Nature Neuroscience December 2001

2001 Nature Publishing Group http://neurosci.nature.
com
contents
volume 4 no 12 december 2001
http://neurosci.nature.com
2001 Nature Publishing Group http://neurosci.nature.com
editorial
Brain structure and function are Great expectations.........................................................................................1151
determined by the interplay of
genes and environment.
Thompson and colleagues now
news and views
report that the amount of gray Genes, brain and cognition.............................................................................1153
matter in several brain regions, Robert Plomin and Stephen M. Kosslyn
including language areas and SEE ARTICLE, PAGE 1253
frontal cortex, is more similar
between identical twins than Spreading synapsins.......................................................................................1155
between fraternal twins. These Venkatesh Murthy
heritable differences in brain SEE ARTICLE, PAGE 1187
structure were correlated with
measures of cognitive Synaptic connectivity and computation..........................................................1157
performance, suggesting a Anthony M. Zador
possible link to general SEE ARTICLE, PAGE 1230
cognitive ability.
See pages 1153 and 1253. Synchronicity: when youre gone Im lost without a trace?..............................1159
Anthony D. Wagner
SEE ARTICLES, PAGE 1259
book review
Yes, but am I free?..........................................................................................1161
Neurophilosophy of Free Will
by Henrik Walter, translated by Cynthia Klohr
REVIEWED BY ADINA L. ROSKIES
brief communications
Induction of photoreceptor-specific phenotypes in
adult mammalian iris tissue.............................................................................1163
Regulation of hindbrain M Haruta, M Kosaka, Y Kanegae, I Saito, T Inoue, R Kageyama, A Nishida,
organizer FGF8.
Page 1175.
Y Honda and M Takahashi
Melanopsin in cells of origin of the retinohypothalamic tract...........................1165
J J Gooley, J Lu, T C Chou, T E Scammell and C B Saper
Does bouton morphology optimize axon length?............................................1166
J C Anderson and K A C Martin
Passive eye displacement alters auditory spatial receptive fields
of cat superior colliculus neurons....................................................................1167
J C Zella, J F Brugge and J W H Schnupp
Nature Neuroscience (ISSN 1097-6256) is published monthly by Nature America Inc., headquartered at 345 Park Avenue South, New York, NY 10010-1707. Editorial Office: 345 Park
Avenue South, New York, NY 10010. Telephone 212 726 9200, Fax (212) 696 9635. North American Advertising: Nature Neuroscience, 345 Park Avenue South, New York, NY 10010-
1707. Telephone (212) 726-9200. Fax (212) 696-9006. European Advertising: Nature Neuroscience, Porters South, 4 Crinan Street, London N1 9SQ. Telephone (0171) 833 4000. Fax
(0171) 843 4596. New subscriptions, renewals, changes of address, back issues, and all customer service questions in North America should be addressed to Nature Neuro-
science Subscription Department, PO Box 5054, Brentwood, TN 37024-5054. Telephone (800) 524-0384 or (615) 850 5315, Fax (615) 377 0525. Outside North America:
Nature Neuroscience, Macmillan Magazines Ltd., Houndsmill, Brunel Road, Basingstoke, RG21 6XS, U.K.. Tel: +44-(0)1256-329242. Fax: +44-(0)1256 812358. Email:
subscriptions@nature.com. Annual subscription rates: U.S./Canada: U.S. $650, Canada add 7% for GST (institutional/corporate), U.S. $199, Canada add 7% for GST (individual
making personal payment BN: 14091 1595 RT); U.K./Europe:435 (institutional/corporate), 185 (individual making personal payment), 99 (student); Rest of world (excluding
Japan): 480 (institutional/corporate), 195 (individual making personal payment), 110 (student); Japan: Contact Japan Publications Trading Co. Ltd., 2-1 Sarugaku-cho 1 chome,
Chiyoda-ku, Tokyo 101, Japan, phone (03) 292-3755. Back issues: U.S./Canada, $45, Canada add 7% for GST; Rest of world: surface U.S. $43, air mail U.S. $45. Reprints: Nature
Neuroscience Reprints Department, 345 Park Avenue South, New York, NY 10010-1707. Subscription information is available at the Nature Neuroscience homepage at http://neu-
rosci.nature.com. POSTMASTER: Send address changes to Nature Neuroscience Subscription Department, P.O. Box 5054, Brentwood, TN 37024-5054. Application to mail periodicals
postage rate is pending at New York, NY. Executive Officers of Nature America Inc: Annette Thomas, President; Edward Valis, Secretary-Treasurer. Printed by Publishers Press, Lebanon
Junction, KY, USA. Copyright 2001 Nature America Inc.
nature neuroscience volume 4 no 12 december 2001 i

contents
Memory retrieval impairment induced by hippocampal CA3 lesions is

blocked by adrenocortical suppression............................................................1169
B Roozendaal, R G Phillips, A E Power, S M Brooke, R M Sapolsky and
J L McGaugh
Visual stimuli activate auditory cortex in the deaf............................................1171
E M Finney, I Fine and K R Dobkins
articles
Distinct regulators control the expression of the mid-hindbrain

organizer signal FGF8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1175
W Ye, M Bouchard, D Stone, X Liu, F Vella, J Lee, H Nakamura, S Ang,
The wiring problem in M Busslinger and A Rosenthal
cortical axons. A receptor is required for response to the sugar trehalose in taste
Page 1166.
neurons of Drosophila.....................................................................................1182
A Dahanukar, K Foster, W M van der Goes van Naters and J R Carlson
Synapsin dispersion and reclustering during synaptic activity...........................1187
P Chi, P Greengard and T A Ryan
SEE NEWS AND VIEWS, PAGE 1155
TNF contributes to the death of NGF-dependent neurons
during development.......................................................................................1194
V Barker, G Middleton, F Davey and A M Davies
Wallerian degeneration of injured axons and synapses is delayed by a
Ube4b/Nmnat chimeric gene..........................................................................1199
T G A Mack, M Reiner, B Beirowski, W Mi, M Emanuelli, D Wagner,
D Thomson, T Gillingwater, F Court, L Conforti, F S Fernando, A Tarlton,
C Andressen, K Addicks, G Magni, R R Ribchester, V H Perry and M P Coleman
Auditory responses and
eye position. GABAB receptor activation enhances mGluR-mediated responses at
Page 1167. cerebellar excitatory synapses.........................................................................1207
M Hirono, T Yoshioka and S Konishi
Long-term depression in the nucleus accumbens: a neural correlate of
behavioral sensitization to cocaine..................................................................1217
M J Thomas, C Beurrier, A Bonci and R C Malenka
Endogenous nicotinic cholinergic activity regulates dopamine release
in the striatum................................................................................................1224
F Zhou, Y Liang and J A Dani
Differential synaptic processing separates stationary from transient
inputs to the auditory cortex...........................................................................1230
M Atzori, S Lei, D I P Evans, P O Kanold, E Phillips-Tansey, O McIntyre
and C J McBain
A putative receptor for the SEE NEWS AND VIEWS, PAGE 1157
sugar trehalose.
Page 1182. Inducible, pharmacogenetic approaches to the study of learning and
memory.........................................................................................................1238
M Ohno, P W Frankland, A P Chen, R M Costa and A J Silva
Inferotemporal neurons represent low-dimensional configurations of
parameterized shapes.....................................................................................1244
H Op de Beeck, J Wagemans and R Vogels
Genetic influences on brain structure..............................................................1253
P M Thompson, T D Cannon, K L Narr, T van Erp, V Poutanen, M Huttunen,
J Lnnqvist, C Standertskjld-Nordenstam, J Kaprio, M Khaledy, R Dail,
C I Zoumalan and A W Toga
Human memory formation is accompanied by rhinalhippocampal
coupling and decoupling................................................................................1259
J Fell, P Klaver, K Lehnertz, T Grunwald, C Schaller, C E Elger and G Fernndez
Protecting axons from

errata.......................................................................................................1265
Wallerian degeneration.
Page 1199. classifieds...................................................................................see back pages
nature neuroscience volume 4 no 12 december 2001 ii

2001 Nature Publishing Group http://neurosci.nature.com editorial
Great expectations
The invention of the vacuum cleaner, the washing machine and data according to broadly accepted standards. Current web mark-
the dishwasher were supposed to free housewives from drudgery. up languages, such as HTML, label words according to their posi-
Instead, commonly accepted standards for housekeeping became tion and appearance on the page, but newer languages such as XML
higher with each new labor-saving device, while the time spent permit tagging for content, so that search engines can identify a
on housework remained the same or even increased. New tech- string of words as article title or gene or keyword. Taking
nologies create new expectations, and electronic publishing is advantage of these new capabilities should improve the usefulness
unlikely to be an exception. of scientific journals, but it is not likely to reduce their cost.
In the last decade, many journals have taken advantage of the Reducing publishers costs significantly would require saving
opportunities presented by the internet. Two-thirds of all jour- money on labor. One possible solution to this problem would
nals are now available both online and in print. Over 1000 peer- involve major changes to the current system of peer review. Some
reviewed journals are published solely on the web, approximately electronic journals have experimented with automated peer review,
20% of them in the life sciences1,2. Despite this rapid evolution- in which the author supplies key words and a database program
ary change, online publication has recently become a lightning randomly assigns referees who are described with similar key words.
rod for a variety of more revolutionary ideas about scientific pub- More radical proposals include posting unreviewed papers on a
lishing3. However, many of the concerns discussed in this context, web site and allowing readers to comment online, or selecting
such as library journal pricing, restrictions on access to the litera- papers based on the number of hits they receive from browsing sci-
ture and streamlining the peer review system, are only loosely entists. Such approaches would undoubtedly result in substantial
related to one another. Separating these arguments would improve cost savings, but they also run the risk of forfeiting quality control.
the clarity of the debate. Perhaps greater efficiency can be realized without loss of qual-
One component of this discussion is the belief that the web ity. Another effect of electronic publishing is that it has become
should greatly reduce the costs of publishing a research article. How- easier to test new business models, because the web lowers barri-
ever, it cannot be taken for granted that switching from paper to ers to entry into the publishing business. New electronic-only jour-
electronic publishing alone would reduce publishers expenses. For nals such as Journal of Vision and the Biomed Central (BMC)
most journals, the main cost of publication under the present system journals are peer reviewed and provide free access to primary
is not paper or distribution, but skilled labor. The amount of edi- research articles. J. Vision charges a processing fee to authors, and
torial input varies considerably among journals, but in general, the BMC journals plan to do so from January 2002. A more radi-
high-profile journals spend more on the peer review process, includ- cal attempt at quality control is the Faculty of 1000 initiative4, a
ing the costs of considering papers that are ultimately rejected. Such soon-to-be-launched subscription service that will separate the
filtering adds value to the highly selected papers that appear in pres- filtering from the publishing function, by rating individual arti-
tigious journals, but having papers reviewed at multiple journals cles from other publishers based on the recommendations of well-
also increases the overall cost of scientific publication. Accepted known scientists. It remains to be seen which of these approaches
papers then must be copy edited and formatted for publication, and will prove economically viable.
this process is largely independent of whether the content eventually The current system of scientific publication has evolved over
appears on paper or as a pdf file on a computer screen. Printing and time in response to a variety of selective pressures, and like a bio-
distribution are a surprisingly low percentage of the cost of publi- logical system, aspects of it are undoubtedly shaped by history
cation1. Electronic submission can save money on administrative rather than function. Because there is no central authority to impose
labor and the costs of shipping manuscripts to referees, but these a solution, even revolutionary ideas will have to compete with other
expenses are typically even lower than the cost of printing. Because models for acceptance among authors, readers, referees, librarians
most of the expenses of producing a journal article do not scale with and publishers. In a time of rapid change, perhaps the best strate-
the number of copies produced, a journals circulation is a strong gy will be to occupy as many ecological niches as possible. Whether
determinant of the publishers cost per copy. However, revenues the future evolution of scientific publishing will be gradual or cat-
(from both subscriptions and advertising) do increase as more astrophic remains to be seen, but it seems likely that the changes
copies are sold, explaining why many prestigious journals are cheap- will include more competition among a larger number of publish-
er than their more specialized counterparts. ers, which should be good news for scientists.
Web publication also brings with it substantial new expenses
for establishing and maintaining electronic archives, along with 1. Tenopir, C. & King, D. W. Nature 413, 672674 (2001).
any additional searching or linking services that the publisher may 2. http://dsej.arl.org/dsej/2000/mogge.html.
choose to provide. In the future, sophisticated search options will 3. http://www.nature.com/nature/debates/e-access/index.html
be greatly facilitated if electronic articles can be tagged with meta- 4. http://www.facultyof1000.com
nature neuroscience volume 4 no 12 december2001 1151

news and views
Genes, brain and cognition

Robert Plomin and Stephen M. Kosslyn
By making maps of the differences in cortical gray matter volume between twins, Thompson
et al. describe which brain regions are strongly determined by genetic factors; they further
investigate how these brain differences correlate with measures of cognitive performance.
The 1990s were declared the Decade of This distinction is in essence the dif- One exciting finding from the Thomp-
the Brain for good reason, but the preference between means and variance, son et al. study1 is the high heritability for
sent decade might yield even more fun- which have no necessary connection, gray matter volume in several cortical
damental discoveries as neuroscience either descriptively or etiologically. regions. The remarkably high correlations
begins to capitalize on developments in Despite its name, the analysis of variance (about 0.95) for MZ twins mean that MZ
genetics. The report by Thompson et al.1 (the most widely used statistical test in co-twins are virtually identical in their
in this issue represents an important step science) is actually an analysis of mean volume of gray matter. The same mea-
forward because it bridges these two effects, with individual differences sures for DZ twins, who like any brother
fields. The authors used magnetic reso- included in the error term. Most species- and sister are only 50% similar genetical-
nance imaging (MRI) to create three- universal research is experimental in the ly, are much less correlated. Although pre-
dimensional maps of gray matter and then sense that it manipulates an independent vious twin studies reported high
computed correlations between these variablesuch as genes, lesions, drugs heritability for brain region volumes
measures and general cognitive ability or tasksand asks whether the manipu- assessed by MRI (reviewed in ref. 4), the
(g), derived from diverse cognitive tests lation can have an effect. Individual present study1 goes beyond mere size to
for 40 individuals. What makes this study differences research, in contrast, is cor- the more specific measure of gray matter
special is that the subjects were twins10 relational in the sense that it investigates volume, thus ruling out differences in
pairs of monozygotic (MZ or identical) factors that do have an effect in the world white matter volume. Gray matter con-
twins and 10 pairs of dizygotic (DZ or fra- outside the laboratory. sists of neural cell bodies, whereas white
ternal) twinsallowing the authors to Not all genetic research informs us matter consists of axons. Connections
estimate the genetic contribution to indi- about the basis for naturally occurring among neurons reflect, at least in part, the
vidual differences in gray matter volume differences within a species. For example, results of learningwhich might be
in various brain regions. although knocking out a gene can have expected to differ among individuals as a
The new study1 focuses on the influ- major effects, such experiments do not result of experience. In contrast, the new
ence of naturally occurring genetic varia- imply that the gene has anything to do findings1 suggest that density of neurons
tion on normal interindividual variation, with the variation responsible for hered- may not be easily modified by experience.
that is, the standard deviation found for itary transmission of individual differ- Studies of individual differences have
nearly any characteristic assessed sensi- ences within a species. In contrast, much greater demands for statistical
tively enough. Heredity is not only about quantitative genetic methods such as the power than studies of mean differences.
passing species-general characteristics twin method used by Thompson et al.1 Statistical power refers to the likelihood
from parent to offspring, but also about are rooted in the study of naturally occur- of detecting a true difference (more accu-
transmitting variation in such character- ring variation. Although 99.9% of the rately, of rejecting the null hypothesis). A
istics (Fig. 1). Indeed, inheritance of vari- human DNA sequence is identical for all rule of thumb is to consider the power
ation is the mainspring of evolution, and people, the 0.1% that differs3 million required to detect a true result of a spec-
thus a central focus of genetics. In con- base pairsis ultimately responsible for ified effect size 80% of the time (in other
trast, most neuroscience research focuses the ubiquitous hereditary differences words, in four of five studies). Ten pairs
on universal characteristics. Although per- found for nearly all complex dimensions of MZ twins, as used by Thompson et al.,
spectives are not right or wrong, just more and disorders, including cognitive abili- confers 80% power to detect a correlation
or less useful for particular purposes, the ties and disabilities2. only if the correlation is greater than 0.70
species-universals perspective and the As the new study1 demonstrates, valu- (one-tailed test, p < 0.05). If correlations
individual-differences perspective can able information can be gained by exam- for gray matter density are as high as 0.95
arrive at different answers because they ining individual differences instead of for MZ twins, as suggested by this study
ask different questions. averaging across groups and treating the (and in studies of brain volume as well5),
differences as error. Indeed, such studies they can be detected reliably with just 10
can provide a crucial bridge between neu- twin pairs. However, MZ twins could be
Robert Plomin is in the Institute of Psychiatry,
roscience and genetics, leading to new similar not simply because they have
Social, Genetic & Developmental Psychiatry
Research Center, 111 Denmark Hill, London
insights not only about how genes affect identical genes, but also because they
SE5 8AF, UK. Stephen Kosslyn is in the
cognition but also about how the brain were raised (and continue to live) in sim-
Department of Psychology, Harvard University, works3. A full understanding of the rela- ilar environments. To remove the
William James Hall, 33 Kirkland Street, tionships among genes, brain and cogni- coarsest contributions of common envi-
Cambridge, Massachusetts 02138, USA. tion needs to encompass events at both ronment, heritability estimates are based
e-mail: r.plomin@iop.kcl.ac.uk or levels of analysis (Fig. 1) and discover the on the difference in correlations for MZ
smk@wjh.harvard.edu links between them. and DZ twins. The essence of any esti-
nature neuroscience volume 4 no 12 december 2001 1153

news and views
mate of heritability is to ferent brain regions are

subtract the correlation intercorrelated, as is like-
Distributions Levels Cognitive examples Genes
for DZ twins from that ly. For example, an MRI
for MZ twins and double study of total volume of
the difference. 13 brain regions found
Trying to detect such that the brain regions
a difference in correla- Species Language intercorrelated substan-
universals learning Non-varying
tions more than doubles tially and that a general
the demands for statisti- factor (first unrotated

cal power. For example, principal component in a
even if heritability is factor analysis) accounted
Rare severe Severe retardation
0.90 based on an MZ disorders Early-onset Alzheimer's Single for 48% of the variance5.
correlation of 0.90 and a Thus the simple correla-
DZ correlation of 0.45, tions between gray matter
power is less than 40% to volume in different brain
detect the heritability Common Mild retardation regions and g should be
with 10 pairs of each type mild disorders Learning disabilities Multiple considerably higher than
of twin. This means that suggested by Thompson
a true heritability of 90% et al. Moreover, simple
would not be detected as correlations would prob-
significant more than Normal Specific cognitive abilities ably show that all brain
variation General cognitive ability QTLs
half the time. For more regions correlate with g,
typical heritabilities of Amy Center not just the frontal region.
0.50 (such as MZ and DZ Further analyses of these
Fig. 1. Levels of analysis from species to individual differences. Interposed between
correlations of 0.75 and these extremes are rare severe disorders, often caused by a single gene necessary and data could also examine
0.50, respectively), 80 sufficient for the disorder, and common mild disorders, called complex disorders whether gray matter den-
pairs of each type of twin because they are influenced by multiple genes and environmental factors. Many sity correlates positively
are needed to achieve researchers now believe that common mild disorders are often merely the quantita- with different cognitive
80% power. Comparing tive extreme of the same factors that create normal variation. In other words, there abilities, not just with a
heritabilitiesfor exam- may be no common disorders, just dimensions of normal variation. Genes in such composite g score. That
ple, asking whether heri- multiple-gene (polygenic) systems are called quantitative trait loci (QTLs) because is, the correlation between
tability differs for brain they are likely to result in 15
dimensions (quantitative continua) rather than disorders gray matter volume and
regionsagain raises the (qualitative dichotomies) . the g composite could be
ante substantially. The due to certain abilities
Thompson et al. sample (such as verbal abilities)
of 40 twin individuals is large for a neu- Moreover, multivariate genetic analysis, correlating highly and other abilities (such
roimaging study; studying many hun- which investigates the extent of genetic as spatial abilities) correlating less well. In
dreds of individuals is daunting and may basis for associations between variables, contrast, the hypothesis of genetic g
require multi-site collaborative efforts. shows that most of the genetic action on that the same genetic factors affect diverse
The second important feature of the diverse cognitive abilities involves g8. A cognitive abilitiesleads to the predic-
new study1 is that it shows an association key issue for neuroscience is to under- tion that gray matter volume should cor-
between individual differences in gray stand the brain mechanisms that medi- relate not just with a g composite but
matter volume in the frontal cortex and ate this genetic effect. with all cognitive abilities.
g or general cognitive ability. The con- Studies of total brain volume antici- The old workhorse of the twin design
cept of g is controversial; not all pated the interesting finding by Thomp- (comparing MZ and DZ twins) can be
researchers are comfortable with the idea son et al. of an association between gray used to ask questions that go beyond esti-
that a single factor may influence all matter volume and g. In 14 studies of mating heritability. For example, the twin
types of intelligence6. Although g is not about 700 individuals, correlations design can trace the developmental course
the whole story of cognitive abilities between brain volume and g are about of genetic and environmental influences.
group factors representing specific abili- 0.40 (ref. 4), indicating that individuals One of the most fascinating findings
ties are also important level of with larger brain volumes have higher g about g is that its heritability increases
analysistrying to tell the story without scores. These correlations are similar in almost linearly from infancy (about 20%)
g loses the plot entirely. The Thompson magnitude to the correlations found for to childhood (about 40%) to adulthood
et al. results suggest that g is not simply frontal gray matter volume in the new and old age (about 60%)9. Does the her-
a statistical abstraction that emerges from study1. However, Thompson et al. under- itability of gray matter follow a similar
factor analyses of psychometric tests; it estimate the extent to which gray matter developmental course?
also has a biological substrate in the volume in each brain region correlates In addition, a multivariate genetic
brain. Dozens of studies, including more with g. They report partial regressions or analysis suggests that the association
than 8,000 parentoffspring pairs, 25,000 correlations that indicate the association between total brain volume and intelli-
pairs of siblings, 10,000 twin pairs and between each brain region and g inde- gence is substantially mediated genetical-
hundreds of adoptive families, all con- pendent of other brain regions. Such ly 5 . Although the extent to which
verge on the conclusion that genetic fac- analysis will miss associations with g to correlations between brain and cognition
tors contribute substantially to g 7 . the extent that gray matter volumes in dif- are genetic must be assessed rather than
1154 nature neuroscience volume 4 no 12 december 2001

news and views
assumed, given the high heritability of (ERP) measures yield widely varying her- 3. Kosslyn, S. & Plomin, R. in Psychiatric
gray matter volume in the new paper1, it itability estimates across cortical sites, Neuroimaging Research: Contemporary
Strategies (eds. Dougherty, D., Rauch, S. L. &
seems likely that its association with g is measurement conditions and age, some Rosenbaum, J. F.) 491515 (American
also mediated genetically rather than envi- researchers have reported that ERP (espe- Psychiatric Press, Washington, DC, 2001).
ronmentally. Multivariate analysis can also cially the P-300 component) is related to 4. Vernon, P. A., Wickett, J. C., Banzana, P. G. &
help with the next step: discovering what g14. Other researchers have reported cor- Stelmack, R. M. in Handbook of Intelligence
underlies this association and what other relations between g and brain function- (ed. Sternberg, R. J.) 245264 (Cambridge
Univ. Press, 2000).
aspects of brain anatomy and physiology ing as assessed by positron emission
5. Pennington, B. C. et al. J. Cogn. Neurosci. 12,
give rise to individual differences in g. For tomography, single photon emission

223232 (2000).
example, could differences in the number tomography and functional MRI13, but
6. Gardner, H. Frames Of Mind: The Theory of
of specific types of receptors or the den- we are not aware of genetic studies using Multiple Intelligences (Basic, New York,
sity of neuromodulatory pathways be these techniques. 1983).
responsible for the observed correlations Finding high heritability for g-relat- 7. Plomin, R., DeFries, J. C., Craig, I. W. &
with intelligence? Magnetic resonance ed brain measures paves the way for mol- McGuffin, P. (eds.) Behavioral Genetics in a
spectroscopy provides measures of meta- ecular genetic studies to harvest the fruits Postgenomic World (APA Books, Washington,
DC, in press).
bolic byproducts that can serve as markers of the Human Genome Project. Armed
for some of these variables. with such information, we are poised to 8. Plomin, R. Nat. Rev. Neurosci. 2, 136141
(2001).
Although it is possible that a single identify the specific DNA variation
9. McClearn, G. E. et al. Substantial genetic
fundamental brain characcteristics such responsible for high heritability. However, influence on cognitive abilities in twins 80+
as frontal gray matter volume is respon- identifying specific genes associated with years old. Science 276, 15601563 (1997).
sible for g, it seems more likely that many complex traits has proven more challeng- 10. Posthuma, D., Neale, M. C., Boomsma, D. I. &
brain processes are involved. However, so ing than expected, largely because many de Geus, E. J. C. Behav. Genet. (in press).
far, the pickings are slim other than brain genes are probably involved, each with 11. van Beijsterveldt, C. E., Molenaar, P. C. M.,
volume measures. For example, although small effects7. Nevertheless, finding spe- de Geus, E. J. C. & Boomsma, D. I. Behav.
EEG alpha peak frequency10, EEG coher- cific genetic variation is a high priority for Genet. 28, 443453 (1998).
ence (which has been taken as a measure research because it will provide a very 12. Rijsdijk, F. V. & Boomsma, D. I. Behav. Genet.
27, 8798 (1997).
of brain interconnectivity11) and periph- sharp scalpel for dissecting pathways relat-
eral nerve conduction velocity12 are all ing genes, brain and cognition. 13. Deary, I. J. Looking Down on Human
Intelligence: From Psychometrics to the Brain
highly heritable, these measures do not (Oxford Univ. Press, 2000).
relate to g13. Thus, g does not seem to 1. Thompson, P. et al. Nat. Neurosci. 4,
12531258 (2001). 14. van Beijsterveldt, C. E. & Boomsma, D. I.
involve speedier brains, at least as assessed Hum. Genet. 94, 319330 (1994).
2. Plomin, R., DeFries, J. C., McClearn, G. E. &
by these physiological measures. McGuffin, P. Behavioral Genetics 4th edn. 15. Plomin, R., Owen, M. J. & McGuffin, P.
Although event-related brain potential (Worth, New York, 2001). Science 264, 17331739 (1994).
Although it remains to be seen whether

Spreading synapsins removing all three synapsin genes has a
more profound effect on survival, knock-
Venkatesh N. Murthy out mice alone may not reveal subtle reg-
ulatory roles; mechanistic studies are
Fluorescent synapsins were used to study the dissociation important in this regard.
Previous experiments using bio-
reassociation cycle of this synaptic vesicle protein in situ, and
chemical methods have suggested the fol-
how this process relates to regulation of exocytosis. lowing sequence of events. At rest,
synapsins are associated with synaptic
vesicles and, perhaps, with any actin fil-
Three decades ago, Greengard and col- with molecular biology to provide new aments that may be present in presynap-
leagues identified an abundant brain insight into the involvement of synapsins tic sites 1 . Synapsins do not have a
protein that is a substrate for the cAMP- in neurotransmitter release. membrane-spanning sequence; there-
dependent protein kinase1. In the ensu- Synapsins are abundant at nerve ter- fore, their observed association with
ing years, this family of proteins, called minals and are highly conserved, and their synaptic vesicles must arise from bind-
synapsins, has been investigated intense- biochemical properties are regulated by ing to vesicle components. Synapsins also
ly. Somewhat surprisingly, their precise activity. For this reason, investigators have form homo- and heterodimers, which
role in synaptic transmission is still anticipated that synapsins are critical in may assist in crosslinking neighboring
unclear. Now, Chi and colleagues2 ele- synaptic transmission. Synapsins have vesicles. During action potential stimu-
gantly combine fluorescence microscopy been implicated in a variety of func- lation, synapsins dissociate from vesicles
tionssynaptic vesicle clustering, mobi- and disperse into the cytosol79. Synapsin
The author is in the Department of Molecular lization and even exocytosisbased on dissociation from vesicles, controlled by
and Cellular Biology, Harvard University, 16 their dynamic affinity for synaptic vesi- phosphorylation, frees the vesicles to
Divinity Avenue, Cambridge, Massachusetts cles 36 . Mice with two of the three move toward the active zone to replen-
02138, USA synapsin genes knocked out are viable, but ish spent vesicles. Upon termination of
e-mail: vnmurthy@fas.harvard.edu have abnormal synaptic transmission4. stimulation, synapsins are dephospho-

news and views
Synaptic bouton Fig. 1. The relationship of the synapsinsynaptic vesicle cycle to exocytosis dur-
ing nerve activity. Vesicles are initially clustered with synapsin bound in the
unphosphorylated state. Stimulation with action potentials results in release of
Clustered
vesicles and phosphorylation of synapsins, which dissociates from vesicles and
diffuses throughout the axon. After the termination of stimulation, vesicles are
Axon shaft
endocytosed, and dephosphorylated synapsin reassociates with the vesicles.
Active zone
synapsin dissociation rate is does not regulate the reassociation.
causally related to vesicle Interestingly, synapsin reclustering and
P P P P
Stimulation mobilization. As the rate of endocytosed vesicle reclustering follow
P
P P P P
unbinding of synapsins from a similar time course 10. This suggests
P P
P
P
P P
P vesicles is slowed, vesicle that synapsins rebind to vesicles as they
P movement toward the active transit from their sites of endocytosis
P
P P zone is also slowed, resulting back to the vesicle cluster. Therefore, the
in reduction of the rate of rate of synapsin reclustering might be
exocytosis. The strong evi- controlled in part by the rate of vesicle
dence for dispersion preced- reclustering. A simple experiment could
P P P ing vesicle mobilization address this hypothesis: if synapsin dis-
Re-clustering
P
would be further bolstered if persion is induced in the absence of
P
P it could be shown that dis- exocytosis (and therefore minimal sub-
P persion occurs even in the sequent endocytosis), the reclustering of
absence of exocytosis. For synapsin is independent of vesicle traf-
example, synapses treated fic. This experiment would determine
with tetanus toxin (which whether reclustering could be accounted
essentially abolishes evoked for by diffusional return of synapsin to
Synaptic
=
vesicle
= Synapsin P = Phosphate vesicle release) should the synaptic vesicle cluster, or whether
Bob Crimi exhibit synapsin dispersion it is governed by the traffic of vesicles
rates similar to control back to the vesicle cluster.
rylated, and they reassociate with vesi- synapses upon stimulation. The value of Chi and colleagues
cles. Phosphorylation of synapsins can A further important finding was that study arises in part from its unexpected
occur at several sites by a variety of kinas- the effects of mutations at CaMK sites 2/3 results and the resulting predictions.
es, including calcium calmodulin-depen- are more severe when expressed in First, all mutant EGFP-synapsins can dis-
dent kinases (CaMK) I, II and IV, MAP synapsin I/II-null neurons than in wild- sociate from vesicles. A previous bio-
kinases and protein kinase A. type neurons. In contrast, CaMK site 1 chemical investigation of synapsin
Despite a wealth of biochemical evi- mutation has similar effects on vesicle phosphorylation and vesicle affinity
dence, a real-time view of synapsin mobilization rates whether expressed in found that dissociation of synapsin from
dynamics has been lacking until null background or in wild type. The vesicles strongly depended on phospho-
now. Chi and colleagues 2 examined authors conclude that phosphorylation at rylation at the CaMK site 1 (ref. 9). In
synapsin Ia in living hippocampal site 1 influences direct binding of the present study, synapsins that lack all
synapses by tagging it with enhanced synapsins to vesicles, whereas phospho- CaMK phosphorylation sites could dis-
green fluorescent protein (EGFP). At rest, rylation at sites 2 and 3 regulates interac- sociate from vesicles upon stimulation,
EGFP-synapsin localized to synaptic vesi- tion of synapsins among themselves. although the rate of dissociation
cle clusters, but upon stimulation, it lost What happens to the dispersed was slower. Thus, there must be anoth-
its clustered appearance and dispersed synapsins once stimulation is terminat- er switch controlling the affinity of
along the axon (Fig. 1). Synapsin disper- ed? EGFP-synapsins return to the synap- synapsins for vesicles, and Chi and col-
sion was faster than dispersion of an inte- tic vesicle cluster at a rate slower than leagues remind us of other phosphory-
gral vesicle membrane protein or of the observed dispersion rate (t 1/2 of lation sites and the ATP binding site in
vesicles labeled with the fluorescent dye about 100 seconds versus about 15 sec- the central domain. Examination of the
FM4-64. Therefore, the redistribution of onds). To recluster, synapsins need to dynamics of EGFP-synapsin with muta-
synapsin is due to its dissociation from regain their strong affinity for synaptic tions in these sites will be informative. A
vesicles and its movement into axonal vesicles, which is thought to be due to second surprising finding is that the
regions. EGFP-synapsins carrying muta- the removal of the phosphate on specif- reclustering dynamics are not altered by
tions in three CaMK phosphorylation ic serine residues. However, this idea is the mutations studied. Perhaps the
sites (serine to alanine substitutions) dis- not supported by Chi and colleagues rebinding of synapsins to vesicles fol-
persed more slowly upon stimulation. data 2 , which indicate not only that lowing cessation of activity is not rate
Concomitantly, vesicle exocytosis was reclustering occurs in all their mutants, limiting for reclustering, and therefore
also slowed, presumably because of but also that the reclustering is quanti- is not visible to the assay used. Also,
reduced mobility of vesicles. tatively similar. This means either that there may be a stronger effect of muta-
The key finding that phosphorylation other phosphorylation sites are involved tions on dispersion and reclustering in
of synapsins alters dispersion rates and in governing the kinetics of reassocia- inhibitory synapses. It is likely that Chi
vesicle exocytosis rates in a correlated way tion of synapsins with vesicles, or that and colleagues studied excitatory synaps-
led Chi and colleagues to suggest that the phosphorylation state of synapsins es, which are numerically dominant in

news and views
the type of cultures they used. An earli- synapses in real time will no doubt 6. Humeau, Y. et al. Neuroscience 21, 41954206
er electrophysiological study found that facilitate analysis of mechanisms in (2001).
the depletion of releasable vesicles vesicle trafficking. 7. Sihra, T. S., Wang, T. K., Gorelick, F. S. &
occurred more readily in inhibitory Greengard, P. Proc. Natl. Acad. Sci. USA 86,
81088112 (1989).
synapses from synapsin I null mice11. 1. Greengard, P., Valtorta, F., Czernik, A. J. &
Benfenati, F. Science 259, 780785 (1993). 8. Torri-Tarelli, F., Bossi, M., Fesce, R.,
Real-time visualization of synapsins Greengard, P. & Valtorta, F. Neuron 9,
and other proteins, as developed by Chi 2. Chi, P., Greengard, P. & Ryan, T. A. Nat. 11431153 (1992).
and colleagues2, adds a powerful analyt- Neurosci. 4, 11871193 (2001).
9. Hosaka, M., Hammer, R. E. & Sudhof, T. C.
ical method to the existing array of 3. Li, L. et al. Proc. Natl. Acad. Sci. USA 92, Neuron 24, 377387 (1999).
techniques used to study presynaptic 92359239 (1995).
10. Li, Z. & Murthy, V. N. Neuron 31, 593605
function. It permits analysis of the 4. Rosahl, T. W. et al. Nature 375, 488493 (2001).
dynamics of presynaptic molecules in situ (1995).
11. Terada, S., Tsujimoto, T., Takei, Y., Takahashi,
while monitoring synaptic function. 5. Hilfiker, S. et al. Nat. Neurosci, 1, 2935 T. & Hirokawa, N. J. Cell Biol. 145, 10391048
Imaging biochemistry inside living (1998). (1999).
gest that the two classes of synaptic con-

Synaptic connectivity and nection might provide two distinct, par-
allel channels within the cortex for
computation processing information about these two

kinds of auditory stimuli, just as the
magnocellular and parvocellular path-
Anthony M. Zador ways provide separate pathways in the
visual system for processing different
kinds of visual stimuli.
A new study finds two classes of synapses between layer 2/3 The core of the present findings
neurons in auditory cortex, and suggests they may be involved relates to the physiological properties
in processing transient versus sustained acoustic stimuli underlying excitatory coupling between
neuronal pairs of layer 2/3 of the cortex.
According to the classical quantal
What endows a cortical circuit with its neurons in turn make connections both model 2 , synaptic transmission is a
unique identity? How does a bit of cor- to other layer 2/3 neurons, and to layer probabilistic process in which the presy-
tex implement the computation that it 5 neurons. Atzori et al. found that these naptic terminal is coupled to its postsy-
must perform? The simple answer, of layer 2/3 connections fell into two naptic target through a set of n release
course, is that function arises from the classes, weak and strong, which sites. When an action potential invades
pattern of synaptic connection between differed in a number of important the presynaptic terminal, each of the
neurons and the strengths of these con- characteristics, including average sites releases a vesicle of neurotransmit-
nections. This view motivates much amplitude, failure rate and paired pulse ter with a probability p (and therefore
research on synaptic function and plas- ratio. Most notably, these connections fails to release with probability 1 p);
ticity, and is enshrined in formal neural differed in their temporal dynamics, as the postsynaptic response due to the
network models of computation. There assessed by the response to a sustained vesicle is given by the quantal size q
is, however, remarkably little experi- train of action potentials. Strong con- (which has units of mV or pA). The
mental evidence detailing how a partic- nections decayed during the train, total postsynaptic response R following
ular matrix of synaptic connectivity gives while weak connections retained their an action potential is thus given by the
rise to a particular computation. efficacy at a constant, albeit lower, level simple equation:
Atzori and colleagues 1 advance an throughout the train. Strong connec-
intriguing hypothesis that begins to tions thus gave their most robust R=npq (1)
bridge this gap between cortical com- responses to the transient portion of
putation and synaptic mechanism. stimuli, while weak connections Together, the product of the three
Using dual whole-cell patch-clamp responded equally well to both tran- quantal variables n, p and q governs the
recordings, they examined the proper- sient and sustained stimuli. By con- average synaptic response R, with weak
ties of synaptic connections between trast, synaptic characteristics between synapses having a smaller product than
pairs of coupled neurons in layer 2/3 of pairs of layer 2/3 neurons in the barrel strong synapses.
acute slices of auditory cortex. In the cortex fell into a third category, which Although the quantal model was
rat, these layer 2/3 neurons (along with might be called very strong. originally developed to describe events
neurons in layer 4) receive direct input The authors noted an interesting at the neuromuscular junction, the
from the auditory thalamus; layer 2/3 parallel between these two synaptic same framework, with relatively minor
classes identified in auditory cortex and modifications, has been able to account
The author is at Cold Spring Harbor
the two types of firing patternstran- for transmission at central synapses as
Laboratory, 1 Bungtown Rd., Marks Bldg., Cold sient and sustainedwith which thal- well. One important difference is quan-
Spring Harbor, New York, 11724, USA. amic inputs to the auditory cortex titative. At the neuromuscular junction,
e-mail: zador@cshl.org respond to acoustic stimuli. They sug- the number of release sites is typically

news and views
quite large (n 104), whereas the dynamics in both vertical and

number of release sites mediat- horizontal excitatory path-
ing the coupling between a pair ways 12 . These changes might
of neurons in the cortex be due to changes in the tem-
(including both neocortex and poral dynamics of individual
hippocampus) is much smaller, synapses, or to a change in
often near one. The smaller the relative contribution of
value at central synapses has populations with different
characteristic dynamics13. The
made it easier to study not

just the aggregate statistical possibilities for temporal
properties of release sites, as processing offered by dynam-
at the neuromuscular junc- ic synapses are only beginning
tion, but differences among Statistical Statistical n = 4 to be explored in the context
Bob Crimi
them as well. n=1 of formal neural network
Strong and weak connec- models14.
Fig. 1. Number of release sites varies among different synapses.
tions differ in their release Atzori and colleagues1 have
probability. Part of this differ- proposed a thought-provoking
ence may be due to differences and testable hypothesisthat
in the number of release sites, n, get (as in autaptic cultures), have dif- the two classes of synaptic connection
between those two connections (Fig. 1); ferent release probabilities7,8. between layer 2/3 neurons in auditory
even small differences in the number of Central synapses differ in their tem- cortex provide a substrate for differen-
release sites among synaptic popula- poral dynamics as well. Synaptic effica- tial processing of transient versus sus-
tions can alter the interpretation of cy during a train of action potentials can tained acoustic stimuli. It should be
experimental results. In the hippocam- increase or decrease, depending on the emphasized that, by virtue of the prepa-
pus, the well-studied Schaffer collateral properties of the synapse and the tem- ration they used (in vitro recording in
connection between neurons in regions poral dynamics of the input spike train. acute slices), their experimental results
CA3 and CA1 is usually mediated by The complexity of these synaptic provide no direct support for this idea,
only a single release site3, whereas in the dynamics arises from the interplay of a not even correlative. Testing this
neocortex a single axon from one neu- host of physiologically distinct mecha- hypothesis will require an experimental
ron may make several contactsas nisms, including paired pulse facilita- approach that can link synaptic and sen-
many as a dozenonto its target 4 . A tion, paired pulse depression and sory physiology. The potential payoff for
complete failure of synaptic transmis- post-tetanic potentiation, operating on such challenging experiments is the
sion following an action potential is an characteristic time scales ranging from opportunity to understand how net-
exponentially rare event when the milliseconds to seconds or more10. Most works of cortical neurons implement
synaptic coupling between two neurons forms of short-term plasticity are medi- their computations.
involves multiple release sites: for n ated by changes in release probability.
release sites, each with a probability p Indeed, there is an inverse relation 1. Atzori, M. et al. Nat. Neurosci. 4, 12301237
(2001).
of release and 1 p of failure, the prob- between the initial synaptic release prob-
2. del Castillo, J. & Katz, B. J. Physiol. (Lond.)
ability that all n sites will fail simulta- ability and the amount of short-term 124, 560573 (1954).
neously is given by (1 p)n, a quantity facilitation at single release sites (that is, 3. Sorra, K. E. & Harris, K. M. J. Neurosci. 13,
that diminishes rapidly with increasing sites with high release probability 37363748 (1993).
n. Thus the difference in the failure rate depress 5 ). This is consistent with the 4. Markram, H. Cereb. Cortex 7, 523533
between weak, strong and very strong tendency of strong, high probability (1997).
synapses may arise in part through dif- connections in layer 2/3 of auditory cor- 5. Dobrunz, L. E. & Stevens, C. F. Neuron 18,
ferences in the number of release sites tex to show depression in response to 9951008 (1997).
n, rather than through differences in sustained stimulation . 1 6. Markram, H., Wang, Y. & Tsodyks, M.
Proc. Natl. Acad. Sci. USA 95, 53235328
the release sites themselves. Heterogeneity in the temporal (1998).
Strong and weak connections may dynamics of synaptic responses pro-
7. Murthy, V. N., Sejnowski, T. J. & Stevens, C. F.
differ not only in the number of release vides a rich substrate for cortical Neuron 18, 599612 (1997).
sites n, but also in the release probabil- circuits to implement different compu- 8. Rosenmund, C., Clements, J. D. &
ity p at each release site. Recent studies tations. Synaptic dynamics are them- Westbrook, G. L. Science 262, 754757
of central synapses have revealed selves subject to plasticity. Changes in (1993).
remarkable heterogeneity among release temporal processing in the auditory 9. Hessler, N. A., Shirke, A. M. & Malinow, R.
Nature 366, 569572 (1993).
sites. Heterogeneity of release probabil- cortex have been observed following
ity has been demonstrated by a wide perturbations of sensory experience11, 10. Koch, C. Biophysics of Computation (Oxford
Univ. Press, New York, 1999).
range of experimental techniques, but the cellular and synaptic mecha-
11. Kilgard, M. P. & Merzenich, M. M. Nat.
including minimal stimulation5, paired nisms underlying these changes have Neurosci. 1, 727731 (1998).
6
neuronal recording , and optical and 7 not been examined. In the somatosen- 12. Finnerty, G. T., Roberts, L. S. & Connors, B. W.
pharmacological 8,9 methods. Even sory (whisker) system of the rat, Nature 400, 367371 (1999).
synapses within an ostensibly homoge- sensory deprivation leads to a reorgani- 13. Poncer, J. C. & Malinow, R. Nat. Neurosci. 4,
neous population, such as those arising zation of the pattern of cortical respon- 989996 (2001).
from a single presynaptic axon and ter- siveness, and to corresponding changes 14. Natschlager, T., Maass, W. & Zador, A. M.
minating on a single postsynaptic tar- in the average characteristics of synaptic Network 12, 7587 (2001).

