietic stem
cells capable of long-term multilineage engraftment. Science
333(6039):218–221.
Obara-Ishihara, T, et  al., 1999. The surface ectoderm is essential for
nephric duct formation in intermediate mesoderm. Development
126(6):1103–1108.
Oikawa, T, et al., 2009. Sall4 regulates cell fate decision in fetal hepatic
stem/progenitor cells. Gastroenterology 136(3):1000–1011.
Olins, DE, and AL Olins, 2003. Chromatin history: our view from the
bridge. Nature Reviews Molecular Cell Biology 4(10):809–814.
Osawa, M, et al., 1996. Long-term lymphohematopoietic reconstitution
by a single CD34-low/negative hematopoietic stem cell. Science
273(5272):242–245.
Ostuni, R, et al., 2013. Latent enhancers activated by stimulation in dif-
ferentiated cells. Cell 152(1–2):157–171.
Pan, FC, and C Wright, 2011. Pancreas organogenesis:  from bud to
plexus to gland. Developmental Dynamics 240(3):530–565.
Pan, GJ, et  al., 2002. Stem cell pluripotency and transcription factor
Oct4. Cell Research 12(5–6):321–329.
Pan, G, et al., 2007. Whole-genome analysis of histone H3 lysine 4 and
lysine 27 methylation in human embryonic stem cells. Cell Stem
Cell 1(3):299–312.
Park, Y, and S L Gerson, 2005. DNA repair defects in stem cell func-
tion and aging*. Annual Review of Medicine Annual Review of
Medicine 56(1):495–508.
Pekowska, A, et al., 2011. H3K4 tri-methylation provides an epigenetic sig-
nature of active enhancers. The EMBO Journal 30(20):4198–4210.
Perea-Gomez, A, et al., 2002. Nodal antagonists in the anterior visceral
endoderm prevent the formation of multiple primitive streaks.
Developmental Cell 3(5):745–756.
Piña, B, U Brüggemeier, and M Beato, 1990. Nucleosome positioning
modulates accessibility of regulatory proteins to the mouse mam-
mary tumor virus promoter. Cell 60(5):719–731.
Poy, MN, et al., 2009. miR-375 maintains normal pancreatic alpha- and
beta-cell mass. Proceedings of the National Academy of Sciences of
the United States of America 106(14):5813–5818.
Preger-Ben Noon, E, et  al., 2009. Interplay between activin and Hox
genes determines the formation of the kidney morphogenetic field.
Development 136(12):1995–2004.
Rada-Iglesias, A, et al., 2010. A unique chromatin signature uncovers early
developmental enhancers in humans. Nature 470(7333):279–283.
Rada-Iglesias, A, et al., 2012. Epigenomic annotation of enhancers pre-
dicts transcriptional regulators of human neural crest. Cell Stem
Cell 11(5):633–648.
Ramalho-Santos, M, et al., 2002. “Stemness”: transcriptional profiling of
embryonic and adult stem cells. Science 298(5593):597–600.
Roberts, DJ, et al., 1995. Sonic hedgehog is an endodermal signal induc-
ing Bmp-4 and Hox genes during induction and regionalization of
the chick hindgut. Development 121(10):3163–3174.
Roberts, DJ, et  al., 1998. Epithelial-mesenchymal signaling during the
regionalization of the chick gut. Development 125(15):2791–2801.
Rosa, A, FM Spagnoli, and AH Brivanlou, 2009. The miR-430/427/302
family controls mesendodermal fate specification via species-specific
target selection. Developmental Cell 16(4):517–527.
Rossi, DJ, CHM Jamieson, and IL Weissman, 2008. Stem cells and the
pathways to aging and cancer. Cell 132(4):681–696.
Rossi, JM, et  al., 2001. Distinct mesodermal signals, including BMPs
from the septum transversum mesenchyme, are required in combina-
tion for hepatogenesis from the endoderm. Genes & Development
15(15):1998–2009.
Rothenberg, EV, 2011. T cell lineage commitment: identity and renun-
ciation. Journal of Immunology 186(12):6649–6655.
Samstein, RM, et  al., 2012. Foxp3 exploits a pre-existent enhancer
landscape for regulatory T cell lineage specification. Cell
151(1):153–166.
Sangiorgi, E, and MR Capecchi, 2008. Bmi1 is expressed in vivo in intes-
tinal stem cells. Nature Genetics 40(7):915–920.
Sato, T, and H Clevers, 2013. Growing self-organizing mini-guts from
a single intestinal stem cell:  mechanism and applications. Science
(New York, NY) 340(6137):1190–1194.
Sato, T, et al., 2011. Paneth cells constitute the niche for Lgr5 stem cells
in intestinal crypts. Nature 469(7330):415–418.
Sato, T, et  al., 2009. Single Lgr5 stem cells build crypt-villus
structures in vitro without a mesenchymal niche. Nature
459(7244):262–265.
Scadden, DT, 2006. The stem-cell niche as an entity of action. Nature
441(7097):1075–1079.
Schaffer, AE, et al., 2010. Nkx6 transcription factors and Ptf1a function
as antagonistic lineage determinants in multipotent pancreatic pro-
genitors. Developmental Cell 18(6):1022–1029.
Schepers, AG, et al., 2012. Lineage tracing reveals Lgr5+ stem cell activ-
ity in mouse intestinal adenomas. Science 337(6095):730–735.
Schofield, R, 1978. The relationship between the spleen colony-forming
cell and the haemopoietic stem cell. Blood cells 4(1–2):7–25.
Schultz, E, 1996. Satellite cell proliferative compartments in growing
skeletal muscles. Developmental Biology 175(1):84–94.
Seita, J, et  al., 2012. Gene expression commons:  an open platform for
absolute gene expression profiling. PLoS ONE 7(7):e40321.
Shin, D, et al., 2011a. Restriction of hepatic competence by Fgf signaling.
Development 138(7):1339–1348.
Shin, M, et  al., 2011b. Activin/TGF-beta signaling regulates Nanog
expression in the epiblast during gastrulation. Mechanisms of
Development 128(5–6):268–278.
Silva, J, and A Smith, 2008. Capturing pluripotency. Cell 132(4):532–536.
Skreb, N, D Solter, and I Damjanov, 1991. Developmental biology of
the murine egg cylinder. International Journal of Developmental
Biology 35(3):161–176.
7 6 4   •   G enomics in C linical P ractice
Smith, AG, et  al., 1988. Inhibition of pluripotential embryonic
stem cell differentiation by purified polypeptides. Nature
336(6200):688–690.
Smith, A, 2006. A glossary for stem-cell biology. Nature 441:1060.
Smith, AG, 2001. Embryo-derived stem cells: of mice and men. Annual
Review of Cell & Developmental Biology 17(1):435–462.
Smith, J, and PN Schofield, 1997. Stable integration of an mdx skeletal mus-
cle cell line into dystrophic (mdx) skeletal muscle: evidence for stem cell
status. Cell Growth & Differentiation: The Molecular Biology Journal
of the American Association for Cancer Research 8(8):927–934.
Snippert, HJ, et al., 2010. Intestinal crypt homeostasis results from neu-
tral competition between symmetrically dividing Lgr5 stem cells.
Cell 143(1):134–144.
Snow, MH, 1975. Embryonic development of tetraploid mice during the
second half of gestation. Journal of Embryology & Experimental
Morphology 34(3):707–721.
Snow, MH, 1976. The immediate post-implantation development of tet-
raploid mouse blastocysts. Journal of Embryology & Experimental
Morphology 35(1):81–86.
Soufi, A, G Donahue, and KS Zaret, 2012. Facilitators and impediments
of the pluripotency reprogramming factors’ initial engagement with
the genome. Cell 151(5):994–1004.
Spangrude, GJ, S Heimfeld, and IL Weissman, 1988a. Purification
and characterization of mouse hematopoietic stem cells. Science
241(4861):58–62.
Stanger, BZ, AJ Tanaka, and DA Melton, 2007. Organ size is limited by
the number of embryonic progenitor cells in the pancreas but not
the liver. Nature 445(7130):886–891.
Strahl, BD, and CD Allis, 2000. The language of covalent histone modi-
fications. Nature 403(6765):41–45.
Sumi, T, et  al., 2013. Epiblast ground state is controlled by canonical
Wnt/β-catenin signaling in the post-implantation mouse embryo
and epiblast stem cells. PLoS One 8(5):e63378.
Sun, C, et  al., 2009. Dax1 binds to Oct3/4 and inhibits its transcrip-
tional activity in embryonic stem cells. Molecular and Cellular
Biology 29(16):4574–4583.
Suzuki, A, et  al., 2008. Tbx3 controls the fate of hepatic progeni-
tor cells in liver development by suppressing p19ARF expression.
