496
f o ~'b//~/
choice of topics, although it was
obvious that careful thought had
gone into scheduling the sessions to
avoid obvious clashes. There was not
much competing interest between
'Automation of Enzyme Receptor-
Binding Assays' and 'Retrofitting
Industrial Waste Treatment Facili-
ties', for example, and I found no
problems whatsoever in my choice
of symposia. However, a rather
shocking experience was the early
start of the programme (0800 h). This
undoubtedly resulted in the relatively
low early morning attendance at
some of the symposia, although this
did mean that there was a larger
break at midday for posters and infor-
mal contact.
Many contributions dealt with
microbe-mediated processes for
bioremediation of aquatic and terres-
trial wastes and, as mentioned above,
presentations relevant to agricultural,
chemical, petroleum, paper, metals
and mining industries, as well as con-
taminated waste-waters in general,
received ample exposure. It is interest-
ing and pleasing to note that many
of these contributions were from
industrial research laboratories, often
major companies, for example,
American Cyanamid, Monsanto,
Celgene, Ciba Geigy and BASF.
Newer techniques o f molecular
biology and genetics were more in
evidence in the medically relevant
and diagnostic symposia, such as
those dealing with therapeutics,
Neurotrophic factors - novel
therapeutics?
meeting
report
Specific proteins, known as neuro-
trophic factors, are essential for the
development, survival and mainten-
ance of neurons of both the per-
ipheral- (PNS) and central nervous
systems (CNS). The discovery of
these proteins, and the observation
that several previously characterized
peptide growth-factors also show
neurotrophic activities, has led to the
hope that it may be possible to use
neurotrophic factors in the treatment
of neurological disease and injury:
recombinant growth-factors are
already in clinical trials for several
indications. The current status of
research into the mechanisms of
neurotrophic-factor action, and the
TIBTECHDECEMBER1993 (VOL11)
virology, drug discovery and the
identification and characterization of
cells, as well as in those dealing with
recombinant D N A technology as
applied to insects and plants. The
considerable significance ofbiofilms
in environmental, medical and
biotechnological contexts was again
reinforced, as well as newer tech-
niques for the study of microbially
catalysed processes, for example
biosensors, which are of relevance to
many processes including toxicity
assessment, bioremediation and
remote sensing. More-routine sub-
ject areas that were covered included
fermentations and other conversions,
while the search for novel products
and antibiotics from organisms such
as actinomycetes and fungi, con-
tinues apace. It is evident that all
these applied areas of work still rely
on a fundamental reservoir of basic
microbiological knowledge.
Among the more arcane lecture
titles were 'E = mc 2 (Energy =
microbial cellulases)'; 'The Ultimate
Biomass Conversion: Strain Devel-
opment for Beer Production'
(appropriately or inappropriately,
depending on your palate, from the
Coors Brewing Company); 'The
Fungi' (in only 30 minutes, includ-
ing questions), 'Home on the Range:
An Ecological Perspective'; 'Coal
Dust and Microorganisms: An Inti-
mate Relationship'; 'Bioelectrodes:
What, Why, How?'; 'Gene Transfer
in Maize' (a large green m o n o -
potential therapeutic applications of
these factors were the subject of a
recent conference*.
The neurotrophic factors ident-
ified to date were reviewed by Franz
Hefti (Genentech Inc., South San
Francisco, CA, USA). The factors
have specific actions on specific tar-
get-cell populations, but there is
overlap in both the spectra of their
activities and the spectra of target
cells. Nerve growth factor (NGF;
reviewed in Ref. 1), the first neuro-
*The conference 'Neurotrophic Factors',
organizedby IBC TechnicalServicesLtd, was
held in Cambridge, UK, 22-24 September
1993.
© 1993, Elsevier Science Publishers Ltd (UK)
cotyledonous microbe) and 'Lower-
ing your Microbes' Inhibitions and
Increasing their Productivity'. A final
light-hearted session was entitled 'If
I had to do it over...' with various
eminences recounting the experi-
ences that contributed to their
success. The stark truth about scien-
tific research was revealed by the
following titles: 'Expect the U n -
expected'; 'Serendipity'; 'Retrospec-
tive of a Microbiologist: Adventures
of Being in the Right Place at the
W r o n g Time' and ' I f I Have to Live
My Life Again, I'd Make The Same
Mistakes Only Sooner'. As the final
speaker of this session (Arnold
Demain, Massachusetts Institute of
Technology, Cambridge, MA, USA)
amply illustrated, attendance, com-
munication and interchange at scien-
tific meetings is an integral and im-
portant part o f scientific research
and enjoyment. This meeting was
very enjoyable and clearly planned to
appeal to a diverse range of interests
and to scientists from academia,
industry and government research
establishments. 1 have no doubt that
collaborative relationships were re-
inforced and initiated between these
differing sectors as a result of this
meeting.