news and views
recalled words. These synchronization dif-

Synchronicity: when youre ferences partially overlapped in time with
transient reductions in gamma power in
gone Im lost without a both rhinal and hippocampal regions dur-

ing effective encoding trials.
Due to the correlational nature of these
trace? results, we cannot conclude that early rhi-
nalhippocampal gamma synchronization,
Anthony D. Wagner
later desynchronization, or decreased

gamma power in these regions is necessary
for declarative memory formation. Never-
theless, these novel findings suggest that
Recordings from the human medial temporal lobe suggest that more effective encoding may emerge when
synchronization of oscillations between rhinal cortex and rhinal and hippocampal neurons synchro-
hippocampus may contribute to building declarative memories. nously oscillate and then desynchronize,
and further suggest that decreased gamma
power in these regions during encoding
may aid learning. Fell and colleagues pro-
Conscious memory for everyday events pal and hippocampal activation during the pose that increased gamma phase coupling
depends on learning mechanisms in the encoding of an experience correlates with may reflect a change in the functional con-
medial temporal lobe1,2, where neocorti- whether the experience will be later nectivity between rhinal and hippocampal
cal inputs converge on the hippocampus remembered or forgotten7, with these subregions that is important for initiating
by way of rhinal cortex. A key to under- sequent memory effects emerging in the encoding, for instance by facilitating the
standing medial temporal contributions rhinal cortex before the hippocampus. transmission of information between these
to learning is determining how these Although considerable insights into medi- regions. Subsequent desynchronization may
regions interact during the building of al temporal lobe function have emerged mark termination of these regional inter-
memories. One proposed mechanism for from these and related studies, evidence actions following information transfer. The
functional integration across different regarding the nature of human rhinalhip- authors further suggest that the negative
brain regions is gamma-band phase syn- pocampal interactions during encoding correlation between gamma power and
chronization, in which distinct popula- and their relationship to effective learning effective encoding may reflect adverse con-
tions of neurons fire at around 40 Hz and has been lacking. sequences of noise-like ambient gamma
in synchrony3,4. In this issue, Fell and col- Fell and colleagues recorded field activity that could interfere with stimulus-
leagues 5 report that intracranial elec- potentials from the seizure-free rhinal specific activity and thus encoding. Alter-
troencephalograms from human rhinal (perirhinal and entorhinal) and hip- natively, they suggest that decreases in
cortex and hippocampus tend to demon- pocampal regions of patients with gamma could reflect the suppression of
strate greater synchrony while subjects intractable epilepsy while the patients were components of the rhinalhippocampal cir-
learn words later remembered than words attempting to learn individually present- cuit and that failure of such suppression
later forgotten. This brainbehavior cor- ed words (Fig. 1). After encoding, the may hinder encoding. Although these inter-
relation suggests that rhinal-hippocampal patients were asked to freely recall the pretations are speculative, they provide
interactions may contribute to effective words that had been studied, and EEG important directions for further investiga-
memory formation. data acquired during learning were sorted tion that undoubtedly will continue to
Lesion evidence from humans and by whether the words were subsequently unravel the mysteries of memory forma-
experimental animals indicates that the recalled or forgotten. Rhinalhippocam- tion in the medial temporal lobe.
medial temporal lobe circuit is necessary pal interactions were indexed separately Fell and colleagues results raise a num-
for declarative memory. It is well estab- for later remembered and later forgotten ber of fundamental questions regarding
lished that damage to these regions trials by assessing the phase synchroniza- rhinalhippocampal synchronization. First,
decreases the ability to consciously tion of gamma-frequency oscillations in how do these changes in synchrony
remember events that occur after the neur- the EEG signals from these regions. Crit- emerge? One possibility is that they emerge
al insult1,2. Moreover, recent functional ically, Fell and colleagues observed that the directly through the dynamics of the medi-
imaging and electroencephalographic constancy of the phase lag between rhinal al temporal circuit. However, it is also pos-
(EEG) studies in humans implicate and hippocampal gamma oscillations dif- sible that an attentional signal beyond the
medial temporal lobe computations in fered depending on whether the words medial temporal lobe may serve as a pace-
mnemonic encoding6. For example, the were later remembered or forgotten. maker through inputs that entrain rhinal
degree of rhinal, posterior parahippocam- Gamma synchronicity was initially greater and hippocampal neurons. Functional
for remembered words from 100 to MRI studies of encoding consistently show
300 ms and from 500 to 600 ms following greater prefrontal cortical activation dur-
The author is in the Department of Brain &
word onset. These changes reflected a ing learning trials that are later better
Cognitive Sciences and Center for Learning &
Memory, Massachusetts Institute of Technology,
decrease in the phase differences between remembered7,8. A number of theorists have
NE20-463, Cambridge, Massachusetts 02139
rhinal and hippocampal oscillations dur- posited that encoding depends on interac-
and the Athinoula A. Martinos Center for ing these periods (Fig. 1). Later during tions among prefrontal, posterior neocor-
Biomedical Imaging, Massachusetts General stimulus processing, decreased syn- tical and medial temporal computations,
Hospital, USA. chronicity was observed from 1,000 to with prefrontal cortex initiating a cascade
e-mail: awagner@psyche.mit.edu 1,100 ms after the onset of subsequently of events that can modulate effective trace

news and views
Fig. 1. EEG recordings from human medial temporal lobe revealed greater gamma phase syn-
a chronization and desynchronization during the encoding of words later remembered com-
pared to words later forgotten. (a) Approximate location of Fell and colleagues recordings
from rhinal cortex and hippocampus. The dashed black line represents the angle of electrode
insertion along the long axis of the hippocampus. (b) Relationship between rhinalhippocam-
pal coupling and subsequent memory performance. Encoding of events that were subse-
quently remembered first evoked increased gamma-phase synchronization between rhinal
and hippocampal regions (blue shading) and then decreased synchronization (yellow shading)
relative to the encoding of events later forgotten. (Note that, for visualization purposes, the
presently rendered oscillations are slower than the observed gamma frequency.)
Hippocampus Rhinal
encountered stimulus, such as a per- results of Fell and colleagues mark a signifi-
b cortex son you recently met at a conference, cant advance in understanding the tempo-
Subsequently remembered
can be based on recollection of spe- ral dynamics of activity within these regions
cific details about the past encounter and their relationship to memory formation.
with the stimulus or on a general Moreover, their study highlights the leverage
sense of stimulus familiarity. For that can be gained by assessing temporal
Subsequently forgotten
example, when subsequently encoun- characteristics of neuronal responses, both
tering the person, you may recall her within and across distinct structures. This
name or professional affiliation or investigation may prove to be the first of
you may simply have the subjective many influential efforts to specify how neur-
Time
Bob Crimi sensation that the face is familiar and, al coupling across brain regions affects mem-
hence, that you must have met her ory behavior. Such future efforts may also
formation8,9,10. Fell and colleagues propose before. Recently, extensive attention has emerge from the integration of scalp-record-
that the early onset of increased rhinalhip- focused on whether rhinal and hip- ed magnetoencephalography or EEG with
pocampal synchronization may preclude a pocampal subregions differentially sub- fMRI. The findings by Fell and colleagues5
prefrontal source, perhaps pointing to thal- serve recollection and familiarity. From may well stand as a landmark along the road
amic modulation of the circuit. However, one perspective, the hippocampus is to specifying the neurocognitive processes
given the hypothesized role of prefrontal thought to specifically mediate processes that allow us to remember our past.
cortex in representing goal statesrepre- that underlie subsequent conscious recol-
sentations that may be on-line before lection of event details13,14. Within this 1. Squire, L. R. Psychol. Rev. 99, 195231 (1992).
stimulus presentationand in biasing pos- framework, hippocampally derived traces 2. Eichenbaum, H. & Cohen, N. J. From
Conditioning to Conscious Recollection:
terior processes in favor of task-relevant do not subserve memory based on item Memory Systems of the Brain (Oxford Univ.
codes and pathways11, assessment of pre- familiarity in the absence of recollection. Press, New York, 2001).
frontal contributions to the emergence of Rather, perirhinal cortex is posited to sub- 3. Engel, A. K., Fries, P. & Singer, W. Nat. Rev.
rhinalhippocampal synchrony would serve the acquisition of item traces that Neurosci. 2, 704716 (2001).
appear to be a promising direction for support subsequent familiarity-based 4. Varela, F. J., Lachaux, J.-P., Rodriguez, E. &
further investigation. memory14. Fell and colleagues assessed Martinerie, J. Nat. Rev. Neurosci. 2, 229239
(2001).
Second, is subsequent memory selec- subsequent memory using a free recall
5. Fell, J. et al. Nat. Neurosci. 4, 12591264
tively associated with changes in rhi- test. Thus, their results demonstrate that (2001).
nalhippocampal synchronization in the synchronous rhinalhippocampal activi- 6. Schacter, D. L. & Wagner, A. D. Hippocampus
gamma range? Fell and colleagues do not ty is correlated with subsequent recollec- 9, 724 (1999).
specify whether changes in frequency tion. However, these findings need not 7. Wagner, A. D., Koutstaal, W. & Schacter, D. L.
bands outside of gamma were associat- imply that rhinal and hippocampal struc- Phil. Trans. R. Soc. Lond. B Biol. 354,
ed with memory performance. Prior tures subserve the same form of declara- 13071324 (1999).
intracranial EEG recordings in humans tive memory. That is, although rhinal 8. Wagner, A. D. in Neuropsychology of Memory
3rd edn. (Guilford, New York, New York, in
have shown theta (48 Hz) oscillations inputs to the hippocampus are likely press).
during spatial navigation12. These results important for successful hippocampal for- 9. Moscovitch, M. J. Cognit. Neurosci. 4,
converge with animal studies that demon- mation of traces that ultimately yield rec- 257267 (1992).
strate a relationship between theta rhythm ollection, within rhinal cortex the 10. Buckner, R. L., Kelley, W. M. & Petersen, S. E.
and hippocampal place codes, with theta resultant traces may simply support sub- Nat. Neurosci. 2, 311314 (1999).
modulation being associated with pro- sequent item memory. It should prove 11. Miller, E. K. & Cohen, J. D. Annu. Rev.
cessing stages that may strengthen mem- informative in future investigations to Neurosci. 24, 167202 (2001).
ory representations13. To fully appreciate derive separate behavioral measures of 12. Kahana, M. J., Sekuler, R., Caplan, J. B.,
the role of gamma-band synchronization, recollection and familiarity, and to exam- Kirshen, M. & Madsen, J. R. Nature 399,
781784 (1999).
it may be critical to determine whether ine the relationship between each of these
13. Louie, K. & Wilson, M. A. Neuron 29, 145156
memory-related rhinalhippocampal forms of declarative memory and rhi- (2001).
coupling is selective to this oscillatory fre- nalhippocampal synchronization (and
14. Brown, M. W. & Aggleton, J. P. Nat. Rev.
quency or derives from broader coupling. gamma power). Neurosci. 2, 5161 (2001).
Third, what form of declarative mem- Although questions remain regarding 15. Eldridge, L. L., Knowlton, B. J., Furmanski, C. S.,
ory emerges from rhinalhippocampal how the medial temporal lobe circuit sup- Bookheimer, S. Y. & Engel, S. A. Nat. Neurosci. 3,
coupling? Memory for a previously ports declarative memory formation, the 11491152 (2000).

Induction of ically expressed in the photoreceptors of the mature retina and

is crucial in photoreceptor differentiation. To examine whether
photoreceptor-specific iris-derived cells could acquire photoreceptor-specific pheno-
types as a result of ectopic expression of Crx, iris-derived cells
phenotypes in adult were infected with a replication-defective recombinant aden-
ovirus6,7. The iris-derived cells infected with Crx-transducing
mammalian iris tissue adenovirus expressed rhodopsin (Fig. 2a and d), whereas none
of the infected cells with enhanced green fluorescent protein
(EGFP)-transducing adenovirus did (Fig. 2c). Similar results were

Masatoshi Haruta1, Mitsuko Kosaka2, Yumi Kanegae3, obtained by using the anti-recoverin antibody8 that detects pho-
3 4
Izumu Saito , Tomoyuki Inoue , Ryoichiro Kageyama , 4 toreceptors and subpopulation of bipolar cells (Fig. 2b and d).
Akihiro Nishida1, Yoshihito Honda1 and Masayo Takahashi1 These results indicate that iris-derived cells have the potential to
differentiate into photoreceptors in response to Crx.
1 Department of Ophthalmology and Visual Sciences, Kyoto University Whereas the adenovirus infection study demonstrated the
Graduate School of Medicine, Sakyo-ku, Kyoto 606-8507, Japan rhodopsin-inductive effect of Crx in the iris-derived cells, it was
2 Japan Science and Technology Corporation/TOREST, Okayama Technology not clear whether these rhodopsin-expressing cells were differ-
Center, 5301 Haga, Okayama 701-1221, Japan entiated from mitotic progenitor-like cells or postmitotic cells
3 Laboratory of Molecular Genetics, Institute of Medical Science, University of that were committed to the neural fate. To examine whether Crx
Tokyo, Minato-ku, Tokyo 108-8639, Japan can induce rod generation in dividing cells, Crx was misex-
4 Institute for Virus Research, Kyoto University, Sakyo-ku, Kyoto 606-8507, Japan pressed with a retrovirus, which was infectious only to mitotic
Correspondence should be addressed to M.T. (masataka@kuhp.kyoto-u.ac.jp) cells. The iris-derived cells infected with the control retrovirus
CLIG9 (Fig. 3a), which regulates expression of only EGFP, were
Published online: 12 November 2001, DOI: 10.1038/nn762 large and flat, and none of them expressed rhodopsin (Fig. 3ce).
In contrast, the iris-derived cells infected with the CLIG-Crx
We show that iris tissue in the adult rat eye, which is embryoni- (Fig. 3b), which regulates expression of both Crx and EGFP, were
cally related to the neural retina, can generate cells expressing small and round, characteristic of the rod photoreceptors in
differentiated neuronal antigens. In addition, the Crx gene trans- monolayer culture (Fig. 3fh). Most EGFP-positive cells also
fer induced the specific antigens for rod photoreceptors in the expressed rhodopsin, although none of the uninfected (EGFP-
iris-derived cells, which was not seen in the adult hippocampus- negative) cells did. These results showed that Crx induction is
derived neural stem cells. Our findings demonstrate a remark- necessary for the iris-derived cells to express rhodopsin under
able plasticity of adult iris tissue with potential clinical these culture conditions. Moreover, as retrovirus-infected cells
applications, as autologous iris tissue can be feasibly obtained represent the population of mitotic cells at the time of infection,
with peripheral iridectomy. our findings suggest that it is possible to expand the cell source
During vertebrate eye development, the inner layer of the for rod photoreceptors before virus infection.
optic cup differentiates into the neural retina and iris- As a control, we examined if neural stem cells have the capac-
pigmented epithelium. This common developmental origin ity to differentiate into rod photoreceptors. A neural stem cell line
prompted us to test whether iris tissue could give rise to retinal derived from the adult rat hippocampus10 is one of the few proven
neurons. When the iris tissue of adult rats was plated and main- to have the potential to self-renew and differentiate into neurons
tained in serum-free media containing basic fibroblast growth and glial cells. When grafted to the retina, these neural stem cells
factor, many cells migrated out of the iris tissue and proliferat- are well integrated into the host retina, where they differentiate
ed as a monolayer of cells (Fig. 1a). Some of these iris-derived into neurons and glial cells, but cannot acquire rhodopsin
cells became immunoreactive for neurofilament 200 when trans- immunoreactivity1113. To determine whether Crx gene transfer to
ferred to an environment that promotes retinal cell differentia- the hippocampus-derived neural stem cells can also induce rod
tion1 (Fig. 1b). However, none of these cells expressed rhodopsin, photoreceptors, these cells were infected with recombinant ade-
a specific marker for rod photoreceptors2. novirus. EGFP-positive cells were detected when infected with
Both extrinsic cues and intrinsic properties regulate the choice EGFP-transducing adenovirus, but none of the cells acquired
of photoreceptor cell fates3. Crx4,5 is the homeobox gene specif- rhodopsin immunoreactivity when infected with Crx-transduc-
ing adenovirus (data not shown). These results suggest
that adult rat hippocampus-derived neural stem cells
a b are intrinsically restricted in their response to Crx under
these culture conditions. The common developmental
origin of the iris-pigmented epithelium and neural reti-
na may account for the iris-derived cells expressing
rhodopsin in response to a Crx cue.
Single pigmented ciliary margin cells, but not sin-
gle pigmented iris cells, can proliferate in vitro to form
spherical colonies of cells that can differentiate into
rod photoreceptors1,14. To test if the ciliary tissue could
also give rise to retinal neurons in adherent monolay-
Fig. 1. Iris-derived cells expressing neural antigens. (a) Iris tissue cultured in the er culture, we transferred the ciliary-derived cells pro-
presence of basic fibroblast growth factor for 3 days. (b) Some of the iris-derived liferated as an adherent monolayer of cells to a
cells expressed neurofilament 200 (green) when cultured in a differentiating envi-
ronment. Nuclei in cells stained with DAPI (4, 6-diamidino-2phenylinedole; blue).
differentiating environment. None of these ciliary-
Scale bar, 200 mm (a), 50 mm (b). derived cells showed rhodopsin immunoreactivity.

a b a
b
c f
c d
d g
Fig. 2. Induction of rod photoreceptor-specific antigens with Crx gene

transfer. (a, b, d) Iris-derived cells infected with Crx-transducing aden-
e h
ovirus expressed rhodopsin (a, red; 523 of 5011, 10.6 1.2%) and
recoverin (b, red; 336 of 2854, 11.8 2.5%). (c) Iris-derived cells
infected with EGFP-transducing adenovirus expressed enhanced green
fluorescent protein (green; 1896 of 5394, 35.4 11.4%); however, none
of the infected cells expressed rhodopsin. (ac) Nuclei in cells stained
with DAPI (blue). Scale bar, 25 m.
Fig. 3. Induction of rhodopsin expression from mitotic progenitor-like
cells. (a, b) Structures of the recombinant retrovirus CLIG and CLIG-
Crx. The internal ribosomal entry sequence (IRES) allows for bicistronic
However, when these cells were infected with Crx-transducing expression. LTR, long terminal repeat. (ce) Infection with the control
adenovirus, they became immunoreactive for rhodopsin (data retrovirus CLIG. The virus-infected cells (EGFP positive, 212 of 2158,
not shown). These results indicate that iris- and ciliary-derived 9.8%) were flat and large. None of them expressed rhodopsin.
cells behave similarly under the culture conditions in this study. (fh) Infection with the retrovirus CLIG-Crx. The virus-infected cells
Rod photoreceptors fail to develop in dissociated-cell cultures of (EGFP positive, 187 of 2106, 8.9%) were small and round. Most of the
virus-infected cells expressed rhodopsin (179 of 187, 96%).
the neonatal neural retina, although they develop in large num-
(e, h) Nuclei in cells stained with DAPI (blue). Scale bar, 50 m.
bers in pellet cultures15. The spherical colonies obtained from
ciliary margin cells may provide a differentiating environment
similar to that of pellet cultures for neonatal retinal cells. If the hippocampus-derived neural stem cells, J.F. McGinnis and R.J. Elias for anti-
iris-derived cells could be cultured to yield spherical colonies, recoverin antibody and C.J. Barnstable for the information on the anti-
such cells may also give rise to photoreceptors without the need rhodopsin antibody. This study was supported by a Grant-in-Aid from the
for homeobox gene transfer. Ministry of Education, Science, Sports and Culture of Japan (No. 13671834,
Clinically, using iris tissue as the source of retinal regeneration 13210077) and a grant from Japan Science and Technology Corporation.
offers the ability to feasibly obtain autologous tissue with periph-
eral iridectomy, a long established method of ophthalmic surgery. RECEIVED 17 JULY; ACCEPTED 11 OCTOBER 2001
Obtaining autologous ciliary tissue, on the other hand, necessi-
tates such a traumatic surgical intervention as to make this 1. Tropepe, V. et al. Science 287, 20322036 (2000).
approach unrealistic. The present study shows that iris tissue in 2. Barnstable, C. J. Nature 286, 231235 (1980).
the adult mammalian eye retains a remarkable plasticity to give 3. Cepko, C. L., Austin, C. P., Yang, X., Alexiades, M. & Ezzeddine, D. Proc. Natl.
Acad. Sci. USA 93, 589595 (1996).
rise to cells expressing neuronal antigens. In addition, iris-derived 4. Furukawa, T., Morrow, E. M. & Cepko, C. L. Cell 91, 531541 (1997).
cells can express photoreceptor-specific antigens with Crx gene 5. Chen, S. et al. Neuron 19, 10171030 (1997).
transfer, raising the possibility that these cells constitute a poten- 6. Miyake, S. et al. Proc. Natl. Acad. Sci. USA 93, 13201324 (1996).
7. Kanegae, Y., Makimura, M. & Saito, I. Jpn. J. Med. Sci. Biol. 47, 157166
tial source of retinal transplantation in patients with retinal degen- (1994).
erative diseases or damaged retinas. (For Methods, see the 8. McGinnis, J. F. et al. J. Neurosci. Res. 55, 252260 (1999).
9. Hojo, M. et al. Development 127, 25152522 (2000).
supplementary information page of Nature Neuroscience on line.) 10. Palmer, T. D., Takahashi, J. & Gage, F. H. Mol. Cell Neurosci. 8, 389404
(1997).
Note: Supplementary methods are available on the Nature Neuroscience web site 11. Takahashi, M., Palmer, T. D., Takahashi, J. & Gage, F. H. Mol. Cell Neurosci.
12, 340348 (1998).
(http://neurosci.nature.com/web_specials). 12. Nishida, A. et al. Invest. Ophthalmol. Vis. Sci. 41, 42684274 (2000).
13. Young, M. J., Ray, J., Whiteley, S. J., Klassen, H. & Gage, F. H. Mol. Cell
Neurosci. 16, 197205 (2000).
14. Ahmad, I., Tang, L. & Pham, H. Biochem. Biophys. Res. Commun. 270,
ACKNOWLEDGEMENTS 517521 (2000).
We thank C.L. Cepko and T. Furukawa for Crx cDNA and F.H. Gage for adult 15. Watanabe, T. & Raff, M. C. Neuron 4, 461467 (1990).

Melanopsin in cells of lateral and contralateral to the FG injection. Although the extent
of retrograde labeling differed between cases, approximately 70%
origin of the of RGCs that were intensely labeled for melanopsin mRNA were
also retrogradely labeled. Both calculations are likely to under-
retinohypothalamic tract estimate the actual percentage of colocalization, because techni-
cal factors limit the efficiency of the combined labels. Therefore,
most RGCs that project to the SCN express melanopsin, and a
Joshua J. Gooley, Jun Lu, Thomas C. Chou, majority of melanopsin-containing RGCs project to the SCN.
Thomas E. Scammell and Clifford B. Saper These observations suggest that RGCs that contain
melanopsin are particularly well poised to provide photic infor-
Department of Neurology and Program in Neuroscience, Harvard Medical mation to the SCN. Melanopsin in these retinohypothalamic
School, Beth Israel Deaconess Medical Center, Boston, Massachusetts 02215, RGCs may therefore mediate the photic entrainment of circa-
USA
dian rhythms in mice lacking rods and cones. Although a high
Correspondence should be addressed to C.B.S. (csaper@caregroup.harvard.edu) percentage of RHT RGCs express melanopsin, RHT cells may
also receive other photic signals through rods and cones in intact
Published online: 19 November 2001, DOI: 10.1038/nn768 animals. In addition, the photopigments cryptochrome 1 and 2
have been localized to RGCs of the mouse retina7. Further exper-
All known eukaryotic organisms exhibit physiological and behav- iments will be necessary to determine whether cryptochromes
ioral rhythms termed circadian rhythms that cycle with a near-24- are involved in circadian photic entrainment. However,
hour period; in mammals, light is the most potent stimulus for melanopsin may now be considered a primary candidate pho-
entraining endogenous rhythms to the daily light cycle. Photic topigment for mediating circadian entrainment.
information is transmitted via the retinohypothalamic tract
(RHT) to the suprachiasmatic nucleus (SCN) in the hypothala- Acknowledgements
mus, where circadian rhythms are generated, but the retinal pho- This work was supported by USPHS grants HL60292, MH62589 and
topigment that mediates circadian entrainment has remained HL07901.
elusive. Here we show that most retinal ganglion cells (RGCs)
that project to the SCN express the photopigment melanopsin. Competing interests statement
The phase of circadian rhythms in rodents is shifted most The authors declare that they have no competing financial interests.
effectively by light ranging from 480511 nm, consistent with an
opsin-based photopigment13. However, mice lacking rods and RECEIVED 11 OCTOBER; ACCEPTED 1 NOVEMBER 2001
cones have normal circadian entrainment, suggesting that a novel
photopigment mediates phase-shifting in response to light4. 1. Takahashi, J. S., DeCoursey, P. J., Bauman, L. & Menaker, M. Nature 308,
Recently, melanopsin, an opsin-based photopigment, was local- 186188 (1984).
ized to the RGC layer of rodents and primates5. We therefore test- 2. Provencio, I. & Foster, R. G. Brain Res. 694, 183190 (1995).
3. Yoshimura, T. & Ebihara, S. J. Comp. Physiol. A 178, 797802 (1996).
ed whether RGCs that express melanopsin project to the SCN. 4. Freedman, M. S. et al. Science 284, 502504 (1999).
We injected the right SCN of 10 rats with FluoroGold (FG) 5. Provencio, I. et al. J. Neurosci. 20, 600605 (2000).
to retrogradely label the retinohypothalamic RGCs. Four of the 6. Moore, R. Y., Speh, J. C. & Card, J. P. J. Comp. Neurol. 352, 351366 (1995).
7. Miyamoto, Y. & Sancar, A. Proc. Natl. Acad. Sci. USA 95, 60976102 (1998).
injections were confined to the SCN and did not include the optic
chiasm or optic tract (Fig. 1a). In these animals, FG labeled a dis-
tinct subset of widely distributed RGCs, corresponding to
type III or W cells, as previously reported6.
For in situ hybridization, we used a 957-base-pair mouse
melanopsin riboprobe5. Melanopsin transcript occurred in a pat-
tern similar to that previously described5, with a scattered pop-
ulation of cells showing intense hybridization, predominantly in
the RGC layer (Fig. 1b).
In doubly labeled sections, 74.2 0.3% (mean s.e.m.) of
retrogradely labeled RGCs also expressed melanopsin mRNA
(Fig. 1c), with a similar percentage of double labeling in eyes ipsi-
a
Fig. 1. Colocalization of retrogradely labeled FluoroGold (FG) and
melanopsin transcript in retinal ganglion cells of rat. (a) The suprachi-
asmatic nucleus (asterisk) was injected by glass micropipette with 3 nl
of 5% FG, resulting in retrograde labeling of the contralateral
suprachiasmatic nucleus (arrow) due to reciprocal innervation. The
injection avoided the optic chiasm. (b) In situ hybridization for
melanopsin localized with NTB-2 emulsion autoradiography, demon-
b
strating a group of three intensely labeled cells (arrows) in the gan-
glion cell layer. Light diffuse labeling over all three cellular layers5 was
similar to labeling seen with sense probe. (c) All three intensely
labeled RGCs were retrogradely labeled with FG (arrows). 3v, third
ventricle; oc, optic chiasm; scn, suprachiasmatic nucleus; gcl, ganglion
cell layer; inl, inner nuclear layer; onl, outer nuclear layer. Scale bar,
200 m (a), 50 m (b, c).
c
Does bouton morphology We examined the axons of 3 spiny neurons, which were
recorded in area 17 of anesthetized cats (protocols approved by
optimize axon length? the Veterinary Department of the Canton of Zurich; for details,
see ref. 10). The neurons were filled intracellularly with horse-
radish peroxidase. Two were layer 3 pyramidal cells with extensive
John C. Anderson and Kevan A. C. Martin axon collaterals in layers 2, 3 and 5. The third cell was a spiny
stellate cell from layer 4A whose axon arborized in layers 2, 3 and
Institute for Neuroinformatics, University of Zurich and ETH Zurich, 4. We selected collateral segments located in layer 3 that had a
Winterthurerstr. 190, 8057 Zurich, Switzerland

mix of both bouton types.
Correspondence should be addressed to J.C.A. and K.A.C.M Eighty-five boutons were serially sectioned, photographed
(jca@ini.phys.ethz.ch and kevan@ini.phys.ethz.ch)
in the electron microscope (EM), and reconstructed together
Published online: 19 November 2001, DOI: 10.1038/nn772 with their targets. The morphological type of each bouton was
determined from the serial EM reconstructions. The synaptic
The total length of cortical axons could be reduced if the parent targets of the labeled boutons were positively identified using
axons maintained straight trajectories and simply connected to established criteria. Spines and dendritic shafts were the only
dendritic shafts via spine-like terminaux boutons and to den- synaptic targets and were readily distinguished. Forty-five of
dritic spines via bead-like en passant boutons. Cortical axons the boutons were en passant and formed 49 synapses, and
from cat area 17 were reconstructed from serial electron micro- 40 boutons were terminaux and formed 43 synapses. As is typ-
graphs and their bouton morphology was correlated with their ical for spiny cortical neurons, spines were the major target. Of
synaptic targets. En passant or terminaux boutons did not differ the 34 synapses formed by the spiny stellate axon, 24 (70%)
in the proportion of synapses they formed with dendritic spines were with spines and similarly, of 58 synapses formed by the
and shafts, and thus, the two morphological variants of synap- two pyramidal cell axons, 37 (64%) were spines.
tic bouton do not contribute directly to optimizing axon length. In one reconstruction of a 20-m length of the spiny stel-
The bulk of the brain consists of axonal wiring. In the gray late axon (Fig. 1a), all but one of the boutons were located in
matter of the neocortex, each cubic millimeter contains about a tight cluster, where they formed synapses with dendritic
4 km of axon1, so removing axonal zigzags is not trivial: if only shafts and spines. This cluster illustrates the morphological
0.1 mm were pruned off each axon in cat area 17, it would save variety of terminaux boutons and the size of their necks rela-
3 km of wire. It has been proposed that one role of dendritic tive to the parent axon. The target spines were of the same
spines is to help optimize the length of axonal wire2,3. The idea dimension as the presynaptic boutons. At the center of the
is that instead of zigzagging through the neuropil to contact cluster was a single en passant bouton. Terminaux boutons
their specific target dendrites, axons simply grow in economi- were formed even when the parent axon actually touched the
cally straight trajectories and form en passant synaptic boutons, target. In the example in Fig. 1b, the synaptic target (a spine)
leaving to dendrites the task of emitting spines to catch the of the spiny stellate axon was directly on the path of the parent
passing axons. This gives spiny dendrites the active role in select- axon. Yet, instead of forming an en passant bouton, the axon
ing particular axons during development and learning4. New formed a terminaux bouton, whose slender neck had to wrap
evidence pointing to the involvement of dendritic spines in around a myelinated axon (approximately 0.8 m diameter)
forming connections has come from reports of motile spines5 to reach its target spine. Even when the targets were aligned
and of new spines appearing during long-term potentiation68. along the same path as the axon, multiple synapses were rarely
However, one-fifth of cortical neurons lack spines and thus can- formed. In the only two cases found, two closely spaced en
not exploit such a mechanism, so some degree of axonal zigzag- passant boutons formed two synapses with the dendritic shafts
ging would seem inevitable. of smooth neurons, before the trajectories of the axon and
In all these discussions of spines, it has been completely over- dendrites deviated (see supplementary reconstructions, avail-
looked that cortical axons too can produce spine-like structures, able on the Nature Neuroscience web site).
called terminaux boutons, whose dimensions match those of In the analysis of the distribution of all the targets of the
dendritic spines. Do terminaux boutons exist to prevent 3 axons (Fig. 1c), spines formed 63% of the targets of terminaux
zigzags in the parent axons by connecting to dendritic shafts by boutons and 69% of en passant boutons. Dendritic shafts formed
means of axonal spines? Evidence in sup-
port of this comes from the pyramidal cells a b
of layer 6, which, unusually for pyramidal c Bouton morphology and target specificity
cells, form synapses mainly with dendritic 80
70 n = 43 n = 49
shafts9 and whose axons bear mainly ter- 60
minaux boutons 9,10 . This is the mirror- 50
Percent
image of the pattern for spiny cells in other 40
cortical layers whose axons bear mainly 30
en passant boutons and form synapses 20
10
mainly with spines. These observations sug-
0
gest an only connect hypothesis3 in which Spines - Dendrites
terminaux rmin Spines Dendrites
Terminaux
es dendrite En passant
both dendritic and axonal spines are spe-
cializations that allow axons to maintain Fig. 1. Reconstructions from serial electron microscope sections of segments of axon and
economically straight trajectories. We test- summary histograms. (a) Axon, light blue; target dendritic shafts, red; target spines, orange;
dendritic postsynaptic densities, yellow. (b) Axon, light blue; spine, transparent brown; post-
ed this hypothesis by correlating the target
synaptic density, yellow; unlabeled myelinated axon, transparent mauve. Scale bars (a, b), 1 m.
(dendritic spine/shaft) with the bouton type (c) Histograms of all identified targets of the two bouton types for all 92 synapses. Electron
in the same local region of neuropil. microscope reconstructions created using Nuages and Blue Moon Rendering Tools.
the remainder of the targets. The distributions for the two bou- Note: Supplementary reconstructions are available on the Nature Neuroscience
ton types were not significantly different (chi-square test). The web site (http://neuroscience.nature.com/web_specials).
morphology of the bouton is not correlated with the type of
synaptic target. It seems that axons and dendrites have no trou-
ble finding each other, regardless of bouton type or whether the ACKNOWLEDGEMENTS
target is dendritic spine or shaft. Thus, the only connect hypoth- We thank T. Binzegger for help with calculations and reconstructions. Additional
esis fails to account for the data. support from EU (QULG3-1999-01064) and HFSP (RG0123/2000-B) grants to
It is tempting to write off the differences between the two K.A.C.M.
bouton types as being of any particular significance. Are they

simply an expression of some developmental quirk that pro- RECEIVED 3 AUGUST; ACCEPTED 24 OCTOBER 2001
duces one or the other kind of bouton? However, the morpho-
logical similarity of the terminaux boutons to dendritic spines 1. Braitenberg, V. & Schz, A. Anatomy of the Cortex (Springer, Berlin, 1991).
prompts another interpretation. Perhaps, as with dendritic 2. Peters, A. & Kaiserman-Abramof, I. R. Am. J. Anat. 127, 321356 (1970).
spines, which compartmentalize calcium1113, the fine necks of 3. Swindale, N. V. Trends Neurosci. 4, 240241 (1981).
4. Gray, E. G. Trends Neurosci. 5, 56 (1982).
these axonal spines also prevent calcium from diffusing rapid- 5. Fischer, M., Kaech, S., Knutti, D. & Matus, A. Neuron 20, 847854 (1998).
ly into the parent axon during an action potential? Because each 6. Engert, F. & Bonhoeffer, T. Nature 399, 6670 (1999).
7. Maletic-Savatic, M., Malinow, R. & Svoboda, K. Science 283, 19231927 (1999).
bouton would retain more residual calcium after each impulse, 8. Toni, N., Buchs, P.-A., Nikonenko, I., Bron, C. R. & Muller, D. Nature 402,
synapses formed by terminaux boutons might show more presy- 421425 (1999).
naptic facilitation than the en passant boutons. The layer 6 pyra- 9. McGuire, B. A., Hornung, J.-P., Gilbert, C. D. & Wiesel, T. N. J. Neurosci. 4,
30213033 (1984).
midal synapses, which are formed mainly by terminaux boutons, 10. Martin, K. A. C. & Whitteridge, D. J. Physiol. (Lond.) 353, 463504 (1984).
show strong facilitation due to an increased probability of trans- 11. Muller, W. & Connor, J. A. Nature 354, 7376 (1991).
mitter release14. If bouton morphology does indeed influence 12. Guthrie, P. B., Segal, M. & Kater, S. B. Nature 354, 7680 (1991).
13. Majewska, A., Tashiro, A. & Yuste, R. J. Neurosci. 20, 82628268 (2000).
synaptic dynamics, it will be a revision of the present view that 14. Tarczy-Hornoch, K., Martin, K. A. C. & Stratford, K. J. Cereb. Cortex 9,
cortical axons exist only to connect. 833843 (1999).
Extraocular proprioceptive signals reach the SC8,9, but mod-

Passive eye displacement eling studies have so far emphasized the importance of effer-
alters auditory spatial ence copy and visual feedback in SC motor function (for
example, see ref. 10). Furthermore, the motor aspects of SC
receptive fields of cat function can operate accurately when proprioceptive feedback
is abolished11. We assessed the involvement of proprioceptive
superior colliculus neurons feedback in sensory processing within the SC by testing
whether passive eye displacement alters auditory SRFs in anes-
thetized (1% halothane, 66% N2O), paralyzed (pancuroni-
J. Curtis Zella1, John F. Brugge1 and Jan W. H. Schnupp1,2 um bromide, 1 mg/kg every 3 h) cats in complete darkness. In
this preparation, neither efference copy nor visual input could
1 Department of Physiology and Waisman Center, 627 Waisman Center, contribute to any observed changes. Experimental protocols
University of Wisconsin, Madison, Wisconsin 53705
were approved by the University of Wisconsin Institutional
2 Laboratory of Physiology, Oxford University, Parks Road, OX1 3PT
Animal Care and Use Committee.
Correspondence should be addressed to J.S. (jan.schnupp@physiol.ox.ac.uk) Twenty neurons from 11 cats were tested for the effects of
passive eye movements. The position of the contralateral (left)
Published online: 19 November 2001, DOI: 10.1038/nn773 eye was manipulated by tension on four sutures (6-0 silk)
passed through the nasal, temporal, ventral and dorsal margin
The superior colliculus (SC) is thought to use a set of superim- of the sclera. The sutures were attached to a mechanical device
posed, topographically organized neural maps of visual, audito- which held the eye securely either in its straight-ahead (con-
ry, somatosensory and motor space to direct the eyes toward trol) position, or displaced by 23 in a contralateral (tempo-
novel stimuli1,2. Auditory spatial response fields (SRFs) of SC ral) and downward direction. This displacement was chosen
neurons may change when an animal moves its eyes, presumably for maximal proprioceptive activation (stretching both the
to compensate for the resulting misalignment of visual and audi- superior and medial recti, and possibly the superior oblique,
tory sensory spatial reference frames36, but the mechanisms while relaxing the inferior and lateral recti). Tension on the
responsible for these SRF changes remain unknown. Here we sutures was maintained for the control and the deviated con-
report that passive deviation of the eye in anesthetized, paralyzed ditions. We used short (10100 ms) Gaussian noise bursts
animals can profoundly affect the auditory responsiveness of SC delivered in virtual acoustic space (VAS) and standard extra-
neurons, but seems insufficient by itself to provide adaptive shifts cellular recordings12 to map auditory SRFs of single SC neu-
of auditory SRFs. rons. SRFs were constructed from 3 to 5 randomly interleaved
In awake animals, changes in eye position either shift the stimulus presentations at each of 324 different VAS directions.
rostral edge or the center of auditory SRFs in the SC 36, or In some cases, we repeated SRF measurements in the eye-
modulate the overall strength of auditory responses while not deviated and the control position several times, and at different
systematically shifting the SRF5,6. It is unknown whether these sound intensities. Recording sites were confirmed histologi-
SRF changes are mediated through efference copy of eye move- cally to be distributed evenly throughout the central two-thirds
ment commands, through sensory feedback from the oculo- of the intermediate and deep layers of the SC. With the eyes in
motor plant, or through some combination of the two. their control position, SC neurons exhibited circumscribed

Fig. 1. Contraction of SRF resulting Control 252 spikes C296
a 90
Elev Start
from passive eye displacement. To 60 2
the left of each SRF is a dot raster 30

1 .3
0 .7
taken for all azimuthal positions 0
0
-3 0
along an elevation of +10 (arrow) 180 135 90 45 0 -4 5 -9 0 -1 3 5 -1 8 0
through the SRF. Right, raw spatial 1 A zim S tart
response profile data. (Diameters of

the filled circles are proportional to
the mean spike count evoked at the Eye displaced
b C296
65 spikes 90
virtual sound direction indicated by
Elev S tart
60 2
the axes.) The SRFs in the center 30

1 .3
0 .7
were generated from the raw data 0
0
-3 0
by interpolation (Delaunay triangu- 180 135 90 45 0 -4 5 -9 0 -1 3 5 -1 8 0
lation), smoothed (30 wide running 2 A zim S tart
averaging filter) and plotted using

an equal area projection. Zero
degrees azimuth is at the midline, c Return 379 spikes 90
C 2 96
Elev S tart
negative directions to the cats 60 2
1 .3
left. All data were recorded 30
0 .7
0
in complete darkness. Number 0
-3 0
below each SRF marks the order in 1 80 1 35 90 45 0 -4 5 -9 0 -1 35 -1 80
A zim S tart
which the SRF was recorded. The 3
black contour line delineates the
best area (>75% of maximum). The
d
cross indicates SRF centroid direc- 180 Eye displaced C296
81 spikes 90
tion. To test for statistical signifi-
Elev S tart
Elev
60 2
cance, the mean-squared-difference 30

1 .3
0 0 .7
between two SRFs was compared 0
0
against a bootstrapped null hypothe- -3 0
180 135 90 45 0 -4 5 -9 0 -1 3 5 -1 8 0
sis distribution. Z-scores outside 180 4 A zim S tart
the interval (2.33, 2.33) indicate 0 10 20 30 40 50 0021-3

ms
statistical significance ( = 0.01). Z-
0 1 .6
scores (a) versus (b), Z = 9.97; (b)
Spikes/stimulus
versus, (c), Z = 12.26; (c) versus (d),
Z = 13.44; (a) versus (c), Z = 1.42; (b) versus (d), Z = 0.0126. Z-scores, shown for each intensity. Not shown because of space constraints are return
SRFs obtained at 20 and 35 dB, which are very similar to controls. (For supplementary Methods, see the Nature Neuroscience web site.)
SRFs with a defined best area and a gradient of response (Fig. 1). Systematic shifts in the centroids of the SRF were not
strength radiating from it, as described by others13,14. In 8 units observed (Fig. 3). SRFs were stable over hours of recording,
(40%), static eye displacement resulted in a significant change and effects were independent of the order in which the SRFs
in SRF size and in overall spike count (Figs. 13). When the were obtained. Eye displacement alone generated neither
eye was returned to its resting position (return), the SRF evoked activity nor changes in background firing level. Six
returned toward its original control configuration. These additional neurons exhibited reversible changes in SRFs, but
changes could be observed over a range of stimulus intensities the results were not statistically significant. The remainder
(Fig. 2). Half the units that exhibited a significant dependence showed no systematic relationship to changes in eye position.
on eye position responded with an increase of SRF size and The finding that passive eye displacement produced no sys-
magnitude, whereas the other half showed a marked decrease tematic shifts in SRF centroids suggests that mechanisms
underlying the compensatory shifts of auditory SRFs described
Control Eye Displaced previously in awake monkey SC3,4 may not have been engaged
10 dB
Z = 0.08 in our preparation. Our data are, however, very similar to pre-
vious observations showing a modulation of overall respon-
siveness of SC5,6 and IC7 neurons in awake animals which is
3 5
not accompanied by SRF shifts. We could rule out corollary
25 dB
Z = 5.65 discharge or efference copy of motor signals due to voluntary
gaze shifts, as well as visual input, as our animals were anes-
thetized, paralyzed and in the dark. Thus, we conclude that
2 4 periorbital proprioceptive feedback seems capable of pro-
45 dB
Z = 7.70
foundly influencing auditory responsiveness, providing a mod-
ulating influence over brainstem circuits underlying auditory
SRFs. These modulations of auditory SRF size and changes in
1 6 overall response magnitude may, in the alert animal, work in
9939-15
concert with other mechanisms underlying compensatory shifts
Fig. 2. Expansion of spatial response field (SRF) resulting from passive eye
in the location of an SRF.
displacement obtained at three sound intensities above the neurons
acoustic threshold. SRF representation as in Fig. 1. To compare changes in
SRF across intensity, data were normalized to the maximum response at Note: Supplementary Methods are available on the Nature Neuroscience web
each intensity. Maximal spike counts, 10 dB, 0.81; 25 dB, 2.66; 45 dB, 1.43. site (http://neuroscience.nature.com/web_specials).

Fig. 3. Population results. (a) Shifts in spatial response field (SRF) cen-
a troids with changes in eye position. Data shown for 12 neurons. (Eight
neurons were excluded because their relatively poor spatial tuning made
centroid direction an unreliable measurement.) Some neurons were
tested at multiple sound levels, leading to a total of 17 data sets. Crosses,
15o control and eye return conditions; triangles, eye displacement condition.
Displacement direction shown by arrow. (b) Histograms summarizing dis-
tribution of centroid shifts shown in (a) and Z-scores for all SRFs in our
b 10 Azimuth N = 37
10
N = 64 sample (see Fig. 1 legend). Yellow shading denotes significant differences
5 between control and eye-displaced conditions. (c) Azimuthal functions

Count
Count
0 5 through the SRF center for four neurons showing SRF expansion or con-
5 traction with relatively small shifts in the response peak. Red, control or
10 Elevation
0
return; blue, eye displaced; open, original condition; closed, repeat.
20 10 0 10 20 20 10 0 * 10 20
Degrees Z-Score
c 1
0.8
EL 10o 9903-12
2
EL 10o 0021-3
RECEIVED 8 JUNE; ACCEPTED 9 OCTOBER 2001
1.5
0.6
1
0.4
Average Spike Count
0.2 0.5 1. Jay, M. F. & Sparks, D. L. J. Neurophysiol. 57, 2234 (1987).