Development 135(9):1589–1595.
Takahashi, K, and S Yamanaka, 2006. Induction of pluripotent stem cells
from mouse embryonic and adult fibroblast cultures by defined fac-
tors. Cell 126(4):663–676.
Takeda, N, et  al., 2011. Interconversion between intestinal stem
cell populations in distinct niches. Science (New  York, NY)
334(6061):1420–1424.
Tam, PP, and RS Beddington, 1987. The formation of mesodermal tis-
sues in the mouse embryo during gastrulation and early organogen-
esis. Development 99(1):109–126.
Tarkowski, AK, A Witkowska, and J Opas, 1977. Development of
cytochalasin in B-induced tetraploid and diploid/tetraploid
mosaic mouse embryos. Journal of Embryology & Experimental
Morphology 41:47–64.
Tay, YMS, et al., 2008a. MicroRNA-134 modulates the differentiation
of mouse embryonic stem cells, where it causes post-transcriptional
attenuation of Nanog and LRH1. Stem Cells 26(1):17–29.
Tay, YMS, et  al., 2008b. MicroRNAs to Nanog, Oct4 and Sox2 cod-
ing regions modulate embryonic stem cell differentiation. Nature
455(7216):1124–1128.
ten Berge, D, et al., 2011. Embryonic stem cells require Wnt proteins to
prevent differentiation to epiblast stem cells. Nature Cell Biology
13(9):1070–1075.
Teo, AKK, et  al., 2011. Pluripotency factors regulate definitive endo-
derm specification through eomesodermin. Genes & Development
25(3):238–250.
Terskikh, AV, et  al., 2001. From hematopoiesis to neuropoiesis:  evi-
dence of overlapping genetic programs. Proceedings of the National
Academy of Sciences 98(14):7934–7939.
Tesar, PJ, et  al., 2007. New cell lines from mouse epiblast share
defining features with human embryonic stem cells. Nature
448(7150):196–199.
Thomas, ED, et al., 1959. Supralethal whole body irradiation and isol-
ogous marrow transplantation in man. International Journal of
Clinical Investigation 38:1709–1716.
Thomas, ED, et  al., 1957. Intravenous infusion of bone marrow in
patients receiving radiation and chemotherapy. New England
Journal of Medicine 257(11):491–496.
Thomson, JA, et al. 1998. Embryonic stem cell lines derived from human
blastocysts. Science 282(5391):1145–1147.
Tian, H, et al., 2011. A reserve stem cell population in small intestine ren-
ders Lgr5-positive cells dispensable. Nature 478(7368):255–259.
Uchida, N, and IL Weissman,1992. Searching for hematopoietic stem
cells:  evidence that Thy-1.1lo Lin-Sca-1+ cells are the only stem
cells in C57BL/Ka-Thy-1.1 bone marrow. Journal of Experimental
Medicine175(1):175–184.
Udvardy, A, E Maine, and P Schedl, 1985. The 87A7 chromomere.
Identification of novel chromatin structures flanking the heat shock
locus that may define the boundaries of higher order domains.
Journal of Molecular Biology 185(2):341–358.
Urso, P, and CC Congdon, 1957. The effect of the amount of isologous
bone marrow injected on the recovery of hematopoietic organs, sur-
vival and body weight after lethal irradiation injury in mice. Blood
12(3):251–260.
Valenzuela, L, and RT Kamakaka, 2006. Chromatin insulators. Annual
Review of Genetics 40(1):107–138.
Vallier, L, et al., 2009. Activin/Nodal signalling maintains pluripotency
by controlling Nanog expression. Development 136(8):1339–1349.
van Es, JH, et al., 2012. Dll1+ secretory progenitor cells revert to stem
cells upon crypt damage. Nature Cell Biology 14(10):1099–1104.
Waddington, CH, 1940. Organisers and Genes. Cambridge,
UK: Cambridge University Press.
Waghmare, SK, et  al., 2008. Quantitative proliferation dynamics and
random chromosome segregation of hair follicle stem cells. The
EMBO Journal 27(9):1309–1320.
Wamstad, JA, et  al., 2012. Dynamic and coordinated epigenetic regu-
lation of developmental transitions in the cardiac lineage. Cell
151(1):206–220.
Wang, JCY, and JE Dick, 2005. Cancer stem cells: lessons from leukemia.
Trends in Cell Biology 15(9):494–501.
Wang, Y, et  al., 2007. DGCR8 is essential for microRNA biogenesis
and silencing of embryonic stem cell self-renewal. Nature Genetics
39(3):380–385.
Wang, Z, et  al., 2012. Distinct lineage specification roles for Nanog,
OCT4, and SOX2 in human embryonic stem cells. Cell Stem Cell
10(4):440–454.
Weissman, IL, 2000. Stem cells: units of development, units of regenera-
tion, and units in evolution. Cell 100(1):157–168.
Weissman, I, V Papaioannou, and R Gardner, 1978. Fetal hematopoietic
origins of the adult hematolymphoid system. In Differentiation of
Normal and Neoplastic Hematopoietic Cells. Clarkson B, Marks
PA, and Till JE, eds., pp. 33–47. New  York:  Cold Spring Harbor
Laboratory.
S tem C ell G enomics : D evelopmental C ompetence   •   7 6 5
Weissman, IL, 2002. The road ended up at stem cells. Immunological
Reviews 185:159–174.
Williams, RL, et  al., 1988. Myeloid leukaemia inhibitory factor main-
tains the developmental potential of embryonic stem cells. Nature
336(6200):684–687.
Wilson, A, et  al., 2008. Hematopoietic stem cells reversibly switch
from dormancy to self-renewal during homeostasis and repair. Cell
135(6):1118–1129.
Wolff, E, 1968. Specific interactions between tissues during organogen-
esis. Current Topics in Developmental Biology 3:65–94.
Wray, J, T Kalkan, and AG Smith, 2010. The ground state of pluripo-
tency. Biochemical Society Transactions 38(4):1027–1032.
Xie, R, et al., 2013. Dynamic chromatin remodeling mediated by poly-
comb proteins orchestrates pancreatic differentiation of human
embryonic stem cells. Cell Stem Cell 12(2):224–237.
Xu, R-H, et  al., 2008. NANOG is a direct target of TGFβ/
activin-mediated SMAD signaling in human ESCs. Cell Stem Cell
3(2):196–206.
Yao, S-N, and K Kurachi, 1993. Implanted myoblasts not only fuse with
myofibers but also survive as muscle precursor cells. Journal of Cell
Science 105(4):957–963.
Yilmaz, OH, et al. 2005. SLAM family markers are conserved among
hematopoietic stem cells from old and reconstituted mice and mark-
edly increase their purity. Blood 107(3):924–930.
Ying, Q-L, et  al., 2008. The ground state of embryonic stem cell
self-renewal. Nature 453(7194):519–523.
Young, RA, 2011. Control of the embryonic stem cell state. Cell
144(6):940–954.
Yu, P, et al., 2011. FGF2 sustains Nanog and switches the outcome of
BMP4-induced human embryonic stem cell differentiation. Cell
Stem Cell 8(3):326–334.
Zaret, K, 1999. Developmental competence of the gut endo-
derm:  genetic potentiation by GATA and HNF3/fork head pro-
teins. Developmental Biology 209(1):1–10.
Zaret, KS, et al., 2008. Pioneer factors, genetic competence, and induc-
tive signaling: programming liver and pancreas progenitors from the
endoderm. Cold Spring Harbor Symposia on Quantitative Biology
73:119–126.
Zeitlinger, J, et al., 2007. RNA polymerase stalling at developmental con-
trol genes in the Drosophila melanogaster embryo. Nature Genetics
39(12):1512–1516.
Zhang, J, et al., 2006. Sall4 modulates embryonic stem cell pluripotency
and early embryonic development by the transcriptional regulation
of Pou5f1. Nature Cell Biology 8(10):1114–1123.
Zhang, K, et  al., 2010a. Distinct functions of BMP4 during differ-
ent stages of mouse ES cell neural commitment. Development
137(13):2095–2105.
Zhang, X, et al., 2010b. Pax6 is a human neuroectoderm cell fate deter-
minant. Cell Stem Cell 7(1):90–100.
Zhu, J, et al., 2013. Genome-wide chromatin state transitions associated
with developmental and environmental cues. Cell 152(3):642–654.
Zhu, Y, D van Essen, and S Saccani, 2012. Cell-type-specific control
of enhancer activity by H3K9 trimethylation. Molecular Cell
46(4):408–423.
 48.
GENOMIC APPLICATIONS IN CRITICAL
CARE MEDICINE
Matthew C. Frise, Charles Hinds, and Julian C. Knight
INTRODUCTION
The development of intensive care units (ICUs) has been
driven by the need to support and effectively treat the most