Geoffrey M. Gadd
Department of BiologicalSciences,
University of Dundee,
Dundee, UK DD1 4HN.
trophic factor identified, is a member
o f the neurotrophin family, which
includes brain-derived neurotrophic
factor (BDNF), neurotrophin 3
(NT-3) and neurotrophin 4/5
(NT4/5). Ciliary neurotrophic fac-
tor (CN TF), unrelated to the neuro-
trophin family, was originally ident-
ified by its ability to support the
survival in vitro of ciliary ganglion
neurons. Glial cell line derived
neurotrophic factor (GDNF), iso-
lated from conditioned cell-culture
medium, shows a low degree of
homology to members of the trans-
forming growth factor [3 (TGF-]3)
family. Midkine and heparin-bind-
ing neurotrophic factor are two
related proteins that promote neurite
outgrowth from embryonic brain
cells.

Several previously identified
growth factors have been shown to
have neurotrophic activity in vivo or
on neurons in culture. These include
members of the fibroblast growth
factor (FGF) and epidermal growth
factor (EGF) families, interleukins,
insulin-like growth factors-1 and -2,
TGF-[31, granulocyte-macrophage
colony-stimulating factor (GMCSF),
erythropoietin (EPO), and leukaemia
inhibitory factor (LIF), which has
now been shown to be identical to
cholinergic neuronal differentiation
factor. In addition to the peptide fac-
tors that act directly on neurons,
there are growth factors, such as glia
maturation factor ~ and heregulin,
which act on glial cells and, thus,
influence neurons indirectly.
The neurotrophin family
The neurotrophins exert their
effects on neuronal survival and dif-
ferentiation by binding to members
of the Trk family of tyrosine kinase
receptors. (The role of the low-
affinity N G F receptor, p75 LNoFP',has
not yet been elucidated.) Site-
directed mutagenesis studies and the
construction of chimaeric N G F -
BDNF molecules as described by
Carlos Ib~i~ez (Karolinska Institute,
Stockholm, Sweden), combined
with the three-dimensional crystal
structure of NGF, indicate that the
NGF dimer has two Trk receptor
binding-sites, each comprising a
cluster o f discontinuous stretches of
amino acid residues from both N G F
monomers. The fact that the recep-
tor binding-sites are formed by
residues from both monomers
implies that the dimeric molecule is
required for biological activity, and
this suggests a possible mechanism for
ligand-induced receptor dimeriz-
ation.
Mutated neurotrophins that
induce similar levels o f receptor
phosphorylation can show different
biological activities, indicating that
structural differences in the ligand
may result in different responses in
the receptor. This may offer possi-
bilities for designing therapeutic
neurotrophin molecules with
specific activities. The different
members ofthc neurotrophin family
bind to different Trk receptors and
induce effects only in the cell types
expressing the appropriate receptors.
At the Karolinska Institute, a
chimaeric molecule comprising
elements from NGF, B D N F and
N T - 3 has been engineered and is
able to activate Trk receptors A, B
497
f o rlJlq~
and C. This molecule, named pan-
neurotrophin 1, shows that the
design of therapeutic molecules with
multiple activities and target-cell
populations is feasible.
One of the actions of N G F is to
promote the survival and differen-
tiation of peripheral sympathetic and
sensory neurons. These neuronal
populations are affected in peripheral
neuropathy of toxic or metabolic
aetiology. The most important cause
of peripheral neuropathy is diabetes
mellitus. In rats suffering from
streptozotocin-induced diabetes,
David Tomlinson (Queen Mary and
Westfield College, London, UK)
found reduced levels of expression of
N G F and NGF-responsive genes.
Administration of N G F (but not
BDNF) prevented the decrease in
expression of substance P and calci-
tonin gene-related peptide (CGR.P),
without affecting the severity of
the diabetes. Stuart Apfel (Albert
Einstein College of Medicine, New
York, NY, USA) showed that the
loss of sensitivity to thermal pain
induced in mice by the antitumour
agents, taxol or cisplatin, is prevented
by NGF, as is the loss of thermal sen-
sitivity in streptozotocin-diabetic
rats. Clinical trials o f N G F in the
treatment of peripheral neuropathy
have already commenced.