2. Stein, B. E. & Meredith, M. A. (MIT Press, Cambridge, 1993).
0 0
3. Jay, M. F. & Sparks, D. L. Nature 309, 345347 (1984).
2
EL 10o 0021-2
2.5
EL 20o 9939-15 4. Jay, M. F. & Sparks, D. L. J. Neurophysiol. 57, 3555 (1987).
1.5 2 5. Peck, C. K., Baro, J. A. & Warder, S. M. Exp. Brain Res. 103, 227242 (1995).
1.5 6. Hartline, P. H., Pandey Vimal, R. L., King, A. J., Kurylo, D. D. &
1 Northmore, D. P. M. Exp. Brain. Res. 104, 402408 (1995).
1
0.5 7. Groh, J. M., Trause, A. S., Underhill, A. M., Clark, K. R. & Inati, S. Neuron 29,
0.5
509518 (2001).
0 0
180 92.57 0 92.57 180 180 92.57 0 92.57 180
8. Abrahams, V. C. Prog. Brain. Res. 50, 325334 (1979).
9. Nelson, J. S., Meredith, M. A. & Stein, B. E. J. Neurophysiol. 62, 13601374
Azimuth (deg)
(1989).
10. Quaia, C., Lefevre, P. & Optican, L. M. J. Neurophysiol. 82, 9991018 (1999).
ACKNOWLEDGEMENTS 11. Guthrie, B. L., Porter, J. D. & Sparks, D. L. Science 221, 11931195 (1983).
12. Brugge, J. F., Reale, R. A. & Hind, J. E. J. Neurosci. 16, 44204437 (1996).
Supported by NIH grants DC00116 and HD03352 and a Defeating 13. Middlebrooks, J. C. & Knudsen, E. I. J. Neurosci. 4, 26212634 (1984).
Deafness/Dunhill Research Fellowship to J.W.H.S. 14. King, A. J. & Hutchings, M. E. J. Neurophysiol. 57, 596624 (1987).
Memory retrieval synthesis inhibitor metyrapone, administered shortly before

water-maze retention testing, blocks the impairing effects of the
impairment induced by lesion on memory retrieval. These findings suggest that elevated
adrenocortical activity is critical in mediating memory retrieval
hippocampal CA3 lesions is deficits induced by hippocampal damage.
Male SpragueDawley rats (1012 weeks old) from Charles
blocked by adrenocortical River Laboratories (Wilmington, Massachusetts) were housed
suppression individually with food and water ad libitum (lights on 7.00
19.00 h). All procedures were done in compliance with NIH
guidelines and were approved by UC Irvines Institutional Ani-
Benno Roozendaal1, Russell G. Phillips2, Ann E. Power1, mal Care and Use Committee. Rats were anesthetized with sodi-
Sheila M. Brooke2, Robert M. Sapolsky2 and um pentobarbital (50 mg/kg, intraperitoneally), and a low dose
of kainic acid (0.12 g per 3.0 l of phosphate buffer,
James L. McGaugh1
pH 7.4; Sigma, St. Louis, Missouri) was infused bilaterally over
1 Center for the Neurobiology of Learning and Memory, and Department of 180 seconds at the following coordinates: 3.3 mm from breg-
Neurobiology and Behavior, University of California, Irvine, California ma; 2.0 mm from midline; 3.5 mm from skull surface4. For
92697-3800, USA sham lesions, the needle tip was inserted but no infusion was
2 Department of Biological Sciences, Stanford University, Stanford, California given. Lesion size was determined in 35-m brain slices by mea-
94305-5020, USA suring the length and width of damage with a calibrated ocular
Correspondence should be addressed to B.R. (broozend@uci.edu) grid. Areas of damage in seven coronal sections equidistantly
spaced along the rostrocaudal axis were used to determine the
Published online: 19 November 2001, DOI: 10.1038/nn766 volume of damage. In agreement with previous evidence5, dam-
age as identified by this technique was restricted to the pyrami-
There is evidence that in rats, partial hippocampal lesions or dal neurons of CA3 region (Fig. 1). Right and left hemispheric
selective ablation of the CA3 subfield can disrupt retrieval of spa- lesion volumes were averaged to calculate total lesion volume.
tial memory1 and that hippocampal damage disinhibits hypo- The four groups of lesioned rats did not differ in total lesion vol-
thalamic-pituitary-adrenocortical (HPA)-axis activity, thereby umes (ANOVA, F3,56 = 0.42; p = 0.74; data not shown).
elevating plasma levels of adrenocorticotropin and corticos- Water-maze training commenced 810 days after surgery and
terone2,3. Here we report evidence that attenuation of CA3 lesion- consisted of 4 trials on each of 2 consecutive days. Training and
induced increases in circulating corticosterone levels with the testing were conducted between 10.00 and 14.00 hours, at the

Fig. 1. Kainic acid infusions induce selective damage to the CA3 field
a b with pyramidal neuron loss and gliosis. (a) Lower magnification. Scale
bar, 500 m. (b) Higher magnification. Scale bar, 200 m. DG, dentate
gyrus; CA1, CA3, Ammons horn. Arrowheads point to damaged area.
(Fishers, p < 0.01). Lesioned rats also had longer latencies to

cross the platform location (Fishers, p < 0.01; Fig. 2c). CA3
lesions did not influence total path lengths on the probe trial
(Fishers, p = 0.32; data not shown). These results indicate that

the CA3 lesions impaired retention of spatial information and
that the probe trial performance impairment was not due to
attenuation of motor performance or swimming pattern.
Corticosterone levels were determined by radioimmunoassay
in plasma from trunk blood collected within 90 seconds after the
nadir of the rat circadian cycle for corticosterone. Four starting probe trial (see ref. 6, for Methods). Corticosterone levels were
positions were spaced equally around a black circular water tank significantly elevated in vehicle-injected rats with CA3 damage
(25C; diameter, 1.83 m; height, 58 cm). An escape platform relative to sham-operated controls (Fishers, p < 0.01; Fig. 2d).
(20 25 cm) was submerged in a fixed location at a depth of Consistent with previous findings6,7, metyrapone treatment did
2.5 cm below the water surface. After mounting the platform, the not affect plasma corticosterone levels or retention performance
rats were retained on it for 10 seconds and were then placed in a in sham-operated controls. However, metyrapone blocked the
holding cage for 20 seconds between trials. A two-way ANOVA lesion-induced elevation of corticosterone levels (Fishers,
of training escape latencies indicated that the sham-operated and p < 0.01; Fig. 2d) and blocked the retention impairment induced
CA3-lesioned rats did not differ in acquisition performance by CA3 lesions, as assessed both by time spent in the target quad-
(F1,672 = 1.08; p = 0.30; Fig. 2a). The lesioned animals were also rant (Fishers, p < 0.05; Fig. 2b) and by platform-crossing laten-
not impaired on the first trial of the second day of training (which cies (p < 0.01; Fig. 2c). Metyrapone did not affect total path
is a retention test), indicating that on day 1, they had acquired lengths (Fishers, p = 0.23 and p = 0.29, respectively; data not
procedural aspects of the task (swimming away from the sides of shown). Metyrapone is known to induce anxiolytic and/or seda-
the tank) without acquiring explicit spatial information. tive effects and may also directly influence adrenal steroid recep-
Memory for the platform location was tested on day 3 with a tors, which could potentially alter water-maze probe trial
60-second free-swim probe trial using a novel starting position. performance7. However, such alternative explanations of the find-
Ninety minutes before the probe trial, they received an injection ings were excluded, as metyrapone did not block retention
of vehicle or metyrapone, a corticosterone-synthesis inhibitor impairment in CA3-lesioned rats given corticosterone supple-
that prevents corticosterone secretion above basal levels, even mentation (2.0 mg/kg, subcutaneously; Sigma) 30 minutes before
in response to major stressors6,7. Metyrapone (35 mg/kg, sub- the probe trial. Time spent in the target quadrant (two-tailed t-
cutaneously; Sigma) was dissolved in 40% polyethylene glycol test, p < 0.05; Fig. 2b) and platform-crossing latencies (p < 0.01;
and 60% saline. Vehicle-injected control rats spent significantly Fig. 2c) were both impaired in corticosterone-injected rats rela-
more time in the target quadrant than chance level (Fig. 2b). tive to those of rats given a vehicle injection (4% ethanol in
Retention of vehicle-treated rats with CA3 damage was impaired saline), and their performance resembled that of CA3-lesioned
relative to that of sham-operated controls, as indi-
cated by less time spent in the target quadrant
a b
Fig. 2. Effect of metyrapone (35 mg/kg) injections

90 min before the retention test on CA3 lesion-induced
changes in water-maze spatial performance and plasma
corticosterone. (a) Training escape latencies
(mean s.e.m.) on a water-maze spatial task. (b) Time
spent in the target quadrant (mean s.e.m.) on the probe
trial (ANOVAs, lesion effect, F1,65 = 6.80, p = 0.01;
metyrapone effect, F1,65 = 1.06, p = 0.31; lesion
metyrapone effect, F1,65 = 4.50, p = 0.04). (c) Latencies
(mean s.e.m.) to cross the platform location on the
probe trial (ANOVAs, lesion effect, F1,65 = 6.32, p = 0.01; c d
metyrapone effect, F1,65 = 5.41, p = 0.02; lesion
metyrapone effect, F1,65 = 6.96, p = 0.01). (d) Plasma
corticosterone levels (mean s.e.m.) as assessed immedi-
ately after the probe trial (ANOVAs, lesion effect,
F1,65 = 8.01, p = 0.006; metyrapone effect, F1,65 = 7.63,
p = 0.008; lesion metyrapone effect, F1,65 = 14.03,
p = 0.0004). Double asterisk, p < 0.01 CA3 lesion com-
pared to sham lesion; Single diamond, p < 0.05; double dia-
mond, p < 0.01 metyrapone compared to vehicle; single
circle, p < 0.05; double circle, p < 0.01 corticosterone
compared to vehicle. veh, vehicle; cort, corticosterone
(n = 1320/group).

rats without metyrapone treatment. These findings strongly sug- glucocorticoids can induce atrophy of CA3 pyramidal neu-
gest that the metyrapone-induced attenuation of retention rons and reduce hippocampal volume, changes associated with
impairment is due to inhibition of corticosterone synthesis. cognitive impairments1012. Our findings suggesting that such
The retention-probe trial findings confirm previous reports cognitive impairments result directly from elevated glucocor-
that CA3 lesions impair retention of water-maze spatial train- ticoid levels or HPA-axis dysregulation may contribute to
ing (H.-A. Steffenach, E.I. Moser & M.-B. Moser, Soc. the development of new strategies in the treatment of mem-
Neurosci. Abstr. 25, 649.4, 1999) 5 and elevate plasma corti- ory disorders following hippocampal damage or sustained
costerone2,3. We found that normalizing corticosterone levels hypercortisolemia.
at the time of the probe trial is sufficient to block the reten-

tion impairment induced by CA3 damage. These findings ACKNOWLEDGEMENTS
strongly suggest that CA3 damage-induced retention deficits of We thank Q. Griffith, G. Hui and J. Buranday for technical assistance.
this magnitude are due to influences on memory retrieval Supported by USPHS NIMH Grant MH12526 (J.L.M.), NIH RO1 MH53814
mediated by HPA-axis dysregulation. We reported previous- (R.M.S.) and the Adler Foundation (R.M.S.).
ly that acute exposure to stress of glucocorticoids shortly
before retention testing impairs memory retrieval6,8. Impor- RECEIVED 24 JULY; ACCEPTED 29 OCTOBER 2001
tantly, the plasma corticosterone elevation obtained with CA3
lesions was highly comparable to levels produced by doses of 1. Moser, M.-B. & Moser, E. I. J. Neurosci. 15, 75357542 (1998).
acutely administered corticosterone that induce memory 2. Fischette, C. T., Komisaruk, B. R., Edinger, H. M. & Siegel, A. Brain Res. 195,
retrieval impairment in intact rats6,8. Furthermore, our find- 105109 (1980).
3. Herman, J. P. et al. J. Neurosci. 9, 30723082 (1989).
ing that the CA3 lesions did not impair acquisition perfor- 4. Paxinos, G. & Watson, C. The Rat Brain in Stereotaxic Coordinates (Academic,
mance is congruent with previous evidence indicating that New York, 1997).
glucocorticoids do not impair acquisition or immediate 5. McLaughlin, J. et al. Proc. Natl. Acad. Sci. USA 97, 1280412809 (2000).
6. de Quervain, D. J.-F., Roozendaal, B. & McGaugh, J. L. Nature 394, 787790
recall6,8. The effects of hippocampal damage on hypersecre- (1998).
tion of glucocorticoids are temporary and dissipate after sev- 7. Roozendaal, B., Bohus, B. & McGaugh, J. L. Psychoneuroendocrinology 21,
eral weeks (rodents) or months (primates) 9 . It is not yet 681693 (1996).
8. de Quervain, D. J.-F., Roozendaal, B., Nitsch, R. M., McGaugh, J. L. & Hock,
known whether the cognitive deficits induced by lesioning of C. Nat. Neurosci. 3, 313314 (2000).
the CA3 recover with the same time course. The present find- 9. Jacobson, L. & Sapolsky, R. Endocr. Rev. 12, 118134 (1991).
10. Sapolsky, R. M. J. Neurosci. 6, 22402244 (1986).
ings may have implications for understanding the complex 11. Arbel, I., Kadar, T., Silbermann, M. & Levy, A. Brain Res. 657, 227235
relationship between glucocorticoids, hippocampal integrity (1994).
and memory function. Prolonged exposure to stress levels of 12. McEwen, B. S. Ann. Rev. Neurosci. 22, 105122 (1999).
Previous brain imaging studies have demonstrated responses to

Visual stimuli activate tactile and auditory stimuli in visual cortex of blind subjects, sug-
gesting that removal of one sensory modality leads to neural reor-
auditory cortex in the deaf ganization of the remaining modalities13. To investigate whether
similar cross-modal plasticity occurs in human auditory cortex,
Eva M. Finney, Ione Fine and Karen R. Dobkins we used functional magnetic resonance imaging (fMRI) to mea-
sure visually evoked activity in auditory areas of both early-deaf-
Psychology Department-0109, University of California, San Diego, ened and hearing individuals. Here we find that deaf subjects
La Jolla, California 92093, USA exhibit activation in a region of the right auditory cortex, corre-
Correspondence should be addressed to K.R.D. (kdobkins@ucsd.edu) sponding to Brodmanns areas 42 and 22, as well as in area 41
(primary auditory cortex), demonstrating that early deafness
Published online: 12 November 2001, DOI: 10.1038/nn763 results in the processing of visual stimuli in auditory cortex.
Fig. 1. Visual stimuli activate auditory cortex in the deaf. Shown is an anatomical scan averaged across all deaf and hearing subjects. Auditory
regions of interest (ROIs, green regions) and voxels activating differentially in deaf versus hearing subjects in response to the visual motion stimu-
lus (colors defined in scale bar) are shown on axial (left), coronal (middle) and sagittal (right) sections of an averaged anatomical brain, trans-
formed into the standard stereotaxic space of Talairach and Tournoux5. The area of visual responsiveness falls within Brodmanns areas 41, 42 and
22 in the right auditory ROI. Crosshairs highlight a voxel within the area of main effect that maps to Brodmanns area 41 (primary auditory cor-
tex). Scale bar indicates the functional intensity (FIT) value, or magnitude of activation. L, left; R, right; A, anterior; P, posterior side of brain.
Fig. 2. Mean activation in deaf versus hearing subjects to visual stimuli.

Mean functional intensity (FIT) values, corresponding to magnitude of
activation, for deaf (white bars) and hearing (gray bars) subjects within
the area of main effect shown in Fig. 1. Visual stimuli produced signifi-
cant activation in deaf but not hearing subjects. For comparison, audi-
tory-evoked activity in this region of hearing subjects is also shown.
Error bars denote standard errors of the means.
auditory stimuli in these same voxels within hearing subjects was

6.90 2.65 (p = 0.0069). Based on Talairach and Tournoux coor-
dinates, this region of visual activation in the deaf corresponds to
Brodmanns areas 42 and 22 (secondary and association auditory
areas, respectively), which includes part of the planum temporale.
Six profoundly deaf and six hearing subjects were tested (3 In addition, several voxels (0.22 cm3, 23% of the total region of
females in each group; deaf, 27.0 5.7 years old; hearing, effect) fell within area 41 (primary auditory cortex), which encom-
26.8 2.6 years old). All subjects were right handed with nor- passes the medial portion of Heschls gyrus. We also cross-checked
mal or corrected-to-normal vision. All protocols were conduct- these coordinates against probabilistic atlases7,8 and confirmed
ed in compliance with the University of California at San Diegos that our region of effect included both A1 and the planum tem-
Human Subjects Committee Institutional Review Board. Using porale. There was no main effect of visual field or interaction
fMRI (AFNI software4, 1.5 T BOLD; 28 slices; TR, 4 s; voxel size, between visual field and subject group within this region.
3 3 6 mm), we defined an auditory region of interest (ROI) by In a second version of these experiments, we collected fMRI
measuring responses elicited by auditory stimuli (music responses when subjects were instructed to ignore the motion
sequences) in our hearing subjects. In Fig. 1, auditory ROIs (green stimulus and instead perform a dimming task on the fixation
regions) are plotted on an anatomical scan averaged across all spot. Here, deaf and hearing subjects differed significantly
deaf and hearing subjects, after transforming individual (F1,10 = 4.09, p = 0.036) within a region of the right auditory ROI
anatomies into standard Talairach and Tournoux coordinate (0.54 cm3, a subset of the region of difference obtained from the
space5. Based on Talairach and Tournoux coordinates, auditory attend condition) that mapped onto area 42. Deaf subjects exhib-
stimuli were found to activate regions in both right and left audi- ited significant activity in this region (FIT coefficient, 2.85 2.91,
tory cortex, including Brodmanns areas 41, 42 and 22, although, p = 0.031) and, as before, hearing subjects did not (FIT coeffi-
consistent with known hemispheric asymmetries for music pro- cient, 0.175 1.43, p = 0.78). The smaller region of effect
cessing6, the total volume of the right auditory ROI (26.1 cm3) observed under the ignore condition is consistent with the gen-
was larger than that of the left (14.5 cm3). Analysis of visually eral tendency for sensory cortical areas to activate less strongly
evoked fMRI responses was limited to within these functionally to ignored than attended stimuli9,10. Nonetheless, the fact that
defined auditory ROIs. visual activation in auditory cortex was observed in deaf subjects
Our visual stimulus consisted of a moving dot pattern (size, even when the motion stimulus was ignored attests to the robust-
10 diameter; speed, 7/s; dot luminance, 590 candelas/m2 against ness of the cross-modal plasticity effect.
a black background; dot size, 0.2; dot density, 2.7%; percent dots Related to the present findings, results from previous fMRI
moving coherently, 87%). On alternate runs, the stimulus was and positron emission tomography studies have suggested that
presented in either the right or left visual field (15 eccentric to a the auditory regions in which we found visual activation in the
central fixation spot). The two subject groups performed com- deaf include areas that may also be involved in visual language
parably on a dimming task on the motion stimulus to control processing in both deaf and hearing subjects. Specifically, Brod-
for attentional state (deaf, 91.5 12.2% correct; hearing 94.0 manns areas 42 and 22 in deaf subjects are activated to visual
2.9%; deaf, 609 107 ms reaction time; hearing, 654 255 ms). images of sign language11,12, and these same areas are activated
Visually evoked activity within the right and left auditory ROIs in hearing subjects during a silent lip reading task13 (and during
was computed by correlating fMRI signal amplitude in individ- auditory speech tasks13). In addition, there has been previous
ual ROI voxels to a reference function corresponding to the time suggestion that these auditory areas may be used for processing
course of the visual stimulus, after first correcting for individual purely visual (that is, non-linguistic) stimuli in the deaf. How-
subject movements in six dimensions by realigning images to a ever, these earlier studies used techniques with poor spatial res-
single reference image. olution, such as electroencephalography, that could not
Visually evoked activity significantly differed between deaf and distinguish whether visual responses in deaf subjects originated
hearing subjects in the right auditory ROI (Fig. 1, colors from auditory cortex or nearby visual areas14.
defined in scale bar, main effect of subject group, In sum, the results of the present study using fMRI demon-
F1,10 = 11.12, p = 0.0038), encompassing a volume of 0.95 cm3. strate the recruitment of auditory cortex in the deaf for the pro-
Although differences were also observed in the left auditory ROI, cessing of purely visual stimuli. The cross-modal plasticity
the region of effect was extremely small (0.054 cm3) and did not observed in the present study appeared predominantly in the
survive stringent statistical standards for safeguarding against false right auditory cortex. Because our experiment used moving visu-
positives. Within the right ROI region of main effect, the visual al stimuli, this hemispheric asymmetry may simply reflect a pre-
stimulus caused significant activation in deaf subjects (Fig. 2), with disposition for motion processing in the right auditory cortex.
a mean functional intensity (FIT, which reflects the magnitude of This possibility is supported by the finding that in hearing sub-
activation) value of 2.26 1.37 (p = 0.0049), whereas no signifi- jects, the right auditory cortex (specifically, the planum tempo-
cant activation occurred in hearing subjects (mean FIT, 0.31 rale) is specialized for processing auditory motion15. Thus, right
1.30, p = 0.59). In comparison, the mean FIT value produced by auditory cortex in the deaf, devoid of its normal auditory input,

may come to serve motion processing in the visual modality. RECEIVED 11 JUNE; ACCEPTED 17 OCTOBER 2001
Remarkably, the reciprocal result has recently been reported in
blind subjects. Here, responses to moving auditory stimuli are 1. Sadato, N. et al. Nature 380, 526528 (1996).
observed predominantly in the right visual cortex of the blind2, 2. Weeks, R. et al. J. Neurosci. 20, 26642672 (2000).
again suggesting a predisposition toward motion processing in 3. Kujala, T. et al. Trends Neurosci 23, 115120 (2000).
the right hemisphere. Most importantly, our demonstration of 4. Cox, R. W. Computers and Biomedical Research 29, 162173 (1996).
5. Talairach, J. & Tournoux, P. Co-Planar Stereotaxic Atlas of the Human Brain
cross-modal plasticity in deaf subjects, in conjunction with that (Thieme Medical, New York, 1988).
observed in the blind, attests to the robust ability of the human 6. Christman, S. Cerebral Asymmetries in Sensory and Perceptual Processing
brain to reorganize in response to the removal early in develop- (Elsevier Science B.V., 1997).
7. Westbury, C. F. et al. Cereb. Cortex 9, 392405 (1999).
ment of input from one sensory modality. 8. Rademacher, J. et al. Neuroimage 13, 669683 (2001).
9. Martinez, A. et al. Nat. Neurosci. 2, 364369 (1999).
ACKNOWLEDGEMENTS 10. Gandhi, S. P. et al. Proc. Natl. Acad. Sci. USA 96, 33143319 (1999).
Supported by an NSF grant to K.R.D. and an NRSA grant to E.M.F. We thank 11. Petitto, L.A. et al. Proc. Natl. Acad. Sci. USA 97, 1396113966 (2000).
12. Nishimura, H. et al. Nature 397, 116 (1999).
G. Boynton and M. Sereno for helpful discussions, G. Brown, L. Eyler Zorrilla, P.
13. Calvert, G. A. et al. Science 276, 593596 (1997).
Goldin and S. Tapert for assistance with data analysis, and D. Cai for assistance 14. Neville, H. J. Ann. NY Acad. Sci. 608, 7191 (1990).
with subject recruitment. 15. Baumgart, F. et al. Nature 400, 724726 (1999).

articles
Distinct regulators control the

expression of the mid-hindbrain
organizer signal FGF8
Weilan Ye1, Maxime Bouchard2, Donna Stone1, Xiaodong Liu1, Francis Vella4, James Lee1,
Harukazu Nakamura3, Siew-Lan Ang4, Meinrad Busslinger2 and Arnon Rosenthal5
1 Department of Molecular Biology, Genentech, 1 DNA Way, South San Francisco, California 94080, USA
2 Research Institute of Molecular Pathology, Vienna Biocenter, Dr. Bohr-Gasse 7, A-1030 Vienna, Austria
3 Department of Molecular Neurobiology, Institute of Development, Aging & Cancer, Tohoku University, Seiryo-machi 4-1, Aoba-ku, Sendai 980-8575 Japan
4 IGBMC, 1 rue Laurent Fries BP163, 67404 Illkirch Cedex CU de Strasbourg, France
5 Rinat Neuroscience, 3155 Porter Drive, Palo Alto, California 94304, USA
The first two authors contributed equally to this work

Correspondence should be addressed to A.R. (ar@rinatneuro.com)
Published online: 12 November 2001, DOI: 10.1038/nn761
Local expression of FGF8 at the mid/hindbrain boundary (MHB) governs the development of
multiple neurons and support cells. Here we show that the paired-domain protein Pax2 is
necessary and sufficient for the induction of FGF8 in part by regulating the expression of Pax5&8.
A network of transcription and secreted factors, including En1, Otx2, Gbx2, Grg4 and Wnt1&4,
that is established independently of Pax2, further refines the expression domain and level of
FGF8 at the MHB through opposing effects on Pax2 activity. Our results indicate that the
expression of local organizing factors is controlled by combinatorial interaction between
inductive and modulatory factors.
During development, distinct classes of neurons and support cells MHB of the mouse9,10 or chick11 embryos leads either to a shift
appear in stereotypic locations and subsequently define the brain in the position of FGF8 expression or to its induction. In con-
architecture and function. The identity and position of these cells trast, mutation of Gbx2 or Otx2 results abnormal positioning
are controlled in part by local organizing centers that emit secret- and eventual loss of FGF8 (refs. 1214). However, the actual role
ed inductive factors. A well-characterized local brain organizer of Otx2 and Gbx2 in the control of FGF8 expression and the spa-
is found at the boundary between the midbrain (MB) and hind- tial limits of their activities has not been defined. Furthermore,
brain (HB), the isthmus. This organizer is required for the devel- the role of other regulatory genes (such as Pax2, En1 and Wnt1)
opment of all cell types and nuclei in the MB as well as for the in the induction and stable expression of FGF8 at the MHB is
development of the cerebellum in the anterior HB. Transplanta- poorly understood. To determine the pathways controlling the
tion of the isthmic organizer to the diencephalon leads to the for- induction and expression pattern of FGF8 and to investigate the
mation of an ectopic MB1, whereas transplantation to the HB roles of Otx2, Gbx2, En1, Pax2, 5, 8 and Wnt in this process, we
results in the development of an ectopic cerebellum2. The isth- performed multiple gain- and loss-of-function studies in chick
mic organizer activity seems to be mediated by FGF8 (ref. 3). and mouse embryos.
Despite the seminal role of FGF8 in MB and cerebellum devel- Here we provide evidence that Pax2 is necessary and sufficient
opment, the mechanism that delineates its induction and stable for the induction of FGF8 as well as for the expression of Pax5
expression in a narrow band of isthmus cells is not fully under- and Pax8 at the MHB. In contrast, Otx2 and Gbx2 are neither nec-
stood. The isthmus develops within the expression domains of essary nor sufficient for the induction of FGF8 and can only define
the paired domain protein Pax2 (ref. 4) and the homeodomain its position within the Pax2 expression domain. We further found
protein Engrailed-1 (En1)5. It is further demarcated by the jux- that Otx2 and Gbx2 act in part by having opposing influences on
taposition of the homeodomain transcription factors Otx2, which the expression of Grg4 (ref. 15), which is a transcriptional core-
is expressed in the MB6, and Gbx2, which is expressed in the HB7. pressor of Pax and En proteins16,17. Otx2 also has a non-cell-
FGF8 first appears in the anterior HB adjacent to the secreted autonomous role in the maintenance of FGF8 expression in part
proteins Wnt1 (ref. 4) and Wnt4 (in the chick) 8, which are through the induction of Wnt1 (and Wnt4), which in turn act
expressed in the posterior MB. through En1 (ref. 18). These findings outline a general mecha-
It has been suggested that the positioning and/or induction nism for the establishment of local organizing centers in the ver-
of FGF8 are regulated by interaction between Otx2+ and Gbx2+ tebrate brain, which involves combinatorial interactions between
cells. This is because ectopic expression of Gbx2 or Otx2 in the transcriptional activators, repressors and secreted proteins.

articles
a b
Fig. 1. Expression pattern of endogenous Otx2, Gbx2, Pax2, FGF8 and
Wnt1 in the chick embryos. (a) Otx2 (red) and Gbx2 (blue) at stage 6 of
chick embryogenesis. (b) Gbx2 (red) and Pax2 (blue) at stage 9.
(c) Gbx2 (red) and FGF8 (blue) at stage 11. (d) Otx2 (red) and FGF8
(blue) at stage 20. (e) Pax2 (red) and Wnt1 (blue) at stage 16. (f) Wnt1
(blue) and FGF8 (red) at stage 15. (a) and (b) are whole-mount samples;
(cf) are flat-mount specimens. All are dorsal views with anterior to
the left. Scale bar, 0.14 mm (a); 0.23 mm (b, e); 0.26 mm (c); 0.29 mm
(d); 0.2 mm (f). ANR, anterior neural ridge; R1, rhombomere 1; MHB,
c d mid/hindbrain boundary; FB, forebrain; MB, midbrain; HB, hindbrain.
appeared at the newly generated anterior but not posterior

Otx2/Gbx2 border and only in the Otx2 cells (Fig. 2a). Like-
wise, scattered expression of Otx2 extending into the entire
HB resulted in a posterior expansion of FGF8 expression in
the adjacent Otx2 cells (Fig. 2b; also see ref. 11). By using
e f Hoxa2 as a reference to define the R1/R2 boundary, we found
that Otx2 can induce FGF8 expression in a non-cell-
autonomous manner only in the anterior two thirds of R1
(Fig. 2e). At the stage when the ectopic Otx2 protein is trans-
lated (around stages 1316), this region coincides with the
caudal limit of the Pax2 domain (Fig. 1e).
Uniform expression of Otx2 slightly posterior to its endoge-
nous domain resulted in a posterior shift of FGF8 expression
Results as previously observed in transgenic mice10 or chick embryos
Control of FGF8 expression by Otx2 ectopically expressing Otx2 (ref. 11, data not shown). In contrast,
Otx2, a vertebrate homologue of the Drosophila orthodenticle gene, uniform expression of a large patch of Otx2 from the MB to R2
is expressed in the forebrain and MB before FGF8 expression and led to ablation of endogenous FGF8 expression and did not result
delimits the anterior boundary of the isthmus as early as the head- in the induction of FGF8 at any other location (data not shown).
fold stage (Fig. 1a and d)6. Ectopic expression of Otx2 under the These findings suggest that Otx2 stimulates FGF8 expression in
En1 promoter in the anterior HB of the mouse embryo relocates non-cell-autonomous manner and only in a region that is demar-
the MHB to the newly generated Otx2/Gbx2 border10. In addi- cated by Pax2 expression. Otx2 also inhibits FGF8 expression cell-
tion, ectopic expression of Otx2 or Gbx2 in the chick embryo11 autonomously (Fig. 2b).
resulted in ectopic induction of FGF8. These findings led to the
suggestion that Otx2+ cells co-define the MHB and thus the site Control of FGF8 expression by Gbx2
of FGF8 induction. We have re-examined the role of Otx2 by Gbx2, a homeobox gene related to Drosophila unplugged, is
expressing it at various locations along the anterior-posterior (A/P) expressed in the HB7 as early as the head-fold stage, preceding
axis of the neural tube of the chick embryo.
Local uniform expression of a small patch of ectopic Otx2 a
straddling rhombomeres 1 (R1) and 2 (R2) caused the
b
appearance of a second stripe of FGF8+ cells. This stripe
Fig. 2. Ectopic expression of Otx2 and Pax2 in the chick embryo.

(a) Ectopic expression of mouse Otx2 (red) at the anterior HB leads
to the formation of a second FGF8 (blue) expression domain at the
anterior ectopic Otx2/Gbx2 boundary. Endogenous chick Otx2 is not
stained. (b) Ectopic scattered expression of mouse Otx2 (red, c d
endogenous chick Otx2 is also stained) in the HB leads to induction
of FGF8 (blue) in the anterior R1. (c) Ectopic expression of Pax2
(red) leads to weak induction of FGF8 (blue) in the diencephalon.
(d) Ectopic expression of Pax2 (red) from the midbrain to R4 leads
to weak induction of tiny patches of FGF8 (blue, arrowhead) in the
R4, but not in the midbrain or R1R3. (e) Ectopic expression of
scattered Otx2 (not stained) in the HB (same as in b) leads to induc-
tion of FGF8 (blue) in the anterior two thirds of R1 as judged by the
expression domain of Hoxa2 (red). (f) Ectopic expression of scat- e f
tered Otx2 (not stained, same as in b) together with Pax2 (not
stained, same as in d) in the HB leads to the expansion of FGF8
(blue) beyond R3, as judged by Hoxa2 (red) expression. The upper
side of the embryo is the electroporated side, whereas the lower
side serves as control. All samples are flat-mounts shown in dorsal
view with anterior to the left. Scale bar, 0.5 mm (a, f); 0.3 mm (b, c);
0.4 mm (d); 0.32 mm (e). Ec, ectopic; Ep, electroporation.
articles
Fig. 3. Ectopic expression of Gbx2 and Pax2 in the chick embryo.

(a) Ectopic expression of Gbx2 (red) in the MB very close to the MHB
a b
leads to ectopic FGF8 (blue) expression at the posterior ectopic
Otx2/Gbx2 boundary. (b) Ectopic expression of Gbx2 (not stained, simi-
lar to a) leads to the induction of FGF8 (blue) within the endogenous
Pax2 (red) domain. (c) Ectopic expression of Gbx2 (red) in the midbrain
outside of Pax2 (not stained) fails to induce ectopic FGF8 (blue).
(d) Ectopic expression of Gbx2 (not stained) in the MHB slightly anterior
to the endogenous Gbx2 leads to anterior shift of FGF8 (blue) and the
elimination of Otx2 (red) within the ectopic Gbx2 domain. (e) Local c d
ectopic expression of Pax2 (red) together with Gbx2 (not stained) in the
midbrain leads to the formation of a second FGF8 (blue) expression
domain within the midbrain, well beyond the endogenous Pax2 expres-
sion domain. Expression of the endogenous Pax2 in the MB is very weak
at stage 20 shown here. (f) Ectopic expression of Gbx2 (red) together
with Pax2 (not stained) in the midbrain leads to expansion of the FGF8
(blue) expression domain throughout the midbrain. All samples are flat-
mounts shown in dorsal view with anterior to the left. Top sides, elec-
troporated sides; bottom sides, control sides. Scale bar, 0.34 mm.
e f
the expression of FGF8 (Fig. 1ac). In the mouse, ectopic

expression of Gbx2 in the posterior MB under the Wnt1 pro-
moter9 results in a relocation of the MHB to the newly gener-
ated Otx2/Gbx2 border. Likewise, ectopic expression of Gbx2
in the chick embryo11 leads to the expression of FGF8 in ectopic
locations. These findings suggested that Gbx2 also defines the
MHB and thus the site of FGF8 expression19. To determine the examine this possibility, Pax2 was ectopically expressed at vari-
roles of Gbx2+ cells in the induction of FGF8, we next examined ous locations along the A/P axis in the chick embryo either alone
the consequence of expressing it at various locations along the or together with Otx2 or Gbx2. As reported22, electroporation of
A/P axis of the chick neural tube. Pax2 into the neural tube of the chick embryo led to weak expres-
Forced expression of Gbx2 in the MB in close proximity and sion of FGF8 mRNA in the diencephalon, telencephalon (Fig. 2c)
anterior to the endogenous MHB led to a small patch of FGF8 and the posterior HB (R4) (Fig. 2d), suggesting that Pax2 alone is
expression at the ectopic Gbx2/Otx2 border (Fig. 3a). Ectopic FGF8 capable of inducing FGF8. However, Pax2 did not induce FGF8 in
expression could only be achieved within the Pax2+ MB zone near the MB or anterior HB (R1 to R3) (Fig. 2c and d), indicating that
the endogenous MHB (Fig. 3b). In contrast, ectopic expression of additional regulators are required in these brain regions.
Gbx2 outside of the Pax2+ territory in the MB failed to induce We then examined whether the interaction between Pax2 and
FGF8 (Fig. 3c). In addition, expression of Gbx2 within and slight- Otx2 would allow the induction of FGF8 in the MB and anterior
ly anterior to the MHB resulted in the suppression of Otx2 HB. As shown above, FGF8 was not detected in the posterior R1
(Fig. 3d) and in an anterior shift of the FGF8+ cells to the new through R3 following ectopic expression of Otx2 (Fig. 2b and e)
Gbx2/Otx2 border (Fig. 3d) as observed in Gbx2 transgenic mouse9 or Pax2 (Fig. 2d) alone. In contrast, scattered expression of Otx2
and chick embryos11. Finally, homogeneous high-level expression (similar to that shown in Fig. 2b) in conjunction with uniform
of Gbx2 throughout the MB and anterior HB, which shifted the expression of Pax2 in R1R3 (similar to that shown in Fig. 2d)
Gbx2/Otx2 boundary outside of the Pax2+ domain, led to the led to ectopic expression of FGF8 in Pax2+ Otx2 cells throughout
extinction of endogenous FGF8 expression and did not result in R1R3 (Fig. 2f and data not shown). Thus, induction or sus-
ectopic induction of FGF8 (data not shown). The endogenous Pax2 tained expression of FGF8 in the anterior HB requires, in addition
domain did not change in response to the new Gbx2/Otx2 bound- to Pax2, non-cell-autonomous signal from Otx2+ cells.
ary (Fig. 3b). Taken together, these findings support the idea that We next determined whether interaction between Pax2 and
Gbx2 controls the expression of FGF8 only in a narrow region in Gbx2 is required for the induction or sustained expression of
the posterior MB defined by the presence of Pax2+ cells. FGF8 in the MB and anterior HB. Although no ectopic expres-
sion of FGF8 was detected in the central and posterior MB fol-
Control of FGF8 expression by Pax2 lowing electroporation of Pax2 (Fig. 2c and d) and Gbx2 (Fig. 3a
Expression of the paired domain protein Pax2 also precedes FGF8 and b) alone, co-electroporation of Pax2 and Gbx2 led to the
expression in the MHB and occupies both the posterior Otx2+ widespread expression of FGF8 in Pax2+ Gbx2+ MB cells (Fig. 3e
and anterior Gbx2+ domains (Fig. 1b and e)20. Pax2 expression is and f). These findings indicate that Pax2 alone can induce FGF8
activated in the absence of Otx2 (ref. 13) or Gbx2 (ref. 12) and is outside of the MHB region but that it requires the presence of
eventually refined to a narrow band across the MHB, covering a Gbx2 and a non-cell-autonomous signal from the Otx2+cells for
region immediately anterior to the Wnt1 and the anterior part induction or stable expression of FGF8 in the MHB region.
of R1 (Fig. 1e). After the onset of Pax2 expression, Pax5 and Pax8
(ref. 21) are also activated in the MHB region, and FGF8 is Pax2 is essential for FGF8 induction
induced in a broad band in the anterior Gbx2+ domain, occupy- The gain-of-function experiments suggested that Pax2, Otx2 and
ing a large portion of R1 (Fig. 1c). Gbx2 might all be significant in the induction or stable expres-
Our findings suggested that Pax2 rather than the Otx2/Gbx2 sion of FGF8. To discern the exact roles and relationships between
boundary might directly control the initial expression of FGF8. To these genes in vivo, we analyzed mice deficient in Otx2, Gbx2 or

articles
Fig. 4. Absence of FGF8 expression in Pax2/ mouse embryos.

(a, b) Pax2/ embryos on the C3H/He background do not express FGF8 in
a b the MHB at the 4-somite (a) and 7-somite (b) stages in contrast to the
control Pax2+/ embryos. (c) Pax2/ C3H/He embryos still express
-galactosidase in the MHB at the 10-somite stage, indicating that the MHB
tissue is still present at this time point. (d, e) Otx2 is expressed normally in
Pax2/ embryos at 0-somite (d) and 5-somite (e) stages. (f) Gbx2 is
expressed normally in 3-somite Pax2/ embryos. (g) En1 is expressed at a
slightly reduced level in 8-somite Pax2/embryos. (h) Wnt1 is expressed
normally in 7-somite Pax2/ embryo. (i) Pax5 is expressed only at a very

low level in Pax2/ embryo at the 7-somite stage. (j) Pax8 is not expressed
c d in Pax2/ embryos at the 8-somite stage. Arrowhead, MHB. Som, somite.
We further examined the consequence of Pax2 deficiency for

the expression of Otx2, Gbx2, Pax5, Pax8, En1 and Wnt1. The
expression of Otx2 was unchanged in Pax2/ embryos at either
the 0- or 5-somite stages (Fig. 4d and e). Likewise, no change in
e the expression pattern of Gbx2 was detected at the 3-somite stage
f (Fig. 4f), supporting the idea that the regulation of Otx2 and
Gbx2 is independent of Pax2. In addition, Pax2/ embryos dis-
played normal expression of Wnt1 (Fig. 4h) and slightly reduced
expression of En1 (Fig. 4g), arguing that Pax2 is not a general
regulator of gene expression in the MHB region. In contrast, only
a very low level of Pax5 mRNA (Fig. 4i) and no Pax8 expression
(Fig. 4j) were detected in Pax2/ embryos. These findings argue
g h that Pax2 regulates Pax5 and Pax8 either directly or indirectly.
Otx2 and Gbx2 have opposite effects on Grg4

The combined loss- and gain-of-function data obtained with Otx2
and Gbx2 suggested that these two proteins control the precise
position (compare Fig. 1c with d)4 and stable expression rather
than the initial induction of FGF8 at the MHB. As Otx2 prevents
and Gbx2 stimulates FGF8 expression in a cell-autonomous man-
i j ner, we first examined possible cell-intrinsic mechanisms by which
these two proteins could determine the domain of FGF8 expres-
sion. Specifically, we investigated the possibility that Otx2 and
Gbx2 control the position of FGF8 by regulating the expression
of the transcriptional corepressor Grg4. Grg4 is a vertebrate homo-
logue of the Drosophila Groucho gene26, which is expressed in the
MB and very weakly in the dorsal HB and diencephalon15,17.
Moreover, it seems to be excluded from the FGF8 expression
domain in the anterior HB (Fig. 5a and b), it changes Pax5 from
a transcriptional activator into a repressor16 and it suppresses the
Pax2. Chimeric Otx2/ mice, which do not express Otx2 in the expression of En2 and FGF8 in the MHB region17.
neural tube13, still transiently expressed FGF8 at the MHB with- Consistent with the hypothesis that Otx2 and Gbx2 define the
in the Pax2 domain until the 8-somite stage (data not shown and expression domain of FGF8 through their influence on Grg4,
ref. 23). Likewise, Gbx2/ mice displayed expanded and stable ectopic expression of Otx2 in the chick HB led to induction of
expression of FGF8 in the MHB region 12 (data not shown). Grg4 in R1 to R3 (Fig. 5a). Conversely, ectopic expression of Gbx2
Therefore, Otx2 or Gbx2 are not required for the initial induc- in the MB resulted in suppression of Grg4 (Fig. 5b) and Otx2
tion of FGF8 at the MHB, but instead determine its precise posi- (Fig. 3d) in this region. To further examine whether Grg4 nega-
tion and long-term expression. tively regulates Pax2 during MHB development, we generated a
We then analyzed C3H/He Pax2-deficient embryos, in which mutant Pax2-OP protein that fails to bind Grg4 due to deletion
Pax2 was replaced by the -galactosidase gene24. Although FGF8 of its octapeptide (OP) interaction motif16. Ectopic expression
expression is initiated in wild-type embryos at the 3-somite stage of the mutant but not wild-type Pax2 protein led to ectopic
before the onset of Pax8 and independently of Pax5 (ref. 25), it is expression of FGF8 in the midbrain (Fig. 5c and d). Thus, in the
never induced in the MHB of Pax2/ embryos at the 34 somite diencephalon and posterior HB, where Grg4 is not expressed,
stage (Fig. 4a) or later developmental stages (Fig. 4b). Important- Pax2 can alone induce FGF8. In contrast, in the MHB, where
ly, the MHB tissue is still present in these Pax2/ embryos at the Grg4 is present, the transcriptional activity of Pax2 depends on
10-somite stage, as evidenced by a normal complement of -galac- antagonistic effects of Otx2 and Gbx2 on Grg4 expression.
tosidase-expressing cells (Fig. 4c) and by the presence of a well-
defined Gbx2/Otx2 boundary (Fig. 4df). Thus, the correct Wnt mediates non-cell-autonomous function of Otx2
temporal expression of Pax2, but not that of Otx2 or Gbx2, is essen- As Otx2 is important for maintaining high-level expression of
tial for the induction of FGF8 in the MHB of the C3H/He embryos. FGF8 through non-cell-autonomous mechanism, we set out to

articles
Fig. 5. Regulation and activity of Grg4, En1 and Wnt1. (a) Ectopic
mouse Otx2 (red) induces the expression of Grg4 (blue) in the HB. a b
(b) Ectopic Gbx2 (red) suppresses Grg4 (blue) in the MB. (c) Wild-type
Pax2 (red) fails to induce FGF8 (blue) in the MB. (d) The mutant Pax2-
OP protein (red), which does not bind Grg4 due to deletion of the
octapeptide (OP) motif16, induces FGF8 (blue) in the MB and HB.
(e) Ectopic Otx2 (red) induces the expression of Wnt1 (blue) in the HB.
(f) Wnt1 (red) induces the expression of FGF8 (blue) in the anterior R1.
(g) Ectopic expression of En1 (red) from the midbrain to R4 does not
lead to efficient ectopic expression of FGF8 (blue). (h) Ectopic expres- c d

sion of En1 (not stained) together with Pax2 (not stained) in the MHB
region leads to ectopic expression of FGF8 throughout the hindbrain
(blue). Dorsal views are shown with the anterior end to the left. Scale
bar, 0.34 mm (a, c, d); 0.36 mm (b, e, f).
determine the molecular nature of the Otx2-derived signal. One

candidate we examined was the secreted protein Wnt1. The e f
expression of Wnt1 was apparently not sufficient for the initial
induction of FGF8 (compare Fig. 4h and b), but is essential for
the stability of FGF8 expression and the maintenance of the MHB
organizer18,27. In the chick embryo, Wnt1 and Wnt4 are induced
concurrently with FGF8, in a small stripe of cells at the posterior
edge of the Otx2+ boundary (Fig. 1e and f; data not shown). FGF8
is initially expressed in a wide domain in the anterior R1
(Fig. 1c) and later refined to a narrow band immediately next to g h
the Otx2 (Fig. 1d) and Wnt1 expression domains (Fig. 1f). Wnt1
is never induced in Otx2/ chimeric mouse embryos14, but is
induced by forced expression of Otx2 in the chick HB (Fig. 5e
and data not shown). More importantly, ectopically expressed
Wnt1 can induce the expression of FGF8 in the anterior R1
(Fig. 5f) in a manner similar to Otx2 (Fig. 2b).
Because genetic epistasis experiments have shown that Wnt1
maintains the MHB and the expression of FGF8 indirectly
through the homeodomain protein En1 (ref. 18), we also exam- The mechanism controlling the formation of an FGF8-express-
ined the ability of En1 to control FGF8 expression. En1 was shown ing MHB is not well understood. One possibility is that the MHB
to induce FGF8 in the diencephalon28 and MB29. However, in the is formed following juxtaposition of different neural plate (pla-
absence of both En1 and En2, FGF8 is still induced in the MHB nar) signals. This hypothesis was supported by the finding that
region 5. En1 by itself was unable to efficiently induce FGF8 the MHB can regenerate in vivo between the MB and HB follow-
expression in the chick HB (Fig. 5g) but strengthened the inducing surgical removal of part of the MB33 and that an FGF8-pro-
tion of FGF8 by Pax2 in R13 (Fig. 5h), indicating that En1 has an ducing MHB will develop following juxtaposition of R1 and MB
auxiliary function in regulating the level of FGF8 expression. tissue in explant culture34. Because the juxtaposition of the MB
and HB is demarcated early by the expression of the transcription
DISCUSSION factors Otx2 and Gbx2, respectively, it has been further suggest-
Local inductive centers that emit secreted proteins determine the ed that the mutually antagonistic action of these two proteins is
identity and initial position of most cell types in the vertebrate central in the development of the MHB. Here we used chimeric
brain. Here we provided evidence that the expression of FGF8, mouse embryos, which are Otx2-deficient only in the neural
the key mediator of the MHB organizer, is controlled by combi- tube23, to confirm that Otx2/Gbx2 boundary is not required for
natorial interactions between the transcription factors Pax2, En1, the initial induction of FGF8 (data not shown). Instead, our find-
Grg4, Otx2 and Gbx2. Pax2 is the key inducer, whereas Otx2, Gbx2, ings argue that the Otx2/Gbx2 boundary is involved in the posi-
Grg4, Wnt1 and En1 have modulatory roles. Similar combinator- tioning of FGF8 expression rather than in its induction.
ial and sequential interactions between transcriptional repressors,
activators and secreted proteins are likely to control the position of The role of Pax2 in the induction of the MHB organizer
other local organizing centers in the vertebrate brain. Although En1, En2, Wnt1, FGF8, Pax5, Pax8 and multiple other
genes are expressed in overlapping or adjacent domains of the
The role of Otx2 and Gbx2 in the control of the MHB MHB region, their individual deletion in the C57BL/6 genetic back-
Transplantation experiments in the chick and mouse embryos grounds did not lead to ablation of FGF8 expression. We show here
revealed that the MHB possesses organizing activity, as it can that Pax2 alone is essential either directly or indirectly for the
induce an ectopic MB in the diencephalon or an ectopic cere- induction of FGF8 as well as of Pax5 and Pax8 at the MHB in
bellum in the HB30. The molecule responsible for the patterning C3H/He mouse and is therefore the master regulator of FGF8.
activity of the MHB seems to be FGF8. This is evident from the Although Pax2 is essential and sufficient for the induction of
finding that FGF8 beads induce a new MHB, MB and cerebel- FGF8, stable expression and precise positioning of FGF8 at the
lum3, whereas the reduction of FGF8 activity31,32 leads to defects MHB still requires the regulation of Pax2 activity through the
in MB and anterior HB development. opposing action of Gbx2 and Otx2 on the expression of the tran-

articles
Fig. 6. Putative interactions between negative and positive regulators

of FGF8 in the MHB. Pax2 induces FGF8 expression, whereas Otx2 and
Gbx2 further refine the position and stable expression of FGF8
through their opposing effects on the expression of Grg4 and through
the induction of Wnt proteins by Otx2. En, which may be induced by
the mesoderm47 and sustained by Wnt1, supports the induction of
FGF8 by Pax2. This model suggests that the induction and positioning
of FGF8 expression in the MHB are regulated by integration of vertical
(En inducer) and planar (Otx2, Gbx2, Wnt1) signals. Whether Pax2
expression is induced by vertical or planar signals remains to be

determined. Arrowheads indicate positive regulation (not necessarily
direct), and T-shaped bars denote negative regulation (not necessarily
direct). Dashed lines indicate transient expression, and contiguous
lines denote stable expression. Grg4 negatively regulates the activity
of Pax and En proteins16. Di, diencephalon.
scriptional corepressor Grg4. Precise FGF8 expression also cephalon but an ectopic cerebellum in the MB suggests that it
depends on the cooperative interaction of Pax2 with En1, whose either acts instructively in conjunction with other patterning
expression in the HB may be controlled by the non-cell- signals or functions as a permissive or mitogenic signal for
autonomous Wnt1 signal from the MB18 (Fig. 6). committed progenitor cells41. In addition, although a relay
mechanism has not been entirely excluded, the finding that
Involvement of non-neural tissues in FGF8 expression distinct genes and cell types are induced at stereotypic distances
It has been suggested that vertical signals from underlying meso- from a point source of FGF8 (ref. 42) argues for a morpho-
dermal tissue may be responsible, at least in part, for the induc- genetic function of FGF8. The specific function of the Pax, En
tion of FGF8 in the MHB. For example, FGF8, which is secreted and Wnt genes, apart from regulating FGF8 expression, is not
by the underlying cardiac mesendodermal cells3, may induce fully understood either. Gene ablation studies suggested that
FGF8 in the MHB. However, FGF8 mRNA is transcribed in the Wnt1 is required for cell proliferation in the MB43,44. Howev-
absence of functional FGF8 protein in neural plate explant cul- er, ectopic expression of Wnt1 in the chick embryo did not lead
tures31. Furthermore, FGF8 expression is abolished in Pax2/ to abnormal proliferation, suggesting that Wnt1 is not rate lim-
mice in which the cardiac mesoderm expression of FGF8 is nor- iting for this process. In contrast, En1 seems to be essential for
mal (data not shown). Alternatively, it has been proposed that the induction of axon guidance cues41 and possibly for cell sur-
En1, which could be activated by FGF4 from the notochord, may vival in the MB and HB45.
induce FGF8 in the MHB29, possibly in conjunction with En2, In summary, we have provided evidence that the precise
which is also activated by vertical signals35. However, En1 does expression domain of at least one local organizer in the verte-
not seem to be required for the induction of FGF8 in vivo5. brate nervous system is determined in a stepwise process involv-
Finally, it is possible that Pax2 is activated in the MHB region ing initial induction and subsequent refinement by a feedback
in response to vertical signals and then, in turn, induces FGF8. mechanism between multiple transcription factors and secreted
However, we found that explants of chick embryos from the proteins (Fig. 6).
early-streak stage (stages 34) induce Pax2 in the presumptive
MHB in the absence of both axial and paraxial mesodermal tis- METHODS
sues (data not shown). Likewise, we found that Pax2 was acti- Electroporation of chick embryos. Electroporation was done as described
vated in mouse explants that were separated from the head previously46.The chick Pax2-OP mutant was generated by PCR (poly-
mesenchyme and axial mesoderm as soon as the neuroecto- merase chain reaction)-based mutagenesis, which resulted in deletion of
derm was induced (early E7; data not shown). These findings the entire octapeptide motif (YSINGILG; amino acids 185192; ref. 22).
All animal experiments were reviewed and approved by Genentech insti-
suggest that a subset of cells is committed to express Pax2 as tutional animal use committee.
soon as the neural ectoderm is induced. Whether these cells are
committed to express Pax2 in response to vertical or planar sig- In situ hybridization. Chick or mouse embryos were processed for in situ
nals remains to be determined. hybridization as described46. In most cases, neural tube of the chick
embryos was cut open along the roof plate, and the mesoderm was par-
The maintenance and function of FGF8 in the MHB tially removed to allow flat-mount preparation for photography. All chick
FGF8, Wnt, En1, En2, Pax2 and Pax5 are dependent on each other figures shown in this paper are dorsal views of flat-mount specimens.
for stable expression and, when one of them is missing, the expres-
sion of the remaining genes is extinguished over time36. Conversely, Mutant mice. The Pax2 mutant24, Otx2 chimeric13 and Gbx2 mutant12
mice were generated, maintained and genotyped as described.
these genes can induce and maintain each others expression at
ectopic locations3,22,29,37,38. This feedback regulation involves direct
interactions between Pax2 and the MHB-specific enhancer of Pax5 ACKNOWLEDGEMENTS
(ref. 21), En proteins and regulatory sequences on the FGF8 gene39 We thank A. Joyner for sharing unpublished results and for the chick En1 and
as well as Pax proteins and a regulatory element of the En2 gene5,40. En2 cDNAs, G. Martin, J. Rubenstein and E. Miyashita for Gbx2 mutant mice,
This cross-regulatory network provides a mechanism to stabilize, A. Simeone for the full-length mouse and partial chick Otx2 cDNAs, A. Leutz for
maintain and sharpen the expression domain of FGF8. the full-length chick Gbx2 cDNA, G. Martin for the chick FGF8 cDNA,
Although FGF8 is capable of inducing an ectopic MHB, MB A. Lumsden for the chick Hoxa2 cDNA, A. McMahon for the chick Wnt1 and
and anterior HB, its mechanism of action is not understood. Wnt4 cDNAs and S. Greenwood for reading the manuscript. This research was
The finding that FGF8 induces an ectopic MB in the dien- supported in part by Boehringer Ingelheim.