seriously ill individuals within the hospital population, in
particular those with impairment or failure of vital organs.
There are remarkably few specific effective therapies for
such patients, and their recovery depends primarily on sup-
portive care while their physiological and biochemical equi-
librium is restored. Although skills in supporting failing
organ systems have advanced significantly in recent decades,
mortality rates, particularly from severe infection, remain
disappointingly high. Furthermore, the morbidity associ-
ated with prolonged organ support is considerable, and
many patients who survive do so with significant long-term
sequelae.1
A general ICU will admit patients with a broad range
of conditions, but the most important single cause of death,
irrespective of admission diagnosis, remains sepsis. The
underlying infection may have been the primary reason for
hospital admission or may have been acquired in the hospi-
tal or ICU. Most patients who acquire an infection in the
community or hospital, such as community acquired pneu-
monia (CAP), will not become critically ill. In some cases,
recognizable factors such as significant co-morbidity or a
particularly virulent pathogen lead to greater illness sever-
ity. Often, however, the individuals who fare worse are out-
wardly no different from the vast majority of those infected,
leading clinicians to suspect that genetic factors must play
an important role. Genetic influence on infectious disease
in general is discussed elsewhere (see Chapters  11  & 37).
Particular questions relevant to the critically ill include,
firstly, whether identifiable genetic factors lead to an
increased likelihood of requiring ICU admission, which
might permit early identification and treatment of high-risk
individuals. Secondly, once admitted to ICU, do genetic fac-
tors predict a greater likelihood of developing a particular
766
complication such as acute respiratory distress syndrome
(ARDS) or septic shock, such that therapy could be tailored
to be of maximum benefit? Thirdly, does genomic medicine
help us assess how patients might respond or are responding
to a given treatment, so that interventions can be individu-
alized and modified appropriately? If the answer to any of
these questions is yes, then there is the potential to signifi-
cantly reduce morbidity, mortality, and costs.
S E P S I S : A N I M MU N O L O G I C A L A N D
CLINICAL CHALLENGE
Microbial invasion presents a challenge for the innate
immune system:  the host must initiate an immediate,
aggressive response to limit and eradicate the pathogen, but
at the same time avoid deleterious effects such as circulatory
failure, vital organ dysfunction, and immune paresis. When
disseminated, this inflammatory reaction is defined clini-
cally by consensus2 as the systemic inflammatory response
syndrome (SIRS). As well as infection, SIRS can be precipi-
tated by a range of non-infectious insults including burns,
severe pancreatitis, and tissue injury from trauma or surgery
(Figure 48.1). There are, however, important immuno-
pathophysiological differences between sepsis and SIRS in
the absence of infection.
Sepsis is increasingly common,3 accounting for 27%
of adult ICU admissions in the United Kingdom.4 When
severe, mortality rates remain unacceptably high, at up to
50%,5 and severe sepsis now ranks as the second leading
cause of death, with 36,800 people estimated to die per
annum in the United Kingdom at a cost to the National
Health Service (NHS) of £1.5 billion.6
Sepsis is the epitome of a complex polygenic disorder.
Our understanding of human sepsis is at present undoubt-
edly inadequate,7 and it had been hoped, perhaps naïvely,
that outcome could be improved by targeting just one
component of the innate immune response. Unfortunately,
 Local
Sterile causes SIRS
– Fever
– Tachycardia
Trigger
– Tachypnea
– Leucocytosis
Infection Sepsis
(proven or suspected)
Figure 48.1 
Relationship between local and systemic inflammatory responses, the systemic inflammatory response syndrome (SIRS) and sepsis which
may progress to severe sepsis and septic shock. Reproduced with permission from Dejager et al. 2011.114
such interventions have usually been applied unselectively
to heterogeneous patient groups, without considering the
potential influence of their genetic diversity on response.
Sepsis has been described as the “graveyard” of pharmaceu-
tical discovery8 because most initially promising drugs have
proved ineffective.
G E N ET I C VA R I AT I O N, D I S E A S E
S US C E P T I B I L IT Y, A N D O U TC O M E
Most genomic research in critical illness has focused on
analysis of disease susceptibility or outcome based on asso-
ciation with polymorphisms of single “candidate” genes.
In addition to the inherent problems of adopting such an
approach discussed elsewhere in this book (see Chapter 1),
a major challenge remains the heterogeneity of the patient
population, making it very difficult to define and study a
clear disease phenotype. This is likely to confound analy-
sis and accounts in part for the often conflicting and
inconclusive literature in this field.9–11 The application
of more recently developed genomic approaches, includ-
ing genome-wide association studies (GWAS), may prove
to be more informative. Expression quantitative trait loci
(eQTL) analysis12 may also ameliorate some of these prob-
lems by viewing expression of a particular mRNA or protein
as the outcome of interest rather than a complex phenotypi-
cal characteristic such as the development of ARDS fol-
lowing traumatic injury. The value of functional genomic
techniques based on high-throughput sequencing for
large-scale analysis of multiple genes including their tran-
scription, translation, and epigenetic factors such as histone
acetylation and chromatin remodeling, is highlighted by
recent work from the ENCODE project.13 In the context of
sepsis and critical illness, these techniques may help resolve
G e no m ic A p p l ication s in C ritica l C ar e M e dicin e   •   7 6 7
Systemic
Severe
Shock
SIRS
Organ failure Hypotension
Severe Septic
sepsis shock
novel biomarkers and disease mechanism. Our understand-
ing is therefore evolving from considering just single genes
and single-nucleotide polymorphisms (SNPs) into a much
more global and integrated picture that resolves DNA
sequence variation, the transcriptome, and the proteome in
the context of a specific disease such as sepsis, pneumonia
or peritonitis.
ET H I C A L C O N S I D E R AT I O N S
For an area of healthcare so resource- and cost-intensive,
and for many patients regrettably ultimately futile, it is
no surprise that there should be considerable debate sur-
rounding the ethics of collecting and using genomic data.14
Genetic studies of critically ill patients raise particular issues
not encountered elsewhere, including the issue of con-
sent, as typically patients will lack capacity.15 The value of
genomic approaches is controversial, but the potential ben-
efits to this group of patients should not be underestimated.
Moreover, the opportunities presented by the depth of phe-
notypical data that can be obtained through already estab-
lished electronic data-capture and management systems in
many ICUs illustrate the scope and complexity that may be
possible in the future.16
G E N ET I C A SS O C I AT I O N S I N T H E
C R I T I C A LLY I LL :   T H E S TO RY S O FA R
A comprehensive review of gene association studies in infec-
tious diseases in general, and indeed in critical illness, is out-
side the scope of this chapter. What follows aims to give an
overview of the field in recent years, illustrating some of the
challenges encountered. The disproportionate number of
 LPS and other
bacterial components