Alzheimer's disease
In the brain, N G F acts selectively
on forebrain cholinergic neurons,
which are involved in cognitive
functions and degenerate in
Alzheimer's disease. Heidi Phillips
(Genentech Inc., South San Francisco,
CA, USA) described the study of
mice carrying a deletion mutation in
the N G F gene. Although the
homozygotes ('NGF-knockouts') do
not show any gross morphological
defects at birth, they do show some
loss of sympathetic and sensory
neurons. The basal forebrain cholin-
ergic system does not appear to be
impaired. However, these mice do
not thrive due to a failure to feed -
perhaps because o f sensory re-
duction. They may die before the
effects of NGF deficiency on the basal
forebrain cholinergic system become
apparent. Mice that are heterozygous
for the N G F deletion mutation do
survive, but they exhibit reduced
levels of NGF and poorer cognitive
function than wild-type mice.
Franz Hefti discussed data from a
number of studies showing that the
administration of N G F is effective in
preventing cholinergic degeneration
following either experimental injury,
or associated with normal ageing in
rats. In primates, it has been shown
that N G F treatment protects cholin-
ergic cell bodies from axotomy-
induced degeneration. It is not clear
how levels of NGF, N G F receptor
or signal transduction are altered in
Alzheimer's disease, but these animal
studies suggest that the adminis-
tration of N G F will counteract
cholinergic atrophy, irrespective of
its cause. N G F treatment of patients
suffering from Alzheimer's disease
would not be a replacement therapy,
but rather a pharmacological attempt
to induce hypertrophy o f the sur-
viving cholinergic neurons. Clinical
trials of N G F in patients with
Alzheimer's disease have begun.
Parkinson's disease
Parkinson's disease is characterized
by a degeneration of the substantia
nigra dopaminergic neurons that
innervate the striatum, and a corre-
sponding decrease in striatal
dopamine content. Mary Hynes
(Genentech Inc., South San Francisco,
CA, USA) reported that N T 4 / 5 and
B D N F promote the survival of rat
embryo dopaminergic neurons in
culture. Transcripts for N T 4 / 5 and
B D N F are present in the striatum,
and m R N A encoding the TrkB
receptor for N T 4 / 5 and B D N F are
present in neurons of the substantia
nigra. These observations indicate
that administration o f N T 4 / 5 or
B D N F might prevent, or slow the
degeneration of dopaminergic
neurons in Parkinson's disease.
G D N F promotes the survival and
morphological differentiation ofdopa-
minergic neurons in rat embryo mid-
brain cultures and increases their
dopamine uptake (Lev Lin, Synergen,
Boulder, CO, USA). In vivo studies
in rats suggest that G D N F may also
be a potential therapeutic agent for
treating Parkinson's disease.
ALS and ciliary neurotrophic
factor
C N T F promotes the survival of
several neuronal cells types in vitro
and regulates the synthesis of trans-
mitter molecules and the low-affinity
neurotrophin receptor (Silvio Varon,
School of Medicine, San Diego, CA,
USA). Responses to C N T F can
require, or be modified by, other
agents such as extracellular matrix
laminin or (as yet unidentified) prod-
ucts of Schwann cells. In vivo, in a
mouse model of motor nero-on dis-
ease (MND), C N T F has been shown
TIBTECH DECEMBER1993 (VOL 11)

to alleviate pathological and be-
havioural progression of the disease.
Peter DiStefano (Regeneron
Pharmaceuticals Inc., Tarrytown,
NY, USA) described the evidence
that C N T F is a neurogenic
myotrophic factor. C N T F receptor
components are expressed in skeletal
muscle as well as in neurons, and the
expression of C N T F receptor
m R N A is up-regulated after dener-
vation of muscle. C N T F adminis-
tration reduces denervation-induced
atrophy in a dose-dependent fashion
and reduces the degree of functional
impairment. The trophic action of
C N T F on both motor neurons and
their targets, muscles, may prove
important in the treatment of amyo-
trophic lateral sclerosis (ALS or Lou
Gehrig's disease) and nerve injury
where degeneration of the nerve and
its target occurs. C N T F is already in
clinical trials for the treatment of ALS
(Ref. 2), as is IGF-1, which has been
shown to enhance functional survival
and regeneration of motor and sen-
sory function in mice with mechan-
ical or chemotoxin-induced injuries
(Patricia Contreras, Cephalon Inc.,
Westchester, PA, USA).