articles
RECEIVED 26 SEPTEMBER; ACCEPTED 21 OCTOBER 2001 transcription factors Pax2 and Pax5 in mouse development. Development
127, 37033713 (2000).
25. Urbanek, P., Fetka, I., Meisler, M. H. & Busslinger, M. Cooperation of Pax2
1. Martinez, S., Wassef, M. & Alvarado-Mallart, R. M. Induction of a and Pax5 in midbrain and cerebellum development. Proc. Natl. Acad. Sci.
mesencephalic phenotype in the 2-day-old chick prosencephalon is preceded USA 94, 57035708 (1997).
by the early expression of the homeobox gene en. Neuron 6, 971981 (1991). 26. Mallo, M., Franco del Amo, F. & Gridley, T. Cloning and developmental
2. Marin, F. & Puelles, L. Patterning of the embryonic avian midbrain after expression of Grg, a mouse gene related to the groucho transcript of the
experimental inversions: a polarizing activity from the isthmus. Dev. Biol. Drosophila Enhancer of split complex. Mech. Dev. 42, 6776 (1993).
(Orlando) 163, 1937 (1994). 27. McMahon, A. P., Joyner, A. L., Bradley, A. & McMahon, J. A. The
3. Crossley, P. H., Martinez, S. & Martin, G. R. Midbrain development induced midbrainhindbrain phenotype of Wnt-1-/Wnt-1- mice results from
by FGF8 in the chick embryo. Nature 380, 6668 (1996). stepwise deletion of engrailed-expressing cells by 9.5 days postcoitum. Cell
4. Hidalgo-Sanchez, M., Millet, S., Simeone, A. & Alvarado-Mallart, R. M. 69, 581595 (1992).
Comparative analysis of Otx2, Gbx2, Pax2, Fgf8 and Wnt1 gene expressions 28. Araki, I. & Nakamura, H. Engrailed defines the position of dorsal di-
during the formation of the chick midbrain/hindbrain domain. Mech. Dev. mesencephalic boundary by repressing diencephalic fate. Development 126,
81, 175178 (1999). 51275135 (1999).
5. Liu, A. & Joyner, A. L. EN and GBX2 play essential roles downstream of FGF8 29. Shamim, H. et al. Sequential roles for Fgf4, En1 and Fgf8 in specification and
in patterning the mouse mid/hindbrain region. Development 128, 181191 regionalisation of the midbrain. Development 126, 945959 (1999).
(2001). 30. Nakamura, H., Nakano, K. E., Igawa, H. H., Takagi, S. & Fujisawa, H.
6. Boncinelli, E., Gulisano, M. & Broccoli, V. Emx and Otx homeobox genes in Plasticity and rigidity of differentiation of brain vesicles studied in quail-
the developing mouse brain. J. Neurobiol. 24, 13561366 (1993). chick chimeras. Cell Differ. 19, 187193 (1986).
7. Bouillet, P., Chazaud, C., Oulad-Abdelghani, M., Doll, P. & Chambon, P. 31. Ye, W., Shimamura, K., Rubenstein, J. L., Hynes, M. A. & Rosenthal, A. FGF
Sequence and expression pattern of the Stra7 (Gbx-2) homeobox-containing and Shh signals control dopaminergic and serotonergic cell fate in the
gene induced by retinoic acid in P19 embryonal carcinoma cells. Dev. Dyn. anterior neural plate. Cell 93, 755766 (1998).
204, 372382 (1995). 32. Meyers, E. N., Lewandoski, M. & Martin, G. R. An Fgf8 mutant allelic series
8. Hollyday, M., McMahon, J. A. & McMahon, A. P. Wnt expression patterns in generated by Cre- and Flp-mediated recombination. Nat. Genet. 18, 136141
chick embryo nervous system. Mech. Dev. 52, 925 (1995). (1998).
9. Millet, S. et al. A role for Gbx2 in repression of Otx2 and positioning the 33. Nieuwkoop, P. D. The successive steps in the pattern formation of the
mid/hindbrain organizer. Nature 401, 161164 (1999). amphibian central nervous system. Dev. Growth Differ. 32, 149154 (1989).
10. Broccoli, V., Boncinelli, E. & Wurst, W. The caudal limit of Otx2 expression 34. Irving, C. & Mason, I. Regeneration of isthmic tissue is the result of a specific
positions the isthmic organizer. Nature 401, 164168 (1999). and direct interaction between rhombomere 1 and midbrain. Development
11. Katahira, T. et al. Interaction between Otx2 and Gbx2 defines the organizing 126, 39813989 (1999).
center for the optic tectum. Mech. Dev. 91, 4352 (2000). 35. Ang, S.-L. & Rossant, J. Anterior mesendoderm induces mouse Engrailed
12. Wassarman, K. M. et al. Specification of the anterior hindbrain and genes in explant cultures. Development 118, 139149 (1993).
establishment of a normal mid/hindbrain organizer is dependent on Gbx2 36. Wurst, W. & Bally-Cuif, L. Neural plate patterning: upstream and
gene function. Development 124, 29232934 (1997). downstream of the isthmic organizer. Nat. Rev. Neurosci. 2, 99108 (2001).
13. Rhinn, M. et al. Sequential roles for Otx2 in visceral endoderm and 37. Funahashi, J. et al. Role of Pax-5 in the regulation of a mid-hindbrain
neuroectoderm for forebrain and midbrain induction and specification. organizers activity. Dev. Growth Differ. 41, 5972 (1999).
Development 125, 845856 (1998). 38. Liu, A., Losos, K. & Joyner, A. L. FGF8 can activate Gbx2 and transform
14. Rhinn, M., Dierich, A., Le Meur, M. & Ang, S. Cell autonomous and non-cell regions of the rostral mouse brain into a hindbrain fate. Development 126,
autonomous functions of Otx2 in patterning the rostral brain. Development 48274838 (1999).
126, 42954304 (1999). 39. Gemel, J., Jacobsen, C. & MacArthur, C. A. Fibroblast growth factor-8
15. Koop, K. E., MacDonald, L. M. & Lobe, C. G. Transcripts of Grg4, a murine expression is regulated by intronic engrailed and Pbx1-binding sites. J. Biol.
groucho-related gene, are detected in adjacent tissues to other murine Chem. 274, 60206026 (1999).
neurogenic gene homologues during embryonic development. Mech. Dev. 59, 40. Li Song, D. & Joyner, A. L. Two Pax2/5/8-binding sites in Engrailed2 are
7387 (1996). required for proper initiation of endogenous mid-hindbrain expression.
16. Eberhard, D., Jimenez, G., Heavey, B. & Busslinger, M. Transcriptional Mech. Dev. 90, 155165 (2000).
repression by Pax5 (BSAP) through interaction with corepressors of the 41. Lee, S. M., Danielian, P. S., Fritzsch, B. & McMahon, A. P. Evidence that FGF8
Groucho family. EMBO J. 19, 22922303 (2000). signalling from the midbrainhindbrain junction regulates growth and
17. Sugiyama, S., Funahashi, J. & Nakamura, H. Antagonizing activity of chick polarity in the developing midbrain. Development 124, 959969 (1997).
Grg4 against tectum-organizing activity. Dev. Biol. 221, 168180 (2000). 42. Martinez, S., Crossley, P. H., Cobos, I., Rubenstein, J. L. & Martin, G. R. FGF8
18. Danielian, P. S. & McMahon, A. P. Engrailed-1 as a target of the Wnt-1 induces formation of an ectopic isthmic organizer and isthmocerebellar
signalling pathway in vertebrate midbrain development. Nature 383, 332334 development via a repressive effect on Otx2 expression. Development 126,
(1996). 11891200 (1999).
19. Simeone, A. Positioning the isthmic organizer where Otx2 and Gbx2 meet. 43. McMahon, A. P. & Bradley, A. The Wnt-1 (int-1) proto-oncogene is required for
Trends Genet. 16, 237240 (2000). development of a large region of the mouse brain. Cell 62, 10731085 (1990).
20. Dressler, G. R., Deutsch, U., Chowdhury, K., Nornes, H. O. & Gruss, P. Pax2, a 44. Thomas, K. R. & Capecchi, M. R. Targeted disruption of the murine int-1
new murine paired-box-containing gene and its expression in the developing proto-oncogene resulting in severe abnormalities in midbrain and cerebellar
excretory system. Development 109, 787795 (1990). development. Nature 346, 847850 (1990).
21. Pfeffer, P. L., Bouchard, M. & Busslinger, M. Pax2 and homeodomain 45. Wurst, W., Auerbach, A. B. & Joyner, A. L. Multiple developmental defects in
proteins cooperatively regulate a 435 bp enhancer of the mouse Pax5 gene at Engrailed-1 mutant mice: an early mid-hindbrain deletion and patterning
the midbrainhindbrain boundary. Development 127, 10171028 (2000). defects in forelimbs and sternum. Development 120, 20652075 (1994).
22. Okafuji, T., Funahashi, J. & Nakamura, H. Roles of Pax-2 in initiation of the 46. Hynes, M. et al. The seven-transmembrane receptor smoothened cell-
chick tectal development. Brain Res. Dev. Brain Res. 116, 4149 (1999). autonomously induces multiple ventral cell types. Nat. Neurosci. 3, 4146
23. Acampora, D. et al. Visceral endoderm-restricted translation of Otx1 (2000).
mediates recovery of Otx2 requirements for specification of anterior neural 47. Ang, S.-L., Conlon, R. A., Jin, O. & Rossant, J. Positive and negative signals
plate and normal gastrulation. Development 125, 50915104 (1998). from mesoderm regulate the expression of mouse Otx2 in ectoderm explants.
24. Bouchard, M., Pfeffer, P. & Busslinger, M. Functional equivalence of the Development 120, 29792989 (1994).

articles
A Gr receptor is required for

response to the sugar trehalose in
taste neurons of Drosophila
Anupama Dahanukar, Kara Foster, Wynand M. van der Goes van Naters and John R. Carlson
Department of Molecular, Cellular, and Developmental Biology, Yale University, PO Box 208103, New Haven, Connecticut 06520-8103, USA
The first three authors contributed equally to this work
Correspondence should be addressed to J.R.C. (john.carlson@yale.edu)
We recently identified from the Drosophila genome database a large family of G proteincoupled
receptor genes, the Gr genes, and predicted that they encode taste receptors on the basis of their
structure and specificity of expression. The expression of Gr genes in gustatory neurons has
subsequently been confirmed and 56 family members have been reported. Here we provide
functional evidence that one Gr gene, Gr5a, encodes a taste receptor required for response to the
sugar trehalose. In two different mutants that carry deletions in Gr5a, electrophysiological and
behavioral responses to trehalose were diminished but the response to sucrose was unaffected.
Transgenic rescue experiments showed that Gr5a confers response to trehalose. The results correlate
a particular taste ligand with a Gr receptor and indicate a role for G proteinmediated signaling in
the transduction of sweet taste in Drosophila.
The ability of the gustatory system to detect and discriminate are eliminated16. The expression of Gr genes in gustatory neu-
different taste stimuli is a universal feature of animals. In the rons has subsequently been confirmed by other researchers17,18,
adult fly, gustatory receptor neurons are housed in hair-like and 56 family members have so far been reported. However, the
structures called sensilla14. These are located on the external functional significance of Gr genes has not been demonstrated.
and internal mouth parts, the tarsal segments of the legs and Genetic studies in Drosophila have identified a locus that
the anterior margins of the wings. The primary gustatory organ mediates response to the disaccharide trehalose19. Behavioral
in the adult is the labellum, which is located at the tip of the tests have shown that the Tre locus alters the taste sensitivity
proboscis and bears approximately 66 taste hairs1. Most label- to trehalose without affecting the response to other sugars. The
lar taste hairs are innervated by a single mechanosensory neu- Tre locus has been mapped to cytogenetic region 5A using var-
ron and four bipolar chemosensory neurons, which have been ious wild-type strains and chromosomal rearrangement stocks
classified as a sugar cell, a water cell and two salt cells based on that show differences in trehalose response19,20. A GPCR gene,
their electrophysiological responses to various stimuli4. The CG3171, in region 5A has recently been identified; it has been
axons of the chemosensory neurons project to the sub- named Tre1 and proposed to encode the taste receptor for tre-
esophageal ganglion in the brain, where gustatory information halose21. Tre1 shows no homology to the Gr genes; rather, it
is processed1,5,6. Stimulation of these neurons by chemical cues has 2022% sequence identity to vertebrate melatonin recep-
elicits a variety of behaviors, such as proboscis extension, inges- tors. The two genes in the Drosophila genome to which Tre1 is
tion and proboscis retraction4. most similar both have been annotated in the fly sequence data-
Perception of sweetness is a critical taste modality for terres- base as probable hormone receptors (http://flybase.bio.indi-
trial animals such as Drosophila that ingest sweet substances for ana.edu version 7/01).
nutrition. G proteincoupled receptors (GPCRs) have been impli- We found that a member of the Gr family, Gr5a, is tightly
cated in the reception of monosodium glutamate (umami)7 and linked to the Tre locus. To determine whether Gr5a plays a role
bitter taste810 in mammals. Recently, two vertebrate GPCRs, in trehalose reception, we studied the response to trehalose in
T1R2 and T1R3, have been shown to mediate responses to several mutant flies that have deletions in Gr5a. We find that in flies
sweet-tasting molecules, including sucrose1114. mutant for Gr5a, the electrophysiological as well as behavioral
We recently identified from the Drosophila genome database response to trehalose is diminished. Notably, this defect is res-
a large family of GPCR genes, the Gr genes, and predicted that cued by supplying a wild-type copy of Gr5a on a transgene, but
they encode taste receptors on the basis of their structure and not by supplying a mutant copy of Gr5a. Rescue is not depen-
specificity of expression15. Of the first 19 Gr genes identified, 18 dent on the presence of a wild-type copy of Tre1. Our results thus
were found to be expressed in the labellum. Gr gene expression indicate that Gr5a, a member of the Gr gene family, is required
was not detected in a variety of other tissues or in the proboscis for the reception of trehalose and validate the functional role of
of pox-neuro flies, mutants in which chemosensory taste neurons a Gr gene as a gustatory receptor.

articles
Fig.1. Molecular structure of the Gr5a locus. Both Tre1

(open boxes) and Gr5a (filled boxes) transcription units are
indicated. The inverted triangle depicts the insertion site of
the P element in strain 496. Sequences deleted in 5 and
19 are indicated below.
strain, the translation initiation codon of Gr5a is

removed (Fig. 1). Flies that are either homozygous
or hemizygous for these deletions are viable.

We next asked whether the deletion mutants lack-
ing Gr5a were defective in trehalose response, using
two independent assays. First, we measured electro-
physiological response to sugars in individual label-
lar taste hairs of the large (L) or medium (M) class22,
RESULTS using the tip-recording method23. L and M sensilla each have four
Gr5a is tightly linked to the Tre locus chemosensory neurons, classified according to the stimuli that
To test the hypothesis that Gr genes encode taste receptors, we elicit a response from them: a sugar cell, two salt cells and a water
asked whether any Gr genes reside at loci implicated in taste per- cell19,24,25. The sugar cell responds to a number of different sugars,
ception. The trehalose response locus19, whose alleles confer dif- including disaccharides as sucrose and trehalose26 , which is pre-
ferent levels of response to the disaccharide trehalose, has been sent in yeast (an important food source of Drosophila). Individual
mapped to cytogenetic region 5A on the X chromosome20. One sensilla were stimulated by placing an electrode containing the
Gr gene, Gr5a, is also located in region 5A and is thus tightly sugar over the tip of the sensillum. Action potentials are elicited in
linked to this locus. Gr5a is a member of a small subfamily of both the sugar and the water cell, but the activities of the two cells
eight Gr genes18. We determined the intronexon structure of can be distinguished by their different amplitudes. Doseresponse
Gr5a (Fig. 1) by 5 and 3 RACE and RT-PCR experiments. The curves for sucrose and trehalose in the various strains are shown
experiments also showed expression of Gr5a in the proboscis and in Fig. 2a. Sucrose response was the same for all strains across a
the legs (not shown), both of which contain sensilla that respond broad range of concentrations. In the same cells, however, the tre-
to trehalose (ref. 21 and data shown below). The 5 end of Gr5a halose responses of 5 and 19 were drastically reduced com-
lies less than 900 base pairs from CG3171, which has previously pared to that of the parental strain 496.
been reported to encode the trehalose receptor and has been
named Tre1 (ref. 21 and Fig. 1). Gr5a mutants show sugar-specific behavioral defects
We next tested the behavioral response to trehalose using the two-
Gr5a mutants show sugar-specific physiological defects choice preference test19. This protocol compares the consump-
To investigate the possible role of Gr5a in trehalose reception, tion of two sugars offered simultaneously to populations of flies.
we used two strains of flies, each with a deletion in the Gr5a-
Tre1 genomic region. The two deletion lines, EP(X)-19 and a
EP(X)-5, henceforth referred to as 19 and 5, were generat-
ed by imprecise excision of a P element that lies in the region
between the two genes in strain EP(X)0496, henceforth referred
to as 496. We determined by sequence analysis that in 19, the
proximal endpoint of the deletion lies within the second exon
of Gr5a (3 to the codon specifying Ala 62), and in 5, the prox-
imal endpoint lies in the first intron (which lies 3 to the codon
specifying Arg 36) (CG15779 in Gadfly, Release 1, http://hedge-
hog.lbl.gov:8002/cgi-bin/annot/query/). Thus, in each deletion
Fig. 2. Gr5a mutants show sugar-specific physiological and behavioral

defects. (a) Neural responses to trehalose are diminished in the deletion
lines. Shown here are concentrationresponse plots of mean impulse
rates in recordings from labellar sensilla (n = 1213). Error bars indicate b
s.e.m. The response to trehalose in 496 differs significantly from the
responses in 19 and 5 (ANOVA with Scheff post hoc tests; p < 0.001),
whereas the strains do not differ significantly in their response to
sucrose. (b) The behavioral response to trehalose is defective in the
deletion lines. Shown are dosesensitivity curves determined using the
two-choice preference test. The tested sucrose concentrations corre-
spond to 1, 2 and 3 mM. The dashed line indicates PI50; error bars indi-
cate s.e.m. There is no significant difference in the sucrose response of
496, 5 and 19. For comparison of the trehalose response between the
different strains, PI50 values were calculated for each experiment (n = 10
experiments; 3 concentrations per experiment) and used for statistical
analysis. There is no significant difference between 19 and 5, but both
are significantly different from 496 (p < 0.001). Arcsine-transformed data
were analyzed using ANOVA and Sidak post hoc tests.

articles
Fig. 3. Gr5a rescues the physiological defect of Tre mutants.

(a) Genomic rescue constructs. The wild-type construct (T+G+) con- a
tains both the Tre1 and the Gr5a transcription units. Asterisks indicate
the position of stop codon mutations in either Tre1 (TG+) or Gr5a
(T+G). (b) Sample traces of physiological recordings 200700 ms after
stimulus onset. (c) Mean physiological responses. Shown are results for
the indicated concentrations, which give a response closest to half-max-
imum in Fig. 2a (7 n 15). Error bars indicate s.e.m.s. Trehalose was
also tested at 101.5 M (31.6 mM) and 100.5 M (316 mM) and sucrose
was also tested at 101 M (100 mM) and 100.5 M (316 mM), with similar
results. For trehalose, Gr5a+ lines differ significantly from Gr5a lines
(ANOVA with Sidak post hoc tests, p< 0.001). Mean responses to tre-
halose in Canton S wild-type and 496 parental lines are also indicated.
Mean responses to sucrose do not differ significantly among the lines. b
denotes the 19 or 5 line in the absence of a transgene.
In the first control experiment, we tested the preference for dif-

ferent concentrations of sucrose versus a standard concentration
of 2 mM sucrose (Fig. 2b, left panel). Flies of all strains tested pre-
ferred 3 mM (102.5 M) sucrose to 2 mM sucrose, with a prefer-
ence index (PI) of 0.9, and they preferred 2 mM sucrose to 1 mM
sucrose (PI = 0.2). Consistent with the physiological data, there
was no difference in the response to sucrose between strain 496
and the strains that are mutant for Gr5a. To investigate trehalose
response, we tested several concentrations of trehalose against
2 mM sucrose. The PI50 value, the concentration of trehalose at
which flies consumed as much trehalose as sucrose, was 11 mM
for 496, but 77 mM for 5 and 76 mM for 19 (Fig. 2b, right).
Thus, in this behavioral protocol both mutants were severely
defective in their response to trehalose.
Transformation rescue of trehalose response with Gr5a c

To determine whether loss of Gr5a expression is responsible for
the reduced response to trehalose in the deletion mutants, we
engineered a 10-kb genomic DNA rescue construct that includ-
ed both the Gr5a and the Tre1 coding regions (Fig. 3a). To assess
the contributions of each gene separately, we also generated deriv-
atives of this 10-kb construct that included a stop codon near the
N-terminus of either Gr5a or Tre1. We generated transgenic flies
containing these constructs and crossed them into 5 and 19
genetic backgrounds. The resulting flies were tested for rescue of
the trehalose response defect using the physiological and behav-
ioral assays described above.
The physiological response of the sugar cell to trehalose was
rescued by the construct with wild-type copies of both Tre1 and
Gr5a (T +G + in Fig. 3) in both the 19 and 5 backgrounds
(Fig. 3c). Although rescue occurred in both the Tre1+Gr5a+ and
the Tre1-Gr5a+ transgenic flies, it did not occur in the Tre1+Gr5a-
transgenic flies (Fig. 3b and c), indicating that rescue required a
wild-type Gr5a but not a wild-type Tre1. None of the transgenes
affected response to sucrose (Figs. 3b and c). pelling evidence that Gr5a is responsible for the trehalose sensi-
We then tested the behavioral phenotype of the transgenic tivity phenotype is that deletion mutations affecting Tre func-
flies. We measured the preference for trehalose at a 10 1.5 M tion can be rescued by a transgenic construct containing a
(31.6 mM) concentration, which gave the maximal difference in wild-type copy of Gr5a, but not by a construct containing a
PI between the parental and deletion strains (Fig. 2b). Rescue mutant copy of Gr5a. Rescue has been shown both by measure-
again occurred, and the pattern of rescue was consistent with the ments of single-cell electrophysiology and by behavioral assays.
physiological data: it was dependent on Gr5a and not on Tre1 in The equivalence of Gr5a and the Tre locus is also consistent with
both the 19 and 5 flies (Fig. 4). mapping data indicating that all reported Tre deletions remove
portions of Gr5a (Fig. 1 and ref. 21).
DISCUSSION The simplest interpretation of our results is that Gr5a
We have shown here that Gr5a, a member of a large family of encodes a taste receptor for trehalose. The absence of a func-
candidate taste receptors, corresponds to the Tre locus, a genetic tional Gr5a affects the physiological response of a taste neuron
locus that affects response to the sugar trehalose. The most com- to trehalose, but not the response of the same neuron to anoth-

articles
Fig. 4. Gr5a rescues the behavioral defect of Tre mutants. Shown are
the results of a two-choice test that measures preference between
101.5 M (31.6 mM) trehalose and 2 mM sucrose. Here 8 n 10,
except that n = 5 for 19;T+G and n = 4 for 5;T+G. Error bars indi-
cate 95% confidence intervals. denotes the 19 or 5 line in the
absence of a transgene.
nation might be achieved by insulation of different signaling

pathways, as probably happens within individual chemosensory
neurons of Caenorhabditis elegans28.
Although studies in mammals have implicated G pro-
teinmediated signaling in the transduction of sweet taste, the
er disaccharide sugar, sucrose. Thus, the specificity of Gr5a func- mechanism in invertebrates is largely unknown29. There is evi-
tion is consistent with that expected of a taste receptor, and dence that sweet taste in larger flies is mediated at least in part
inconsistent with that expected of a GPCR playing a general through a cGMP second messenger30; however, there is also a
role in taste neuron development or function. Trehalose recep- report of a channel that is activated directly by sucrose with-
tion could also be mediated in part by additional receptors; out the mediation of second messengers or G proteins31. Our
however, the severely reduced trehalose response in flies lack- results support a role for G proteinmediated transduction of
ing Gr5a and the rescue data (for example, Fig. 4) suggest that the disaccharide trehalose in Drosophila, as is found for sweet
the response to trehalose is mediated primarily by Gr5a. taste in mammals.
A previous report identified the product of Tre1 as the tre- In summary, the simplest interpretation of our data is that a
halose receptor21. That report used a heat shockinducible Tre1 member of the Gr gene family encodes a taste receptor required
cDNA construct to rescue the phenotype of a deletion mutant for response to the sugar trehalose, as indicated by both electro-
similar to 5. Although our results do not formally exclude a physiological and behavioral analysis of mutant and transgenic
role for Tre1 in trehalose response, they do not support it. In flies. The association of a particular ligand with a particular Gr
considering the differing conclusions of these two studies, we taste receptor now allows for a variety of studies, including
note that although Tre1 mutants were shown to be abnormal detailed functional studies of the receptor and of the mechanism
in both a two-choice preference test and a proboscis extension by which it transduces gustatory information. It will be interest-
test, rescue was described only in the two-choice preference test ing to determine whether the other genes in the Gr5a subfamily
and only for a single pair of concentrations (2 mM sucrose and encode sweet taste receptors for other sugars.
80 mM trehalose)21. Moreover, although some limited data are
reported to indicate a physiological phenotype for a Tre1 METHODS
mutant, rescue of the peripheral physiological defect by the Constructs and flies. A 10-kb genomic fragment that includes both the
hs-Tre1 transgene has not been shown. We note that Tre1 is Tre1 and Gr5a coding regions was amplified by Expand Hi-Fi PCR
expressed in embryonic EST collections, suggesting that it is (Boehringer Ingelheim, Germany) using genomic DNA from Canton S
expressed in early Drosophila development. Tre1 is also flies as template and primers 5-GAGAGGTTGCGTTAAGCCAC and
5-CGCAGCTGTGTAGCATAGTG. Amber mutations were engineered
expressed ubiquitously in adult tissue, according to a paper pub- by introducing a linker (CTAGCTAGCTAG) in either the Not1 site in
lished while the current manuscript was under review27. This Tre1 or the HindIII site in Gr5a (see Fig. 3a). Transgenic flies were gen-
study also reports that polymorphisms in the sequence of Gr5a, erated by standard means. At least two independent lines for each trans-
but not Tre1, correlate with the trehalose phenotype. gene were tested in both assays with similar results. Transgenic flies were
The sugar cells in all wild-type L- or M-class sensilla from heterozygous for the transgene in each case. Strains 496, 19 and 5 were
which we have taken recordings respond to both sucrose and tre- obtained from K. Isono (Tohoku University, Japan).
halose. In our recordings from Gr5a deletion mutants, we found
no sensilla of these types whose sugar cells respond at wild-type Electrophysiology. Action potentials were recorded23 from single label-
levels to trehalose. Though not exhaustive, our data suggest that lar sensilla of male flies aged 515 days. Sugars were obtained from
Aldrich (D-(+)-trehalose, >99%) or from Sigma (sucrose, >99.5%; St.
Gr5a is expressed in the sugar cells of all the L and M sensilla on
Louis, Missouri). Sugars were dissolved in water, with tricholine citrate
the labellum. These two sensillar classes together constitute 30 (Sigma) added as electrolyte at 0.03 M. Tricholine citrate does not elicit
of the 66 sensilla on the labellum; thus, Gr5a seems to be a response from the salt cells, nor does it appreciably affect the sugar cell
expressed in at least 30 neurons. Although this number has not responses26. Neural response was quantified as impulse rate by counting
been confirmed by in situ hybridization17, 30 is a larger number the number of impulses generated in the 500-ms period beginning 200
than is seen for the several members of the Gr family whose ms after onset of stimulation.
expression has been detected by in situ hybridization or reporter-
gene expression17,18. For these receptors, the number of cells Behavioral assay. The two-choice preference test was carried out as
exhibiting expression ranges from 4 to 22 in the labellum17,18. described elsewhere with minor modifications19,20. For each trial, 50 flies
Because response of all tested Gr5a mutant sugar cells was abnor- aged 35 d were fed on standard molasses culture medium for a day, then
starved on water-saturated tissues for 21 h. Flies were fed on sugars in
mal to trehalose but normal to sucrose, it seems likely that each of 60-well plates for 2 h at 25 C in the dark and then killed by freezing at
these cells expresses at least two receptor genes: Gr5a, which 20C. Indigo Carmine dye (Spectra Colors, Kearny, New Jersey) was
mediates response to trehalose, and another receptor that medi- present in the 2 mM sucrose standard at a concentration of 0.125 mg per
ates response to sucrose. Colocalization of receptors within a sin- ml. Acid red 106 (Sigma) was present at a concentration of 0.5 mg per
gle sugar cell might constrain the ability of the animal to ml in the wells containing variable sugar concentrations. Ten indepen-
discriminate among sugars through differential activation of dis- dent trials were carried out for each concentration of the sugar, and results
tinct taste neurons. It remains possible, however, that discrimi- were scored blind. Only those trials in which at least 20% of the flies had

articles
fed were included for statistical analysis. Preference index (PI) values 13. Max, M. et al. Tas1r3, encoding a new candidate taste receptor, is allelic to the
were calculated according to the formula (NR + 0.5NP)(NR + NB + NP)1, sweet responsiveness locus Sac. Nature Genetics 28, 5863 (2001).
where NR, NB and NP are the number of flies with red, blue and purple 14. Sainz, E., Korley, J. N., Battey, J. F. & Sullivan, S. L. Identification of a novel
member of the T1R family of putative taste receptors. J. Neurochemistry 77,
abdomens, respectively19,20. In preliminary experiments, males and 896903 (2001).
females were tested separately and found not to differ; in all subsequent 15. Clyne, P., Warr, C. & Carlson, J. Candidate taste receptors in Drosophila.
experiments males were used exclusively. Science 287, 18301834 (2000).
16. Dambly-Chaudiere, C. et al. The paired box gene pox neuro: a determinant of
ACKNOWLEDGMENTS poly-innervated sense organs in Drosophila. Cell 69, 159172 (1992).
17. Scott, K. et al. A chemosensory gene family encoding candidate gustatory and
We thank K. Isono for providing strains and members of the Carlson laboratory olfactory receptors in Drosophila. Cell 104, 661673 (2001).
for comments on the manuscript. The research was supported by an NRSA to 18. Dunipace, L., Meister, S., McNealy, C. & Amrein, H. Spatially restricted
A.D. and NIH grants and a McKnight Investigator Award to J.R.C. expression of candidate taste receptors in the Drosophila gustatory system.
Curr. Biol. 11, 822835 (2001).
19. Tanimura, T., Isono, K., Takamura, T. & Shimada, I. Genetic dimorphism in
RECEIVED 16 AUGUST; ACCEPTED 24 OCTOBER 2001 the taste sensitivity to trehalose in Drosophila melanogaster. J. Comp. Physiol.
147, 433437 (1982).
20. Tanimura, T., Isono, K. & Yamamoto, M. Taste sensitivity to trehalose and its
1. Nayak, S. & Singh, R. N. Sensilla on the tarsal segments and mouthparts of alteration by gene dosage in Drosophila melanogaster. Genetics 119, 399406
adult Drosophila melanogaster Meigen (Diptera: Drosophilidae). Int. J. Insect (1988).
Morphol. & Embryol. 12, 273291 (1983). 21. Ishimoto, H., Matsumoto, A. & Tanimura, T. Molecular identification of a
2. Stocker, R. The organization of the chemosensory system in Drosophila taste receptor gene for trehalose in Drosophila. Science 289, 116119 (2000).
melanogaster: a review. Cell Tissue Res. 275, 326 (1994). 22. Ray, K., Hartenstein, V. & Rodrigues, V. Development of the taste bristles on
3. Singh, R. N. Neurobiology of the gustatory systems of Drosophila and some the labellum of Drosophila melanogaster. Dev. Biol. 155, 2637 (1993).
terrestrial insects. Microsc. Res. Tech. 39, 547563 (1997). 23. Hodgson, E., Lettvin, J. & Roeder, K. Physiology of a primary chemoreceptor
4. Dethier, V. The Hungry Fly (Harvard Press, Cambridge, Massachusetts, unit. Science 122, 417418 (1955).
1976). 24. Rodrigues, V. & Siddiqi, O. A gustatory mutant of Drosophila defective in
5. Stocker, R. F. & Schorderet, M. Cobalt filling of sensory projections from pyranose receptors. Mol. Gen. Genet. 181, 406408 (1981).
internal and external mouthparts in Drosophila. Cell Tissue Res. 216, 513523 25. Tanimura, T. & Shimada, I. Multiple receptor proteins for sweet taste in
(1981). Drosophila discriminated by papain treatment. J. Comp. Physiol. 141,
6. Shanbag, S. R. & Singh, R. N. Functional implications of the projections of 265269 (1981).
neurons from individual labellar sensillum of Drosophila melanogaster as 26. Wieczorek, H. & Wolff, G. The labellar sugar receptor of Drosophila. J. Comp.
revealed by neuronal marker horseradish peroxidase. Cell Tissue Res. 267, Physiol. A 164, 825834 (1989).
273282 (1992). 27. Ueno, K. et al. Trehalose sensitivity in Drosophila correlates with mutations in
7. Chaudhari, N., Landin, A. & Roper, S. A metabotropic glutamate receptor and expression of the gustatory receptor gene Gr5a. Curr. Biol. 11, 14511455
variant functions as a taste receptor. Nature Neurosci. 3, 113119 (2000). (2001).
8. Adler, E. et al. A novel family of mammalian taste receptors. Cell 100, 693702 28. LEtoile, N. & Bargmann, C. Olfaction and odor discrimination are mediated
(2000). by the C. elegans guanylyl cyclase ODR-1. Neuron 25, 575586 (2000).
9. Matsunami, H., Montmayeur, J. & Buck, L. A family of candidate taste 29. Glendinning, J., Chaudhari, N. & Kinnamon, S. Transduction and molecular
receptors in human and mouse. Nature 404, 601604 (2000). biology. in Neurobiology of Taste and Smell II (eds. Finger, T. & Restrepo, D.)
10. Chandrashekar, J. et al. T2Rs function as bitter taste receptors. Cell 100, 315351 (Academic Press, New York, 2000).
703711 (2000). 30. Amakawa, T., Ozaki, M. & Kawata, K. Effects of cyclic GMP on the sugar taste
11. Nelson, G. et al. Mammalian sweet taste receptors. Cell 106, 381390 (2001). receptor cell of the fly Phormia regina. J. Insect Physiol. 36, 281286 (1990).
12. Montmayeur, J. P., Liberles, S. D., Matsunami, H. & Buck, L. B. A candidate 31. Murikami, M. & Kijima, H. Transduction ion channels directly gated by
taste receptor gene near a sweet taste locus. Nature Neurosci. 4, 492498 (2001). sugars on the insect taste cell. J. Gen. Physiol. 115, 455466 (2000).

articles
Synapsin dispersion and

reclustering during synaptic activity
Ping Chi1,2, Paul Greengard2 and Timothy A. Ryan1
1 Department of Biochemistry, Weill Medical College of Cornell University, 1300 York Avenue, New York, New York 10021, USA
2 Laboratory of Molecular and Cellular Neuroscience, The Rockefeller University, 1230 York Avenue, New York, New York 10021, USA
Correspondence and requests for materials should be addressed to T.A.R. (taryan@med.cornell.edu)
Published online: 29 October 2001, DOI: 10.1038/nn756
Presynaptic modulation of synaptic transmission provides an important basis for control of synaptic
function. The synapsins, a family of highly conserved proteins associated with synaptic vesicles, have
long been implicated in the regulation of neurotransmitter release. However, direct physiological
measurements of the molecular mechanisms have been lacking. Here we show that in living
hippocampal terminals, green fluorescent protein (GFP)-labeled synapsin Ia dissociates from synap-
tic vesicles, disperses into axons during action potential (AP) firing, and reclusters to synapses after
the cessation of synaptic activity. Using various mutated forms of synapsin Ia that prevent phospho-
rylation at specific sites, we performed simultaneous FM 4-64 measurements of vesicle pool
mobilization along with synapsin dispersion kinetics. These studies indicate that the rate of synapsin
dispersion is controlled by phosphorylation, which in turn controls the kinetics of vesicle pool
turnover. Thus synapsin acts as a phosphorylation-state-dependent regulator of synaptic vesicle
mobilization, and hence, neurotransmitter release.
Synaptic vesicle availability and mobilization are crucial elements relatively limited behavioral phenotypes2527. Although there are
in the regulation of synaptic transmission and synaptic plastici- clear differences in synaptic physiology and presynaptic ultra-
ty. Synapsins, a family of highly conserved neuronal phospho- structure in the mutant animals, the lack of lethality and the
proteins that are specifically associated with synaptic vesicles1, modest effect on a number of learning protocols in the mutants
have been implicated in the regulation of neurotransmitter release have cast some doubt on the original hypothesis of synapsin func-
by controlling the number of vesicles available for exocytosis. tion. However, as the primary role of synapsins that has been
Synapsins exist in all organisms with a nervous system, and proposed is a regulatory one, it would be difficult to predict how
are encoded by three distinct genes, synapsin I, II and III, in most loss of a layer of regulation would be manifested.
vertebrates26. Synapsins are the most abundant synaptic vesicle To more fully determine the mechanisms by which synapsins
proteins, with synapsin I alone accounting for 6% of total vesi- regulate synaptic function, we GFP-labeled synapsin Ia to examine
cle protein1,7. They are present in nearly all presynaptic nerve the dynamic behavior of synapsin in response to action potential
terminals, but different neurons have a distinct repertoire of dif- (AP) firing. We combined this approach with kinetic measurements
ferent synapsins1,6,810. The high abundance, the specific associ- of vesicle-pool turnover monitored by FM 4-64 to examine vari-
ation with synaptic vesicles, the highly conserved features, as well ous mutants at the calcium-dependent phosphorylation sites of
as the widespread distribution at nerve terminals, all signify synapsin Ia in both wild-type and synapsin double knockout back-
synapsins as important and evolutionarily conserved regulatory grounds. Here we demonstrate an activity-dependent dissociation
proteins in synaptic transmission. of synapsin from synaptic vesicles and subsequent dispersion into
Biochemical studies have revealed multiple regulatory roles the axon. The rate of the dispersion is regulated by calcium-depen-
of synapsins. In vitro studies have shown that synapsins can dent phosphorylation, which in turn controls the efficiency of vesi-
interact with lipid and protein components of synaptic vesicle pool turnover. These data thus provide a direct demonstration
cles, as well as various cytoskeletal proteins, such as actin, spec- of the mechanism by which synapsins regulate the availability of
trin and microtubules, in a phosphorylation-dependent synaptic vesicles for neurotransmitter release.
manner1,1124. These studies suggest that synapsins would move
dynamically in response to physiological stimuli, and have led RESULTS
to the following hypothesis: binding of synapsins to synaptic Synapsin dissociates from vesicles during activity
vesicles prevents neurotransmitter release, and during synaptic A variety of in vitro biochemical studies imply that synapsins
activity, synapsins are phosphorylated, dissociate from synap- dynamically associate with and dissociate from synaptic
tic vesicles and allow vesicles to mobilize and fuse with the plas- vesicles in response to synaptic activity1,20,21,24,2830. To study
ma membrane1,20,23,24. synapsin dynamics in vivo, we performed immunolocalization
Given the abundance and conservation of the synapsin fam- of synapsin and an integral synaptic vesicle protein, synapto-
ily of proteins, it is interesting that mice with genetic perturba- physin. Double immunostaining of synapsin and synapto-
tions to specifically remove synapsin genes are viable and have physin in hippocampal cell cultures at rest showed a typical

articles
Fig. 1. Synapsin disperses from synaptic vesicles Rest Stimulated Recovery
Anti-synapsin Anti-synaptophysin
during activity. (a, d, g) Co-immunolocalization of
synaptophysin (red; a) and synapsin (green; d) in a b c
hippocampal cell cultures fixed at rest show punc-
tate staining for both markers (g). (b, e, h) In a
parallel culture, fixed immediately after a train of a
900-AP stimulation at 10 Hz, synapsin staining
(e, h) is much more diffuse than synaptophysin
staining (b, h) indicating that synapsins dissociate
from synaptic vesicles and redistribute into axons

(arrows) in response to electrical stimulation. d e f
(c, f, i) In a different specimen fixed 10 min after
stimulation, immunostaining of synapsin returns to
a punctate pattern colocalized with synaptophysin
(c, i), indicating a dynamic relocalization of
synapsins from axons to synapses post-stimulation
(f, i). Scale bar, 5 m.
g h i
colocalized punctate staining pattern at the
Overlay
presynaptic nerve terminals (Fig. 1a, d

and g). In parallel cultures fixed immedi-
ately following a train of 900 AP, the
immunostaining of synaptophysin remained
largely punctate, indicating that synaptic
vesicles remained at nerve terminals (Fig. 1b and h). However, Real-time measurements of synapsin dispersion
the immunostaining of synapsins became much more diffuse To measure the real-time dynamic movement of synapsin Ia
during stimulation (Fig. 1e and h), demonstrating a physical at living hippocampal nerve terminals, we transfected neu-
separation of synapsins from synaptophysin and suggesting ronal cell cultures with a plasmid encoding GFP-labeled
that a substantial fraction of synapsins have dissociated from synapsin Ia. To test whether GFP-synapsin Ia behaves similar-
synaptic vesicles and redistributed into axons. The immunolo- ly to endogenous synapsins, we examined the localization of
calization of synapsins returned to a punctate pattern in cul- GFP-labeled synapsin Ia before, during and after synaptic
tures fixed 10 minutes after stimulation (Fig. 1f and i), and it activity. GFP-synapsin Ia, when expressed transiently in cul-
colocalized well with the immunostaining of synaptophysin tured rat hippocampal neurons, was targeted to nerve termi-
(Fig. 1c and i). These observations demonstrate a dissociation nals (Fig. 2a), and appeared to have the same localization as
of synapsin from synaptic vesicles and a dynamic movement native synapsins at rest (Fig. 1d). During a train of 900 AP
in response to synaptic activity at nerve terminals. stimulation, GFP-synapsin Ia fluorescence decreased at the
nerve terminal and increased in the inter-bouton
(axonal) region in a manner similar to that of endoge-
a Rest Stimulus c nous synapsins (Fig. 2a and b). Ten minutes after
2500
stimulation, GFP-synapsin Ia reclustered within the
0.2 nerve terminals, and appeared virtually indistin-
F/F0 (GFP-Syn Ia)
10 Hz
0.0
0.2
= 12.9 s
Fig. 2. Kinetics of GFP-synapsin Ia dispersion and recovery.
Bouton
0.4
(a) GFP-synapsin Ia fluorescence at synapses (green arrows)
decreases upon stimulation. (b) GFP-synapsin Ia fluorescence
0
1100
d in axons increases upon stimulation (red arrow). Same image
10 Hz as in (a) with different color scale to emphasize the fluores-
b 2.0
F/F0 (GFP-Syn Ia)
cence in the axon. The color scales show fluorescence inten-

sity in arbitrary fluorescence units. Scale bar, 2 m. (c, d) Time
1.5
Axon
1.0 courses of fluorescence intensity at synapses (c) and in axons
= 13.0 s
0.5 (d), during and following a train of 900 AP at 10 Hz (black bar).
The decay of fluorescence intensity during stimulation, aver-
0.0
aged over 43 boutons expressing GFP-synapsin Ia, is fit by a
single exponential with = 12.9 s (c). During the same period
0 100 200 300
0 Time (s)
of AP stimulation, GFP-synapsin Ia redistributes into axons
e with a similar rate constant, = 13.0 s (d). GFP-synapsin Ia
0.2 10 Hz reclusters within the terminal (c), and decreases in the axonal
F/F0(GFP-Syn Ia)
region (d) in approximately 4 min. (e) Time-lapse measure-

0.0
ment of the total integrated fluorescence in the field of view
Ftotal
0.2
Fbouton
and the fluorescence of GFP-synapsin Ia at synapses during a
0.4 train of 900 AP field stimulation at 10 Hz (black bar). The total
0.6
fluorescence remains relatively constant during stimulation,
whereas the fluorescence intensity of GFP-synapsin Ia at
0 40 80 120
Time (s)
synapses (n = 43) decreases.
articles
Fig. 3 Comparison of the dynamics of GFP-synapsin Ia dis-

persion and recycling vesicle turnover. (a) Colocalization a b 10 Hz
of GFP-synapsin Ia-expressing boutons (green) with FM 4- 1.00
FM 4-64 destaining
64- loaded terminals (red), giving rise to yellow puncta, Non-transfected
indicating that GFP-synapsin Ia-positive terminals function- 0.75 = 20.7 s
GFP-Syn Ia +
ally recycle vesicles. In addition, in the same field, there are 0.50
= 21.0 s
boutons that do not express GFP-synapsin Ia, labeled by
FM 4-64 only (red), which are used as internal controls in 0.25
FM 4-64 destaining experiments. Scale bar, 5 m.
(b) Overexpression of GFP-synapsin Ia (wild type) does 0.00

not affect the kinetics of vesicle pool turnover as assayed 0 30 60 90 120
using FM 4-64 destaining. Destaining kinetics of FM 4-64- c d Time (s)
labeled synaptic vesicle pools during a train of 900 action 50 1.2 10 Hz
potentials at 10 Hz (black bar) in GFP-synapsin Ia-positive
F/F0 (GFP-VAMPaxon)
Acid (pH 4.0)
(n = 40) and GFP-synapsin Ia-negative (n = 50) boutons are 40 0.8
GFP-Syn Ia (s)
shown. Average time constants of FM 4-64 destaining in
30 0.4 Acid (pH 4.0)
GFP-synapsin Ia-positive and negative boutons were
= 21.0 s and = 20.7 s, respectively. Similar results were 0.0
20
obtained in five of five experiments. (c) Frequency depen- 5 Hz
dence of GFP-synapsin Ia dispersion and FM 4-64 destain- 10 Hz 0.4
10 20 Hz
ing measured in GFP-synapsin Ia expressing nerve
0.8
terminals. GFP-synapsin Ia dispersion kinetics and FM 4-64 0
destaining kinetics at different frequencies are well corre- 0 20 40 60 80 100 0 200 300 400 500
lated over a 520 Hz stimulus frequency range (r = 0.99). FM4-64 (s) Time (s)
Time constants of GFP-synapsin Ia dispersion and FM 4-64

turnover at different frequencies were averaged over e
10 Hz f
1.2 10 Hz 1.2
1.2 1.2
1333 boutons pooled from 3 experiments. GFP-synapsin
F/F0 (GFP-Syn Iaaxon)
F/F0 (GFP-VAMPaxon)
FM 4-64 fluorescence
FM 4-64 fluorescence
Ia disperses with statistically faster kinetics than FM 4-64 0.9 0.9