Endothelium Neutrophils Monocytes

Cytokines

Oxygen radicals Lipid mediators

Complement
Increased
TF and PAI-1
Chemotaxis
Procoagulant effect Lysosomal enzymes

Microvascular occlusion Vascular instability

Coagulopathy Fever Vasodilation Capillary leak

Sepsis and multiple organ failure

Figure 48.2 
The major pathways and inflammatory mediators of sepsis. Lipopolysaccharide (LPS); plasminogen-activator inhibitor-1 (PAI-1); tissue
factor (TF). Reproduced with permission from Cohen 2002.115

studies related to sepsis as opposed to other causes of critical
illness, for the reasons already discussed, will be obvious. An
overview of the major pathways and inflammatory media-
tors of sepsis is given (Figure 48.2) to provide context for
the genetic findings presented below.
P RO -I N FL A M M ATO RY M E D I ATO R S
IFN-γ promotes TNF production; studies suggest an
effect of the intron 1 (CA)n repeat of IFNG on sepsis
risk after traumatic injury,19 and intronic SNP rs2430561
(IFNG+874T/A) on infectious complications following
surgery for esophageal cancer.20 These two polymorphisms
are in linkage disequilibrium.

Tumor Necrosis Factor Alpha (TNF) and Interferon
Gamma (IFN-γ)
The TNF locus is highly polymorphic, with many SNPs
and microsatellites having been identified and investi-
gated using a candidate-gene approach. Studies of the
promoter SNP rs1800629 (TNF-308G/A) have reported
conflicting results (Table 48.1). A meta-analysis suggested
an association with the development of sepsis, but not
mortality.17 In contrast, a U.S.  study of over a thousand
patients from 12 ICUs found a significant association
with mortality and duration of mechanical ventilation.18
Interleukins (IL-1, IL-1ra, IL-6, IL-8)
Of three IL-1 genes, two encode the pro-inflammatory
cytokines IL-1α and IL-1β; a third produces the naturally
occurring inhibitor IL-1ra. Within IL-1ra intron 2, there
is a variable number tandem repeat (VNTR) that has
been extensively studied in sepsis (Table 48.1), as has an
IL-6 promoter SNP rs1800795 (IL6-174G/C). The SNP
rs2814778 (DARC-46T/C) in the gene encoding Duffy
antigen/receptor for chemokines is associated with worse
clinical outcomes among African Americans with acute
lung injury (ALI), perhaps related to increased IL-8 levels.21

7 6 8   •   G e no m ic s in C l inica l Practic e
Table 48.1 SUMMARY OF SELECTED PRO-INFLAMMATORY MEDIATOR GENE ASSOCIATION STUDIES