Peptide-drug delivery to the
brain
The delivery of neurotrophic fac-
tors to the brain presents special
problems. The peptides are hydro-
philic and are poorly transported
Recombinant therapeutic proteins
and translational errors
letter
Many cloned human genes can now
be readily expressed in easily manipu-
lated production organisms, such as
Escherichia coli and yeast, and their
protein products are therefore avail-
able in far greater quantities than ever
before. One consequence of this is
the increasing use o f recombinant
human proteins in medical practice.
Recombinant proteins already in
clinical use include insulin, erythro-
poietin (EPO), interferons, inter-
leukins and vaccines. Many more are
still in intensive development. In this
letter, we wish to address a potential
problem in the clinical use of recom-
binant proteins, which has so far
received less attention than it
TIBTECH DECEMBER1993 (VOL 11)
498
foYt~i~
across the brain capillary endothelial
wall that forms the blood-brain
barrier (BBB). William Pardridge
(UCLA School of Medicine, Los
Angeles, CA, USA) described a var-
iety of strategies for drug delivery
across the BBB 3. lntraventricutar
infusion and intracerebral implan-
tation of a reservoir containing the
peptide drug are two possible
approaches, but their use is limited
by poor drug penetration from the
cerebrospinal fluid (CSF) into the
brain. Hydrophilic molecules can be
lipidized or incorporated into lipo-
somes. However, the transport of
lipid-soluble substances across the
BBB is limited to molecules with
molecular weights < 800-1000 Da,
so even lipidized peptide-based
molecules and liposomes penetrate
the BBB poorly.
The therapeutic peptide can be
conjugated to a protein that under-
goes receptor-mediated transcytosis
across the BBB, such as a monoclonal
antibody to the transfen'in receptor,
or one that undergoes absorptive-
mediated transcytosis, such as cation-
ized albumin. Conjugation to the
delivery vector can be achieved by
chemical coupling, genetic fusion or
avidin-biotin technology. A disul-
phide bond is commonly used as a
chemical crosslink: it is relatively
stable in plasma, but labile in cells
where it can be cleaved to liberate
pharmacologically active peptide.
deserves. This problem involves
translational errors, which occur
during the overexpression of het-
erologous genes in convenient pro-
duction organisms, and the possible
effects of these contaminant, erron-
eous translation products.
The large-scale synthesis of recom-
binant proteins almost always
involves overexpression o f the prod-
uct. For example, to produce a
human protein in E. coli, the intro-
duced genes are placed under the
control of a strong, inducible bac-
terial promoter and inserted into
multicopy plasmids. After promoter
induction, the protein specified by
the plasmid will constitute 20-40%
© 1993, Elsevier Science Publishers Ltd (UK)
Cleavage of genetic fusion proteins,
if required for biological activity, carl
be more problematic. A biotinylated
vasoactive intestinal peptide ana-
logue has been transported success-
fully across the BBB, linked via
avidin to a monoclonal antibody to
the transferrin receptor. Disulphide
cleavage of the biotin linker released
active peptide. This technique may
prove successful in delivering neuro-
trophic factors across the BBB.
Preclinical and clinical studies
suggest that the neurotrophic factors,
and agents that modulate their ac-
tivities, may provide powerful tools
for the treatment of neurological
disease and injury. As our under-
standing of the underlying path-
ologies and the mechanisms of
neurotrophic-factor action improve,
it will become possible to design
rational treatment regimes and
specifically modified or chimaeric
neurotrophic peptides.
References
1 Levi-Montalcini, R. (1987) Science 237,
1154-1162
2 SCRIP (1993) 1856, 23
3 Pardridge, W. M. (1991) Peptide Drug
Deliveryto the Brain, Raven Press
C. Jane Robinson
Division of Endocrinology, National Institute for
Biological Standards and Control,
Blanche Lane, South Mimms, Potter's Bar,
Hefts., UK E N 6 3QG.
of the total protein synthesized. Such
high levels of heterologous protein
can induce severe nutritional stresses
and create imbalances in the charged
t R N A supply 1-3. When a charged
t R N A becomes rate-limiting for
protein synthesis, the ribosomes are
forced to pause at the codons requir-
ing that particular type o f t R N A , and
all types of translational errors are
increased at such 'hungry' codons 1-4.
Thus, the rate at which translational
errors occur is almost certain to
increase during high-level recombi-
nant-protein synthesis. However, at
present, only limited information is
available regarding the frequencies of
mistranslation.
A major aim during product
recovery is to remove any abnormal