0.9 0.9
destaining at all frequencies tested (p < 0.002). (d) GFP-
VAMP dispersion into the axon during AP firing is 0.6
FM 4-64
0.6 0.6 0.6
= 26.9 s FM 4-64
restricted to the plasma membrane. Before stimulation, a GFP-SYNIAaxon = 25.5 s
large fraction of GFP-VAMP in the axon is on the plasma 0.3 = 15.9 s
0.3 GFP-VAMP axon
0.3 = 25.4 s 0.3
membrane, and is quenchable by brief application of sur-
face impermeant acid (pH 4.0). During stimulation (1,200 0.0 0.0
0.0 0.0
action potentials at 10 Hz), the concentration of GFP-
VAMP increases at the axon. This additional stimulation- 0.3 0.3
0 30 60 90 120 30 0 30 60 90 120
induced fluorescence remained quenchable with Time (s) Time (s)
impermeant acid, indicating that the dispersion consisted
entirely of surface GFP-VAMP and not vesicles simply dis-
persing into the axon. Axon regions analyzed, n = 51. (e) Comparison of GFP-synapsin Ia fluorescence (right y-axis) in the axon regions (n = 39) and simul-
taneous measurement of FM 4-64 turnover (left y-axis) at the GFP-synapsin Ia expressing nerve terminals (n = 40) during a train of 900 AP stimulation
(black bar). The kinetics of the rise of GFP-synapsin Ia fluorescence in the axon ( = 15.9 s) is much faster (p < 0.001) than that of FM 4-64 turnover
( = 26.9 s), suggesting a physical separation of GFP-synapsin Ia from synaptic vesicles during activity. (f) Comparison of GFP-VAMP fluorescence (right
y-axis) along the axon regions (n = 51) and simultaneous measurement of FM 4-64 turnover (left y-axis) at the GFP-VAMP-
positive nerve terminals (n = 14) during a train of 900 AP stimulation (black bar). The kinetics of the rise of GFP-VAMP fluorescence along the axon
( = 25.4 s) is very similar to that of FM 4-64 turnover ( = 25.5 s). Similar results were obtained in three of three experiments.
guishable from GFP-synapsin Ia in a prestimulated culture Synapsin Ia disperses faster than vesicle pool turnover
(data not shown). GFP, when expressed alone, uniformly To examine the physiological relevance of the dynamic movement
labeled the cell body, axons and dendrites, and showed no of synapsins at nerve terminals, we used FM 4-64
activity-dependent concentration change (data not shown). (ref. 31), a red-shifted variant of FM 1-43, to simultaneously mon-
Quantitative analyses of the dynamic movement indicated that itor synapsin dispersion and synaptic vesicle turnover in GFP-
the concentration of GFP-synapsin Ia decreased by approxi- synapsin Ia-expressing synapses. Loading of FM 4-64 by AP
mately 45% and reached a steady state at synaptic boutons stimulation resulted in labeling all GFP-synapsin Ia-expressing
during AP firing (Fig. 2a and c), coincident with a steep rise of synapses (Fig. 3a, yellow puncta), indicating that these synapses
GFP-synapsin Ia concentration in axonal regions (Fig. 2b and functionally recycle synaptic vesicles. The extent of FM 4-64 load-
d); the total fluorescence in the whole imaging field remained ing was similar in synapses either expressing or not expressing
relatively constant (Fig. 2e). Following stimulation, GFP- GFP-synapsin Ia (data not shown). Subsequent AP stimulation of
synapsin Ia reclustered within nerve terminals within approx- FM 4-64-loaded nerve terminals resulted in dye destaining as
imately four minutes (Fig. 2c), with a gradual decrease of synaptic vesicles fused with the plasma membrane and released
synapsin concentration in axons over the same time course the dye. GFP-synapsin Ia-expressing and non-expressing synapses
(Fig. 2d). Stimulation in the absence of external calcium failed in the same culture exhibited similar rates of FM 4-64 destaining
to disperse GFP-synapsin Ia (data not shown). These results (Fig. 3b), indicating that transient overexpression of GFP-synapsin
indicate that this imaging approach provides a high-fidelity Ia (wild-type) did not affect synaptic vesicle turnover at synapses.
real-time measurement of the dynamic movement of synapsins The kinetics of GFP-synapsin Ia dispersion at 10 Hz stimulation
in response to synaptic activity. ( = 12.9 s, Fig. 2c) was significantly faster than synaptic vesicle

articles
Non-transfected boutons
a 0.2 10 Hz
b 10 Hz c < > = 23.1 s
F/F0 (GFP-Syn Ia, S2/3-A) 1.0
10
Number of events
0.0
FM 4-64 destaining
5
0.2
= 20.0 s 0
Non-transfected GFP-Syn Ia-S2/3A +
= 24.0 s
10 < > = 31.4 s
0.4
GFP-Syn Ia-S2/3A +
= 31.7 s 5
0.6
0
0 100 200 300 400 0 20 40 60 80 0 10 20 30 40 50
Time (s) Time (s) FM4-64 (s)
Fig. 4. Synapsin Ia is a phosphorylation-dependent negative regulator of vesicle pool turnover. (a) Mutations of serine to alanine at CaM kinase II sites
2 and 3 (S2/3A) slow the rate of GFP-synapsin Ia dispersion when expressed in rat hippocampal cell culture. Time course of fluorescence intensity at
synapses, averaged over 20 boutons expressing GFP-synapsin Ia-S2/3A, during (dark bar) and following a train of 900 AP at 10 Hz. The dispersion
kinetics of the GFP-synapsin Ia mutant during stimulation is much slower ( = 20.0s) than wild type (Fig. 2c, = 12.9 s; p < 0.001). The time course
of reclustering is similar to wild-type GFP-synapsin Ia (Fig. 2c). (b) Synaptic vesicle pool turnover monitored by FM 4-64 destaining is significantly
slowed by the mutant form of GFP-synapsin Ia as compared to non-transfected boutons. A typical example is shown in this panel on a semi-log plot.
The average time constant for FM 4-64 destaining during 1,200 AP (dark bar) in GFP-synapsin Ia-S2/3A-positive boutons ( = 31.7 s, n = 21) is signif-
icantly larger than that in GFP-synapsin Ia-negative boutons ( = 24.0 s, n = 35; p < 0.001). (c) Frequency distribution of FM 4-64 destaining time con-
stants (FM 4-64) measured at individual GFP-synapsin Ia-S2/3A expressing and non-expressing boutons.
turnover assayed by FM 4-64 ( = 21.0 s, Fig. 3b, p < 0.001). Stim- strongly indicate that the measured synapsin Ia dispersion cor-
ulation at higher (20 Hz) or lower (5 Hz) frequency also revealed responds to dissociation of synapsin from synaptic vesicles and
that kinetics of GFP-synapsin Ia dispersion were significantly faster redistribution into the axon during activity in a step that pre-
than kinetics of FM 4-64 turnover (Fig. 3c, p < 0.002), and that cedes fusion of vesicles with the plasma membrane.
the kinetics of synapsin Ia dispersion and vesicle pool turnover
were well correlated in this frequency range (Fig. 3c). The difference Phosphorylation regulates synapsin movement
in time scale between vesicle pool turnover and GFP-synapsin Ia Previous studies indicate that phosphorylation at site 1, a cal-
dispersion provides additional evidence for a dynamic dissocia- cium-calmodulin-dependent kinase I/IV (CaM kinase I/IV)
tion of synapsin Ia from synaptic vesicles during synaptic activity. and protein kinase A (PKA) site, and at sites 2 and 3,
To further test the physical separation of synapsin from CaM kinase II sites33, leads to dissociation of synapsin Ia from
synaptic vesicles during stimulation, we specifically compared synaptic vesicles14,21,29. Therefore, we specifically investigated
the increase of GFP-synapsin Ia fluorescence in axons with the physiological functions of CaM kinase phosphorylation sites
simultaneously monitored FM 4-64-labeled vesicle turnover. in synapsin Ia by directly mutating each of them from serine to
The kinetics of GFP-synapsin Ia appearing in axons ( = 15.9 s) alanine, an amino acid of similar size that cannot be phospho-
was much faster than that of FM 4-64 turnover (Fig. 3e, rylated. When both CaM kinase II sites were mutated to ala-
= 26.9 s, p < 0.001), indicating a physical separation of GFP- nine (S2/3A), the kinetics of dissociation and dispersion of
synapsin Ia from synaptic vesicles that precedes the fusion of GFP-synapsin Ia-S2/3A from boutons during AP stimulation
synaptic vesicles with plasma membrane. We performed the was slowed significantly (p < 0.001), with a of 20.0 s (Fig. 4a)
same measurements for an integral-membrane protein of as compared to a of 12.9 s observed with wild-type GFP-
synaptic vesicles, VAMP, labeled with GFP on the lumenal synapsin Ia (Fig. 2c). The kinetics of synaptic vesicle turnover as
domain along with FM 4-64 detaining kinetics. We previously monitored by FM 4-64 was also significantly slowed in boutons
showed that during stimulation, a small portion of the pool of expressing GFP-synapsin Ia-S2/3A ( = 31.7 s) as compared to
VAMP disperses onto the axonal surface following stimula- non-expressing boutons ( = 24.0 s; Fig. 4b and c, p < 0.001).
tion32. Similarly, the GFP-VAMP signal increased to a small Similar observations were obtained from 7 other experiments
degree along the axonal region during stimulation. The with S2/3A; average dispersion was 20.9 s and there was an
kinetics of the appearance of GFP-VAMP in the axon, howev- approximately 22% slowing of FM 4-64 turnover (Fig. 5a and b,
er, matched that of vesicle pool turnover very well (Fig. 3f), p < 0.0001). These results are consistent with studies that show
and was much slower than GFP-synapsin Ia dispersion decreased and increased synaptic transmission in squid giant
(p < 0.001). Furthermore, the VAMP axonal signals measured synapse injected, respectively, with dephosphorylated synapsin
in these experiments during stimulation were fully quenched by Ia and CaM K II34,35. The time course of reclustering of GFP-
transient application of an impermeant acidic buffer (pH 4.0; synapsin Ia-S2/3A to nerve terminals was not significantly dif-
Fig. 3d), in agreement with previous studies32. As similar appli- ferent from wild-type when averaged over eight experiments
cations of impermeant acidic buffer did not quench intracel- (data not shown).
lular GFP (such as GFP-synapsin Ia fluorescence, data not To determine whether the different CaM kinase phosphory-
shown), we conclude that the elevation in VAMP-GFP fluo- lation sites have differential effects on synaptic transmission, we
rescence along the axon during stimulation is confined to the measured the rate of synapsin dispersion and FM 4-64 destaining
plasma membrane and does not correspond to vesicles break- in rat hippocampal cultures transfected with various permuta-
ing away from presynaptic clusters. These findings, along with tions of serine to alanine at the CaM kinase sites: S1A, S2A, S3A,
the differential immunolocalization of synaptophysin and S2/3A, and with all of sites 1, 2 and 3 mutated, S1/2/3A. All of
synapsins in stimulated nerve terminals (Fig. 1b, e and h), these GFP-synapsin mutants except for S2A showed significant-

articles
Fig. 5. Dispersion kinetics of GFP- Wild-type (rat)

synapsin Ia mutants regulate the effi- a b c
Percent slowing FM 4-64 destaining

25 30 30
ciency of vesicle pool turnover in
dispersionGFP-Syn Ia mutants (s)

wild-type rat neurons. (a) Effect of 25
20
different CaM kinase site mutants on
20
the dispersion kinetics of GFP- 20
15
synapsin Ia. Number of experiments
15
for each specific mutation is as fol- 10 wt
10 S1A
lows: wild type (wt), 6; site 1 (S1A), 10 S2A
6; site 2 (S2A), 3; site 3 (S3A), 5 S3A

3; sites 2 and 3 (S2/3A), 8; sites 1, 2 5 0 S2/3A
S1/2/3A
and 3 (S1/2/3A), 12. Except for S2A, 0 0
all GFP-synapsin Ia mutants show 3A 2/3
A
A 2A 3A 2/3A 1/2/3
A 15 20 25
statistically significant difference
A A A
wt S1 S2 S3 S2/ S1/ wt S1 S S S S dispersion GFP-Syn Ia mutants (s)
from wild-type GFP-synapsin Ia
(p < 0.02). (b) Effect of various mutants on FM 4-64 destaining. The effect
is calculated as the percentage increase of the time constant for FM 4-64 cence intensity of GFP-synapsin mutants at individual boutons
destaining of GFP-synapsin Ia-mutant-positive boutons compared to non- was similar (data not shown).
transfected boutons in the same experiment to reduce the effect of vari- The effect of either S1/2/3A or S2/3A mutations in GFP-
ability in different neuronal cultures. Typically, data from 1530 synapsin Ia (Fig. 6a and b) was more pronounced when
transfected boutons and 4050 non-transfected boutons were obtained expressed in synapsin I/II/ neuronal cultures than in wild-
in each experiment. Except for S2A, all other GFP-synapsin Ia mutants type mouse (data not shown) or rat cultures (Fig. 5a and b,
show statistically significant difference from wild-type GFP-synapsin Ia p < 0.0001 for each comparison). This presumably reflects an
(p < 0.03). (c) Positive correlation between the dispersion kinetics and interaction between GFP-synapsin Ia mutant and either
the effect on FM 4-64 destaining for various GFP synapsin Ia mutations.
endogenous wild-type synapsin I or endogenous wild-type
synapsin II, when expressed in wild-type background. In con-
trast, the S1A mutation had a similar effect on FM 4-64
ly decreased rates of dispersion and correspondingly decreased turnover in synapsin I/II / (Fig. 6a and b) and wild-type
rates of FM 4-64 destaining compared to the wild type (Fig. 5a (Fig. 5a and b) neuronal cultures. This observation suggests
and b, p < 0.03 for each comparison). Furthermore, there was a that site 1 directly modulates binding affinities of synapsin Ia
strong correlation between how fast GFP-synapsin Ia and mutants to synaptic vesicles, but not through interactions with endoge-
dissociated from vesicles during AP firing and how fast the whole nous synapsins, consistent with the observation that site 1
recycling vesicle pool could be mobilized to fuse with the plas- phosphorylation regulates the phospholipid binding of
ma membrane; that is, as dispersion kinetics of the GFP-synapsin synapsins to synaptic vesicles21. Importantly, in the rat as well
Ia mutant were slower, the effect was bigger on slowing synaptic as in the synapsin I/II/ mouse, there was a strong correlation
vesicle turnover as monitored by FM 4-64 (Fig. 5c). These results between how fast GFP-synapsin Ia and its mutants dispersed
strongly suggest that synapsin Ia is a negative regulator of neu- and how fast FM 4-64 was released (Figs. 5c and 6c). All of
rotransmitter release, and that its function is controlled by cal- these observations not only point to synapsin Ia as a negative
cium-dependent phosphorylation. regulator of synaptic vesicle turnover, but also demonstrate
All three identified synapsins, I, II and III, can form homod- two different levels of regulation exerted by different CaM
imers36. Heterodimerization of synapsin I and II and of synapsin kinase phosphorylation sites. The CaM kinase II sites seem to
II and III (but not of synapsin I and III) have also been demon- modulate the binding affinity of synapsin Ia to synaptic vesicles
strated by glutathione S-transferase (GST)-pull down or co- through interaction among synapsins, whereas the CaM kinase
immunoprecipitation assays36. It seemed possible that some of I/IV site (also a PKA site) seems to regulate the binding of
the effects of GFP-synapsin Ia mutants observed in our trans- synapsin Ia to vesicles directly.
fection experiments in rat neuronal cultures might have been
confounded by the presence of endogenous synapsins. To fur- DISCUSSION
ther determine the physiological function of synapsin Ia, we The studies presented here have allowed us to visualize the
examined the CaM kinase phosphorylation site mutants in hip- dynamic movement of a regulator of neurotransmission,
pocampal cultures from synapsin I/II/ mice in which interac- synapsin Ia, during synaptic activity in living nerve terminals.
tions between transfected mutants and wild-type endogenous This dynamic movement was regulated by the phosphorylation
synapsins are minimal. Wild-type GFP-synapsin Ia dispersed state of various CaM kinase sites of synapsin Ia, revealing the
in synapsin I/II (Fig. 6a) and wild-type mouse (data not shown) importance of the activities of CaM kinase I/IV and II on
neuronal cultures with kinetics similar to those in rat cultures synapsin dynamics during well-defined periods of synaptic activ-
(Fig. 5a). The kinetics of vesicle pool turnover monitored by ity. Further, the correlation between the rate of dispersion of
FM 1-43 was approximately 10% faster in synapsin I/II / synapsin Ia mutants and the kinetics of vesicle pool turnover
synapses (unpublished data), consistent with the hypothesis strongly suggest that synapsin Ia regulates neurotransmitter
that synapsin acts as a negative regulator of pool turnover. GFP- release via its dynamic interaction with synaptic vesicles.
synapsin Ia-S2/3A and S1/2/3A mutants showed significantly Our observation of activity-dependent synapsin dispersion
slowed rates of dispersion and correspondingly decreased rates using immunohistochemistry corroborates a previous report37.
of FM 4-64 turnover in synapsin I/II/ background (Fig. 6a However, in that study, it was concluded that dispersed synapsin
and b, p < 0.03 for each comparison). The differential effect of remained associated with vesicle membranes. Here, using real-
various mutant synapsins on pool turnover kinetics does not time measurements and detailed comparisons of synapsin dis-
seem to arise from differential expression levels or targeting persion, vesicle pool turnover and the redistribution of an
efficiency of these mutant constructs, as the average fluores- integral membrane protein of synaptic vesicles (VAMP), we

articles
/ Fig. 6. Dispersion kinetics of GFP-synapsin

Synapsin I/II mice Ia mutants regulate the efficiency of vesicle
60 pool turnover in synapsin I/II/ neurons.

40
dispersion GFP-Syn Ia mutants (s)
60
S2/3A S1/2/3A
(a) Dispersion kinetics of GFP-synapsin Ia
45
mutated at CaM kinase phosphorylation
30 40
sites. The number of experiments for each
30 mutation is as follows: wt, 2; S1A, 3; S2/3A,
20 S1A 20 4; S1/2/3A, 7. Abbreviations are as in Fig. 5.
15 wt All GFP-synapsin Ia mutants tested showed
S1A significantly slower dispersion kinetics than

10 wt 0 S2/3A
0 S1/2/3A wild-type GFP-synapsin Ia (p < 0.001).
(b) Effect of different mutations on FM 4-64
0 20
wt S1A S2/3A S1/2/3A 15 20 30 40 turnover. All mutants show statistically sig-
a b c dispersion GFP-Syn Ia mutants (s) nificant difference from wild-type
GFP-synapsin Ia (p < 0.03). (c) Positive cor-
relation between the dispersion kinetics
demonstrated that a large fraction of synapsin must dissociate and the effect on FM 4-64 destaining kinetics for various GFP synapsin Ia
from vesicles during activity. mutations when expressed in the absence of endogenous synapsin I and II.
The phenotypes of synapsin I-, II- and I/II-deficient mice
characterized by postsynaptic electrophysiological measure-
ments2527 are considered mild. Further detailed analyses of function by their ability to control vesicle availability at the
presynaptic vesicle recycling of synapsin I-deficient mice by nerve terminal during repetitive AP firing.
FM 1-43 only reveal reduced vesicle pool turnover with brief
trains of AP stimulation and reduced total recycling pool size, METHODS
whereas other parameters including endocytosis and repriming cDNA subcloning and site-directed mutagenesis. EGFP-synapsin Ia fusion
are unaltered 38. However, it is possible that the phenotypic protein was generated by subcloning the rat synapsin Ia cDNA into the
changes in these mice are alleviated by functional redundancy of pEGFP-C1 vector (Clontech, Palo Alto, California). Site-directed muta-
synapsin III4 and by long-term compensatory changes in the lev- genesis was done using the Stratagene (La Jolla, California) QuikChange
els of other nerve terminal proteins as demonstrated in synapsin site-directed mutagenesis kit.
I/II deficient mice26.
Hippocampal cell culture and transfection. Synapsin I/II/ mice were
Our studies demonstrate a link between dissociation of generated by homologous recombination41. Hippocampal CA3CA1
synapsin Ia and the ability of vesicles to mobilize and fuse with regions were dissected from 34-day-old SpragueDawley rats, and
the plasma membrane during AP firing. However, several phe- 01-day wild type and synapsin I/II/ mice. The dissected hippocam-
nomena remain to be explained. First, it is not known if the pal regions were then dissociated, plated and cultured as described42.
dispersion and reclustering of synapsins following phospho- Transfections of GFP-synapsin Ia and its phosphorylation-site mutants
rylation are driven by free diffusion or by an active transport as well as GFP-VAMP were done using calcium phosphate precipita-
process. Second, we do not understand why there is still dis- tion as described43, and performed on 78-day-old neuronal cultures.
persion in a fully mutated S1/2/3A form of synapsin when All animal experiments and use were approved by the Institutional
expressed in a synapsin I/II / background. This result sug- Animal Care and Use Committee of the Weill Medical College of
gests that in addition to the CaM kinase sites, an unknown Cornell University.
activity-dependent modification of synapsin modulates its dis-
sociation from vesicles. This could potentially be achieved Immunocytochemistry. Individual dishes of 23-week-old rat hip-
pocampal cultures were processed for immunofluorescence of both
through other phosphorylation sites, or through calcium-
synaptophysin and synapsins after fixation in 4% paraformaldehyde
dependent binding of ATP 39,40. Third, the finding that the (Electron Microscopy Science, Washington, Pennsylvania), 1 PBS,
concentration of GFP-synapsin Ia decreases only by approxi- and 0.041% sucrose for 15 min. Dishes were separated into groups and
mately 45% before reaching steady state during AP firing at subjected to 1 of 3 conditions immediately before fixation: control,
nerve terminals (Fig. 2c) suggests that a portion of synapsins 900 AP field stimulation, or 900 AP field stimulation followed by
do not dissociate from vesicles, as has been suggested by in 10-min rest period. Cells were then permeabilized in the same fixative
vitro experiments21. Analysis of fluorescence intensity distri- plus 0.25% Triton for 15 min, blocked in 10% BSA for an hour at 37C
bution of a freely diffusing cytoplasmic volume marker (cyto- and incubated overnight with both monoclonal anti-synaptophysin
plasmically expressed GFP) suggests that if all GFP-synapsin antibody (clone SVP-38, Sigma, St. Louis, Missouri) and affinity-puri-
fied anti-synapsin Ia, IIa and IIIa10 polyclonal antibody (G-304). Cells
dissociated, the extent of dispersion would be greater than
were then incubated with an Alexa 488-labeled anti-mouse IgG and
approximately 7580% (data not shown). What prevents the an Alexa 546 anti-rabbit IgG secondary antibody (Molecular Probes,
remaining GFP-synapsin Ia from dispersing? Whether this is Eugene, Oregon) for an hour, and mounted for observation. Image
correlated with a non-recycling pool of vesicles or whether analysis was performed using a blinded procedure without knowledge
there is something else in addition to calcium-dependent of the experimental treatment.
phosphorylation that is required for synapsin dispersion
remains to be determined. Experimental conditions. Experiments were performed on 510-day
Although previous knockout analysis clearly demonstrat- post-transfected cultures. Coverslips were mounted in a perfusion
ed that synapsin I and II alone cannot account for the cluster- chamber equipped with field stimulation electrodes on the stage of a
custom-built laser-scanning confocal microscope as described42. Unless
ing of vesicles at active zones, the studies presented here otherwise noted, cells were perfused at room temperature (24C) in
indicate that synapsins control the availability of vesicles a saline solution consisting of 119 mM NaCl, 2.5 mM KCl, 2 mM
through their ability to dissociate from synaptic vesicles in a CaCl2, 2 mM MgCl2, 5 mM HEPES (pH 7.4), 30 mM glucose, 10 M
phosphorylation-dependent manner. Thus, synapsins provide 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) and 50 M D, L -2-
a phosphorylation-mediated layer of control over presynaptic amino-5-phosphonovaleric acid (APV). Acidic-solution with final pH

articles
of 4.0 was prepared by replacing HEPES in the standard saline with with small synaptic vesicles: distinct sites in synapsin I bind to vesicle
MES (pK a 6.1), all other components in the saline remaining phospholipids and vesicle proteins. J. Cell Biol. 108, 18631872 (1989).
unchanged. Synaptic vesicle pools were labeled by field-stimulating 15. Benfenati, F., Greengard, P., Brunner, J. & Bahler, M. Electrostatic and
hydrophobic interactions of synapsin I and synapsin I fragments with
cultures for 30 s at 20 Hz in the presence of FM 4-64 in normal saline. phospholipid bilayers. J. Cell Biol. 108, 18511862 (1989).
An additional 60 s of dye exposure was allowed to ensure complete 16. Benfenati, F., Valtorta, F., Chieregatti, E. & Greengard, P. Interaction of free
labeling of all recycling vesicles. The cultures were subsequently rinsed and synaptic vesicle-bound synapsin I with F-actin. Neuron 8, 377386
in dye-free solution for 10 min before dye destaining. Unless other- (1992).
wise stated, all reagents were obtained from Sigma. 17. Benfenati, F. et al. Interactions of synapsin I with phospholipids: possible role
in synaptic vesicle clustering and in the maintenance of bilayer structures.
J. Cell Biol. 123, 18451855 (1993).
Optical measurements, microscopy and analysis. Laser-scanning flu-
18. Ceccaldi, P. E. et al. Dephosphorylated synapsin I anchors synaptic vesicles to

orescence images were acquired as described32. Quantitative mea- actin cytoskeleton: an analysis by videomicroscopy. J. Cell Biol. 128, 905912
surements of fluorescence intensity at individual boutons and (1995).
neighboring axonal regions were obtained by averaging a 4 4 area of 19. Chilcote, T. J., Siow, Y. L., Schaeffer, E., Greengard, P. & Thiel, G. Synapsin IIa
bundles actin filaments. J. Neurochem. 63, 15681571 (1994).
pixel intensities. FM 4-64 destaining data were normalized to the total 20. Greengard, P., Browning, M. D., McGuinness, T. L. & Llinas, R. Synapsin I, a
loss of fluorescence during the train of AP, determined by subtracting phosphoprotein associated with synaptic vesicles: possible role in regulation
the average of the final three time points from that of the first five time of neurotransmitter release. Adv. Exp. Med. Biol. 221, 135153 (1987).
points before stimulation for each individual bouton. Time constants 21. Hosaka, M., Hammer, R. E. & Sudhof, T. C. A phospho-switch controls the
dynamic association of synapsins with synaptic vesicles. Neuron 24, 377387
for FM 4-64 destaining were obtained by fitting the destaining curves (1999).
to single exponential decays. The fluorescence change of GFP-synapsin 22. Thiel, G., Sudhof, T. C. & Greengard, P. Synapsin II. Mapping of a domain in
Ia and its mutants either at synaptic boutons or in axons were nor- the NH2-terminal region which binds to small synaptic vesicles. J. Biol.
malized to the starting fluorescence (F0) at individual selected region. Chem. 265, 1652716533 (1990).
The time constants for dispersion of GFP-synapsin Ia wild type and 23. Schiebler, W., Jahn, R., Doucet, J. P., Rothlein, J. & Greengard, P.
Characterization of synapsin I binding to small synaptic vesicles. J. Biol.
mutants were determined similarly by fitting the dispersion curve to Chem. 261, 83838390 (1986).
single exponential decay, or in the case of minimal or very slow dis- 24. De Camilli, P., Benfenati, F., Valtorta, F. & Greengard, P. The synapsins. Annu.
persion by directly taking the time point at 63% of the normalized flu- Rev. Cell Biol. 6, 433460 (1990).
orescence decrease. 25. Li, L. et al. Impairment of synaptic vesicle clustering and of synaptic
transmission, and increased seizure propensity, in synapsin I-deficient mice.
Proc. Natl. Acad. Sci. USA 92, 92359239 (1995).
ACKNOWLEDGEMENTS 26. Rosahl, T. W. et al. Essential functions of synapsins I and II in synaptic vesicle
We would like to thank R. Scheller for providing the GFP-VAMP construct, W. Yan regulation. Nature 375, 488493 (1995).
for technical assistance and members of the Ryan and Greengard labs for discussions. 27. Rosahl, T. W. et al. Short-term synaptic plasticity is altered in mice lacking
synapsin I. Cell 75, 661670 (1993).
This work was supported by grants from the NIH to T.A.R. (NS24692 and 28. Stefani, G. et al. Kinetic analysis of the phosphorylation-dependent
GM61925-01) and P.G. (MH39327). T.A.R. is an Alfred P. Sloan Research Fellow. interactions of synapsin I with rat brain synaptic vesicles. J. Physiol. (Lond.)
504, 501515 (1997).
29. Sihra, T. S., Wang, J. K., Gorelick, F. S. & Greengard, P. Translocation of
RECIEVED 3 OCTOBER; ACCEPTED 12 OCTOBER 2001 synapsin I in response to depolarization of isolated nerve terminals. Proc.
Natl. Acad. Sci. USA 86, 81088112 (1989).
30. Torri, T. F., Bossi, M., Fesce, R., Greengard, P. & Valtorta, F. Synapsin I
1. Greengard, P., Valtorta, F., Czernik, A. J. & Benfenati, F. Synaptic vesicle partially dissociates from synaptic vesicles during exocytosis induced by
phosphoproteins and regulation of synaptic function. Science 259, 780785 electrical stimulation. Neuron 9, 11431153 (1992).
(1993). 31. Henkel, A. W. & Betz, W. J. Staurosporine blocks evoked release of FM1-43
2. Hilfiker, S. et al. Two sites of action for synapsin domain E in regulating but not acetylcholine from frog motor nerve terminals. J. Neurosci. 15,
neurotransmitter release. Nat. Neurosci. 1, 2935 (1998). 82468258 (1995).
3. Hosaka, M. & Sudhof, T. C. Synapsin III, a novel synapsin with an unusual 32. Sankaranarayanan, S. & Ryan, T. A. Real-time measurements of vesicle-
regulation by Ca2+. J. Biol. Chem. 273, 1337113374 (1998). SNARE recycling in synapses of the central nervous system. Nat. Cell Biol. 2,
4. Kao, H. T. et al. A third member of the synapsin gene family. Proc. Natl. Acad. 197204 (2000).
Sci. USA 95, 46674672 (1998). 33. Czernik, A. J., Pang, D. T. & Greengard, P. Amino acid sequences surrounding
5. Kao, H. T. et al. Molecular evolution of the synapsin gene family. J. Exp. Zool. the cAMP-dependent and calcium/calmodulin-dependent phosphorylation
285, 360377 (1999). sites in rat and bovine synapsin I. Proc. Natl. Acad. Sci. USA 84, 75187522
6. Sudhof, T. C. et al. Synapsins: mosaics of shared and individual domains in a (1987).
family of synaptic vesicle phosphoproteins. Science 245, 14741480 (1989). 34. Llinas, R., McGuinness, T. L., Leonard, C. S., Sugimori, M. & Greengard, P.
7. Huttner, W. B., Schiebler, W., Greengard, P. & De Camilli, P. Synapsin I Intraterminal injection of synapsin I or calcium/calmodulin-dependent
(protein I), a nerve terminal-specific phosphoprotein. III. Its association with protein kinase II alters neurotransmitter release at the squid giant synapse.
synaptic vesicles studied in a highly purified synaptic vesicle preparation. Proc. Natl. Acad. Sci. USA 82, 30353039 (1985).
J. Cell Biol. 96, 13741388 (1983). 35. Llinas, R., Gruner, J. A., Sugimori, M., McGuinness, T. L. & Greengard, P.
8. De Camilli, P., Harris, S. M. Jr., Huttner, W. B. & Greengard, P. Synapsin I Regulation by synapsin I and Ca2+-calmodulin-dependent protein kinase II
(Protein I), a nerve terminal-specific phosphoprotein. II. Its specific of the transmitter release in squid giant synapse. J. Physiol. (Lond.) 436,
association with synaptic vesicles demonstrated by immunocytochemistry in 257282 (1991).
agarose-embedded synaptosomes. J. Cell Biol. 96, 13551373 (1983). 36. Hosaka, M. & Sudhof, T. C. Homo- and heterodimerization of synapsins.
9. Mandell, J. W. et al. Synapsins in the vertebrate retina: absence from ribbon J. Biol. Chem. 274, 1674716753 (1999).
synapses and heterogeneous distribution among conventional synapses. 37. Tanaka, H. et al. Molecular modification of N-cadherin in response to
Neuron 5, 1933 (1990). synaptic activity. Neuron 25, 93107 (2000).
10. Mandell, J. W., Czernik, A. J., De Camilli, P., Greengard, P. & Townes- 38. Ryan, T. A., Li, L., Chin, L. S., Greengard, P. & Smith, S. J. Synaptic vesicle
Anderson, E. Differential expression of synapsins I and II among rat retinal recycling in synapsin I knock-out mice. J. Cell Biol. 134, 12191227 (1996).
synapses. J. Neurosci. 12, 17361749 (1992). 39. Esser, L. et al. Synapsin I is structurally similar to ATP-utilizing enzymes.
11. Bahler, M. & Greengard, P. Synapsin I bundles F-actin in a phosphorylation- EMBO J. 17, 977984 (1998).
dependent manner. Nature 326, 704707 (1987). 40. Hosaka, M. & Sudhof, T. C. Synapsins I and II are ATP-binding proteins with
12. Bahler, M., Benfenati, F., Valtorta, F., Czernik, A. J. & Greengard, P. differential Ca2+ regulation. J. Biol. Chem. 273, 14251429 (1998).
Characterization of synapsin I fragments produced by cysteine-specific 41. Ferreira, A. et al. Distinct roles of synapsin I and synapsin II during neuronal
cleavage: a study of their interactions with F-actin. J. Cell Biol. 108, development. Mol. Med. 4, 2228 (1998).
18411849 (1989). 42. Ryan, T. A. Inhibitors of myosin light chain kinase block synaptic vesicle pool
13. Benfenati, F., Valtorta, F., Bahler, M. & Greengard, P. Synapsin I, a neuron- mobilization during action potential firing. J. Neurosci. 19, 13171323
specific phosphoprotein interacting with small synaptic vesicles and F-actin. (1999).
Cell Biol. Int. Rep. 13, 10071021 (1989). 43. Threadgill, R., Bobb, K. & Ghosh, A. Regulation of dendritic growth and
14. Benfenati, F., Bahler, M., Jahn, R. & Greengard, P. Interactions of synapsin I remodeling by Rho, Rac, and Cdc42. Neuron 19, 625634 (1997).

articles
TNF contributes to the death of

NGF-dependent neurons during
development
Victoria Barker1, Gayle Middleton1, Fleur Davey2 and Alun M. Davies1,3
1 Department of Preclinical Veterinary Sciences, Royal (Dick) School of Veterinary Studies, Summerhall Square, Edinburgh EH9 1QH, Scotland
2 School of Biomedical Sciences, Bute Medical Buildings, University of St. Andrews, St. Andrews, Fife KY16 9AT, Scotland
3 Rinat Neuroscience Corporation, 3155 Porter Drive, Palo Alto, California 94304, USA
Correspondence should be addressed to A.M.D. (davies@rinatneuro.com)
Published online: 29 October 2001, DOI: 10.1038/nn755
Many sympathetic and sensory neurons depend on a supply of nerve growth factor (NGF) from their
targets during development, and neurons that fail to obtain sufficient NGF die by apoptosis. Here
we show that tumor necrosis factor alpha (TNF) is involved in bringing about the death of NGF-
deprived neurons. Function-blocking antibodies against either TNF or TNF receptor 1 (TNFR1) res-
cued many sympathetic and sensory neurons following NGF deprivation in vitro. Fewer sympathetic
and sensory neurons died during the phase of naturally occurring neuronal death in TNF-deficient
embryos, and neurons from these embryos survived in culture better than wild-type neurons. These
neurons coexpress TNF and TNFR1 during this stage of development, suggesting that TNF acts by
an autocrine loop.
Neurons are generated in excess in the developing vertebrate with ischemia9, HIV-1 infection10 and axotomy11, and can exert a
peripheral nervous system, and the superfluous neurons are lost neuroprotective effect against glutamate excitotoxicity12,13. It is
during a phase of programmed cell death that occurs shortly after not known, however, if TNF is involved in bringing about the
they innervate their targets. NGF is the founder of a family of death of neurons that fail to obtain adequate trophic support from
structurally related secreted proteins termed neurotrophins that their innervation targets during normal embryonic development.
promote and regulate the survival of many kinds of neurons in
the peripheral nervous system during this stage of development. RESULTS
NGF is required for the survival of sympathetic neurons and a TNF and TNFR1 antibodies rescue neurons
subset of nociceptive sensory neurons. These neurons are lost in To investigate if TNF is involved in promoting apoptosis of
developing rodents treated with function-blocking anti-NGF embryonic neurons deprived of trophic support, we studied the
antibodies1 and in mice that are homozygous for null mutations effects of function-blocking anti-TNF and anti-TNFR1 anti-
in the NGF gene2 or the TrkA gene, which encodes the NGF bodies on the survival of NGF-dependent neurons following NGF
receptor tyrosine kinase3. NGF is synthesized in the peripheral deprivation. We established low-density, dissociated cultures of
target tissues of NGF-dependent neurons in proportion to their sympathetic neurons from the superior cervical ganglia (SCG)
innervation density4,5, and administration of exogenous NGF and sensory neurons from the trigeminal ganglia of mouse
prevents naturally occurring cell death within populations of embryos at E16, when most of these neurons have become depen-
NGF-dependent neurons during development1. dent on NGF for survival in vitro and naturally occurring neu-
Because the binding of NGF to TrkA generates survival signals ronal death is occurring in vivo1416. After 12 hours incubation
within the cell that prevent caspase activation and subsequent with NGF, the cultures were washed extensively to remove this
apoptosis6, it is assumed that neurons die following NGF depri- neurotrophin and were either resupplemented with NGF, or
vation because of the withdrawal of these survival signals. How- grown without NGF and treated with anti-TNF or anti-TNFR1
ever, we show here that this death is due in part to the action of antibodies or TNF. Whereas about 80% of the sympathetic and
TNF, a proinflammatory cytokine that induces apoptosis in sensory neurons survived for at least another 48 hours after
some cell types. TNF exerts its effects by binding to the recep- resupplementation with NGF, over 80% died within 48 hours of
tors TNFR1 and TNFR2. The cytotoxic effects of TNF are medi- NGF deprivation (Fig. 1). Treatment of NGF-deprived neurons
ated via TNFR1, which has a cytoplasmic death domain that with either anti-TNF or anti-TNFR1 antibodies rescued a quar-
interacts with the adapter protein TRADD following ligand bind- ter to a third of the neurons that would have otherwise died
ing. TRADD in turn interacts with another adapter protein FADD, (Fig. 1). Doseresponse analysis revealed that this effect of the
which recruits and activates pro-caspase 8 with the resultant acti- antibodies reached a plateau at a concentration of 2 ng/ml, and
vation of the cell death machinery7. TNFR2 lacks a death domain higher concentrations of antibodies singularly and in combina-
but synergistically enhances TNFR1-induced cytotoxicity8. In the tion did not rescue additional neurons (data not shown). These
nervous system, TNF contributes to neuronal death associated results were not due to non-specific actions of antibodies on

articles
Fig. 1. Survival of E16 SCG and trigeminal neu-

rons cultured with NGF, TNF and function-
blocking TNF or TNFR1 antibodies after NGF
deprivation. The neurons were grown for 12 h
after plating in medium containing 2 ng/ml NGF.
After washing, growth factors and antibodies were
added to a concentration of 2 ng/ml. The numbers
of surviving neurons were counted after a further
48-h incubation and expressed as a percentage of
the number counted after NGF deprivation.

Control cultures received no reagents following
NGF deprivation. The means and standard errors
of six separate experiments are shown.
survival of neurons grown with TNF plus

NGF relative to neurons grown with NGF
alone. These results suggest that endoge-
nously produced TNF is not functioning
to maximum effect in promoting neuronal
death and that only a proportion of the
neurons are susceptible to the cytotoxic
actions of TNF at this stage of develop-
ment. Also, TNF receptors are active in
sympathetic neurons, as TNF treatment
neuronal survival because there was no significant difference promotes nuclear translocation of the RelA subunit of NFB
between the number of neurons surviving in control cultures (data not shown), although we do not know if this signaling path-
and cultures treated with antibodies against an intracellular anti- way is involved in mediating the cytotoxic action of TNF on
gen (III tubulin) even at concentrations as high as 50 ng/ml. these neurons.
Anti-TNF and anti-TNFR1 rescued a similar proportion of
sympathetic and sensory neurons that were plated in medium Enhanced survival of TNF-deficient neurons in culture
containing these antibodies from the outset (data not shown). To provide additional evidence for the involvement of TNF in
These results suggest that TNF synthesized by cells in dissoci- bringing about neuronal death following NGF deprivation, we
ated cultures of embryonic sympathetic and sensory ganglia is compared the survival of embryonic SCG and trigeminal neurons
involved in bringing about neuronal death following NGF depri- cultured from wild-type mice and mice that are homozygous for a
vation, and that this action of TNF is mediated at least in part null mutation in the TNF gene. After 48 hours incubation, there
by TNFR1. were approximately twice as many sympathetic and sensory neu-
TNF treatment also reduced the survival of sympathetic and rons surviving in cultures established from E16 TNF/ mice com-
sensory neurons grown either with or without NGF (Fig. 2). pared with neurons from wild-type embryos in the same litters
Doseresponse analysis revealed that as little as 3.2 pg/ml TNF (Fig. 3). These results further support the idea that endogenously
caused significant reductions in the survival of neurons grown in produced TNF is involved in bringing about the death of embry-
either the presence or absence of NGF (p < 0.005, t-tests). This onic sympathetic and sensory neurons in the absence of NGF.
cytotoxic effect of TNF reached a plateau
at a concentration of 2 ng/ml, and no
additional death was observed at a 25-fold
higher concentration. At these concentra-
tions of TNF, less than 2% of the neu-
rons were remaining in NGF-free cultures
(compared to approximately 15% survival
in NGF-free cultures without TNF) and
there was an approximately 20% drop in
Fig. 2. E16 SCG and trigeminal neuron dose

responses to NGF and TNF. The neurons
were grown for 12 h after plating in medium
containing 2 ng/ml NGF. After washing, the
cultures were supplemented with NGF, TNF
or NGF plus TNF over a range of concen-
trations. The numbers of surviving neurons
were counted after a further 48-h incubation
and expressed as a percentage of the number
counted after NGF deprivation. Control cul-
tures received no reagents following NGF
deprivation. The means and standard errors
of six separate experiments are shown.
articles
Fig. 3. In vitro survival of SCG and trigeminal neurons from TNF-defi-

cient and wild-type embryos. Cultures were established from E16
TNF/ and TNF+/+ embryos. After 48 h incubation in defined, serum-
free medium, the number of surviving neurons was counted, and is
expressed here as a percentage of the number counted 6 h after plat-
ing. The means and standard errors of three separate experiments
are shown.
in the SCG (52% increase, p < 0.0001, t-test) and trigeminal

ganglia (23% increase, p < 0.0001, t-test). By the first postnatal
day (P1), the SCG and trigeminal ganglia of TNF/ embryos
contained less than half as many dying neurons as in wild-type
embryos, and neurons were still significantly more abundant
in the SCG (14% increase, p < 0.001, t-test) and trigeminal gan-
glia (26% increase, p < 0.0001, t-test) of these TNF-deficient
embryos compared with wild-type embryos. These results
demonstrate that TNF is required to stimulate the death of a
proportion of neurons in the developing SCG and trigeminal
ganglia during the period of naturally occurring neuronal death.
Analysis of neuronal death in the nodose ganglia also revealed
significantly reduced neuronal death and increased numbers
of neurons in TNF-deficient mice at these stages of develop-
Reduced neuronal death in TNF-deficient embryos ment (data not shown), suggesting that TNF is also involved in
To ascertain the physiological relevance of our in vitro observa- promoting the death of sensory neurons that depend on BDNF
tions, we estimated the number of surviving neurons and the and NT4 for survival during the period of naturally occurring
extent of neuronal death in the SCG and trigeminal ganglia of neuronal death.
wild-type and TNF/ mice at two stages during the period of
naturally occurring neuronal death by counting the total num- Developing neurons express TNF and TNFR1
ber of neurons and the number of pyknotic neurons in serial sec- To further substantiate the role of TNF in embryonic
tions (Fig. 4). At E16, the SCG and trigeminal ganglia of TNF/ sensory and sympathetic neurons and investigate its mode of
embryos contained only a quarter as many dying neurons as in action, we used immunocytochemistry to identify which cells
wild-type embryos, and there were significantly more neurons express TNF and TNFR1 in dissociated cultures. Because the
Fig. 4. Number of pyknotic neurons and total numbers of neurons in the SCG and trigeminal of E16 and P1 TNF+/+ and TNF/ mice. The means and
standard errors of the data obtained from both sets of ganglia from 3 E16 and 4 P1 TNF+/+ mice and from 4 E16 and 5 P1 TNF/ mice are shown.