STUDY, YEAR , GENE VARIANT OUTCOME POPULATION AND ODDS RATIO NOTES
PMID SAMPLE SIZE (N) (95% CI)
TNF
Stuber TNF -308G/A Development of Severe sepsis (80) Non signifi-
1995 (rs1800629) severe sepsis Healthy controls (153) cant (NS)
8832971
Mira rs1800629 ICU mortality Septic shock (89) 3.7
1999 Healthy blood donors (87) (1.4–10.2)
10450718
Apploni rs1800629 ICU mortality Septic shock (37) 7.4* Excluded 3 patients dying in
2001 (1.2–44.2) first 24 hours; small study
11331061
O’Keefe rs1800629 Severe sepsis Admitted to trauma center 4.6 No association with mortality
2002 (152) (1.9 –10.9)
11988644
Gordon rs1800629 Sepsis Caucasians from eight NS Also examined receptor
2004 Illness severity U.K. and Australian ICUs genes, TNFRSF1A and B
15526005 Outcome (213)
Gill rs1800629 Microchimerism Transfused U.S. trauma 5.7* Immunological consequences
2008 following blood patients (59) (1.2–27.3) uncertain
18199828 transfusion
Shalhub rs1800629 Hospital mortality Severe burns admitted 10.7 Largely accounted for by
2009 U.S. regional center (69) (1.2–95.5) septic shock and multi-organ
19060757 dysfunction (MODS)
Paskulin rs1800629 Sepsis, septic shock Critically ill Caucasian NS
2011 Organ dysfunction general ICU patients from
21670923 Mortality southern Brazil (520)
IL1RN
Arnalich IL-1RN*2 (A2 allele) 30-day mortality Severe sepsis (78) 6.47 Associated with lower PBMC
2002 homozygosity (1.0–41.5) IL-1ra production in vitro
11876758 (VNTR intron 2)
Patwari Absence of Lung injury severity U.S. children of mixed eth- 2.65 Suggestion also of association
2008 IL-1RN*1 (A1 allele) nicity with CAP (847) (1.0– 6.9) with ARDS development
18838927
Hsu IL-1RN*2 Acinetobacter bau- Taiwanese with pneumonia 0.118 Apparent gene dosage effect
2012 homozygosity mannii pneumonia (66) (0.02–0.76) of A1 allele on risk
22558011
IL6
Burzotta IL-6 -174G/C Postoperative death, Italians undergoing elec- 0.24* GG homozygotes had longer
2001 (rs1800795) myocardial infarc- tive coronary artery bypass (0.03–2.2) NS ICU and hospital stays
11703956 tion, or stroke grafting (CABG) (111)
Baier rs1800795 Late bloodstream African-American mechan- 2.6
2006 infections ically ventilated very low (1.1–6.0)
16611358 birthweight infants (233)
Tischendorf rs1800795 Shock German Caucasians with 2.83* -174C conferred low serum
2007 severe sepsis and septic (1.2–7.0) IL6 but high in vitro secre-
18001296 shock (112) tion after LPS stimulation
Martin-Loesches rs1800795 ARDS Caucasian Spanish patients 2.35* Further associations in those
2012 with CAP (1227) (1.2–4.5) with proven S. pneumoniae
22113815
IL8
Hildebrand rs4073 ARDS Germans with blunt mul- NS Increased duration of
2007 IL-8 -251T/A tiple trauma (97) mechanical ventilation
17498967 observed

Patients are adults unless otherwise specified.

* = Odds ratio calculated from data given in publication. PMID = PubMed Identifier.

G e no m ic A p p l ication s in C ritica l C ar e M e dicin e   •   7 6 9

 Macrophage Migration Inhibitory Factor (MIF) levels on admission in survivors.27 Independent of body
mass, differences in allele distribution between survi-
Macrophage migration inhibitory factor (MIF) is encoded
vors and non-survivors of peritonitis have been demon-
by a functionally polymorphic gene; raised levels are seen
strated for SNPs involving LEP (rs7799039) and LEPR
in severe sepsis/septic shock, with non-survivors showing
(rs1137101).28
the greatest elevations.22 A  haplotype that combines SNP
rs755622 (MIF-173G/C) and repeat -794 CATT7, whilst
not predisposing to the development of sepsis, may identify A N T I G E N R EC O G N IT I O N PAT H WAYS
those most at risk of death once sepsis develops.23 A N D I N T R AC E L LU L A R M E D I ATO R S

Toll-like receptors (TLRs) are expressed in a variety of

A N T I-I N FL A M M ATO RY M E D I ATO R S
Interleukin-10 (IL-10)
It was demonstrated some time ago that the magnitude of
IL-10 production by whole blood in response to lipopoly-
saccharide (LPS) stimulation ex vivo is strongly influenced
by genetic variation.24 In critically ill patients with SIRS, a
change in the ratio of IL-6 to IL-10 predicts a poor out-
come.25 Attention has largely focused on three promoter
SNPs with conflicting findings (Table 48.2).
Leptin
The cytokine-like hormone leptin, secreted primarily by
white adipose tissue, is increasingly recognized to have a
role in immune regulation.26 Increased plasma concen-
trations have been reported in sepsis, with the highest
immune and epithelial cells, bind bacterial LPS and other
conserved microbial molecules, and signal via a cascade
of kinases to activate the transcription factor nuclear
factor kappa beta (NF-κB), which plays a central role in
the pathophysiology of sepsis (Figure 48.3). A  SNP of
NF-κB-inducing kinase (rs7222094) has been associated
with mortality in septic shock patients.29 A  variant hap-
lotype of the IL-1 receptor–associated kinase (IRAK-1),
which is involved in intracellular signaling by TLRs, has
been associated with increased NF-κB activation and
worse sepsis outcome.30 Also, the risk of sepsis following
burns has been associated with TLR4 SNP rs4986790.31
It has been suggested that rs8177374, which affects the
downstream adaptor protein Mal, has been selected for
in West Eurasian populations by its protective effect
against septic shock.32 Other studies have examined the
effect of concomitant polymorphisms in multiple cascade

Table 48.2 EXAMPLES OF IL-10 PROMOTER SNP ASSOCIATION STUDIES

STUDY, GENE VARIANTS POPULATION AND SAMPLE SIZE (N) SUMMARY FINDINGS
YEAR , PMID
Shu -592C/A (rs1800872) Severe sepsis (116) rs1800896 possibly associated with susceptibility to severe
2003 -819T/C (rs1800871) Chinese sepsis
14642153 -1082G/A (rs1800896) rs1800872 and rs1800871 in complete linkage
­disequilibrium and not associated with incidence or
­outcome of severe sepsis
Lowe rs1800872 Critically ill (67) rs1800872(A) associated with increased mortality and
2003 Healthy volunteers (132) lower stimulated IL-10 release in vitro
12544990 United Kingdom
Baier rs1800896 Mechanically ventilated very low rs1800896(A) associated with increased incidence of late
2006 birthweight infants (293) bloodstream infection
16611358 Predominantly African-American
Gong rs1800896 Sepsis, trauma, aspiration or massive rs1800896(G) homozygosity associated with lower
2006 transfusion developing ARDS (211) ­mortality and organ failure
16585075 Controls (429)
Caucasian North Americans
Huebinger rs1800872 Burns (265) rs1800872(A) and/or rs1800871(T) associated with
2010 rs1800871 Healthy volunteers (31) reduced risk of death after burn injury
20863526 North Americans
Patients are adults unless otherwise specified.
PMID = PubMed Identifier.