articles
a b Fig. 5. Expression of TNF and TNFR1 by cultured

SCG neurons. E16 SCG neurons from wild-type
embryos were stained for TNF (a) and TNFR1
(b); E16 SCG neurons (arrowheads) from TNF/
embryos did not stain for TNF (c), and no neu-
ronal staining was observed in wild-type neurons
(arrows) if primary antibodies were omitted (d).
The few non-neuronal cells in these cultures
showed a low level of TNF immunoreactivity
(arrowhead, a). Scale bar, 20 M.
these cells and Fas-L is induced following neu-

c d rotrophic factor deprivation. Treatment with
Fas-L causes the death of many neurons even
in the presence of optimal concentrations of
neurotrophic factors. Thus, Fas signaling is
triggered by neurotrophic factor withdrawal
and this in turn brings about the subsequent
demise of many of the neurons. Similarly,
TNF-mediated neuronal death following
brain injury and ischemia in adult rats is asso-
ciated with rapid upregulation in the expres-
sion of TNF21,22.
serum-free medium used in our cultures is not conducive to Immunocytochemical analysis of TNF and TNFR1 expres-
the survival of non-neuronal cells, most cells remaining in E16 sion on cultured embryonic sympathetic neurons suggests
cultures after 48 hours of incubation were neurons (>90%). that virtually all of these neurons express this ligand/receptor
Virtually all of these neurons were immunoreactive for TNF combination. This observation, together with our demon-
protein and for TNFR1 (Fig. 5). The small number of non- stration that inhibition of neuronal death by function-block-
neuronal cells in these cultures displayed only a low level of ing anti-TNF and anti-TNFR1 antibodies was observed in
immunoreactivity for TNF and TNFR1. Neurons in cultures low dissociated density cultures that contained only a very
established from TNF/ embryos showed no immunoreac- small percentage of non-neuronal cells, suggest that TNF
tivity for TNF and no staining was observed in cultures of contributes to the death of NGF-deprived neurons by an
wild-type neurons if the primary antibodies were omitted, autocrine mechanism. However, the low level of TNF
demonstrating specific staining for TNF and its receptor. immunoreactivity observed in non-neuronal cells raises the
These findings indicate that most if not all neurons synthesize possibility that the neurons may additionally obtain some
TNF and express the TNFR1 receptor during the stage of NGF TNF from other cells.
dependence, and suggest that TNF exerts its actions by an In summary, our findings add a twist to the molecular mech-
autocrine mechanism. anisms that bring about the death of developing NGF-dependent
neurons that fail to obtain adequate NGF to support their sur-
DISCUSSION vival. The demise of these neurons seems to be caused in part by
We have used several complementary experimental approach- the cytotoxic actions of TNF that they produce themselves. In
es to demonstrate that TNF is involved in bringing about the future work, it will be important to ascertain the extent to which
death of sympathetic and sensory neurons during the phase of TNF, FasL and possibly other cytotoxic cytokines are involved
programmed cell death in the developing peripheral nervous in bringing about naturally occurring neuronal death in these
system. Function-blocking antibodies to either TNF or and other populations of neurons in the developing vertebrate
TNFR1 but not control antibodies to an unrelated protein pre- nervous system.
vented the death of a significant proportion of these neurons
following NGF deprivation in vitro. TNF treatment promot- METHODS
ed the death of cultured sympathetic and sensory neurons in Neuron culture and survival assays. Dissociated cultures were estab-
lished from the SCG and trigeminal ganglia of E16 CD1 embryos. The
a dose-dependent manner. The sympathetic and sensory gan-
dissected ganglia were trypsinised and dissociated by trituration and were
glia of TNF-deficient mice contained substantially fewer dying grown in defined, serum-free medium on a poly-ornithine/laminin sub-
neurons and significantly greater numbers of neurons than the stratum23 in the 11-mm diameter wells of 4-well dishes (Greiner, Stone-
ganglia of wild-type littermates during the phase of pro- house, UK) at a density of 200400 neurons per well. After an initial 12-h
grammed cell death in vivo. Sympathetic and sensory neurons incubation period with 2 ng/ml NGF, the neurons were washed exten-
from TNF-deficient embryos survived better in culture than sively to remove this factor and were subsequently incubated in medi-
neurons from wild-type embryos. um with either no added factors (control), NGF (gift of A. Rosenthal,
There are some parallels between our findings and recent Rinat Neuroscience Corporation, Palo Alto, California), TNF (Cal-
work on Fas, a member of the TNF receptor family, and its biochem, Nottingham, UK), NGF plus TNF, function-blocking anti-
TNF antibody (Upstate Biotechnology, Buckingham, UK),
ligand Fas-L, which have a well-established role in triggering
function-blocking anti-TNFR1 antibody (Oncogene Research Products,
apoptosis in lymphocytes17. Blocking interaction between Fas Nottingham, UK) or anti-III tubulin antibody (Promega, Southamp-
and Fas-L reduces cell death induced by neurotrophic factor ton, UK). After 48 h, the numbers of surviving neurons in each well were
deprivation in phaeochromocytoma cells, cerebellar granule cells counted and expressed as a percentage of the number of attached neu-
and spinal motoneurons1820. Fas is constitutively expressed in rons after the 12-h wash. Very similar results were obtained if the

articles
cultures were additionally treated with anti-NGF antibody to function- ACKNOWLEDGEMENTS

ally inactivate any residual NGF after washing. This work was supported by grants from the Wellcome Trust and European
To further investigate the role of TNF in promoting neuronal
Commission.
death, we established cultures from mouse embryos that have a null
mutation in the TNF gene. TNF +/ mice (Jackson Laboratories,
RECEIVED 18 AUGUST; ACCEPTED 21 SEPTEMBER 2001
Maine) were crossed to produce TNF / , TNF +/ and TNF +/+
embryos. After 16 days gestation, the pregnant females were killed and
the embryos were genotyped by a PCR-based method. The SCG and 1. Levi-Montalcini, R. The nerve growth factor 35 years later. Science 237,
trigeminal ganglia from wild-type and homozygous embryos were 11541162 (1987).
separately pooled, and low-density, dissociated cultures were estab- 2. Crowley, C. et al. Mice lacking nerve growth factor display perinatal loss of
lished. The neurons were grown in defined medium for 48 h and the sensory and sympathetic neurons yet develop basal forebrain cholinergic
neurons. Cell 76, 10011011 (1994).
number surviving after this time is expressed as a percentage of the 3. Smeyne, R.J. et al. Severe sensory and sympathetic neuropathies in mice
number of attached neurons counted 6 h after plating. All animal work carrying a disrupted Trk/NGF receptor gene. Nature 368, 246249 (1994).
was approved by our institutional animal use committee and by the 4. Korsching, S. & Thoenen, H. Nerve growth factor in sympathetic ganglia and
Home Office. corresponding target organs of the rat: correlation with density of
sympathetic innervation. Proc. Natl. Acad. Sci. USA 80, 35133516 (1983).
5. Harper, S. & Davies, A. M. NGF mRNA expression in developing cutaneous
Immunocytochemistry. Immunocytochemistry was used to visualize epithelium related to innervation density. Development 110, 515519 (1990).
expression of TNF and TNFR1 in cultured neurons. The cultures 6. Kaplan, D. R. & Miller, F. D. Neurotrophin signal transduction in the nervous
were fixed with methanol at 20C followed by 1% H2O2 to quench system. Curr. Opin. Neurobiol. 10, 381391 (2000).
endogenous peroxidase activity and permeabilization in 0.5% Triton 7. Ashkenazi, A. & Dixit, V. M. Death receptors: signaling and modulation.
Science 281, 13051308 (1998).
X-100 in PBS. After incubating at 4C overnight with 1:200 dilutions 8. Grell, M. et al. Induction of cell death by tumour necrosis factor (TNF)
of either anti-mouse TNF goat polyclonal antibody (L-19, Santa Cruz receptor 2, CD40 and CD30: a role for TNF-R1 activation by endogenous
Biotechnology, Wembley, UK) or anti-mouse TNFR1 goat polyclonal membrane-anchored TNF. EMBO J. 18, 30343043 (1999).
antibody (Calbiochem), bound primary antibody was detected using 9. Dawson, D. A., Martin, D. & Hallenbeck, J. M. Inhibition of tumor necrosis
the Vector Elite ABC kit (Vector Labs, Orton Southgate, UK) accord- factor-alpha reduces focal cerebral ischemic injury in the spontaneously
hypertensive rat. Neurosci. Lett. 218, 4144 (1996).
ing to the manufacturers instructions. The stained neurons were 10. New, D. R., Maggirwar, S. B., Epstein, L. G., Dewhurst, S. & Gelbard, H. A.
viewed with a Nikon Diaphot microscope (Nikon, Kingston on HIV-1 Tat induces neuronal death via tumor necrosis factor- and activation
Thames, UK). of non-N-methyl-D-aspartate receptors by a NFB-independent mechanism.
J. Biol. Chem. 273, 1785217858 (1998).
11. Terrado, J. et al. Soluble TNF receptors partially protect injured motoneurons
Quantification of neurons in the SCG and trigeminal ganglia. in the postnatal CNS. Eur. J. Neurosci. 12, 34433447 (2000).
The heads from E16 and P1 TNF/ and wild-type embryos were fixed 12. Houzen, H., Kikuchi, S., Kanno, M., Shinpo, K. & Tashiro, K. Tumor necrosis
for 1 h in Carnoys fluid (60% ethanol, 30% chloroform, and 10% factor enhancement of transient outward potassium currents in cultured rat
glacial acetic acid). Following dehydration through a graded alcohol cortical neurons. J. Neurosci. Res. 50, 990999 (1997).
series, the tissue was paraffin-wax-embedded. Serial sections of 13. Carlson, N. G., Bacchi, A., Rogers, S. W. & Gahring, L. C. Nicotine blocks
TNF--mediated neuroprotection to NMDA by an -bungarotoxin-sensitive
the heads were cut at 8 m and were mounted onto poly- pathway. J. Neurobiol. 35, 2936 (1998).
lysine-coated slides (BDH) or Gold Seal Ultrastick Slides (Erie Sci- 14. Wyatt, S. & Davies, A. M. Regulation of nerve growth factor receptor gene
entific, Loughborough, UK). expression in sympathetic neurons during development. J. Cell Biol. 130,
To identify all neurons in these sections, the sections were stained 14351446 (1995).
15. Davies, A. M. & Lumsden, A. G. S. Relation of target encounter and neuronal
for -tubulin class III. Sections were cleared in xylene and rehydrated death to nerve growth factor responsiveness in the developing mouse
before quenching in 3% hydrogen peroxide in methanol for 20 min. trigeminal ganglion. J. Comp. Neurol. 223, 124137 (1984).
Non-specific antibody binding was blocked in 10% horse serum, 0.5% 16. Francis, N. et al. NT-3, like NGF, is required for survival of sympathetic
Triton X-100 in PBS before incubation with mouse anti-III tubulin neurons, but not their precursors. Dev. Biol. 210, 411427 (1999).
antibody (Promega) diluted 1:10,000 in PBS overnight at 4C. The 17. Nagata, S. Apoptosis by death factor. Cell 88, 355365 (1997).
18. Le-Niculescu, H. et al. Withdrawal of survival factors results in activation of
cells were then labeled using biotinylated secondary antibody (1:200), the JNK pathway in neuronal cells leading to Fas ligand induction and cell
avidin, biotinylated horseradish peroxidase macromolecular complex death. Mol. Cell. Biol. 19, 75163 (1999).
(Vectastain ABC Kit, Vector Labs). The substrate used for the reac- 19. Brunet, A. et al. Akt promotes cell survival by phosphorylating and inhibiting
tion was 1 mg/ml diaminobenzidine tetrachloride (FastDAB, Sigma, a Forkhead transcription factor. Cell 96, 857868 (1999).
St. Louis, Missouri). The sections were then counterstained with Gills 20. Raoul, C., Henderson, C. E. & Pettmann, B. Programmed cell death of
embryonic motoneurons triggered through the Fas death receptor. J. Cell
hemotoxylin before dehydration and mounting. Neuronal number Biol. 147, 10491062 (1999).
was quantified using a digital stereology system that employs a com- 21. Knoblach, S. M., Fan, L. & Faden, A. I. Early neuronal expression of tumor
bination of the optical dissector and volume fraction/Cavalieri meth- necrosis factor- after experimental brain injury contributes to neurological
ods (Kinetics Imaging, Bromborough, UK). For quantification of the impairment. J. Neuroimmunol. 95, 115125 (1999).
number of pyknotic nuclei in the ganglia, the sections were stained 22. Botchkina, G. I., Meistrell, M. E. III, Botchkina, I. L. & Tracey, K. J.
Expression of TNF and TNF receptors (p55 and p75) in the rat brain after
with cresyl fast violet acetate. Pyknotic nuclei were then counted using focal cerebral ischemia. Mol. Med. 3, 765781 (1997).
the digital stereology system. All sections were coded prior to esti- 23. Davies, A. M., Lee, K. F. & Jaenisch, R. p75-deficient trigeminal sensory
mating the number of neurons and pyknotic nuclei to avoid any neurons have an altered response to NGF but not to other neurotrophins.
observer bias. Neuron 11, 565574 (1993).

articles
Wallerian degeneration of injured

axons and synapses is delayed by a
Ube4b/Nmnat chimeric gene
Till G.A. Mack1, Michael Reiner2, Bogdan Beirowski2, Weiqian Mi1, Monica Emanuelli3,
Diana Wagner1, Derek Thomson4, Tom Gillingwater4, Felipe Court4, Laura Conforti5,
F. Shama Fernando6, Andrea Tarlton7, Christian Andressen2, Klaus Addicks2, Giulio Magni3,
Richard R. Ribchester4, V. Hugh Perry8 and Michael P. Coleman1,6
1 Center for Molecular Medicine (ZMMK) and Institute for Genetics, University of Cologne, Zuelpicher Strasse 47, D-50674 Cologne, Germany
2 Department of Anatomy I, University of Cologne, Joseph-Stelzmann Strasse 9, D-50931 Cologne, Germany
3 Institute of Biochemistry, University of Ancona, Via Ranieri, 60131 Ancona, Italy
4 Department of Neuroscience, University of Edinburgh, 1 George Square, Edinburgh, EH8 9JZ, UK
5 Molecular Neurobiology Laboratory, Mario Negri Pharmaceutical Research Institute, Via Eritrea, 62, 20157 Milan, Italy
6 Department of Pharmacology, University of Oxford, Mansfield Road, Oxford, OX1 3QT, UK
7 Department of Biochemistry, University of Oxford, South Parks Road, Oxford, OX1 3QU, UK
8 School of Biological Sciences, University of Southampton, Biomedical Sciences Building, Southampton, SO16 7PX, UK
Correspondence should be addressed to M.P.C. (michael.coleman@uni-koeln.de)
Axons and their synapses distal to an injury undergo rapid Wallerian degeneration, but axons in the
C57BL/WldS mouse are protected. The degenerative and protective mechanisms are unknown. We
identified the protective gene, which encodes an N-terminal fragment of ubiquitination factor E4B
(Ube4b) fused to nicotinamide mononucleotide adenylyltransferase (Nmnat), and showed that it
confers a dose-dependent block of Wallerian degeneration. Transected distal axons survived for two
weeks, and neuromuscular junctions were also protected. Surprisingly, the Wld protein was located
predominantly in the nucleus, indicating an indirect protective mechanism. Nmnat enzyme activity,
but not NAD+ content, was increased fourfold in WldS tissues. Thus, axon protection is likely to be
mediated by altered ubiquitination or pyridine nucleotide metabolism.
The distal segment of an injured nerve normally undergoes Wal- Wallerian degeneration has a prominent causative role in a spec-
lerian degeneration within 2448 hours1. Axon death in diverse trum of human neuropathologies. Axon loss occurs not only in
neurodegenerative diseases follows the same final pathway. The traumatic disorders such as spinal cord injury9, but is an early event
earliest observable events, disruption of the cytoskeleton and in numerous neurological disorders of diverse etiology, such as amy-
blebbing of the axolemma, occur within the axon itself, whereas otrophic lateral sclerosis10, multiple sclerosis11 and toxic neuropa-
later stages also involve the reaction of other cell types, such as thy12. The WldS mutation protects also from vincristine toxicity13
Schwann cells and macrophages. It is not known how Wallerian and its potential for protection of axons in diverse neurological dis-
degeneration is initiated, but the mechanism is clearly distinct eases is an area of considerable current interest. Protection of neu-
from neuronal cell body degeneration2,3. ronal cell bodies often fails to prevent neurological disease14,15, so it
Remarkably, central and peripheral nervous system axons in is also important to find ways to protect axons. Thus, the Wld gene
the slow Wallerian degeneration mutant mouse, C57BL/WldS, may open a new avenue for therapeutic strategies.
survive several weeks after transection46. The neuroprotective An 85-kb tandem triplication16 has been characterized with-
phenotype is dominant5 and intrinsic to the axon2,7,8. Thus, an in the WldS region on distal mouse chromosome 4 (ref. 17). This
unknown protective factor should exist in WldS axons even before led to the discovery of a chimeric gene containing the 5 end of
nerve transection, as the protected distal segment of axon is iso- Ube4b and a gene of previously unknown function, D4Cole1e18.
lated from sites of protein translation. The existence of the puta- Both parent proteins are also expressed in WldS mice. The human
tive regulatory molecule suggests that Wallerian degeneration is homolog of D4Cole1e (F.S. Fernando, unpublished data) is iden-
not a passive process, as previously thought, but an active one tical to human NMNAT19, a key enzyme in the synthetic path-
that removes damaged axons8. How this degenerative process is way of NAD+ (ref. 20). Although the chimeric gene is the most
prevented in healthy axons, and thus the nature of the process plausible candidate for Wld, the triplication also directly affects
itself, should follow from the identification of the Wld gene. retinol binding protein 7 (Rbp7) and could exert a position effect

articles
a b Fig. 1. Preserved axon ultrastructure in distal sciatic

nerve 5 days after transection. Electron micrographs
(3,400) of transverse thin sections of lesioned sciatic
nerve 24 mm distal to the lesion site after 5 days. (a)
Wild-type, (b) 4836 homozygote, (c) 4830 homozygote
(d) WldS homozygote. Insets, 20,000 magnifications of
axons marked by asterisk (top inset) and # (bottom
inset) from main images.
-actin promoter. Four transgenic lines (4830,

4836, 4839 and 4858) expressed the protein from
multi-copy integrations. The transgene expression
level increased in the order 4839 < 4830 4858 <
WldS 4836 (Fig. 3). Line 4858, whose expression
c d level and phenotype was similar to line 4830, is not
discussed further. Like spontaneous WldS mice,
transgenic mice up to at least one year old showed
no unusual overt phenotype.
Structural preservation of transected axons

First we tested the structural preservation of
axons following unilateral sciatic nerve transec-
tion, analyzing a nerve segment 24 mm distal to
the lesion after 35 days. Electron microscopy
revealed that almost all axons in line 4836
homozygotes, separated from their cell bodies for
five days, contained fully preserved cytoskeleton
(Fig. 1 and Supplementary Fig. 1, available on the
Nature Neuroscience web site), thus successfully
on nearby genes. Indeed, mutation of the human homolog of reproducing the WldS phenotype. In contrast, all myelinated
one neighboring gene, kinesin family 1b (Kif1b)21, causes axon and unmyelinated wild-type axons showed clear signs of degen-
loss in Charcot-Marie-Tooth disease type 2A22. eration. Line 4830 axons were partially protected in accordance
To test the hypothesis that the Ube4b/Nmnat chimeric gene with the lower transgene expression level in this line (Fig. 3a)
confers the slow Wallerian degeneration phenotype, we expressed and line 4839 homozygotes also showed partial protection after
it in transgenic mice. The WldS phenotype was reproduced fully
in one line and partially in other lines according to transgene
expression level, thus proving that the chimeric gene is the Wld a b
gene. We detected Wld protein in nuclei of neurons but neither in
axons nor in Schwann cells, implicating the existence of down-
stream factor(s) that mediate the protective effect. Further, we
report that the Wld protein showed Nmnat enzyme activity and
acted in a strongly dose-dependent manner.
RESULTS c e
Generation of transgenic mice
To test the role of the Ube4b/Nmnat chimeric gene, we generated
transgenic mice expressing the Ube4b/Nmnat cDNA from a
Fig. 2. Physiological and morphological evidence for preservation of

axons and neuromuscular junctions. Intracellular recordings (a, b), vital d
staining (c, d) and immunostaining (e, f) in isolated transgenic muscles.
(a) Homozygous 4836 FDB, 3 days after axotomy; (b) hemizygous 4836
FDB after 2 days. (c, d) Lumbrical muscles from 4836 hemizygote,
3 days after axotomy (c); 4836 homozygote, 5 days after axotomy (d)
vitally stained with FM1-43, which indicates intact synaptic transmission
mechanisms, together with TRITC--bungarotoxin staining of acetyl- f
choline receptors. (e, f) Immunostaining for neurofilament/SV2 (FITC)
and TRITC--bungarotoxin. (e) 4836 homozygous lumbrical muscle,
5 days after axotomy, showing occupied endplates and some axon
swelling. (f) Confocal stereo pair of 5-day axotomized 4836 homozy-
gous lumbrical muscle showing almost fully occupied endplates (left and
right), partially occupied (second right) and junction with only a slender
axonal filament tipped by a retraction bulb (second left). Scale bar,
20 m (ce); 50 m (f).
articles
Fig. 3. Dose-dependence of axon protection by the Wld gene.

(a) Top, western blot showing the extent of 200-kD neurofilament a
protein degradation in distal sciatic nerve 3 days after transection
(5 days for 4836 and WldS homozygotes). Complete preservation of
NF-200 in uncut contralateral nerve (lane 1) indicated degradation
occurred only in vivo. Middle, western blot showing the expression
level of the Wld protein in brain homogenates (detected by N70 anti-
serum). Bottom, same western blot probed with control monoclonal
antibody -tub 2.1 against -tubulin. (b) Percentage of intact myeli-
nated axons 24 mm distal to a sciatic nerve lesion. Preserved axons

were counted using light microscopy in nerve segments immediately
proximal to those used in (a). Counts of 9399% in uncut contralat-
eral nerves confirmed that observed degeneration occurred in vivo.
n indicates total number of nerves counted. Mean s.e.m. of the mid-
dle three scores when n is odd; mean s.e.m. of the middle four
when n is even. (c) Axon preservation as a function of the expression
level of Wld protein. Axon preservation after 3 (black circle), 5 b
(white square) and 14 (cross) days as in (b); means s.e.m. where
n > 1. Wld protein expression level is the signal quantified from (a),
standardized against -tubulin and expressed as a percentage of the
expression in WldS homozygotes.
c
three days (Fig. 3a). Presence of the WldS phenotype in inde-
pendent lines confirms that it is caused by the transgene rather
than any integration effect. Thus, Ube4b/Nmnat is the Wld
gene, and it protects both sensory and motor axons in
Wld S and transgenic mutants. No alteration to Rbp7, Kif1b
or any other gene is required to reproduce the phenotype
of WldS mice.
Functionally competent motor axons and synapses

Next we tested whether axons were functionally as well as struc-
turally preserved. Nervemuscle preparations of flexor digito-
rum brevis (FDB) were isolated and stimulated 25 days after
lesion. Axotomized wild-type muscles showed no response at Thus, motor nerve conduction, synaptic transmission, vesi-
these time points. In axotomized transgenics, however, as in the cle recycling and motor nerve terminal morphology were pre-
WldS mutant23, conduction of action potentials and synaptic served in a dose-dependent manner, with line 4836 showing a
transmission at neuromuscular junctions persisted for at least level of protection similar to WldS. Axons evidently persisted
three days (Fig. 2). In a homozygous 4836 mouse, 80% (12/15) of longer than functional motor nerve terminals, supporting the
muscle fibers responded to nerve stimulation three days after sci- hypothesis that synaptic degeneration, at least in WldS mice, dif-
atic nerve section (for example, Fig. 2a). In another 4836 fers from that of axons and neuronal cell bodies25.
homozygote, functional innervation even after 5 days was indi-
cated by 73% (22/30) of FDB muscle fibers showing spontaneous Protection depends on Wld protein expression level
miniature endplate potentials (MEPP) and some responding to The above experiments suggested that the degree of protection
nerve stimulation (data not shown). Weaker, and sometimes depends on the expression level of Wld protein. This was sur-
more variable, responses indicating low quantal content were prising because heterozygous WldS axons at short survival times
observed both in line 4836 hemizygotes (Fig. 2b) and homozy- degenerate only slightly faster than those of homozygotes5. We
gous 4830 mice. In the latter, 50% (12/24) and 30% (8/25) of quantified Wld protein expression in each mutant strain to deter-
fibers from two mice responded to stimulation two days after mine the level required to preserve cytoskeletal protein and axon
lesion. These weaker responses are in accord with the lower structure for 35 days (Fig. 3). At lower expression levels, both
expression levels of the Wld protein (see below). measures of axon preservation indicated a strong dose-depen-
Axotomized 4836 nerve terminals also recycled synaptic vesi- dence. Axon counts at 3 days differed significantly between hem-
cles, visualized by activity-dependent staining with the styryl dye izygous 4836 and hemizygous 4830 mice (p < 0.001). Significant
FM1-43, both upon incubating lumbrical muscles in a depolar- differences between homozygotes and hemizygotes of a single
izing solution (Fig. 2c and d) and upon stimulation of the distal line (p < 0.01 for 4830; p < 0.001 for 4839) indicated that this is
stump of axotomized tibial nerve. Preservation of pre- and post- not a line-specific silencing effect26. Line 4839 hemizygotes were
synaptic structures was confirmed using antibodies against even indistinguishable from wild-type mice in neurofilament
NF165/SV2 and TRITC--bungarotoxin respectively (Fig. 2e and western blotting (NF-200 band, Fig. 3a), light microscopy
f). In two line 4836 homozygotes, 8793% of endplates (n = 154 (Fig. 3b) and electron microscopy (data not shown), despite
and 164) were occupied or partially occupied (8%) by nerve ter- expressing a small amount of the Wld protein (Fig. 3a). Thus, it
minals 5 days after axotomy, similar to WldS mice23,24. Fewer than is necessary for expression of Wld to reach a threshold level to
50% (n = 187 and 210) were occupied in two 4836 hemizygotes exert a significant protective effect.
after 3 days, and fewer still in 4830 homozygotes, again reflect- To investigate whether the protective effect of Wld protein
ing the lower transgene expression. plateaus at higher expression levels, and to test axon protection in

articles
Fig. 4. Axon protection 1014 days after transection. (a) Western blot
a showing the extent of 200-kD neurofilament protein degradation in dis-
tal sciatic nerve 1014 days after transection. Lane 1, 4836 homozygote
after 14 days; lane 2, WldS homozygote after 14 days; lane 3, 4836 hem-
izygote after 14 days; lane 4, 4836 hemizygote after 10 days; lane 5,
C57BL/6J after 12 days; lane 6, C57BL/6J unlesioned. (be) Light micro-
scopic images (scale bar, 10 m) of distal sciatic nerves following tran-
section at a more proximal site 1014 days earlier. (b) 4836
homozygote after 14 days (corresponding intact axon count, 73%),
(c) WldS homozygote after 14 days (73%), (d) 4836 hemizygote after
c 10 days (35%), (e) C57BL/6J after 12 days (0%). Electron microscopy
b (data not shown) indicated that cytoskeleton and myelin was preserved
in 4836 and WldS homozygotes as in Fig. 1.
tion. An antiserum labeling the N-terminal 70 amino acids of

Wld protein and Ube4b detected mainly a punctate nuclear pat-
tern in WldS and transgenic neurons (Fig. 5a, c, g and h) and this
signal was absent in wild-type neurons (Fig. 5b and i). There was
d e no detectable signal either in motor axon terminals (Fig. 5df),
which can be preserved by the Wld gene (Fig. 2e and f), or in sci-
atic nerve axons (Supplementary Fig. 2, available on the Nature
Neuroscience web site). Swollen endbulbs of 24-hour transected
central and peripheral nervous system axons, which accumulate
other axonal proteins, also lacked Wld protein signal (data not
shown). In addition, the antibody weakly labeled the normal
Ube4b protein in cytoplasm of wild-type mice, so additional
labeling of any small amount of cytoplasmic Wld protein might
not be distinguishable.
transgenic mice under still more stringent conditions, we extend- Although the WldS phenotype is intrinsic to neurons2,7,8, Wld
ed these studies to post-lesion times of 1014 days. Fourteen days protein was also detected as a punctate nuclear stain in other cell
after transection, protection of axons in line 4836 homozygotes types in both WldS and transgenic mice (Fig. 5d and e). There
(6973%) was as strong as that in WldS homozygotes (7378%; was no sign of Wld protein in glial cells (Fig. 5ce), although pre-
Figs. 3c and 4), again indicating full reproduction of the WldS phe- vious RT-PCR (reverse transcription-polymerase chain reaction)
notype. Protection in line 4836 hemizygotes, however, was con- data indicate there may be a small amount in Schwann cells18. It
siderably weaker (35% after 10 days and only 7% after 14 days). is highly unlikely that expression in other cell types is required
Thus, the extent of axon protection differed far more between
4836 homo- and hemizygotes after 14 days than after 35 days, a
indicating that dose-dependence still operates at higher Wld
b c
expression levels. It follows that if Wld protein expression could
be raised still further, an even stronger protective effect than in
WldS and line 4836 mice could be achieved.
Wld is a predominantly nuclear protein

To determine whether the Wld protein could itself be the pro-
tective factor in axons, we determined its intracellular loca- d e f
Fig. 5. The intracellular location of the Wld protein. (a) Presence of

the Wld protein in neuron nuclei of isocortex and (b) its absence from
C57BL/6J control tissue. Red, anti-N70 antibody, which detects Wld
and Ube4b. Green, anti-MAP2 antibody markedly outlining neuronal
cell bodies. (c) Absence of Wld protein (red) in astrocytes (arrows,
main picture). Cytoplasmic staining in ependymal cells (central chan- g h i
nel) of WldS thoracic spinal cord could be Ube4b. Inset, absence in
astrocytes and endothelial cells (arrow) of WldS isocortex. Green, anti-
GFAP. Blue, Hoechst Dye nuclear counterstain. (df) Confocal images
of triangularis sterni muscle preparations immunostained with anti-
N70 antibody (green) and motor endplates counterstained with
TRITC -bungarotoxin (red). (d) 4836 homozygote, (e) WldS,
j k l
(f) C57BL/6J. (gi) Motor neurons in thoracic spinal cord of both WldS
(g) and 4836 (h) expressed the Wld protein (red) in their nuclei,
whereas those of C57BL/6J (i) did not. Cytoplasmic signals may be
Ube4b or low level Wld protein. Counterstain, neurofilament (green).
(jl) Red channel (Wld protein plus Ube4b) images corresponding to
(gi). Scale bars, 5 m (a, b), 10 m (c), 50 m (dl).

articles
Fig. 6. NAD+ metabolism in WldS brain. (a) Increased Nmnat activity

and (b) unaltered NAD+ content in WldS brain. Bar charts show Nmnat
a b
specific activity and NAD+ content in homogenates of fresh brain from
homozygous WldS and C57BL/6J mice. Means and standard error are
shown (n = 3).
to protect axons, but it is possible that the Wld protein confers

as yet unknown properties on other cell types. For example, the

delayed response of WldS muscles to denervation could have an
intrinsic component27. We conclude that the Wld protein is pre-
dominantly located in the nucleus; thus, other factor(s) are like-
ly to mediate the protective effect on axons.
The Wld protein has Nmnat enzyme activity axon protection. For example, poly-ADP ribosylation uses NAD+,
Sequence alignment with human NMNAT indicated that influencing protein activity and cellular NAD+ and ATP content,
nucleotides 2821140 of Wld span the entire Nmnat open read- especially in response to stress33,34. Mild activation of PARP with-
ing frame. To detect any intrinsic Nmnat enzyme activity, we out NAD+ depletion can be neuroprotective35. Another metabo-
expressed protein recombinantly and measured the enzyme activ- lite, NADPH, is a coenzyme for nitric oxide synthase, an enzyme
ity of the bacterial lysate. The observed specific activity for Nmnat linked to axon damage36, and synthesis of the signaling molecule
(0.96 U/mg) was comparable with that of a bacterial lysate con- cyclic ADP ribose from NAD+ regulates calcium release from
taining recombinant human NMNAT (1.74 U/mg)19. To deter- intracellular stores37, potentially influencing calcium activated
mine whether there is a corresponding increase in Nmnat activity proteases in Wallerian degeneration.
in WldS mice, we studied Nmnat activity in brain homogenates Both spontaneous and transgenic WldS mice could be used to
and found a fourfold increase in Wld S brain compared to investigate the function of each parent gene. We already show that
C57BL/6J (p < 0.05; Fig. 6a). The total content of NAD+, how- overexpression of Nmnat activity causes no overt phenotype, and
ever, was not significantly altered (p = 0.2; Fig. 6b). Therefore, report an in vivo mutation of a mammalian E4 ubiquitination fac-
the Wld protein confers an increase in Nmnat activity without tor. Despite the critical role played by the ubiquitin-proteasome
altering the steady-state level of NAD+. pathway in neurological disease and many other processes, we
know remarkably little about the function of such proteins.
DISCUSSION Identification of the Wld gene facilitates studies to deter-
We showed that the Ube4b/Nmnat chimeric gene is neces- mine whether it protects axons in clinically relevant situations.
sary and sufficient to protect injured axons for two weeks. We The Wld S mutation is already known to protect neuronal
conclude that the Ube4b/Nmnat chimeric gene is the Wld gene, processes in vitro from the toxic effects of vincristine13, indi-
which encodes a unique neuroprotective factor for axons. It is cating that traumatic and toxic axon death share a common
important now to determine whether protection requires final pathway. Current studies indicate that distal axon loss in
Ube4b sequences, Nmnat sequences, or both. The yeast myelin protein zero knockout mutants38 is rescued by WldS
homolog of Ube4b is required to multi-ubiquitinate proteins30 (M. Samsam and R. Martini, unpublished data), and that WldS
and a direct link between ubiquitination and axon degenera- protects in a mouse model of motoneuron disease (A. Ferri and
tion comes from the Uch-l1 mutation in gracile axonal dystro- A.C. Kato, unpublished data). Studies of WldS in diverse neu-
phy 31. However, the Wld protein contains only 70 of 1,173 rological diseases is facilitated by our identification of the Wld
amino acids from Ube4b, and these are absent from the yeast gene and by recent protocols for tracking the inheritance of
homolog. They are therefore unlikely to confer multi-ubiqui- WldS in crosses with neurological disease mutants39. Models of
tination activity but may have a related role. The protective common and complex neurological disorders such as multiple
mechanism may be linked to the nuclear location of the Wld sclerosis and diabetic neuropathy can now be investigated
protein and perhaps to the non-homogeneous intranuclear through the development of viral vectors for Wld and genera-
distribution. Possibilities include sequestering of ubiquitination of transgenic WldS rats. Mutational analysis in human neu-
tion factors by proteinprotein interactions and ubiquitina- rological disorders with a homologous chromosomal location,
tion within the nucleus altering transcription factor stability such as hereditary Parkinsonism 40 , also becomes possible.
or RNA processing, leading to an axon effect mediated by Delayed Wallerian degeneration also alters the glial response to
unknown proteins. Regulated nuclear transport of other ubiq- injury, as in a mouse model of spinal cord injury, where it delays
uitination factors can control ubiquitin-mediated degradation inflammatory cell and astrocytic responses9,41. It may be that
of nuclear substrates32. However, any Wld protein in the axon this information could be used to optimize tissue destruction
below the detection level of immunostaining could still have a and repair processes.
direct protective role. We conclude that the Wld gene is a chimera of Ube4b and
Nmnat is a nuclear protein and the only known mammalian Nmnat encoding a predominantly nuclear protein in neurons,
enzyme catalyzing the reaction NMN + ATP NAD+ + PPi and we propose that other factors may mediate the protective
(ref. 20), a reaction generally assumed not to be at equilibrium effect on the axon. Axon protection is strongly dose-dependent
because of the constitutive action of pyrophosphatases. Thus, the and pyridine nucleotide metabolism is altered in the WldS mouse.
increase in Nmnat activity in WldS should increase NAD+ syn- These findings open the way to a molecular understanding of
thesis, and the maintenance of normal steady-state levels sug- Wallerian degeneration and to much-needed neuroprotective
gests that the putative additional NAD+ is metabolized. The strategies that target not only the cell body, but also the axons
product of a compensatory reaction could itself be involved in and synaptic terminals.

articles
METHODS bungarotoxin (5 g/ml; 30 min) followed by overnight primary antibody

Generation of transgenic mice. Ube4b/Nmnat cDNA was RT-PCR (1:200 monoclonal anti-165 kDa neurofilament plus anti-synaptic vesi-
amplified from WldS brain using Platinum Pfx polymerase (Life Tech- cle antigen SV2, Developmental Studies Hybridoma Bank, University of
nologies, Karlsruhe, Germany) and cloned downstream of the -actin Iowa; or affinity purified rabbit polyclonal N70; see below). Secondary
promoter in pHAPr-1 (ref. 42; Supplementary Fig. 3). Wld protein antibodies (1:200 dilution; 4 h) were FITC-conjugated sheep-anti-mouse
was thus expressed in neurons, where it has an intrinsic effect2,7,8, and IgG or anti-rabbit IgG (Diagnostics Scotland). Preparations were mount-
other cell types, where WldS mice also express it (Fig. 5). A 6-kb frag- ed in Vektashield (Vector Labs, Burlingame, California) and viewed in a
ment containing promoter, cDNA and polyadenylation signal was fluorescence microscope using a Leitz 50 water immersion objective
released using NdeI and EcoRI and gel-purified using QIAquick extrac- lens (NA 1.00; Wetzlar, Germany). Most images were captured with a
tion (Qiagen, Hilden, Germany). Pronuclear injection into CBA X C57 Hamamatsu C5810 chilled color CCD camera and acquired using Open-
F1 single-cell embryos (G. Kollias, Vari, Greece) resulted in nine lab software (Improvision, Coventry, UK). Confocal images were
founders from 62 pups. obtained using a BioRad Radiance 2000 system (Hemel Hempstead, UK).
Genotyping of transgenic mice. DNA was prepared from a 5-mm tail Western blotting. We analyzed Wld protein expression level in mouse
biopsy using the Nucleon HT kit (Amersham Pharmacia, Freiburg, brains homogenized in two volumes of 20 mM HEPES (pH 7.5), 0.2 M
Germany). The 1.1-kb transgene coding region was detected using CaCl2, 0.2 M MgSO4, 1 ml/20 g tissue protease inhibitor cocktail (Sigma,
alkaline Southern blotting of a BamHI/HindIII double digest on Taufkirchen, Germany) and 1 mg/ml DNase (Sigma). Cytoskeletal protein
Hybond N+ (Amersham Pharmacia) and hybridization with a corre- preservation was determined in lesioned sciatic nerves homogenized in 20
sponding 32P-labeled probe. volumes of this buffer. Proteins were separated using standard SDS-PAGE
and semi-dry blotted onto nitrocellulose. Loading and transfer were
Sciatic nerve lesion. Six- to eleven-week-old mice were anesthetized checked using Ponceau S (Sigma) and Coomassie Blue. Primary anti-
intraperitoneally with Ketanest (100 mg/kg; Bayer, Leverkusen, Germany) bodies were applied (overnight, 4C) followed by horseradish peroxi-
and Rompun (5 mg/kg; Parke Davis/Pfizer, Karlsruhe, Germany). Right dase-coupled secondary antibody (1 h, room temperature;
sciatic nerves (upper thigh) were transected and the wounds were closed goat-anti-mouse 1:3,000, goat-anti-rabbit 1:5,000; Dianova, Hamburg,
with single sutures. Two to fourteen days later, mice were killed, the Germany) and detection using enhanced chemiluminescence (Amer-
swollen first 2 mm of the distal nerve was discarded, the next 2 mm was sham Pharmacia). Chimeric protein expression was quantified using
used for light and electron microscopy, and a segment 410 mm distal affinity-purified N70 antibody (below) and -tub 2.1 (Sigma) control
to the lesion site was used in western blotting. Further distal nerves and and Quantity One software (BioRad). Cytoskeletal protein degradation
muscles were used for electrophysiology. was analyzed using phosphate-independent monoclonal N52 (1:2,000;
Sigma) against heavy neurofilament protein.
Light and electron microscopy. Nerve segments were fixed for 13 days Morphological quantification of axon preservation. We counted 500-
in fresh half-strength Karnovskys fixation (4% paraformaldehyde, 2% 1500 myelinated axons in randomly chosen fields 24 mm distal to a sci-
glutaraldehyde in 0.1 M sodium cacodylate, pH 7.3; ref. 43), extensively atic nerve lesion in transverse semithin sections on a Zeiss Axiophot
buffer-rinsed and osmicated for 4 h with 1% OsO4 in 0.1 M cacodylate. microscope (Gttingen, Germany) coupled to a digital camera. Survival
Samples were taken through a graded ethanol series including a uranylic criteria were normal myelin sheaths, uniform axoplasm and intact mito-
acetate en bloc staining step overnight in 70% ethanol. Before infiltra- chondria, supported by electron microscopy spotchecks. Scoring was
tion with Araldite Cy212 epoxy resin (Serva, Heidelberg, Germany), documented with Meta Imaging software (Universal Imaging Corpora-
propylene oxide was used as intermedium. Tissue blocks were cured for tion, Downingtown, Pennsylvania). Standard errors of the mean and t-
60 h at 60C. Semithin (0.5 m) and thin (60 nm) cross-sections were tests were calculated using SPSS for Windows 10.0.
taken on a Reichert Ultracut UCT ultramicrotome. Semithin sections
for light microscopy were stained with methylene blue and thin sections, Cloning and expression of recombinant proteins. Constructs were gen-
with 1% aqueous uranylic acetate (20 min), and sections were counter- erated to express in bacteria the N-terminal 70 amino acids (N70) of the
stained with Reynolds lead citrate (7 min)44. Thin sections were mount- chimeric protein and full-length chimeric protein. Inserts were PCR-
ed on 150 mesh Formvar coated copper grids and examined with a Zeiss amplified from transgene construct using Pfx polymerase (Life Tech-
EM 902 electron microscope at 80 kV acceleration voltage. nologies). Primers for N70 were5-GACTAGCTAGCATGGAGGAGCTGA
GCGCTGAC-3 and 5-ATCCGCTCGAGCTAGTCTGCTGCACCTATG
Neuromuscular junction electrophysiology and morphology. FDB and GGGGA-3.
lumbrical muscles and contralateral controls were removed in Cologne For full-length chimeric cDNA, the second primer was replaced by 5-
and placed in cold physiological saline (137 mM Na+, 4 mM K+, 2 mM CGCCTCGAGTCACAGAGTGGAATGGTTGTGC-3.
Ca 2+ , 1 mM Mg 2+ , 147 mM Cl , 5 mM glucose, 5 mM HEPES, Products were ligated into NheI/XhoI double-digested pET-28b (+) vec-
pH 7.27.4, equilibrated with air or 100% oxygen). Electrophysiological tor (Novagen, Schwalbach, Germany) and transformed into XL-10 Gold
experiments were done later the same day in Edinburgh, following trans- (Stratagene, Amsterdam, Netherlands). Plasmids isolated using the Plasmid
fer to a medium containing similar concentrations of Na+, K+, Ca2+ and Mini Kit (Qiagen) were retransformed into BL21 (DE3) (Novagen). Pro-
Mg2+, plus 23 mM HCO3, 2 mM H2PO4, equilibrated with 95% O2/5% tein expression was induced with 1 mM IPTG for 3 h at OD600 = 0.8. For
CO2. MEPPs and evoked synaptic responses to tibial nerve stimulation analysis of Nmnat activity, cells were lysed by sonication.
(EPPs) were recorded from FDB using an intracellular glass microelec-
trode and analyzed using WinWCP software45 (J. Dempster, University of Generation of polyclonal antisera. The N70 bacterial pellet was resus-
Strathclyde). Recycled synaptic vesicles of motor nerve terminals were pended in native binding buffer (20 mM sodium phosphate, 500 mM
stained in lumbrical muscles using FM1-43 (Molecular Probes, Leiden, sodium chloride pH 7.8,100 g/ml egg white lysozyme), sonicated on
Netherlands) with 20 Hz nerve stimulation or depolarizing physiological ice (6 15 s with 15-s intervals) and centrifuged (3,000 g, 15 min). N70
solutions, and acetylcholine receptors subsequently stained with TRITC- was purified using a ProBond column (Invitrogen, Groningen, Nether-
-bungarotoxin (Molecular Probes)46,47. Endplates and terminals were lands) and concentrated using a YM-3 Centricon centrifugal filter (Mil-
examined in a Nikon fluorescence microscope (Kingston-upon-Thames, lipore, Bedford, Massachusetts). Antisera were raised by intradermal
UK) using respectively a standard rhodamine filter cube and a customized immunization of two SPF-rabbits by Eurogentec (Seraing, Belgium) with
cube with a 435 nm excitation filter, 455 nm dichroic mirror and a boosts at days 14, 28 and 56, and a final bleed at 80 days.
10 nm bandpass 515 nm emission filter46. For affinity purification 500 g N70 protein bound to ProBond resin
Conventional immunocytochemical and fluorescent bungarotoxin (2 ml wet volume) was blocked with 5% (w/v) dried skimmed milk pow-
probes were used for structural analysis. Muscle preparations, fixed for 60 der plus 1% (w/v) BSA in native binding buffer. Resin was incubated
min in 0.1 M PBS, 4% paraformaldehyde, were incubated in TRITC-- with crude antiserum (2 ml in 8 ml binding buffer), and washed with 10

articles
bed volumes binding buffer. Specific antibodies, binding only Wld pro- 2. Deckwerth, T. L. & Johnson, E. M. Jr. Neurites can remain viable after
tein and Ube4b, were eluted with 100 mM ethanolamine (pH 11.5), neu- destruction of the neuronal soma by programmed cell death (apoptosis).
tralized with 1.5 M Tris and dialyzed against PBS. Dev. Biol. 165, 6372 (1994).
3. Finn, J. T. et al. Evidence that Wallerian degeneration and localized axon
degeneration induced by local neurotrophin deprivation do not involve
Immunocytochemistry of spinal cord, nerves and brain. Perfusion- caspases. J. Neurosci. 20, 13331341 (2000).
fixed tissues were fixed for a further four hours in 4% paraformalde- 4. Lunn, E. R., Perry, V. H., Brown, M. C., Rosen, H. & Gordon, S. Absence of
hyde, 0.1 M PBS before embedding in paraffin. Paraffin sections on Wallerian degeneration does not hinder regeneration in peripheral nerve.
Eur. J. Neurosci. 1, 2733 (1989).
poly-L-lysine coated slides were dewaxed, washed in PBS, rinsed in cit- 5. Perry, V. H., Lunn, E. R., Brown, M. C., Cahusac, S. & Gordon, S. Evidence
rate buffer overnight (60C) and incubated with 0.1% Triton X-100 that the rate of Wallerian degeneration is controlled by a single autosomal
(Sigma). After blocking (5% BSA in PBS), the following primary anti- dominant gene. Eur. J. Neurosci. 2, 408413 (1990).
bodies were applied (overnight, 4C): affinity purified N70 (above), 6. Perry, V. H., Brown, M. C. & Lunn, E. R. Very slow retrograde and Wallerian
plus monoclonal anti-MAP2 (Sigma), mouse polyclonal anti-GFAP degeneration in the CNS of C57BL/Ola mice. Eur. J. Neurosci. 3, 102105
(1991).
(Sigma) or monoclonal anti-neurofilament (Biogenex, San Ramon, 7. Glass, J. D., Brushart, T. M., George, E. B. & Griffin, J. W. Prolonged survival
California). Secondary antibodies (1 h, room temperature) were either of transected nerve fibres in C57BL/Ola mice is an intrinsic characteristic of
Cy3-labeled goat-anti-rabbit (Dako, Hamburg, Germany) or Texas the axon. J. Neurocytol. 22, 311321 (1993).
Red-labeled goat-anti-rabbit (Molecular Probes) together with Cy2- 8. Buckmaster, E. A., Perry, V. H. & Brown, M. C. The rate of Wallerian
labeled goat-anti-mouse (both Dako), diluted according to the degeneration in cultured neurons from wild type and C57BL/WldS mice
depends on time in culture and may be extended in the presence of elevated
manufacturers instructions. K+ levels. Eur. J. Neurosci. 7, 15961602 (1995).
9. Zhang, Z., Fujiki, M., Guth, L. & Steward, O. Genetic influences on cellular
Determination of NMNAT activity. Mice were killed, and whole brains reactions to spinal cord injury: a wound-healing response present in normal
were immediately removed and cut into two equal hemispheres for mice is impaired in mice carrying a mutation (WldS) that causes delayed
Wallerian degeneration. J. Comp. Neurol. 371, 485495 (1996).
NMNAT assay and NAD+ determination, respectively. Tissue was sus- 10. Dal Canto, M. C. & Gurney, M. E. Neuropathological changes in two lines of
pended in 3 volumes of 100 mM Tris-HCl, pH 7.4, 0.5 mM EDTA, 1 mM mice carrying a transgene for mutant human Cu, Zn SOD, and in mice
MgCl2, 1 mM DTT, 1 mM PMSF and homogenized on ice (3 3 s with overexpressing wild type human SOD: a model of familial amyotrophic
10-s intervals) using an Ultra-Turrax homogenizer at medium speed. lateral sclerosis (FALS). Brain Res. 676, 2540 (1995).
NMNAT activity was determined using a reaction mixture of 40 mM 11. Perry, V.H. & Anthony, D.C. Axon damage and repair in multiple sclerosis.
Phil. Trans. R. Soc. Lond. B Biol. Sci. 354, 16411647 (1999).
Tris-HCl pH 7.5, 5 mM nicotinamide mononucleotide (NMN), 3.4 mM 12. Glynn, P. Neural development and neurodegeneration: two faces of
ATP, 18 mM MgCl2, 10 mM NaF, and brain homogenate (final volume of neuropathy target esterase. Prog. Neurobiol. 61, 6174 (2000).
0.5 ml). The reaction was started by the addition of NMN at 37C and 13. Wang, M. S., Wu, Y., Culver, D. G. & Glass, J. D. The gene for slow Wallerian
stopped after 1040 min by adding 100 l of assay mixture to 50 l of degeneration (WldS) is also protective against vincristine neuropathy.
Neurobiol. Dis. 8, 155161 (2001).
ice-cold 1.2 M HClO4. After 10 min, at 0C the mixture was centrifuged 14. Sagot, Y. et al. Bcl-2 overexpression prevents motoneuron cell body loss but
and 130 l of supernatant was neutralized by addition of 35 l 0.8-M not axonal degeneration in a mouse model of a neurodegenerative disease.
K2CO3. NMNAT activity was calculated after HPLC identification and J. Neurosci. 15, 77277733 (1995).
quantification of NAD+ produced48. One unit of enzyme activity cataly- 15. Houseweart, M. K. & Cleveland, D. W. Bcl-2 overexpression does not protect
ses the synthesis of 1 mol of NAD+ per minute at 37C. neurons from mutant neurofilament-mediated motor neuron degeneration.
J. Neurosci. 19, 64466456 (1999).
16. Coleman, M. P. et al. An 85-kb tandem triplication in the slow Wallerian
Determination of NAD+ content. A brain hemisphere in 2 volumes 7.5% degeneration (WldS) mouse. Proc. Natl. Acad. Sci. USA 95, 99859990 (1998).
ice-cold HClO4 was homogenized using an Ultra-Turrax (4 4 s, with 17. Lyon, M. F., Ogunkolade, B. W., Brown, M. C., Atherton, D. J. & Perry, V. H. A
20 s intervals). After 15 min at 0C, the suspension was centrifuged at gene affecting Wallerian nerve degeneration maps distally on mouse
16,000 g for 1 min. The pH was adjusted to 6.0 using 0.8 M K2CO3. After chromosome 4. Proc. Natl. Acad. Sci. USA 90, 97179720 (1993).
18. Conforti, L. et al. A Ufd2/D4Cole1e chimeric protein and overexpression of
centrifugation, the extract was analyzed by HPLC48. Rbp7 in the slow Wallerian degeneration (WldS) mouse. Proc. Natl. Acad. Sci.
Animal experimentation was approved by Stadt Kln Veterinramt, USA 97, 1137711382 (2000).
licence K13,11//00. 19. Emanuelli, M. et al. Human NMN adenylyltransferase: molecular cloning,
chromosomal localization, tissue mRNA levels, bacterial expression, and
GenBank accession numbers. The GenBank accession number for enzymatic properties. J. Biol. Chem. 276, 406412 (2001).
20. Magni, G., Amici, A., Emanuelli, M., Raffaelli, N. & Ruggieri, S. Enzymology
NMNAT is AF312734; for Wld, AF260924. of NAD+ synthesis. Adv. Enzymol. Relat. Areas Mol. Biol. 73, 135182 (1999).
21. Conforti, L. et al. The major brain isoform of Kif1b lacks the putative
mitochondria-binding domain. Mamm. Genome 10, 617622 (1999).
Note: Supplementary figures are available on the Nature Neuroscience web site 22. Zhao, C. et al. Charcot-Marie-Tooth disease type 2a caused by mutation in a
(http://neuroscience.nature.com/web_specials). microtubule motor kif1bbeta. Cell 105, 587597 (2001).
23. Ribchester, R. R. et al. Persistence of neuromuscular junctions after axotomy
in mice with slow Wallerian degeneration (C57BL/WldS). Eur. J. Neurosci. 7,
16411650 (1995).
ACKNOWLEDGEMENTS 24. Gillingwater, T. H. & Ribchester, R. R. Age-dependent synapse withdrawal at
We thank E. Janssen, C. Hoffmann (Department of Anatomy I, University of axotomised neuromuscular junctions in WldS mutant mice. J. Physiol.
(Lond.) 523P, 53P (2000).
Cologne), F. Carnevali, F. Pierella (University of Ancona), S. Fearn and 25. Gillingwater, T. H. & Ribchester, R. R. Compartmental neurodegeneration
M. Botham (University of Southampton) for technical assistance, T. Vogt for and synaptic plasticity in the WldS mutant mouse. J. Physiol. (Lond.) 534,
supplying pHAPr-1 plasmid and R. Martini for critically reading the 627639 (2001).
26. Feng, G. et al. Imaging neuronal subsets in transgenic mice expressing
manuscript. This work was supported by the Federal Ministry of Education and multiple spectral variants of GFP. Neuron 28, 4151 (2000).
Research (FKZ: 01 KS 9502) and Center for Molecular Medicine, University of 27. Brown, M. C., Booth, C. M., Lunn, E. R. & Perry, V. H. Delayed response to
denervation in muscles of C57BL/Ola mice. Neuroscience 43, 279283
Cologne (ZMMK) (T.G.A.M., W.M., D.W., M.P.C.), a Wellcome Trust
(1991).
Biomedical Collaboration Grant (R.R.R., M.P.C.) and Consiglio Nazionale delle 28. Sagot, Y., Tan, S. A., Hammang, J. P., Aebischer, P. & Kato, A. C. GDNF slows
Ricerche Target Project Biotechnology (M.E., G.M.). loss of motoneurons but not axonal degeneration or premature death of
pmn/pmn mice. J. Neurosci. 16, 23352341 (1996).
29. Burne, J. F., Staple, J. K. & Raff, M. C. Glial cells are increased proportionally
RECEIVED 27 SEPTEMBER; ACCEPTED 31 OCTOBER 2001 in transgenic optic nerves with increased numbers of axons. J. Neurosci. 16,
20642073 (1996).
30. Koegl, M. et al. A novel ubiquitination factor, E4, is involved in
1. Waller, A. Experiments on the section of glossopharyngeal and hypoglossal multiubiquitin chain assembly. Cell 96, 635644 (1999).
nerves of the frog and observations of the alternatives produced thereby in the 31. Saigoh, K. et al. Intragenic deletion in the gene encoding ubiquitin carboxy-
structure of their primitive fibres. Phil. Trans. R. Soc. Lond. 140, 423429 (1850). terminal hydrolase in gad mice. Nat. Genet. 23, 4751 (1999).