7 7 0   •   G e no m ic s in C l inica l Practic e
 TLR4 TLR2 IL-1R/IL-18R
MD2

CD14 CD14
CELL MEM
BRANE

TIR

TIR

TIR
MAL
ENDOSOME
TLR3 TLR4 MyD88 MyD88 TLR7/8 TLR9

IRAK1
TIR

TIR

TIR

TIR
TRAM
IRAK4
TRIF TRIF TRAF6 MyD88 MyD88
TAK1

IKKα IKKβ
IKKγ
TRAF3 MAPKs TRAF3
TBK1 TBK1
IκBγ
NFκB
CYTOPLASM

NUCLEUS
DNA IRFs AP-1 NFκB IRFs

Figure 48.3 
Overview of innate imunity signalling pathways examined in critical illness gene association studies. Activating protein (AP)-1; IκB
kinase (IKK); IL-1 receptor-associated kinase (IRAK); interferon regulating factors (IRFs); interleukin (IL)-1; mitogen-activated protein kinase
(MAPK); myeloid differentiation factor 88 (MyD88); MyD88 adaptor-like (MAL); nuclear factor (NF) κB; TANK-binding kinase (TBK); TIR
domain containing adaptor-inducing interferon β (TRIF); TNF-associated factor (TRAF); Toll-like receptor (TLR); transforming growth factor
β-associated kinase (TAK); TRIF-related adaptor molecule (TRAM). Reproduced with permission from Lundberg & Hansson 2010.116

components,33 heat shock proteins,34 and nitric oxide
synthase.35 Figure 48.3 illustrates the ways in which these
molecules interact.
Mannose-binding lectin (MBL) binds carbohydrates
on the surface of bacteria, fungi, and viruses to promote
complement activation and phagocytosis. A  combina-
tion of promoter and structural mutations results in a
deficiency state affecting between a quarter and a third
of the population. Variant alleles were over-represented
in children developing SIRS requiring ICU admission.36
A similar finding in adult patients with sepsis has been
reported,37 and lower circulating MBL levels are asso-
ciated with a poor outcome.38 The importance of this
innate immune-defense protein was highlighted by a
meta-analysis39 showing that risk of death was increased
among MBL-deficient patients with Streptococcus
pneumoniae infection. Importantly, however, a low MBL
level appeared to be a more reliable indicator of defi-
ciency than simply carriage of a low-producing MBL2
genotype.
Other molecules involved in the innate immune
response that have received considerable attention
include:  CD14, a membrane-anchored protein promot-
ing the TLR4 response to ligands; LPS-binding protein
(LBP) and bactericidal-permeability–inducing protein
(BPI), related lipid-transfer proteins that recognize LPS;
myeloid-differentiation protein 2 (MD-2), a glycoprotein
that confers LPS responsiveness to TLR4-expressing cells;
nucleotide-binding oligomerization domain (NOD) pro-
teins; and high-mobility group box (HMGB1), a nuclear
and secreted protein. Brief details of studies of interest are
given in Table 48.3.

G e no m ic A p p l ication s in C ritica l C ar e M e dicin e   •   7 7 1

Table 48.3 EXAMPLES OF ASSOCIATION STUDIES EXAMINING MOLECULES INVOLVED IN ANTIGEN
RECOGNITION AND SIGNALING PATHWAYS

STUDY, YEAR , GENE(S) OF SUMMARY FINDINGS

PMID INTEREST
Hubacek CD14 -159C/T (rs2569190; also known as -260) has no effect on incidence of sepsis or mortality
2000
11196689
Gibot CD14 rs2569190 affects susceptibility to septic shock and mortality
2002
12006789
Heesen CD14 rs2569190 not associated with increased risk of severe sepsis in trauma patients
2002
12185442
Barber CD14 rs2569190(C) associated with increased risk of death after burn injury relative to TT homozygotes
2007
17304102
Jessen CD14 rs2569190 not associated with outcome of Gram-negative sepsis
2007
17877801
Fallavena CD14 rs2569190(T) homozygosity associated with survival among septic and septic shock patients
2009
19860589
Hubacek LBP Several polymorphisms associated with increased risk for development of sepsis and possibly an
2001 ­unfavorable outcome
11373419
Flores LBP A common four SNP haplotype strongly associated with susceptibility to severe sepsis
2009
19707138
Zeng LBP 26877T/C (exon13 rs2232618) SNP significantly associated with LPS-induced activation of peripheral
2012 blood leukocytes in vitro; associated with susceptibility to sepsis and multiple organ dysfunction in
22167001 patients with major trauma
Michalek BPI 545G/C SNP associated with severity of sepsis in children admitted to pediatric ICU; significant
2007 ­predisposition to Gram-negative sepsis in 545 GG homozygotes also carrying 216 AG or GG variant;
17898994 ­haplotypes associated with risk of death due to sepsis-related complications
Zeng MD-2 rs11465996 associated with higher sepsis morbidity rate in major trauma; associated with TNF
2012 ­production by peripheral blood leukocytes in response to LPS in vitro
22266968
Gu MD-2 -1625 SNP affects MD-2 promoter activity; expression of MD-2 mRNA and production of TNF in vitro
2007 significantly increased in subjects with the -1625G, these patients also more likely to experience organ
17592304 dysfunction and sepsis after major trauma; allele–dose dependent effect
Brenmoehl NOD2 Leu1007fsinsC genetic variant an independent risk factor for sepsis-related mortality
2007
17558494
Kornblit HMGB1 Homozygosity and heterozygosity for -1377delA associated with decreased 4-year survival and decreased
2008 number of SIRS criteria; carriage of exon 4 variant 982C>T associated with increased number of SIRS
18577209 criteria, lower serum HMGB1, and higher probability of early death due to infection
PMID = PubMed Identifier.

T H E C OAGU L AT I O N C A S C A D E was in use for a decade as an adjunctive therapy in this group

A N D C O M P L E M E N T
Activated Protein C
Protein C deficiency in severe sepsis (see Figure 48.4) is
associated with lower survival rates and a higher incidence
of shock.40 Recombinant human activated protein C (aPC)
based on the PROWESS study findings,41 but failure to
confirm a survival benefit subsequently led to withdrawal
of the drug.42 In sepsis, a promoter haplotype of PROC
(encoding protein C) rs1799808 (PROC-1654C>T) /
rs1799809 (-1641G>A) was associated with risk of death
and organ dysfunction in Chinese Han patients43 and

7 7 2   •   G e no m ic s in C l inica l Practic e
 Procoagulant pathways Anticoagulant pathways Complement
Sepsis
Plasminogen
Activation of complement is a key feature of sepsis; in
Tissue factor PAI-1
+
activators severe sepsis, C5a elevation is associated with increased
Factor X Factor Vlla Protein C mortality.51 In pediatric critical illness, homozygosity for
–
– Plasminogen complement factor H (CFH) 1277T>C (rs1061170) was
aPC
Factor Xa
–
associated with a reduced incidence of early SIRS and
Factor Va
Plasmin sepsis.52 Factor H inhibits complement activation, and
Prothrombin this variant modifies a protein region involved in binding
heparin and CRP. The meningococcus attempts to evade
Fibrin
Thrombin complement-mediated killing by expressing a protein that
FDP
binds CFH. In the specific setting of childhood menin-
Increased
fibrinogen
Impaired
fibrinolysis
gococcal disease, a GWAS has reported susceptibility to
Enhanced formation infection to be significantly associated with two CFH SNPs
of fibrin blood clots
(rs1065489 and rs11582939).53
Thrombosis of small vessels

Impaired tissue perfusion

Figure 48.4 
Sepsis and the coagulation cascade. aPC, activated form of
protein C; fibrin degradation products (FDP); plasminogen-activator
inhibitor-1 (PAI-1). Reproduced with permission from Cohen 2002.115
North Americans of East Asian ancestry.44 An important
question arises as to whether efficacy of recombinant aPC
might be genetically influenced.
Factor V Leiden
Factor V Leiden (FVL) (rs6025) is a prothrombotic poly-
morphism rendering factor Va partially resistant to inac-
tivation by aPC. Heterozygotes experienced an excess of
complications in meningococcal sepsis, but unchanged
mortality.45 Carriers have increased risk of ICU admis-
sion and death, but normal susceptibility to severe invasive
infections.46 The severity of acute pancreatitis, a disease
often necessitating ICU admission, does not seem to be
affected.47
Plasminogen Activator Inhibitor 1 (PAI-1)
Plasminogen activator inhibitor 1 (PAI-1) is a key player
in the inhibition of fibrinolysis. The homozygous 4G/4G
genotype for a single base-pair promoter insertion/dele-
tion (I/D) polymorphism (rs1799889) correlates with
higher PAI-1 concentrations in vitro and a greater risk
of mortality in meningococcal disease.48 In patients with
severe pneumosepsis, homo- or heterozygosity was associ-
ated with septic shock and organ dysfunction.49 In a gen-
eral adult ICU, homozygotes had higher plasma PAI-1
and risk of death.50
C E L L E N E RG ET I C S A N D S U RVI VA L
Mitochondrial DNA (mtDNA)
Mitochondrial dysfunction is increasingly recognized
as a key pathophysiological process in critical illness.54
The most common European mitochondrial genome
haplogroup, H, has been reported to be associated with
sepsis survival.55 In trauma patients, 4216T>C in the nic-
otinamide adenine dinucleotide dehydrogenase 1 gene,
encoding an electron transport chain member, was asso-
ciated with increased risk for organ dysfunction or septic
shock.56
Apoptosis
Apoptosis plays a central role in sepsis and ARDS.57,58
Repeated activation of T-cell antigen receptors induces
Fas ligand expression and apoptosis. A  haplotype asso-
ciated with increased FAS mRNA levels in response to
LPS stimulation in vitro has recently been identified that
appears to carry an increased risk of developing ALI.59
Caspases mediate proteolytic events in inflammation
and apoptosis; rs497116, which abolishes a stop codon,
thereby producing the functional protein Casp12-L, was
over-represented in African American patients with severe
sepsis.60 Members of the Bcl-2 protein family regulate
the mitochondrial pathway of apoptosis. In a large study
of patients with septic shock, two SNPs, rs12457893
and rs8094315, were associated with a decreased risk of
developing acute kidney injury.61 Another SNP in the
SERPINA4 gene, which encodes kallistatin, showed a
similar association in the same study. Kallistatin appears
to attenuate endothelial apoptosis.62

G e no m ic A p p l ication s in C ritica l C ar e M e dicin e   •   7 7 3

 C E L L A D H E S I O N MO L ECU L E S transform our view of the condition from a physiologi-
cal syndrome to a group of distinct disorders of the host
Adhesion molecules play a pivotal role in cellular interac-
response, thereby improving therapeutic decision making
tions during tissue injury and pathogen invasion. Elevated
and outcomes.74 DNA sequence variants affecting genes
plasma levels of E-selectin, ICAM-1, and VCAM-1 have
encoding biomarkers will be important, then, if they affect
been measured in children with sepsis; persistent eleva-
the expression of a gene in a way that disturbs the relation-
tion of ICAM-1 and VCAM-1 predicts risk of sequential
ship of its product with the outcome of interest. Over two
organ failure and mortality.63 E-selectin rs5361 seems to
decades ago, impaired monocyte expression of HLA-DR
modify the response to endotoxemia in healthy men,64
was shown to identify a group at high risk of infection and
but effects in the setting of critical illness have yet to be
death following trauma.75 In this case, knowledge of the
reported. The PPFIA1 gene encodes liprin alpha, involved
mechanisms of antigen recognition and bacterial phago-
in cell adhesion and integrin expression. A GWAS in
cytosis guided the selection of a target for investigation.
European American trauma patients found significantly
Now, proteomic analysis can identify novel biomarkers
increased risk of ALI in patients carrying rs471931.65
in the absence of prior indication of their importance in
Capillary leak is a fundamental process in critical illness.
the pathophysiology of a condition, though subsequent
Improved understanding of the interaction of cell adhe-
validation and determination of the molecular biology is
sion molecules and the effects of sequence variation in the
essential.76
encoding genes on the risk of complications in critically ill
patients are likely to lead to a better understanding of the
biological pathways involved and perhaps suggest poten- C-R E AC T I VE P ROT E I N (C R P) A N D
tial therapies.7 S E RUM A MY L O I D A ( S A A)