articles
32. Hamilton, M. H., Tcherepanova, I., Huibregtse, J. M. & McDonnell, D. P. 41. Fujiki, M., Zhang, Z., Guth, L. & Steward, O. Genetic influences on cellular
Nuclear import/export of hRPF1/Nedd4 regulates the ubiquitin-dependent reactions to spinal cord injury: activation of macrophages/microglia and
degradation of its nuclear substrates. J. Biol. Chem. 276, 2632426331 (2001). astrocytes is delayed in mice carrying a mutation (WldS) that causes delayed
33. Smith, S. The world according to PARP. Trends Biochem. Sci. 26, 174179 Wallerian degeneration. J. Comp. Neurol. 371, 469484 (1996).
(2001). 42. Gunning, P., Leavitt, J., Muscat, G., Ng, S. Y. & Kedes, L. A human beta-actin
34. Ha, H.C. & Snyder, S.H. Poly(ADP-ribose) polymerase is a mediator of expression vector system directs high-level accumulation of antisense
necrotic death by ATP depletion. Proc. Natl. Acad. Sci. USA 96, 1397813982 transcripts. Proc. Natl. Acad. Sci. USA 84, 48314835 (1987).
(1999). 43. Karnovsky, M. J. A formaldehyde-glutaraldehyde fixative of high osmolarity
35. Nagayama, T. et al. Activation of poly(ADP-ribose) polymerase in the rat for use in electron microscopy. J. Cell Biol. 27, 137A (1965).
hippocampus may contribute to cellular recovery following sublethal 44. Reynolds, E. E. The use of lead citrate at high pH as an electron opaque stain
transient global ischemia. J. Neurochem. 74, 16361645 (2000). in electron microscopy. J. Cell Biol. 17, 208 (1963).
36. Smith, K. J., Kapoor, R., Hall, S. M. & Davies, M. Electrically active axons 45. Costanzo, E. M., Barry, J. A. & Ribchester, R. R. Co-regulation of synaptic
degenerate when exposed to nitric oxide. Ann. Neurol. 49, 470476 (2001). efficacy at stable polyneuronally innervated neuromuscular junctions in
37. Di Lisa, F. & Ziegler, M. Pathophysiological relevance of mitochondria in reinnervated rat muscle. J. Physiol. (Lond.) 521, 365374 (1999).
NAD(+) metabolism. FEBS Lett. 492, 48 (2001). 46. Costanzo, E. M., Barry, J. A. & Ribchester, R. R. Competition at silent
38. Frei, R. et al. Loss of distal axons and sensory Merkel cells and features synapses in reinnervated skeletal muscle. Nat. Neurosci. 3, 694700
indicative of muscle denervation in hindlimbs of P0-deficient mice. (2000).
J. Neurosci. 19, 60586067 (1999). 47. Ribchester, R. R., Mao, F. & Betz, W. J. Optical measurements of activity-
39. Mi, W., Conforti, L. & Coleman, M. P. Genotyping methods to detect a unique dependent membrane recycling in motor nerve terminals of mammalian
neuroprotective factor for axons (WldS). J. Neurosci. Methods (in press). skeletal muscle. Proc. R. Soc. Lond. B Biol. Sci. 255, 6166 (1994).
40. Valente, E.M. et al. Localization of a novel locus for autosomal recessive early- 48. Balducci, E. et al. Assay methods for nicotinamide mononucleotide
onset parkinsonism, PARK6, on human chromosome 1p35p36. Am. J. Hum. adenylyltransferase of wide applicability. Anal. Biochem. 228, 6468
Genet. 68, 895900 (2001). (1995).

articles
GABAB receptor activation enhances

mGluR-mediated responses at
cerebellar excitatory synapses
Moritoshi Hirono1, Tohru Yoshioka1,2 and Shiro Konishi3
1 Department of Molecular Neurobiology, Advanced Research Institute for Science and Engineering, Waseda University, Tokyo 169-8555, Japan
2 Department of Molecular Neurobiology, School of Human Sciences, Waseda University, Tokorozawa 359-1192, Japan
3 Laboratory of Molecular Neurobiology, Mitsubishi Kagaku Institute of Life Sciences and CREST (JST), 11-Minamiooya, Machida-shi, Tokyo 194-8511, Japan
Correspondence should be addressed to S.K. (skonishi@libra.ls.m-kagaku.co.jp)
Metabotropic -aminobutyric acid type B (GABAB) and glutamate receptors (mGluRs) are postsynap-
tically co-expressed at cerebellar parallel fiber (PF)Purkinje cell (PC) excitatory synapses, but their
functional interactions are unclear. We found that mGluR1 agonist-induced currents and [Ca2+]i
increases in PCs were enhanced following co-activation of GABAB receptors. A GABAB antagonist
and a G-protein uncoupler suppressed these effects. Low-concentration baclofen, a GABAB agonist,
augmented mGluR1-mediated excitatory synaptic current produced by stimulating PFs. These results
indicate that postsynaptic GABAB receptors functionally interact with mGluR1 and enhance mGluR1-
mediated excitatory transmission at PFPC synapses. The interaction between the two types of
metabotropic receptors provides a likely mechanism for regulating cerebellar synaptic plasticity.
GABA is the main inhibitory neurotransmitter in the central ner- includes Gq protein-mediated activation of phospholipase C
vous system, and inhibitory GABAergic synapses are endowed (PLC), hydrolysis of phosphatidylinositol-4,5-bisphosphate (PIP2)
with transmitter receptors including ionotropic GABAA and to inositol 1,4,5-triphosphate (IP3) and diacylglycerol, leading to
GABA C receptors and metabotropic GABA B receptors intracellular Ca2+ concentration ([Ca2+]i) increase and protein
(GABABRs)1. Ionotropic GABA receptors exhibit considerable kinase C (PKC) stimulation1820. Interestingly, cellular localiza-
molecular diversity2, whereas recent cloning has revealed that the tion of mGluR1 is similar to that of GABABRs: morphological
GABA BR gene encodes R1a and R2 subunits that form het- studies have shown that mGluR1 is also expressed abundantly at
erodimers to function36. GABABRs are coupled with Gi/o pro- the extra-postsynaptic sites of PFPC synapses2123. It is, there-
teins either to inhibit neurotransmission presynaptically7,8 or to fore, reasonable to expect that GABABRs would interact with
decrease the excitability postsynaptically by opening G-protein- mGluR1 and modulate mGluR1-mediated physiological respons-
coupled inwardly rectifying K + (GIRK) channels 9,10 . The es at these synaptic sites.
GABABR-mediated Gi/o protein activation also depresses adeny- Therefore, the aim of the present study was to explore the
lyl cyclase and reduces Ca2+ channel currents11,12. However, cor- cross-talk between mGluR1 and GABABRs expressed by PCs in
relations between these GABABR-mediated pharmacological acute slices from the mouse cerebellum, using whole-cell record-
actions and synaptic events are not completely understood. ings combined with Ca2+-signal imaging. We found that the acti-
In the cerebellar cortex, radioautographic and immunocyto- vation of GABABRs by the exogenous agonist baclofen enhanced
chemical studies have shown that GABABR binding sites dense- both mGluR1-mediated inward currents and Ca2+ signals in PCs.
ly occur in the molecular layer, where the PCs extend their More importantly, we showed that endogenous GABA released
dendritic branches 13,14 . The postsynaptic localization of by electrical stimulation in the cerebellar cortex mimicked the
GABABRs on PCs was reported by in situ hybridization36. An effect of the GABABR agonist, which augmented the mGluR1-
electron microscopy study showed that GABABRs are present at mediated slow excitatory synaptic current elicited by PF stimu-
the extra-postsynaptic sites of excitatory connections between lation. Therefore, the cross-talk between mGluR1 and GABABR
parallel fibers (PFs) and PCs5,15. GABABRs also suppress a synap- revealed in this study seems to call for a revision of our view that
tic process called rebound potentiation of inhibitory transmis- GABAB receptors serve an exclusively inhibitory role in chemi-
sion following PC depolarization 16 . However, it remains cal signaling at central synapses.
uncertain what physiological role the GABA BRs have at the
PFPC excitatory postsynaptic sites. Long-term depression (LTD) RESULTS
at PFPC synapses has been proposed as a cellular mechanism GABAB activation enhanced mGluR current
of synaptic plasticity closely associated with motor learning17,18. Iontophoretic application of the nonselective mGluR agonist
Induction of LTD requires activation of type 1 metabotropic glu- 1S,3R-ACPD (1S,3R-1-aminocyclopentane-1,3-dicarboxylic acid)
tamate receptors (mGluR1), triggering a signaling cascade that produced an inward current in cerebellar PCs voltage-clamped

articles
a antagonist bicuculline (30 M). Therefore, it is likely that the

enhancement of the 1S,3R-ACPD-induced current in the pres-
ence of GABA is mediated by GABABRs but not by GABAA recep-
tors in PCs, which is consistent with the slow outward current
observed during GABA application.
b We then examined whether the GABABR agonist baclofen repro-
duces the augmentation of the 1S,3R-ACPD response. When applied
by perfusion, baclofen (3 M) increased the amplitude of the 1S,3R-
ACPD-induced current up to 226 11% of the control response

(n = 15; Fig. 2). To test whether the GABABR-induced enhancement
is specific to the mGluR1-mediated response, we compared the
effects of baclofen on inward current responses produced by 1S,3R-
ACPD and AMPA (-amino-3-hydroxy-5-methyl-4-isoxasole pro-
pionic acid), an ionotropic GluR agonist. Baclofen selectively
enhanced the 1S,3R-ACPD response without appreciably affecting
Fig. 1. Enhancement of 1S,3R-ACPD-induced inward current by GABA.
the AMPA-induced current (Fig. 2ac). Baclofen also produced an
(a) Inward current responses were induced by iontophoretic applica-
tion of 1S,3R-ACPD to a Purkinje cell at a constant interval of 40 s. outward current (36 4 pA, n = 15), as observed with GABA appli-
GABA (100 M) was applied by perfusion for 3 min as indicated by hor- cation. The baclofen-induced enhancement of the 1S,3R-ACPD
izontal bar. The record was obtained from a Purkinje cell held at 60 mV response was reversible, subsiding over a period of approximately
in the presence of bicuculline (30 M) and TTX (0.5 M). (b) 1S,3R- 20 minutes after the agonist was washed out. In a current-clamp
ACPD-induced responses indicated by 1, 2 and 3 in (a) are displayed on mode, baclofen hyperpolarized PCs by 2.1 2.2 mV (n = 4,
a fast time base. Fig. 2d) and increased the amplitude of 1S,3R-ACPD-induced depo-
larization from 10.6 0.7 mV to 14.8 1.3 mV (Fig. 2e; n = 4,
p < 0.05). The baclofen-induced enhancement of the mGluR1-medi-
ated response was completely abolished by the selective GABABR
at 60 mV, which is close to the resting potential2426. The 1S,3R- antagonist CGP62349 (Fig. 3a) with a half-maximal inhibitory con-
ACPD-induced current was suppressed by (S)-4-car- centration IC50 of about 15 nM. The 1S,3R-ACPD-induced cur-
boxyphenylglycine (4CPG), a selective antagonist for group I rent remained almost unchanged in the presence of 100 nM
mGluRs including mGluR1 and mGluR5 (ref. 26), whereas CGP62349 (97 2% of the control current, n = 4), which indicates
1S,3R-ACPD did not produce any detectable current in PCs of that tonic activation of GABABRs by endogenous GABA released
mGluR1-deficient mice19, which indicates that the inward cur- spontaneously is insufficient for a detectable enhancement of the
rent resulted from activation of mGluR1. This mGluR1-mediat- mGluR1-mediated response.
ed current increased markedly when GABA, an inhibitory We then investigated whether activation of G-protein-cou-
transmitter substance, was co-applied by superfusion (Fig. 1). pled receptors other than GABABRs also enhances the mGluR1-
The increase of the 1S,3R-ACPD-current amplitude following mediated inward current in PCs. The 1S,3R-ACPD-induced
100 M GABA co-application was 161 8% of the control
response (n = 4) in an artificial cerebrospinal fluid (ACSF) that
contained tetrodotoxin (TTX, 0.5 M) and the GABAA receptor Fig. 2. Selective enhancement of mGluR1-mediated current follow-
ing GABABR activation. (a) 1S,3R-ACPD- and AMPA-current
responses were alternately induced by iontophoretic applications
from separate microelectrodes to a
a d single Purkinje cell. The GABABR
agonist baclofen (3 M) was applied
by perfusion during the period indi-
cated by the horizontal bar.
(b) Sequential 1S,3R-ACPD- and
AMPA-induced currents indicated as
b 13 in (a) are displayed on a fast
time base. (c) Time course of the
e effect of GABABR activation on the
1S,3R-ACPD- and AMPA-induced
current responses. The amplitude of
both responses is given as a percent-
age of the control immediately
before baclofen application. (d) A
current-clamp recording of 1S,3R-
ACPD-induced depolarization in a
c Purkinje cell and the effects of
baclofen (3 M) on the 1S,3R-
ACPD-response and membrane
potential. (e) Expanded and super-
imposed records of 1S,3R-ACPD-
induced depolarizations before and
during baclofen application. Each
response was recorded as indicated
by 1 and 2 in (d).

articles
Fig. 3. Blockade of baclofen-induced enhancement of the mGluR1- a

mediated current response by the GABABR antagonist CGP62349.
(a) 1S,3R-ACPD was iontophoretically applied at a time point indicated
by arrows at a constant interval in the presence of CGP62349 (30 nM).
Responses were recorded from a single PC before (left) and during per-
fusion of baclofen (3 M, middle) and after washing out the GABABR
agonist (right). (b) Comparison of the effects of Gi/o-coupled receptor
agonists on the 1S,3R-ACPD-induced current response in PCs. The
number in parentheses represents the number of PCs in which the
effects of individual agonists were determined: baclofen (3 M), sero-

tonin (5-HT, 30 M), carbachol (CCh, 100 M) and adenosine (10 M). b
***P < 0.001, one-way ANOVA with Tukeys post-test.
current was not significantly altered by application of carbachol

(CCh), serotonin (5-HT) or adenosine (Fig. 3b). However, mus-
carinic, 5-HT and adenosine receptors occur in the cerebellar
cortex2729, and 5-HT and adenosine elicit facilitation and inhi-
bition of cerebellar GABAergic transmission, respectively, when
applied by superfusion, as in this study30,31. It seems, therefore,
that the enhancement of the mGluR1-mediated response requires
selective activation of GABABRs in PCs.
Characterization of GABABR-mediated enhancement the IC50 of PF-EPSC inhibition was approximately 0.77 M,
First, the currentvoltage relationship of the 1S,3R-ACPD- which is consistent with the value reported previously33,34. The
induced response was compared before and during GABABR acti- effective concentration ranges of baclofen for inhibiting PF-EPSCs
vation. The 1S,3R-ACPD-induced current was obtained by as well as causing the mGluR1-current enhancement and out-
subtracting current responses produced by a constant voltage ward currents were very similar (Fig. 4e).
ramp in the absence and presence of 1S,3R-ACPD (Fig. 4a
and b). Baclofen increased the 1S,3R-ACPD-induced current, Effect of baclofen on mGluR1-mediated [Ca2+]i transients
whereas its reversal potential was almost identical before and after Application of the mGluR agonist 1S,3R-ACPD increased [Ca2+]i
baclofen application (5.1 4.2 mV and 2.8 3.6 mV, respec- in PCs as reported previously2426. To determine the effects of
tively, n = 7, p = 0.68). The degree of the GABABR-mediated baclofen on the mGluR1-induced [Ca2+]i increase, we performed
enhancement of the 1S,3R-ACPD current did not change in the simultaneous whole-cell recordings and intracellular Ca2+ imag-
membrane potential range examined (Fig. 4c), indicating that ing in PCs using a Ca 2+ indicator, fura-2. Baclofen (3 M)
the GABABR-mediated enhancement of mGluR1-activated cur- enhanced not only the inward current but also the [Ca2+]i ele-
rents is independent of the membrane potential. Furthermore, vation produced in response to iontophoretic application of
the effect of GABABR activation did not depend on the ampli- 1S,3R-ACPD (Fig. 5): the 1S,3R-ACPD-current and [Ca2+]i tran-
tude of the 1S,3R-ACPD responses, as there was no correlation sients measured in the distal dendrites of PCs were increased to
between the initial amplitude of 1S,3R-ACPD-induced currents 180 22% (n = 4, p < 0.01) and 321 71% (n = 4, p < 0.05) of
(140 to 280 pA) and the extent of the baclofen-induced enhance- the control responses, respectively (Fig. 5c). 1S,3R-ACPD caused
ment of 1S,3R-ACPD currents (Fig. 4d). a larger increase in [Ca2+]i at distal dendrites than at proximal
Extrapolation of the baclofen-induced current response dendrites of the PC.
showed its reversal potential of 93.6 5.0 mV (n = 5), which Furthermore, baclofen increased the basal level of [Ca2+]i in
was close to the K+ equilibrium potential (96.6 mV) predicted three of four PCs tested. The averaged F340/F380 ratio reflect-
from the Nernst equation. In the presence of a GIRK channel ing the basal [Ca2+]i increased during the GABABR agonist appli-
inhibitor, Ba2+ (1 mM), baclofen still induced an outward cur- cation and recovered to the control level after the agonist was
rent, the extent of which was almost comparable to that of the washed out (Fig. 5d). This effect was particularly significant in
control response (43.3 5.8 pA, n = 4, p = 0.46). Furthermore, the recording site at proximal dendrites. The effect of baclofen
Ba2+ had little effect on the GABABR-mediated enhancement of on [Ca2+]i is not compatible with the previous observation that
1S,3R-ACPD-induced current (197 22% of the control GABABR activation inhibits Ca2+ influx through P-type Ca2+
response, n = 4, p = 0.32; example, Fig. 4g). The GIRK current channels in PCs11. A possible involvement of P/Q-type Ca2+
in principal neurons of the amygdala was, in fact, almost com- channel inhibition in the modulation of mGluR1 response was
pletely blocked by 1 mM Ba2+ when applied by superfusion to excluded as the P/Q-type Ca2+ channel blocker -agatoxin IVA
slice preparations (ref. 32 and unpublished observations). There- (100 nM) did not enhance the 1S,3R-ACPD-induced inward cur-
fore, it seems that the GABABR-mediated enhancement and out- rent but, rather, slightly reduced its amplitude to 82.8 6.2% of
ward currents are not due to the activation of GIRK channels. the control response (n = 5, p < 0.05). Thus, it seems that
The baclofen-induced outward current and enhancement of GABABR activation increased the basal [Ca2+]i level by influ-
mGluR1-mediated response were dose-dependent with half max- encing internal Ca 2+ store through mGluR1-mediated and
imal effective concentrations (EC50) of approximately 1.01 and G-protein-coupled signaling pathways. An analogous mechanism
0.94 M, respectively (Fig. 4e), and pooled data showed that there has been proposed for the modulation of the basal [Ca2+]i level
was a modest correlation between both responses (Fig. 4f). In elicited by the activation of other metabotropic receptors, such
addition to the two effects, baclofen markedly inhibited PF-stim- as opioid receptors, and adenosine A1 and NPYY1 receptors35.
ulation-evoked excitatory postsynaptic currents (EPSCs), and Another possibility would be that baclofen enhances the [Ca2+]i

articles
a b c
d e f
Fig. 4. Voltage-independent facilitation of GABABR-mediated enhancement of the 1S,3R-ACPD-induced currents, and comparisons of presynaptic
and postsynaptic actions induced by baclofen. (a) 1S,3R-ACPD was iontophoretically applied to a Purkinje cell (PC) at 40-s intervals, and the hold-
ing potential of the PC (60 mV) was shifted by constant voltage ramps from 130 to 0 mV for 1560 ms at the time points as indicated by (1, 1) and
(2, 2) during and after the 1S,3R-ACPD-induced response, respectively. This sequence of ramp commands was repeated before (1, 2) and during
baclofen (3 M) application (1, 2). (b) 1S,3R-ACPD-induced currents determined by subtraction of currentvoltage relationships produced by
ramp commands as indicated in (a). (c) The amplitude ratios of 1S,3R-ACPD-induced currents before and during baclofen application calculated as
(1 2)/(1 2) 100 (%) are plotted against the membrane potential range determined in PCs (n = 5). (d) Relationship between the extent of
baclofen-induced enhancement of mGluR1-mediated currents and the amplitude of 1S,3R-ACPD-currents before baclofen (3 M) application in indi-
vidual PCs. The straight line indicates a regression line with a correlation coefficient, r = 0.029, indicating no correlation between the two
responses. (e) Comparison of doseresponse relationships for the baclofen-induced enhancement of the mGluR1-mediated current, outward cur-
rent and inhibition of PF-mediated EPSCs. The 50% effective doses(EC50) of baclofen were determined for the 3 responses (n = 315): the increase
of the mGluR current expressed as a percentage of that induced by 100 M baclofen (white circles and solid line, EC50 0.94 M), the amplitude of
baclofen-induced outward current response expressed as a percentage of that induced by 100 M baclofen (black circles and dashed line,
EC50 1.01 M) and percentage inhibition of PF-EPSC by baclofen (white squares and dotted line, IC50 0.77 M). The Hill coefficient (n) deter-
mined from each doseresponse curve was 0.84, 0.92 and 0.94, respectively. (f) Relationship between baclofen-induced enhancement of mGluR1-
mediated current and outward current response. The straight line indicates a regression line with a correlation coefficient was r = 0.558. (g) Effects
of Ba2+ on the baclofen-induced outward current response and enhancement of mGluR1-mediated current. Ba2+ (1 mM) and baclofen (3 M) were
applied by perfusion during the period indicated by horizontal bars. Downward deflections represent the inward currents induced by iontophoretic
application of 1S,3R-ACPD at a constant interval, as in (a).
rise via tonic activation of mGluR1 by glutamate released spon- (89 10% of the control response, n = 4, p = 0.39) and the
taneously from excitatory nerve terminals. AMPA-current (98 6% of the control response, n = 3, p = 0.92).
We next examined the mechanisms underlying the baclofen- This finding suggests that G-proteins, presumably G i/o, are
induced enhancement of the mGluR1 response. First, we inves- responsible for the GABABR-mediated enhancement.
tigated whether G i/o activation is required for the One target of Gi/o proteins linked with GABABRs might be
GABABR-mediated enhancement. Treatment of cerebellar slices adenylyl cyclase, as GABABR agonists reduce the level of intra-
with N-ethylmaleimide (NEM), a Gi/o inhibitor9, at a concen- cellular cyclic AMP by inhibiting adenylyl cyclase12. However,
tration of 50 M for 15 minutes significantly reduced the extent the baclofen-induced increase in the 1S,3R-ACPD current was
of the baclofen-induced enhancement of the 1S,3R-ACPD- not significantly affected by treatment with either forskolin
induced current (127 17% of the baseline after NEM treatment, (30 M) for 20 minutes or 8Br-cAMP (500 M) for 35 minutes
versus 226 11% of the baseline in the control ACSF, n = 4, (data not shown). Treatment with protein kinase inhibitors H-7
p < 0.001; Fig. 6a and b). However, the NEM treatment did not (20 M) or H-8 (20 M) for 20 minutes also had no effect (data
cause any significant effects on the 1S,3R-ACPD-current not shown). Another possible target of Gi/o proteins is PLC. Gi/o

articles
a
b c
Fig. 5. Simultaneous recordings of GABABR activation-mediated enhancement of 1S,3R-ACPD-induced current and [Ca2+]i transients in a PC.
(a) [Ca2+]i transients were fluorometrically measured at proximal (P, white) and distal dendrites (D, red) of the PC held at 60 mV starting 25 min
after loading the Ca2+ indicator fura-2 (1 mM) via the recording electrode. The fura-2-loaded PC was viewed with a fluorescence image produced by
380 nm excitation wavelength (1). Pseudocolor ratio images were recorded before 1S,3R-ACPD application (2), during 1S,3R-ACPD application in the
absence (3) and presence (4) of 3 M baclofen. (b) Effects of baclofen on 1S,3R-ACPD-induced Ca2+ signals (top) and current responses (bottom).
1S,3R-ACPD was iontophoretically applied to a single PC at a constant interval as indicated by dots, and baclofen (3 M) was applied by perfusion dur-
ing the period indicated by a horizontal bar. The numbers 24 indicate the time points where the images in (a) were obtained, and the fluorescence
ratio curves indicated by red and black were measures at distal and proximal dendritic sites, respectively. (c) Time courses of baclofen-induced
enhancement of the mGluR1-mediated current (white circles) and [Ca2+]i transient, F340/F380 ratio (black circles) in PCs. Each response is
expressed as a mean percentage of the control response determined immediately before baclofen (3 M) application (n = 4). (d) Changes in basal
[Ca2+]i level in PCs induced by baclofen. Each plot represents the mean s.e.m. of the F340/F380 ratios determined before (7090 s) and during
baclofen application (150170 s) and after washing out of the drug (310330 s), respectively, in each of four different PCs. *p < 0.05, ***p < 0.001,
one-way ANOVA tested for the values before and during baclofen application.
proteins are linked to [Ca2+]i increases via activation of IP3 recep- from intracellular stores is responsible for the enhancement of
tors35,36, and stimulation of GABABRs in the cerebellar cortex the mGluR1 response. The baclofen-induced outward current
causes activation of PLC via G-proteins, thereby resulting in mod- might be attributable to a Ca2+-activated K+ current.
ulation of GABAA receptor fuctions37. We therefore tested the
possible involvement of PLC and IP3 receptors in the GABABR- Dual effects of baclofen on mGluR1-mediated EPSC
mediated response. A selective PLC inhibitor, U73122 (10 M), In the presence of ionotropic glutamate and GABA receptor
infused into PCs via a patch electrode, significantly suppressed antagonists, repetitive stimulation of PFs produces in PCs a slow
the baclofen-induced enhancement of the 1S,3R-ACPD current excitatory synaptic response that is mediated by group I
when the effect was determined at least 30 minutes after intra- mGluRs26,3840. A crucial test is to determine whether GABABR
cellular application of the PLC inhibitor (n = 5, p < 0.05; activation enhances synaptically evoked mGluR-mediated
Fig. 6d). Application of an IP3 receptor modulator, heparin responses (Fig. 7). Activation of PFs with 10 stimuli at 100 Hz
(300 units/ml), also suppressed the GABABR-mediated enhance- produced fast EPSCs with a gradual increase in amplitudes that
ment (n = 4, p < 0.01). Furthermore, intracellular infusion of the were followed by a slow inward EPSC. Fast and slow components
Ca2+ chelator BAPTA (35 mM) significantly reduced the extent of of the synaptic responses were AMPA-receptor- and mGluR1-
the 1S,3R-ACPD current enhancement (n = 5, p < 0.05). The mediated EPSCs, respectively, because the former was almost
baclofen-induced outward current was also significantly reduced completely blocked by CNQX (30 M) and the latter was abol-
by all the three treatments (data not shown). The pharmacolog- ished by the group I mGluR antagonist 4CPG (500 M)26. Appli-
ical manipulations used here caused only partial suppression of cation of baclofen at a low concentration of 0.3 M increased the
the mGluR1-mediated current response per se as previously amplitude of slow EPSCs to 118 6% of the control (n = 9,
reported 26,38 (Fig. 6c). Taken together, it is likely that the p < 0.01; Fig. 7ac). In contrast, baclofen at the same concentra-
baclofen-induced basal [Ca2+]i rise resulting from Ca2+ release tion decreased the amplitude of fast EPSCs to 89.2 5.7% of the

articles
Fig. 6. Effects of signal transduction

modulators on the GABABR-induced a b
enhancement of the mGluR1-mediated
current. (a) Effects of NEM treatment on
1S,3R-ACPD- and AMPA-current
responses and baclofen-induced
enhancement of 1S,3R-ACPD-response.
NEM (50 M) was applied by perfusion.
The responses were recorded at the
time points indicated by 1 to 3 in (b).

(b) Time courses of the NEM effects on
1S,3R-ACPD-current and GABABR-
induced enhancement of 1S,3R-
ACPD-response. The amplitude of c
1S,3R-ACPD-currents is expressed as a d
percentage of the controls immediately
before NEM application (black squares)
and baclofen application without NEM
treatment (white circles), respectively.
(c) Effects of the PLC inhibitor U73122
(10 M) and the IP3 receptor inhibitor
heparin (300 units/ml) infusions on the
time course change of the 1S,3R-ACPD-
induced current amplitudes. Each data
point represents the mean s.e.m. of a
percentage to the amplitude of initial
1S,3R-ACPD-current determined 2 min
after obtaining whole-cell access in 5 to
12 PCs. (d) Effects of NEM (50 M),
U73122 (10 M), heparin (300 units/ml) and BAPTA (35 mM) infusions on presynaptic GABABRs, respectively, depending on the concen-
the baclofen-induced enhancement of 1S,3R-ACPD-induced current. Each tration of GABA released into the synaptic cleft.
compound except NEM was applied for at least 30 min by infusing into PCs Second, we examined the effect of the GABABR antagonist
via the recording electrode, and then the application of 3 M baclofen was CGP62349 on the mGluR1-mediated EPSC. The slow EPSCs
initiated. NEM was applied by superfusion for 15 min, and the effect of evoked by repetitive stimulation of PFs gradually decreased to
baclofen on 1S,3R-ACPD-induced current response was determined in the
74.5 5.5% of the control after application of CGP62349 at a
presence of NEM and expressed as a percentage of the control 1S,3R-
ACPD-response before the GABABR agonist application. The number in concentration of 300 nM (n = 6, p < 0.01) and the inhibitory
parentheses represents the number of independent experiments. effect persisted over 10 minutes after the drug was washed out
*p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA with Tukeys post-test. (Fig. 8b and c). In contrast, the GABABR antagonist caused a
slight increase in amplitude of the fast EPSC (109 7.9% of the
control, n = 6). These observations are consistent with the pos-
sibility that the mGluR1-mediated slow EPSC could be enhanced
control (n = 9). When the concentration was increased to 130 by the activation of postsynaptic GABABRs due to endogenous
M, baclofen inhibited both fast and slow EPSCs to a similar GABA released during repetitive stimulation of PFs, as PFs form
degree (42.1 8.1% and 43.9 6.5% inhibition at 10 M, respec- excitatory connections with GABA-containing interneurons to
tively, n = 9; Fig. 7d and e). The doseresponse relationships for elicit GABA release. The modulation of fast EPSCs by endoge-
the baclofen-induced enhancement and inhibition of the slow nous GABA seems to be dominated by the inhibitory action of
mGluR1-mediated EPSC are shown in Fig. 8a. The enhancement presynaptic GABABRs.
might be explained by postsynaptic interaction between
GABABRs and mGluR1 activated by the excitatory transmitter DISCUSSION
released from PF terminals. The inhibitory action of baclofen G-protein-coupled GABABRs and mGluR1 display analogous
might be due to a presynaptic mechanism of reducing the distribution profiles at cerebellar PFPC excitatory synapses with
amount of transmitter released39. It is likely, therefore, that post- both receptors being localized mainly at extrasynaptic sites along
synaptic GABABRs are more susceptible to baclofen than are the PC dendrites5,15,2123. The physiological functions of presy-
presynaptic GABABRs. naptic GABABRs in PF terminals have been well studied33,34, but
In the following two series of experiments, we further deter- relatively little is known about the functions of postsynaptic
mined whether endogenous GABA released from cerebellar GABABRs on PCs. Therefore, we investigated functional inter-
interneurons can enhance the mGluR1-mediated slow EPSCs. actions between postsynaptic GABABRs and mGluR1 in PCs. We
First, we examined the effect of a GABA uptake inhibitor, found that activation of GABABRs by the exogenous agonist
SKF89976A, on the slow EPSCs. Application of SKF89976A at a baclofen elicited a profound enhancement of the mGluR1 ago-
concentration of 10 M increased the amplitude of slow EPSCs in nist 1S,3R-ACPD-induced current response and [Ca2+]i rise in
all PCs examined (112 5% of the control, n = 4, p < 0.05), PCs (Figs. 2 and 5) and that endogenous GABA synaptically
whereas the GABA uptake inhibitor at a higher concentration of released from cerebellar interneurons following PF stimulation
100 M decreased the slow EPSC amplitude by 44.5 6.8% could enhance the mGluR1-mediated synaptic response evoked
(n = 4, data not shown). The observations suggest that endoge- by the PF excitatory transmitter (Fig. 8). Our data suggest that
nous GABA causes enhancement and inhibition of the mGluR1- the GABA B R-mediated enhancement of mGluR1 response
mediated transmission via activation of postsynaptic and requires Ca2+ release from internal stores through a signaling

articles
a b d
c e
Fig. 7. Enhancement by baclofen of mGluR1-mediated slow EPSCs produced in response to PF stimulation. (a) Fast and slow EPSCs were evoked by
repetitive stimulation (100 Hz for 100 ms) of the PF at a constant interval of 60 s in the presence of CNQX (10 M), AP5 (30 M) and bicuculline
(50 M) and recorded from a PC. The synaptic responses were recorded before (top) and during baclofen (0.3 M) application (middle), and after the
drug was washed out (bottom). (b) Superimposed PF-mediated slow EPSCs in the control and 0.3 M baclofen-containing ACSF are displayed on a fast
time base. (c) Time course of the effects of baclofen on fast and slow EPSCs produced in PCs by PF stimulation. The amplitude of both EPSCs is
expressed as a percentage of the control amplitude determined immediately before baclofen application (n = 9). (d) Inhibitory effects of a higher con-
centration of baclofen (10 M) on PF-mediated slow EPSCs. (e) Time course of baclofen-induced inhibition of fast and slow EPSCs recorded from PCs.
The amplitude of both responses was expressed as a percentage of the control determined immediately before 10 M baclofen application (n = 4).
mechanism that involves G i/o-coupled PLC activation. This [Ca2+]i40. Studies have suggested that the regulation of Ca2+
GABABR-mediated modulation of synaptic processes seemed to release from internal stores in presynaptic and postsynaptic neu-
be specific to the mGluR1 responses, because the ionotropic rons profoundly influences short- and long-term plasticity at
AMPA-type GluR-mediated current response was not affected cerebellar synapses4244.
by the GABA BR agonist baclofen (Fig. 2). Furthermore, the Another possibility is that the GABABR-mediated enhance-
enhancement of mGluR1-mediated responses is a unique prop- ment of the mGluR1-activated current and Ca 2+ signals are
erty of metabotropic GABABRs, as other G-protein-linked recep- attributable to two independent mechanisms, although there is no
tors including serotonin, adenosine and muscarinic receptors direct evidence supporting this. GABABR activation produced a
were devoid of this capability in PCs. substantial outward current that was resistant to treatment with
Two possible mechanisms may explain the cross-talk between Ba2+, an inhibitor of GABABR-linked GIRK channels. A previ-
GABABR and mGluR1 revealed in this study. First, because not ous study using a genetic knockout of GIRK channels clearly
only the mGluR1-activated current but also the mGluR1-induced demonstrated that presynaptic GABABR-mediated inhibition of
[Ca2+]i increase in PCs were enhanced following GABABR acti- neurotransmission is independent of GABABR-mediated GIRK
vation, Ca2+ mobilization from internal stores through Gi/o- channel activation10. In our study, GABABRs associated with the
linked PLC activation and IP3 formation might be critical in the enhancement of synaptically evoked mGluR1 response exhibited
GABABR-mGluR1 interaction. This notion was supported by the the sensitivity higher than those associated with presynaptic
following observations: treatment with the Gi/o inhibitor NEM inhibitory actions (Fig. 8), although the exogenous mGluR ago-
markedly attenuated the GABABR-mediated enhancement of the nist 1S,3R-ACPD-induced response and PFPC transmission
current response; infusion of the PLC inhibitor U73122 and the were affected by baclofen in a similar concentration range. There-
IP3 antagonist heparin into PCs suppressed these GABABR- fore, it remains to be determined whether separate receptor sub-
mGluR1 interactions; infusion of the Ca2+ chelator BAPTA also types with distinct signaling pathways are involved in the
inhibited the effects of GABABR activation on mGluR1 respons- conventional GABABR-mediated presynaptic inhibition and post-
es; and application of the GABABR agonist baclofen increased synaptic GIRK channel activation, and in the mechanism of
basal [Ca2+]i (Fig. 5). Thus, it seems that GABABR activation may GABABR-mGluR1 cross-talk found in this study.
be linked to intracellular Ca2+ stores to modulate [Ca2+]i and Two possible physiological consequences might be associat-
enhance the mGluR1 functions including the current response ed with the GABABRmGluR1 interaction at PFPC synapses.
and Ca2+ elevation, presumably through a cooperative upregu- First, the interaction may increase the excitability of PCs, as it
lation of G-protein-coupled receptor signaling by G subunits is assumed that the GABABR activation enhances the slow depo-
as reported in -adrenergic receptor-GABABR synergistic inter- larizing synaptic potential mediated through mGluR1 activa-
actions41. This is compatible with the finding that mGluR1-medi- tion by the excitatory transmitter glutamate released from PF
ated excitation at PFPC synapses is markedly increased by nerve terminals. If this is the case, GABA may serve dual func-
activation of the climbing fiber input via a transient increase in tions at PFPC synapses, either as an excitatory modulator caus-

articles
a b
Fig. 8. Dose dependency of baclofen-induced enhancement and inhibition of mGluR1-mediated slow EPSCs, and the effects of synaptic GABABR acti-
vation on fast and slow EPSCs tested by using the GABABR antagonist CGP62349. (a) Doseresponse relationships between baclofen-induced facili-
tatory and inhibitory actions on mGluR1-mediated slow EPSCs following repetitive PF stimulation were determined in PCs. The number in
parentheses represents the number of PCs in which effects of baclofen were tested. (b) CGP62349-induced increase in PF-mediated fast EPSC and
decrease in slow EPSC amplitude. The synaptic responses following repetitive PF stimulation were recorded from a PC in the control (left) and in
300 nM CGP62349-containing ACSF (right). (c) Time courses of CGP62349-induced effects on fast and slow EPSCs. The amplitudes of both EPSCs
are expressed as a percentage of the control amplitude determined immediately before application of the GABABR antagonist (n = 6).
ing the postsynaptic transient facilitation of PF excitatory inputs Interaction between different transmitter receptor systems is
to PCs, or as a classical transmitter eliciting presynaptic inhibi- one emerging feature of neurotransmission at central synapses.
tion of the ionotropic GluR-mediated fast excitatory transmis- For instance, dopamine D and somatostatin SST5 receptors form
sion. Second, the GABABR-mGluR1 cross-talk may be critical heterodimers to create a novel receptor with augmented func-
in synaptic plasticity associated with the cerebellar function, as tional activity48. Another example has been demonstrated for
mGluR1 has been implicated as an important molecule in the dopamine D5 and GABAA 2 receptors49: The GABAA-ligand-
induction of LTD at these synapses19,20, which is proposed as a gated channels complex with D5 receptors via direct binding,
key mechanism underlying motor coordination within in the thereby enabling mutually inhibitory functional interactions
cerebellar system17,18. Furthermore, studies have suggested that between the two receptor systems. Cross-talk between neuro-
mGluR1-mediated Ca2+ release from internal stores in PCs acts transmitter-gated cation channels has been reported for heterol-
as a coincidence detection mechanism for PF and CF activations, ogous expression systems and cultured neurons, where
thereby leading to induction of LTD at PFPC synapses44. There- structurally distinct nicotinic receptor and purinergic P2X2 recep-
fore, simultaneous activation of mGluR1 and GABABRs would tor channels influence each other with regard to cross-inhibition
enhance this mechanism. The physiological significance of between the two channels50. Conformational spread from one
GABABRs in synaptic plasticity has also been shown in other receptor to its neighbors is proposed as a possible mechanism
brain regions. Presynaptic GABABRs seem to be involved in long- for the cross-talk. As exemplified by this cross-talk, together with
term potentiation (LTP) in the hippocampus45,46. In addition, the GABABR-mGluR1 interaction identified in this study, the
postsynaptic GABABRs contribute to long-term regulation of interplay between distinct receptor systems can provide a pow-
synaptic strength at GABAergic inhibitory synapses in the visu- erful mechanism to influence chemical signaling at central ner-
al cortex47. In this case, GABABRs and adrenoceptors seem to vous system synapses.
act in concert to further enhance heterosynaptic monoaminer-
gic LTP, possibly through GABA BR-mediated facilitation of METHODS
monoamine-induced IP 3 formation. Negative regulation of Electrophysiology. BDF1 mice (45 weeks old) were anesthetised with
synaptic plasticity, which involves postsynaptic GABABRs, has pentobarbital, and sagittal slices (180200 m thick) of the cerebellar
been reported at inhibitory synapses in the cerebellar cortex: the vermis were prepared using a vibrating microtome (Microslicer DTK-
activation of GABABRs in PCs downregulates the long-lasting 1000, Dosaka, Kyoto, Japan). Whole-cell recordings were obtained from
PCs visually identified under Nomarski optics using a water immersion
increase, or rebound potentiation, of GABAA receptor sensi-
objective (40, NA 0.75, Zeiss, Germany). Slices were superfused with
tivity following depolarization of PCs 16 . The postsynaptic ACSF containing 138.6 mM NaCl, 3.35 mM KCl, 21 mM NaHCO3,
GABABRmGluR1 interaction identified in this study provides 0.6 mM NaH2PO4, 9.9 mM glucose, 2.5 mM CaCl2, and 1 mM MgCl2
a prominent example of the modulatory role of GABABRs in and were gassed with a mixture of 95% O2 and 5% CO2 (pH 7.4). TTX
synaptic plasticity at excitatory PFPC synapses. Thus, GABABRs (0.5 M) was added to the ACSF to block Na+ spikes and synaptic activ-
localized in presynaptic and postsynaptic neurons seem to be ity except in experiments investigating synaptic responses initiated by
significantly involved in long-term regulation of synaptic effi- focal stimulation within the cerebellar cortex. Patch pipettes (3 to 6 M)
cacy at various synapses in the central nervous system. were filled with an internal solution containing 150 mM KCH3SO3,