C-reactive protein (CRP) is predominantly produced in

T H E R E N I N A N G I OT E NS I N SYS T E M the liver in response to inflammatory stimuli, has a role in
eliminating bacteria, and is a sensitive marker of inflam-
Tissue-specific RAS components are now understood to
mation and tissue damage. Admission CRP, when <100
have pleomorphic effects that influence critical homeo-
mg/L in the setting of CAP, predicts better outcome;
static mechanisms, including the immune response66; for
failure of CRP at least to halve after four days predicts
example, ARDS pathogenesis might be influenced via
mortality and a need for mechanical ventilation or vasoac-
effects on vascular permeability, fibroblast activity, and
tive drugs.77 In Staphylococcus aureus bacteremia, maximal
alveolar epithelial cell apoptosis. An I/D polymorphism
CRP in the first week following diagnosis was influenced
in the angiotensin-converting enzyme (ACE) gene affect-
by rs3091244; however, mortality, degree of leukocytosis,
ing baseline circulating ACE levels was described over two
and time to defervescence were not affected.78 Conversely,
decades ago.67 In children with meningococcal disease, a
in Streptococcus pneumoniae bacteremia, rs2794521 was
DD genotype appears to be associated with increased illness
strongly associated with mortality but did not correlate
severity,68 but in ventilated, very low birthweight infants, no
with plasma CRP concentration.79 Both SNPs affect the
effect on survival from sepsis could be demonstrated.69 The
promoter region. SAA, another acute-phase protein that
DD genotype has been variously associated with develop-
increases in response to tissue injury and inflammation,
ment of ARDS and mortality,70 mortality in ARDS alone,71
seems to be a reliable marker for the diagnosis of pedi-
and neither,72 in different populations.
atric sepsis in a variety of settings.80 A GWAS recently
Septic shock patients homozygous for the rs11121816
identified rs4150642 as associated with SAA concentra-
intronic SNP of angiotensin II type 1 receptor-associated
tions in health;81 implications for critical illness await
protein (AGTRAP) gene showed increased mortality and
investigation.
greater cardiovascular instability; correspondingly higher
mRNA expression was seen in vitro.73
P RO C A L C ITO N I N

BIOMARKER S Of numerous potential biomarkers for the early specific

diagnosis of sepsis in critically ill patients, procalcitonin is
There has been considerable interest in using biomarkers to amongst the most promising and may reduce ICU anti-
aid diagnosis and monitoring in critical care. In the case of biotic use without compromising outcomes.82 In poly-
sepsis, it has been suggested that biomarker research might trauma patients, no association between post-traumatic

7 7 4   •   G e no m ic s in C l inica l Practic e
complications or systemic procalcitonin levels was observed
with a variety of different alleles.83
T R I G G E R I N G R EC E P TO R E X P R E S S E D
O N MY E L O I D C E L L S ( T R E M-1)
Soluble triggering receptor expressed on myeloid cells,
or (s)TREM-1, a member of the immunoglobulin super-
family, is increased in bronchoalveolar fluid before the
development of clinical signs of infection in patients
with ventilator-associated pneumonia.84 In Chinese Han
patients, TREM-1 SNPs rs7768162, rs9471535, and
rs2234237 were not found to be associated with suscep-
tibility to or outcome of severe sepsis.85 Higher sTREM-1
levels in another group of Chinese ICU patients predicted
death, but whilst rs2234237 was significantly associated
with sepsis prognosis, no effect of this SNP on sTREM-1
concentrations was evident.86
FUNCTIONAL GENOMICS
The application of genomics to critical illness, beyond inves-
tigating the relationship between structural genetic diver-
sity and disease susceptibility, is still in its infancy. The hunt
for genetic associations with disease traits or eQTL using
genome-wide scans is a relatively recent development. For
sepsis, gene expression profiling represents a novel approach
to diagnosis and may provide important insights into the
biology of the host response.87,88 For example, a single
intravascular exposure to endotoxin in healthy adults was
shown to differentially affect regulation of many genes not
previously associated with endotoxin-induced inflamma-
tion, whilst inducing those involved in antigen recogni-
tion, intracellular signaling, and cell mobility, as would be
expected.89
D I AG N O S I S A N D T H E R A P Y
The development of functional genomic techniques that
permit the early diagnosis of sepsis and can indicate a likely
pathogen, thus allowing early targeted therapy, would be
a significant step forward. In a longitudinal study of criti-
cally ill but uninfected patients with SIRS, it was possible
to detect gene-expression profile changes in those develop-
ing sepsis prior to expression of the clinical sepsis pheno-
type.90 Genes identified in this way could be functionally
categorized into several themes:  innate immunity, cyto-
kine receptors, T-cell differentiation, and protein synthesis
regulation.
G e no m ic A p p l ication s in C ritica l C ar e M e dicin e   •   7 7 5
Results from a murine sepsis model and cDNA microar-
ray system have suggested a collection of specific genes to be
differentially expressed in Gram-positive and Gram-negative
sepsis.91 In a general pediatric hospital population admit-
ted with acute infections, microarray analysis of peripheral
blood leukocytes appeared able, with a high degree of accu-
racy, to differentiate between influenza A, Escherichia coli,
Streptococcus pneumoniae, and Staphylococcus aureus infec-
tions,92 although blood sampling was performed some days
into the acute illness. In a small study of critically ill adults,
however, no significant differences were demonstrated at
the host transcriptome level between Gram-positive and
Gram-negative sepsis.93 In pediatric septic shock, a biologi-
cally relevant 100-gene expression signature was found to
distinguish three patient groups with important clinical dif-
ferences,94 such as repression of genes involved in adaptive
immunity and glucocorticoid-receptor signaling. Such het-
erogeneity in the genomic response to sepsis might explain
apparent failures of novel therapies in previous studies and
in the future may allow interventions to be targeted to those
most likely to benefit.
PAT H O P H YS I O L O GY A N D P RO G N O S I S
Some studies have attempted to use microarrays to pre-
dict outcome in severe sepsis,95,96 though combining estab-
lished illness-severity scoring systems with biomarkers and
microarray analysis is likely to prove superior to any single
indicator. Use of microarrays or other high-throughput
platforms to analyze changes in differential gene expres-
sion may eventually permit real-time monitoring of disease
evolution, perhaps replacing multiple single laboratory
measurements of discrete biochemical markers.97 In this
regard, genome-wide expression analysis has provided
further insights into the adaptive immune response in
critical illness. A traditional view has been that complica-
tions result from disproportionate SIRS, followed later by
compensatory anti-inflammatory responses (CARS), with
sequential insults, such as ventilator-associated pneumonia
(VAP), leading to recurrent SIRS and organ dysfunction.
However, simultaneous and rapid induction of pro- and
anti-inflammatory innate immunity genes with concomi-
tant suppression of adaptive immunity genes in a “genomic
storm” at the outset of both severe traumatic and burn
injury has recently been described.98 Duration and sever-
ity of the dysregulated acute inflammatory response may
therefore be the decisive factor; to put it another way, “The
problem with inflammation is not how often it starts, but
how often it fails to subside.”99 With this in mind, longi-
tudinal profiling of leukocyte gene expression was able to
identify significant early expression changes that predicted
longer-term complications in trauma patients.100
E P I G E N ET I C S
B ETA-A D R E N O R EC E P TO R S
Vasoactive drugs are the cornerstone of cardiovascular
management of shock. The AA genotype of ADRB2 gene
polymorphism rs1042717 was associated with increased
mortality, more o