articles
5 mM KCl, 0.1 mM K-EGTA, 5.0 mM Na-HEPES, 3.0 mM Mg-ATP and 2. Barnard, E. A. et al. Subtypes of GABAA receptors: classification on the basis
0.4 mM Na-GTP (pH 7.4). KCH3SO3 in the patch pipette solution was of subunit structure and receptor function. Pharmacol. Rev. 50, 291313
(1998).
replaced with an equimolar amount of the calcium chelator BAPTA (1,2- 3. Kaupmann, K. et al. Expression cloning of GABAB receptors uncovers
bis-(2-aminophenoxy)ethane-N,N,N,N-tetraacetic acid; 35 mM) to similarity to metabotropic glutamate receptors. Nature 386, 239246 (1997).
examine the effects of the intracellular application of BAPTA on 4. Jones, K. A. et al. GABAB receptors function as a heteromeric assembly of the
GABABR-mediated responses. The series resistance was 816 M, which subunits GABABR1 and GABABR2. Nature 396, 674679 (1998).
5. Kaupmann, K. et al. GABAB-receptor subtypes assemble into functional
was monitored throughout the recording and was not compensated heteromeric complexes. Nature 396, 683687 (1998).
except in experiments investigating synaptic responses. Membrane cur- 6. Kuner, R. et al. Role of heteromer formation in GABAB receptor function.
rents were recorded in a whole-cell configuration using an EPC-7 ampli- Science 283, 7477 (1999).
fier (List Electronic, Darmstadt, Germany) and pCLAMP software (Axon 7. Thompson, S. M., Capogna, M. & Scanziani, M. Presynaptic inhibition in the
Instruments, Foster City, CA), digitized, and stored on a computer disk hippocampus. Trends Neurosci. 16, 222227 (1993).
8. Wu, L. G. & Saggau, P. Presynaptic inhibition of elicited neurotransmitter
for off-line analysis. release. Trends Neurosci. 20, 204212 (1997).
1S,3R-ACPD (100 mM) and AMPA (10 mM) were applied by ion- 9. Sodickson, D. L. & Bean, B. P. GABAB receptor-activated inwardly rectifying
tophoresis (1020 A, 100200 ms) through microelectrodes placed in the potassium current in dissociated hippocampal CA3 neurons. J. Neurosci. 16,
vicinity of PC soma from which recordings were obtained. As the 1S,3R- 63746385 (1996).
10. Luscher, C., Jan, L. Y., Stoffel, M., Malenka, R. C. & Nicoll, R. A. G protein-
ACPD-induced current response tended to increase during the initial 15 coupled inwardly rectifying K+ channels (GIRKs) mediate postsynaptic but
min after breaking into the whole-cell configuration (Fig. 6c), the effect of not presynaptic transmitter actions in hippocampal neurons. Neuron 19,
GABABR activation was determined thereafter. Furthermore, the effect of a 687695 (1997).
GABABR agonist, baclofen, on 1S,3R-ACPD-responses was tested only once 11. Mintz, I. M. & Bean, B. P. GABAB receptor inhibition of P-type Ca2+ channels
in each slice, because the effect of the GABABR agonist on the mGluR1- in central neurons. Neuron 10, 889898 (1993).
12. Kerr, D. I. & Ong, J. GABAB receptors. Pharmacol. Ther. 67, 187246 (1995).
mediated responses attenuated with repeated applications. Synaptic respons- 13. Martinelli, G. P., Holstein, G. R., Pasik, P. & Cohen, B. Monoclonal antibodies
es were evoked by stimulation (3050 V, 0.10.2 ms) via a glass for ultrastructural visualization of L-baclofen-sensitive GABAB receptor sites.
microelectrode filled with ACSF and placed within the molecular layer. PF- Neuroscience 46, 2333 (1992).
mediated EPSCs were identified based on paired-pulse facilitation. A PF- 14. Turgeon, S. M. & Albin, R. L. Pharmacology, distribution, cellular
localization, and development of GABAB binding in rodent cerebellum.
and mGluR1-mediated slow EPSC was evoked by repetitive stimulation (10
Neuroscience 55, 311323 (1993).
pulses at 100 Hz) in an ACSF that contained 6-cyano-7-nitroquinoxaline- 15. Fritschy, J. M. et al. GABAB-receptor splice variants GB1a and GB1b in rat
2,3-dione (CNQX; 10 M), D-2-amino-5- phosphonopentanoic acid (AP5; brain: developmental regulation, cellular distribution and extrasynaptic
30 M) and bicuculline (50 M). U73122, heparin and BAPTA were added localization. Eur. J. Neurosci. 11, 761768 (1999).
to the patch pipette solution and infused into PCs during whole-cell record- 16. Kawaguchi, S. & Hirano, T. Suppression of inhibitory synaptic potentiation
by presynaptic activity through postsynaptic GABAB receptors in a Purkinje
ings. Other inhibitors and receptor antagonists were applied by perfusion. neuron. Neuron 27, 339347 (2000).
17. Ito, M. Long-term depression. Annu. Rev. Neurosci. 12, 85102 (1989).
Ca2+ imaging. The Ca2+ indicator fura-2 was added to the patch pipette 18. Linden, D. J. & Connor, J. A. Long-term synaptic depression. Annu. Rev.
solution at a concentration of 1 mM and loaded into the cells for Neurosci. 18, 319357 (1995).
2540 min. Fluorescence Ca2+ ratio imaging was done by excitation of the 19. Conquet, F. et al. Motor deficit and impairment of synaptic plasticity in mice
lacking mGluR1. Nature 372, 237243 (1994).
indicator at 340:380 nm and paired emission images were acquired using 20. Ichise, T. et al. mGluR1 in cerebellar Purkinje cells essential for long-term
a cooled CCD camera (C4880, Hamamatsu Photonics, Hamamatsu, Japan) depression, synapse elimination, and motor coordination. Science 288,
at 510 nm. Paired emission images were recorded every 2 s and ratio images 18321835 (2000).
were calculated using a digital image acquisition system and image pro- 21. Nusser, Z., Mulvihill, E., Streit, P. & Somogyi, P. Subsynaptic segregation of
cessing software (ARGUS 50/CA, Hamamatsu Photonics). In all experi- metabotropic and ionotropic glutamate receptors as revealed by
immunogold localization. Neuroscience 61, 421427 (1994).
ments, visual identification of cells and Ca2+ fluorescence imaging were 22. Lujn, R., Roberts, J. D., Shigemoto, R., Ohishi, H. & Somogyi, P. Differential
done using a water immersion objective (LUMPlanFI 60, NA 0.90, Olym- plasma membrane distribution of metabotropic glutamate receptors
pus, Tokyo, Japan) capable of passing 340 nm light efficiently. mGluR1 alpha, mGluR2 and mGluR5, relative to neurotransmitter release
sites. J. Chem. Neuroanat. 13, 219241 (1997).
Drugs. 1S,3R-ACPD and SKF89976A were obtained from Tocris Cook- 23. Mateos, J. M. et al. Immunolocalization of the mGluR1b splice variant of the
metabotropic glutamate receptor 1 at parallel fiberPurkinje cell synapses in
son (Bristol, UK); baclofen, heparin and TTX, from Wako (Osaka, Japan); the rat cerebellar cortex. J. Neurochem. 74, 13011309 (2000).
-agatoxin IVA, from Peptide Institute (Osaka, Japan); and fura-2, from 24. Staub, C., Vranesic, I. & Knpfel, T. Responses to metabotropic glutamate
Dojindo (Kumamoto, Japan). CGP62349 was a gift from Novartis Phar- receptor activation in cerebellar Purkinje cells: induction of an inward
ma (Basel, Switzerland). All other chemicals were from Sigma (St. Louis, current. Eur. J. Neurosci. 4, 832839 (1992).
Missouri). 25. Linden, D. J., Smeyne, M. & Connor, J. A. Trans-ACPD, a metabotropic
receptor agonist, produces calcium mobilization and an inward current in
cultured cerebellar Purkinje neurons. J. Neurophysiol. 71, 19921998 (1994).
Statistics. Each value is expressed as mean s.e.m. Unless otherwise stat- 26. Hirono, M., Konishi, S. & Yoshioka, T. Phospholipase C-independent group I
ed, levels of significance were determined by unpaired Students t-test metabotropic glutamate receptor-mediated inward current in mouse
between groups. Purkinje cells. Biochem. Biophys. Res. Commun. 251, 753758 (1998).
27. Goodman, R. R., Kuhar, M. J., Hester, L. & Snyder, S. H. Adenosine receptors:
autoradiographic evidence for their location on axon terminals of excitatory
ACKNOWLEDGEMENTS neurons. Science 220, 967969 (1983).
28. Jaarsma, D., Levey, A. I., Frostholm, A., Rotter, A. & Voogd, J. Light-
We thank J. Bockaert, T. Murakoshi, D. Rusakov, F. Saitow and K. Yoshioka for microscopic distribution and parasagittal organisation of muscarinic
comments on the manuscript and Novartis Pharma (Basel, Switzerland) for the receptors in rabbit cerebellar cortex. J. Chem. Neuroanat. 9, 241259 (1995).
gift of CGP62349. This work was supported in part by a Grant-in-Aid 0727910 29. Miquel, M. C. et al. Postnatal development and localization of 5-HT1A
receptor mRNA in rat forebrain and cerebellum. Brain Res. Dev. Brain Res.
(T.Y.) from the Ministry of Education, Science, Sports and Culture of Japan, and 80, 149157 (1994).
Grant-in-Aids 96L00310 (T.Y.) and 12780603 (M.H.) from the Japan Society 30. Konishi, S. & Mitoma, H. in The Role of Adenosine in the Nervous System (ed.
for Promotion of Science. S.K. is a research director of CREST, JST (Core Okada, Y.) 8995 (Elsevier, New York, 1997).
31. Mitoma, H. & Konishi, S. Monoaminergic long-term facilitation of GABA-
Research for Evolutional Science and Technology, Japan Science and Technology mediated inhibitory transmission at cerebellar synapses. Neuroscience 88,
Corporation). 871883 (1999).
32. Yamada, J., Saitow, F., Satake, S., Kiyohara, T. & Konishi, S. GABAB receptor-
mediated presynaptic inhibition of glutamatergic and GABAergic
RECEIVED 31 JULY; ACCEPTED 5 OCTOBER 2001 transmission in the basolateral amygdala. Neuropharmacology 38, 17431753
(1999).
33. Takahashi, M., Kovalchuk, Y. & Attwell, D. Pre- and postsynaptic
1. Bowery, N. G. GABAB receptor pharmacology. Annu. Rev. Pharmacol. Toxicol. determinants of EPSC waveform at cerebellar climbing fiber and parallel fiber
33, 109147 (1993). to Purkinje cell synapses. J. Neurosci. 15, 56935702 (1995).

articles
34. Dittman, J. S. & Regehr, W. G. Contributions of calcium-dependent and IPSCs and spontaneous calcium transients. Nat. Neurosci. 3, 12561265
calcium-independent mechanisms to presynaptic inhibition at a cerebellar (2000).
synapse. J. Neurosci. 16, 16231633 (1996). 43. Miyata, M. et al. Local calcium release in dendritic spines required for long-
35. Selbie, L. A. & Hill, S. J. G protein-coupled-receptor cross-talk: the fine- term synaptic depression. Neuron 28, 233244 (2000).
tuning of multiple receptor-signalling pathways. Trends Pharmacol. Sci. 19, 44. Wang, S. S., Denk, W. & Hausser, M. Coincidence detection in single
8793 (1998). dendritic spines mediated by calcium release. Nat. Neurosci. 3, 12661273
36. Jin, W., Lee, N. M., Loh, H. H. & Thayer, S. A. Opioids mobilize calcium from (2000).
inositol 1,4,5-trisphosphate-sensitive stores in NG108-15 cells. J. Neurosci. 45. Davies, C. H., Starkey, S. J., Pozza, M. F. & Collingridge, G. L. GABA
14, 19201929 (1994). autoreceptors regulate the induction of LTP. Nature 349, 609611 (1991).
37. Hahner, L., McQuilkin, S. & Harris, R. A. Cerebellar GABAB receptors 46. Mott, D. D. & Lewis, D. V. Facilitation of the induction of long-term
modulate function of GABAA receptors. FASEB J. 5, 24662472 (1991). potentiation by GABAB receptors. Science 252, 17181720 (1991).
38. Tempia, F., Miniaci, M. C., Anchisi, D. & Strata, P. Postsynaptic current 47. Komatsu, Y. GABAB receptors, monoamine receptors, and postsynaptic
mediated by metabotropic glutamate receptors in cerebellar Purkinje cells. inositol trisphosphate-induced Ca2+ release are involved in the induction of
J. Neurophysiol. 80, 520528 (1998). long-term potentiation at visual cortical inhibitory synapses. J. Neurosci. 16,
39. Batchelor, A. M. & Garthwaite, J. Novel synaptic potentials in cerebellar 63426352 (1996).
Purkinje cells: probable mediation by metabotropic glutamate receptors. 48. Rocheville, M. et al. Receptors for dopamine and somatostatin: formation of
Neuropharmacology 32, 1120 (1993). hetero-oligomers with enhanced functional activity. Science 288, 154157
40. Batchelor, A. M. & Garthwaite, J. Frequency detection and temporally (2000).
dispersed synaptic signal association through a metabotropic glutamate 49. Liu, F. et al. Direct proteinprotein coupling enables cross-talk between
receptor pathway. Nature 385, 7477 (1997). dopamine D5 and GABAA receptors. Nature 403, 274280 (2000).
41. Bourne, H. R. & Nicoll, R. Molecular machines integrate coincident synaptic 50. Khakh, B. S., Zhou, X., Sydes, J., Galligan, J. J. & Lester, H. A. State-dependent
signals. Cell 72 Suppl., 6575 (1993). cross-inhibition between transmitter-gated cation channels. Nature 406,
42. Llano, I. et al. Presynaptic calcium stores underlie large-amplitude miniature 405410 (2000).

articles
Long-term depression in the nucleus

accumbens: a neural correlate of
behavioral sensitization to cocaine
Mark J. Thomas1, Corinne Beurrier1, Antonello Bonci2 and Robert C. Malenka1
1 Nancy Pritzker Laboratory, Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Palo Alto, California 94304, USA
2 Ernest Gallo Clinic and Research Center, Department of Neurology, University of California, San Francisco, California 94110, USA
Correspondence should be addressed to R.M. (malenka@stanford.edu)
A compelling model of experience-dependent plasticity is the long-lasting sensitization to the

locomotor stimulatory effects of drugs of abuse. Adaptations in the nucleus accumbens (NAc), a com-
ponent of the mesolimbic dopamine system, are thought to contribute to this behavioral change.
Here we examine excitatory synaptic transmission in NAc slices prepared from animals displaying sen-
sitization 1014 days after repeated in vivo cocaine exposure. The ratio of AMPA (-amino-3-hydroxy-
5-methyl-4- isoxazole propionic acid) receptor- to NMDA (N-methyl-D-aspartate) receptor-mediated
excitatory postsynaptic currents (EPSCs) was decreased at synapses made by prefrontal cortical affer-
ents onto medium spiny neurons in the shell of the NAc. The amplitude of miniature EPSCs at these
synapses also was decreased, as was the magnitude of long-term depression. These data suggest that
chronic in vivo administration of cocaine elicits a long-lasting depression of excitatory synaptic
transmission in the NAc, a change that may contribute to behavioral sensitization and addiction.
Drug addiction is a pathological behavior characterized by com- induces behavioral sensitization, whereas in sensitized animals,
pulsive drug seeking and drug ingestion despite severe adverse the injection of psychostimulants into the NAc is sufficient to
consequences. Animal models of addiction mimic several of the elicit sensitized responses (for review, see refs. 3, 4).
core features of addiction in humans and therefore can be used Much of the initial work on the adaptations that mediate
to study the neural mechanisms underlying this pathological behavioral sensitization appropriately focused on pre- and
form of experience-dependent behavioral plasticity. Because of postsynaptic changes in dopaminergic transmission. Recent-
advances in molecular neurobiology, much is known about how ly, however, evidence has accumulated that excitatory inputs
drugs of abuse interact with and modify their molecular targets. to the VTA and NAc are critical. Consistent with the VTAs
Furthermore, the molecular adaptations that occur in specific involvement in triggering sensitization, repeated electrical stim-
brain regions in response to acute and chronic administration ulation of excitatory cortical afferents to the VTA induces sen-
of drugs of abuse are being determined at a rapid pace1. There is sitization that, when elicited by systemic drug exposure, is
a relative paucity of information, however, about the changes in blocked by local injection of glutamate receptor antagonists
synapses and circuits that occur as a consequence of these drug- into the VTA5. Indeed, in vivo administration of cocaine elicits
induced molecular changes; this information is critical for a a robust enhancement of excitatory synaptic transmission in
thorough understanding of the neural mechanisms of addiction. this structure6. On the other hand, the expression of behav-
A key feature of addiction is the intensification of drug crav- ioral sensitization is blocked by inhibiting excitatory synaptic
ing that occurs in human addicts with repeated drug exposure. transmission in the NAc7 or by lesions of the excitatory corti-
A prominent model for this behavioral change is the long-last- cal afferents to this structure8 (but see ref. 9).
ing increase in locomotor response to drugs of abuse following What changes occur in the NAc that account for its impor-
repeated exposures. This increased response, termed behavioral tance in behavioral sensitization? The major cell type in the
sensitization, is thought to reflect adaptations in neural circuits NAc (>95% of total) is the medium spiny neuron that receives
that determine the incentive value of external stimuli rendering glutamatergic inputs from a variety of cortical and subcortical
the circuits hypersensitive or sensitized2. Many of the neural limbic areas, including the hippocampus, prefrontal cortex and
adaptations that have been identified following psychostimulant amygdala. Because these neurons have high resting membrane
administration take place in the mesolimbic dopamine system, potentials and are generally quiescent, they depend on these
major components of which are the ventral tegmental area (VTA) excitatory inputs to generate output to their main targets, the
and the nucleus accumbens (NAc). Modifications in the VTA are VTA and ventral pallidum. It is therefore reasonable to
involved in the induction of behavioral sensitization, and modi- hypothesize that changes in the efficacy of these inputs would
fications in the NAc are involved in its long-term maintenance3,4. have a significant effect on the functioning of mesolimbic
For example, repeated injection of psychostimulants into the VTA dopamine circuitry and would thereby alter the behavioral

articles
saline saline cocaine cocaine Synaptic transmission in cocaine-treated mice

( , ) ( , )
To determine whether changes in the efficacy of AMPA receptor
5000 (AMPAR)-mediated synaptic transmission had occurred as a result
of the chronic cocaine treatment, we first measured field EPSPs
Distance traveled / session (cm)
4000
from the NAc10 in slices prepared from the animals one day after
the final cocaine challenge. The NAc is commonly divided into
two components, the shell and the core, which are distinguished
3000
both anatomically and functionally11. Therefore, data were divid-
ed according to the location of the recording. Despite using a range

2000 of stimulus intensities, we could not detect any difference in the
size of the field EPSPs in either the shell or core between cocaine-
1000
and saline-treated groups (data not shown).
Differences in the placement of stimulating and recording
electrodes and the density of afferent fibers may be considerable
0
1 2 3 4 5 6 7
sources of experimental variability when attempting to measure
1014 test
days differences in synaptic strength between slices. To reduce the slice-
Injection day
to-slice variability that might have obscured any cocaine-induced
Fig. 1. Behavioral sensitization induced by repeated cocaine administra- changes in synaptic efficacy, we repeated our analysis using a
tion. Mean ( s.e.m.) locomotor activity in response to saline and
more sensitive assay that compared the relative contributions of
cocaine injections. Locomotor activity was monitored for 15 min imme-
diately following each injection.
AMPARs and NMDA receptors (NMDARs) to EPSCs. This pro-
cedure involved evoking EPSCs while holding cells at +40 mV in
the absence and then in the presence of the NMDAR antagonist
D-APV (50 M)6 (see Methods). Using this assay, we observed a
response to a drug of abuse. To test this hypothesis, we exam- decrease in the AMPAR/NMDAR ratio in cells in the NAc shell
ined excitatory synaptic transmission in slices of NAc prepared from cocaine-treated compared to saline-treated mice (Fig. 2ac;
from mice in which behavioral sensitization was induced by cocaine, 0.68 0.09, n = 12; saline 0.97 0.09, n = 9; p < 0.05)
repeated in vivo administration of cocaine. This in vivo
treatment caused a long-lasting depression of synaptic
strength at synapses made by prefrontal cortical afferents a Shell NMDA
onto medium spiny neurons in the shell of the NAc.

saline cocaine
RESULTS
Long-lasting locomotor sensitization to cocaine
AMPA
To induce behavioral sensitization, we paired repeated
cocaine injections with exposure of the animals to a dis- b Shell c
tinct test environment. After two days of saline injections
to habituate the animals to the activity box, the immediate 1.0 saline 1.0
locomotor response to a fixed dose of cocaine (15 mg/kg) cocaine

Cumulative probability
AMPAR/NMDAR ratio
0.8 0.8 *
increased dramatically across days of testing (Fig. 1; dis-
tance traveled, day 3, saline, 418 22 cm, n = 51; cocaine,
0.6 0.6
1176 90 cm, n = 47; day 7, saline, 462 36 cm, n = 51;
cocaine, 3896 191 cm, n = 47; p < 0.001). To test 0.4 0.4
whether this procedure produced long-lasting sensitiza-
tion, we administered a challenge dose of cocaine to both 0.2 0.2
saline- and cocaine-treated groups 10 to 14 days follow-
ing the last dose of the initial treatment regimen. Mice 0.0 0.0
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 saline cocaine
pretreated with cocaine showed a much greater locomo-
tor response to cocaine than did saline-pretreated animals AMPAR/NMDAR ratio
(cocaine, 4660 204 cm, n = 47; saline, 1556 170 cm,
n = 51; p < 0.001). These results indicate that the initial d Core e
five-day exposure to cocaine caused behavioral sensitiza- 1.0
tion that lasted for at least two weeks. 1.0 saline
cocaine 0.8
AMPAR/NMDAR ratio
0.8
0.6
Fig. 2. Repeated cocaine administration induces a decrease in the 0.6
AMPAR/NMDAR ratio of synaptic currents in the shell but not in
0.4
the core of NAc. (a) Sample EPSCs from saline and cocaine- 0.4
treated animals. Scale bars, 40 ms and 20 pA. (b, d) Cumulative
probability plots of AMPAR/NMDAR ratio in shell (b) and core (d) 0.2 0.2
neurons from saline (n = 9 shell; n = 8 core) and cocaine-treated
(n = 12 shell; n = 7 core) mice. (c, e) Mean AMPAR/NMDAR ratio 0.0 0.0
in shell (c) and core (e) neurons from saline and cocaine-treated 0.2 0.4 0.6 0.8 1.0 1.2 1.4 saline cocaine
mice. *p < 0.05. AMPAR/NMDAR ratio

articles
Fig. 3. Repeated cocaine had no significant effect on paired-pulse facilita-

tion in either shell or core. (a, b) Mean paired-pulse facilitation (PPF) val- a Shell
ues in shell (a; saline, n = 7; cocaine, n = 9) and in core (b; saline, n = 9;
cocaine, n = 8) are shown for different interstimulus intervals. A sample 3.0
trace from a shell neuron in the cocaine group is shown (50 ms ISI; cali-
bration bars are 20 ms, 20 pA). Values did not differ significantly between 2.5
mean PPF ratio

cocaine
groups at any interval tested (p > 0.05; ANOVA). saline
2.0
1.5
but not in cells in the core (Fig. 2d and e; cocaine, 0.76 0.11, 1.0
n = 7; saline 0.78 0.12, n = 8; p > 0.05).
Because a single dose of cocaine can elicit behavioral sensiti- 0.5
zation and an increase in the AMPAR/NMDAR ratio in the VTA6, 0 50 100 150 200
we wondered if the single injection of cocaine received by the Interstimulus interval (ms)
saline-treated mice may have affected the AMPAR/NMDAR ratio
and thereby influenced our results. Therefore, in another set of b Core
experiments, mice were given either one cocaine or one saline 3.0
injection and AMPAR/NMDAR ratios for shell neurons were
2.5
determined the next day. There was no significant difference
mean PPF ratio

between the mean AMPAR/NMDAR ratios of these groups (data 2.0
not shown) indicating that, unlike the VTA, a single cocaine expo-
sure does not affect this measurement in the NAc shell. 1.5
A decrease in the AMPAR/NMDAR ratio likely reflects a
1.0
decrease in AMPAR function and/or number, an increase in
NMDAR function and/or number, or a combination of these. 0.5
Such a change, however, does not rule out that changes in trans- 0 50 100 150 200
mitter release mechanisms may be induced by in vivo cocaine Interstimulus interval (ms)
exposure as well. To assess whether changes in the probability of
transmitter release (Pr) occurred in the cocaine-treated animals,
we compared the magnitude of the facilitation that occurs in changes with changes in Pr12. The paired-pulse ratio did not dif-
response to paired-pulse stimulation, a measure that routinely fer between cocaine- and saline-treated mice when tested at a
variety of intervals in either shell or core neurons following the
a cocaine treatment regimen (Fig. 3), suggesting that significant
changes in Pr did not occur in cocaine-treated mice.
Changes in mEPSC amplitude in cocaine-treated mice

To address whether AMPAR function and/or number was modi-
fied in the NAc shell of cocaine-treated mice, we examined minia-
ture AMPAR-mediated EPSCs (mEPSCs). We found no difference
between cocaine and saline-treated mice in either the frequency
or the amplitude of mEPSCs in shell neurons (Fig. 4). This sug-
b c gested that the decrease in the AMPAR/NMDAR ratio might be
20
due to an increase in the function and/or number of NMDARs.
Mean mEPSC amplitude (pA)
1.0
Testing this possibility is difficult because the relatively high fre-

0.8 15
quency of mEPSCs in this preparation (7 Hz), the increased
0.6 noise generated by opening of NMDARs by ambient glutamate13
10
and the slow rise-time of NMDAR events make it problematic to
0.4
saline
5
reliably detect NMDAR-mediated mEPSCs. To circumvent these
0.2 cocaine problems, we bathed slices in Mg 2+-free solution to record
0.0 0
mEPSCs containing both AMPAR and NMDAR components
0 10 20 30 40 saline cocaine using the fast rise-time of the AMPAR component to reliably
Amplitude (pA) detect the mEPSCs. We then applied D-APV and collected AMPAR
d e mEPSCs from the same cell. Subtracting average mEPSCs in the
10 presence of D-APV from those collected in its absence yielded an
Mean mESPC frequency (Hz)
1.0
8
average NMDAR mEPSC for each cell (Fig. 5a and b). This mea-
0.8
0.6 6
Fig. 4. Repeated cocaine administration had no effect on the amplitude

0.4 4
or frequency of shell AMPAR mEPSCs. (a) Sample traces of mEPSCs in a
0.2 2
shell neuron in a cocaine-treated mouse. Scale bars, 80 ms and 10 pA.
(b, d) Cumulative amplitude (b) and inter-event interval (d) distribu-
0.0 0 tions of shell mEPSCs obtained in cocaine and saline-treated mice
0 200 400 600 800 1000 saline cocaine (n = 400 events per cell, 8 cells in each group). (c, e) Means of mEPSCs
Inter-event Interval (ms) amplitude (c) and frequency (e) p > 0.05.
articles
2+ Fig. 5. Repeated cocaine administration did not affect

a zero Mg b NMDAR mEPSC amplitude or response to NMDA in shell
neurons. (a) Samples of mEPSCs recorded at 65 mV in zero
Mg2+ solution in the absence (top) and presence (bottom) of
D-APV (50 M) in a cocaine-treated mouse. Scale bars, 100 ms
dual
and 20 pA. (b) Sample averaged traces of mEPSCs obtained in
AMPA each condition plus the subtracted trace that yielded an aver-
2+
zero Mg + APV NMDA age NMDAR mEPSC. Scale bars, 10 ms and 2 pA.
(c) Cumulative probability plot for NMDAR mEPSC amplitude
(n = 9 cocaine; n = 7 saline). (d) Means of NMDAR mEPSCs

amplitude (p > 0.05). (e) Bath application of NMDA (10 M)
for 30 s elicited similar change in holding current in neurons
held at +40 mV (cocaine, n = 5; saline, n = 4; p > 0.05).
c d 5
1.0 saline A limitation of our analysis of mEPSCs is that medium
Mean amplitude (pA)
cocaine spiny neurons receive excitatory projections from several

4
0.8
sources and there was no way of knowing from which
3
0.6 synapses the mEPSCs were generated. If the changes in
2
the AMPAR/NMDAR ratio of the evoked EPSC occurred
0.4
primarily at the stimulated synapses, presumably those
0.2 1 made by cortical afferents, we may have missed detecting
a change in mEPSCs. To circumvent this problem, we
0.0 0
1 2 3 4 5 6 7 8 9 10 saline cocaine replaced the Ca2+ in our bathing solution with Sr2+ and
e Amplitude (pA) applied short bursts of afferent stimulation that causes
the asynchronous release of vesicles from the set of
Change in holding current (pA)
120 NMDA synapses that were activated (Fig. 6a)16. Using this tech-
saline
100
cocaine
nique, it was possible to preferentially sample mEPSCs
80 from the same subset of synapses that were activated to
60 yield the evoked EPSC17. Although no significant differ-
40 ence in the frequency of asynchronous quantal events
20 between shell neurons from cocaine and saline groups
0 was observed (cocaine, 25 3 Hz, n = 11; saline,
20 23 2 Hz, n = 12; p > 0.05), the mean amplitude distri-
40 bution of quantal events in the cocaine group was signif-
0 1 2 3 4 5 icantly shifted to the left compared to the saline group
Time (min) (Fig. 6b; n = 11, 12; KolmogorovSmirnov test; p < 0.05)
and their mean amplitude was also decreased (Fig. 6c)
surement did not differ significantly in shell neurons from (cocaine, 12.4 0.4 pA; saline 14.0 0.6 pA, p < 0.05). These
cocaine- and saline-treated mice (Fig. 5c and d), suggesting that results indicate that AMPAR-mediated quantal events generated
NMDAR function and/or number is not altered at synapses that by the synapses activated by cortical afferents are significantly small-
contain both AMPARs and NMDARs. er in the cocaine-treated group, suggesting that the decrease in
Another plausible explanation for the decrease in the AMPAR/NMDAR ratio is due, at least in part, to a reduction in
AMPAR/NMDAR ratio in cocaine-treated mice is that cocaine AMPAR function and/or number specifically at these synapses.
causes an increase in the proportion of synapses that contain only
NMDARs, so-called silent synapses14. Such synapses would not Changes in LTD in cocaine-treated mice
have been sampled in our experiments examining dual-compo- Given that the synapses made by cortical afferents onto medium
nent mEPSCs, but may have been recruited when we evoked spiny neurons can express NMDAR-dependent LTD18, a form of
EPSCs at +40 mV. An initial finding that led to the proposal of synaptic plasticity that is associated with a reduction in AMPAR
silent synapses is that the coefficient of variation (CV) of EPSCs function and number17,19,20, we predicted that the expression of
was higher for the AMPAR component of EPSCs relative to the LTD would be reduced in cocaine-treated mice. We first tested
NMDAR component15. When we examined CV ratios in shell whether the triggering of LTD at NAc synapses was significantly
neurons from cocaine and saline-treated mice, we found that in impaired in cocaine-treated mice by delivering an LTD-inducing
13 of 19 cells examined, the AMPAR/NMDAR CV ratio was train of stimuli (5 Hz, 3 min) while monitoring field EPSPs. This
greater than 1, suggesting that some synapses on a proportion of induced a modest amount of LTD that did not differ between the
medium spiny neurons may contain only NMDARs. However, two groups (cocaine, 79 6% of baseline, n = 6; saline, 85 5% of
there was no significant difference between the groups in this baseline, n = 6; data not shown). To determine whether the mag-
ratio (cocaine, 1.39 0.16, n = 11; saline, 1.15 0.09, n = 8; p > nitude of LTD was reduced, as would be expected if the decrease in
0.05). As a final test to determine whether the function or num- synaptic strength in cocaine-treated mice was due to mechanisms
ber of NMDARs increased in cocaine-treated mice, we bath- shared with LTD, we made whole-cell recordings from shell neu-
applied NMDA (10 M, 30 s) and recorded the change in holding rons and used a strong LTD induction protocol that generated a
current (Fig. 5e). Again, there was no significant difference near-saturating amount of LTD (3 bouts of 5 Hz,
between cocaine and saline-treated groups (cocaine, 82 16 pA, 3 min stimulation paired with depolarization to 50 mV). This
n = 5; saline, 76 17 pA, n = 4; p > 0.05). elicited robust LTD in saline-treated animals (Fig. 7;

articles
Fig. 6. Repeated cocaine treatment decreased the amplitude of 2+ 2+

AMPAR quantal EPSCs in cortical synapses onto NAc shell neu- a Ca Sr
rons. (a) Sample traces of evoked EPSCs in the presence of Ca2+
or Sr2+. Scale bars, 50 ms and 10 pA. (b) Cumulative amplitude
distributions of AMPAR quantal EPSCs in Sr2+ solution for saline
and cocaine-treated mice (n = 300 events per cell; saline, n = 11;
cocaine, n = 12). Distributions are significantly different (p < 0.05;
KolmogorovSmirnov test). (c) Mean amplitude of quantal EPSCs
in Sr2+ solution for saline and cocaine-treated mice. Cocaine
treatment induced a significant decrease in Sr2+-induced AMPAR

mEPSC amplitude (p < 0.05; t-test).
b c 15
66 7% of baseline, n = 7), but much smaller LTD in the
saline
Mean mEPSC amplitude (pA)

1.0
cocaine group (Fig. 7; 84 4% of baseline, n = 8, p < 0.05). cocaine 14
DISCUSSION 0.8 13 *
Psychostimulant-induced behavioral sensitization is thought 12
to model some of the core features of addiction24, as well 0.6
as the development of drug-elicited psychosis21. It is of inter- 11

est not only because it serves as a model for the pathological 0.4
10
changes that occur as a consequence of repeated exposure
to drugs of abuse but also because it is a robust form of 0.2
9
experience-dependent behavioral plasticity. Indeed, sensi-
tized animals can remain hypersensitive to the locomotor 0.0 0
5 10 15 20 25 30 35 saline cocaine
and rewarding actions of drugs for periods of months to
years24. Over the last few years, evidence has accumulated Amplitude (pA)
that modifications in excitatory neural circuitry in the VTA
and NAc may be particularly important in this form of experi- sensitization was induced by repeated in vivo treatment with
ence-dependent plasticity. However, excitatory synaptic trans- cocaine. We found that cocaine treatment decreased synaptic
mission following repeated drug exposure has not been examined strength at excitatory synapses made by cortical afferents onto
directly. In this study, we measured excitatory synaptic respons- medium spiny neurons and that this occurred in the shell region
es in slices of the NAc prepared from animals in which behavioral of the NAc but not in the core. Changes in NMDAR-mediated
synaptic responses and in the probability of transmitter release
were not detected. Consistent with the idea that the cocaine-
saline induced decrease in synaptic strength shares expression mecha-
a 500
nisms with LTD, we found that the magnitude of LTD was

Amplitude (pA)
400
300 significantly reduced in the cocaine-treated mice.

200
Our observations are consistent with several previous find-
ings. First, the single-unit responses of NAc neurons to glutamate
100
are decreased for at least 14 days following a cocaine or ampheta-
0
0 10 20 30 40 50 60 70 mine sensitization regimen22. Although a decrease in AMPAR
cocaine function and/or number, as our results suggest, could explain this
500
result, changes in voltage-dependent conductances following
400
chronic cocaine exposure 23 may also contribute. Second, a
Amplitude (pA)
300
decrease in the levels of the AMPAR subunits, GluR1 and GluR2
200 is observed in the NAc 14 days following amphetamine-induced
100 sensitization24,25 (but see ref. 26). Third, the inward current gen-
0 erated by AMPA/kainate application to acutely dissociated stri-
0 10 20 30 40 50 60 70
atal neurons is decreased in cells from animals that have received
Time (min)
chronic cocaine27. Fourth, viral-mediated overexpression of GluR2
in the NAc, a manipulation that would be expected to decrease
b 140
AMPAR function, increases behavioral sensitivity to cocaine as
EPSC amplitude (% of baseline)
saline measured by place conditioning28. The functional importance of

120 cocaine
100
Fig. 7. Repeated cocaine treatment decreased the magnitude of LTD in
shell neurons. (a) Examples of individual experiments in saline (top) and
80 cocaine-treated mice (bottom) displaying the time course of EPSCs
before and after 3 bouts of 5 Hz, 3 min synaptic stimulation paired with
60 depolarization of the cell to 50 mV. Traces shown were collected during
the baseline period and 20 min following the last bout. Scale bars, 20 ms
0 and 50 pA. (b) Summary graph of the average LTD elicited in saline
0 10 20 30 40 50 60 (n = 7) and cocaine-treated (n = 8) mice. LTD was significantly decreased
Time (min) in cocaine-treated animals compared with saline animals (p < 0.05).
articles
a decrease in the strength of excitatory inputs to the NAc is as psychostimulant-induced behavioral sensitization offer rela-
supported by the findings that virtually all drugs of abuse decrease tively simple models for understanding the neural mechanisms
the firing of NAc neurons29 and that such decreases have been underlying many forms of experience-dependent plasticity,
associated with enhanced locomotor activity30. Furthermore, ani- including learning and memory14,47.
mals will self-administer glutamate receptor antagonists directly
into the NAc shell31, suggesting that decreases in excitatory drive METHODS
are reinforcing. It has been proposed that cocaine-induced behav- Treatment regimen and locomotor activity. Male C57/Bl6 mice (2426
ioral sensitization involves an enhancement, not a depression, of days old) were given intraperitoneal injections of either saline (0.9% NaCl)
excitatory synaptic transmission in the NAc4. This conclusion, or saline with cocaine (15 mg/kg). Immediately following each injection,
however, is based primarily on microdialysis measurements of horizontal locomotor activity was monitored in open-field chambers (Med
Associates, St. Albans, Vermont) for 15 min. After two days of saline injec-
extracellular glutamate concentration, an assay that is not a direct tions, mice were divided into groups that received five daily injections of
measure of synaptic strength. either cocaine or saline. Following 1014 days without injections, both
The NAc is commonly divided into two regions: the core, groups received cocaine injections and locomotor activity was assessed.
which is considered a functional extension of the dorsal striatum, Brain slices were prepared on the following day. All procedures were
and the shell, which is thought to be a transitional region between approved by the Institutional Animal Care and Use Committee.
the striatum and the extended amygdala32. There is evidence that
the shell is particularly important for mediating the effects of Electrophysiology. Sagittal slices of the NAc (200250 m) were pre-
drugs of abuse as well as behavioral sensitization. Certain drugs pared with a vibratome (Leica, Nussloch, Germany) as described10. Slices
of abuse are preferentially self-administered in the shell compared (four per animal) were placed in a holding chamber and allowed to recov-
er for at least 1 h before being placed in the recording chamber and super-
to core31,33,34. Similarly, drug-induced increases in extracellular
fused with bicarbonate-buffered solution (ACSF) saturated with 95%
dopamine levels may preferentially occur in the shell35,36. It has O 2 /5% CO 2 and containing 119 mM NaCl, 2.5 mM KCl, 1.0 mM
also been reported that injections of D1 receptor antagonists into NaH 2 PO 4 , 1.3 mM MgCl 2 , 2.5 mM CaCl 2 , 26.2 mM NaHCO 3 and
the shell reduce the reinforcing effects of cocaine37. In the con- 11 mM glucose (at 2830C). Picrotoxin (100 M) was added to block
text of drug-induced locomotor activity, three weeks after repeat- GABAA receptor-mediated IPSCs. Cells were visualized using infrared-
ed cocaine exposure, injection of amphetamine into the shell, but differential interference contrast video microscopy. Whole-cell voltage-
not the core, elicits sensitized locomotor responses38. Conversely, clamp recordings were made using an Axopatch1D amplifier (Axon
lesions of the shell markedly impair the locomotor effects of Instruments, Foster City, California). Electrodes (38 M) contained
amphetamine as well as its rewarding properties39. There is also 117 mM cesium gluconate, 2.8 mM NaCl, 20 mM HEPES, 0.4 mM
evidence, however, that excitatory synaptic transmission in the EGTA, 5 TEA-Cl, 2.5 MgATP, and 0.25 MgGTP, pH 7.27.4
(285295 mOsm) for whole-cell experiments and ACSF for field record-
core is important for the expression of behavioral sensitization3,4 ings. Series resistance (1040 M) and input resistance were monitored
and thus, our results should not be taken to indicate that modifi- on-line with a 4-mV depolarizing step (50 ms) given with every afferent
cations in this structure are not also important. stimulus. Medium spiny neurons were identified by their morphology
All of our findings are consistent with the hypothesis that chron- and high resting membrane potential (75 to 85 mV).
ic in vivo cocaine administration induces a form of LTD that shares We examined core neurons in slices in which rostral and caudal limbs
expression mechanisms with the NMDAR-dependent LTD previ- of the anterior commisure as well as caudate/putamen were present.
ously described at these synapses18. We were unable to detect any Shell neurons were examined in medial NAc slices that did not contain
change in NMDAR properties but we would caution that our assays dorsal striatal tissue. Stainless steel bipolar microelectrodes were placed
might have been relatively insensitive, especially if such changes are at the prelimbic cortexNAc border to stimulate afferents preferential-
ly from the prelimbic cortex. Afferents were stimulated at 0.1 Hz except
restricted to synapses formed by a subset of NAc afferents. The rela-
where noted. Neurons were voltage-clamped at 80 mV except where
tionship between our findings and the increase in spine density and noted. Data were filtered at 12 kHz, digitized at 25 kHz and collected
proportion of branched spines that occurs following psychostim- on-line using custom software (Igor Pro; Wavemetrics, Lake Oswego,
ulant-induced sensitization40,41 is unclear. Although an increase in Oregon). EPSC amplitudes were calculated by taking the mean of a
spines would suggest that the total afferent input to medium spiny 12 ms (AMPAR EPSCs) or 34 ms (NMDAR EPSCs) window around
neurons was increased, nothing is known about the detailed prop- the peak and comparing this with the mean of an 8-ms window imme-
erties of these new spines. Indeed, if they contained small numbers diately before the stimulation artifact. AMPAR/NMDAR ratios were
of AMPARs or were functionally silent, they could contribute to computed by taking the average of EPSCs at +40 mV (2530 EPSCs) in
our observed decrease in the AMPAR/NMDAR ratio. the absence and presence of D-APV (50 M). The average response in
We have demonstrated that an LTD-like process in the NAc the presence of D-APV (AMPAR-only) was subtracted from that seen in
its absence and an average NMDAR EPSC was calculated. The peak of
may contribute to the reorganization of neural circuitry that the AMPAR EPSC was divided by the peak of the NMDAR EPSC to yield
underlies behavioral sensitization to cocaine, and thus may be an AMPAR/NMDAR ratio. To compute AMPAR and NMDAR CVs,
an important factor in the development of addiction. Also, the short latency (AMPAR; 24 ms following stimulus) and long latency
decrease in the ability of synaptic activity to elicit LTD in the NAc (NMDAR; 4550 ms following stimulus) amplitude measurements
shell may contribute to the drug-induced behavioral changes. (2530 per cell) were made at +40 mV and amplitude variances were
Together with the recent demonstration of a cocaine-induced divided by amplitude means after subtracting the variance of the record-
LTP in the VTA6, these results suggest that like other forms of ing noise. Miniature EPSCs (300600 per cell) were collected in the pres-
experience-dependent behavioral plasticity4246, drug-induced ence of either tetrodotoxin (TTX, 1.5 M) or lidocaine hydrochloride
changes in behavior may be due, at least in part, to modifications (0.60.8 mM). Dual-component mEPSCs were collected in the addi-
tional presence of 20 M glycine, in the absence of added Mg2+ and at a
of synaptic efficacy in critical neural circuits. The sort of
holding potential of 65 mV. In strontium experiments, AMPAR-medi-
approaches taken here should further our understanding of how ated quantal events were collected during a 400-ms period beginning
the molecular adaptations caused by drugs of abuse lead to the 50 ms following each stimulus of a 2-Hz, 10-pulse train delivered once
changes in the behavior of neural circuitry that ultimately must every 30 s in bath solution containing D-APV (50 M), 2.6 mM MgCl2,
underlie addiction. Furthermore, because of shared molecular 2.5 mM Sr2+ and no Ca2+. Quantal events were analyzed using Mini-
and circuit mechanisms, drug-induced behavioral changes such analysis software (Synaptosoft, Decatur, Georgia) with detection para-

articles
meters set at greater than 4 pA amplitude and less than 3 ms rise time 20. Lee, H. K., Barbarosie, M., Kameyama, K., Bear, M. F. & Huganir, R. L.
and verified by eye. For each cell, a random stretch of 300 mEPSCs was Regulation of distinct AMPA receptor phosphorylation sites during
taken for constructing cumulative probability plots and calculating mean bidirectional synaptic plasticity. Nature 405, 955959 (2000).
21. Segal, D. S. & Schuckit, M. A. in Stimulants: Neurochemical, Behavioral and
mEPSC amplitudes. LTD was induced by delivering a single train (5 Hz, Clinical Perspectives (ed. Creese, I.) 131167 (Raven, New York, 1983).
3 min; field experiments) or multiple trains (3 5 Hz, 3 min, 5 min 22. White, F. J., Hu, X. T., Zhang, X. F. & Wolf, M. E. Repeated administration of
inter-train interval) paired with depolarization to 50 mV. cocaine or amphetamine alters neuronal responses to glutamate in the
In over 90% of experiments, data acquisition and analysis were per- mesoaccumbens dopamine system. J. Pharmacol. Exp. Ther. 273, 445454
(1995).
formed blindly without knowledge of the treatment history of the 23. Zhang, X. F., Hu, X. T. & White, F. J. Whole-cell plasticity in cocaine
slices. On over 75% of recording days, a similar amount of data was withdrawal: reduced sodium currents in nucleus accumbens neurons.
collected from both cocaine and saline groups. Results are presented J. Neurosci. 18, 488498 (1998).
as mean s.e.m. Statistical significance was assessed using two-tailed 24. Lu, W., Chen, H., Xue, C. J. & Wolf, M. E. Repeated amphetamine
Students t-tests, ANOVA for paired-pulse experiments, or Kol- administration alters the expression of mRNA for AMPA receptor subunits in
rat nucleus accumbens and prefrontal cortex. Synapse 26, 269280 (1997).
mogorovSmirnov tests for comparing the cumulative probability his- 25. Lu, W. & Wolf, M. E. Repeated amphetamine administration alters AMPA
tograms. Traces in figures have had stimulus artifacts removed and are receptor subunit expression in rat nucleus accumbens and medial prefrontal
averages of 1012 consecutive responses. cortex. Synapse 32, 119131 (1999).
26. Churchill, L., Swanson, C. J., Urbina, M. & Kalivas, P. W. Repeated cocaine
alters glutamate receptor subunit levels in the nucleus accumbens and ventral
tegmental area of rats that develop behavioral sensitization. J. Neurochem. 72,
ACKNOWLEDGEMENTS 23972403 (1999).
This work was supported by grants from NIDA (R.M., M.T.) and the State of 27. Bibb, J. A. et al. Effects of chronic exposure to cocaine are regulated by the
California for medical research on alcohol and substance abuse through the neuronal protein Cdk5. Nature 410, 376380 (2001).
28. Kelz, M. B. et al. Expression of the transcription factor deltaFosB in the brain
University of California, San Francisco (A.B.). We thank T. Robinson, S. Nicola controls sensitivity to cocaine. Nature 401, 272276 (1999).
and G. Hjelmstad for comments on the paper and D. Saal for help with some 29. Wise, R. A. Drug-activation of brain reward pathways. Drug Alcohol. Depend.
experiments. 51, 1322 (1998).
30. Pennartz, C. M., Groenewegen, H. J. & Lopes da Silva, F. H. The nucleus
accumbens as a complex of functionally distinct neuronal ensembles: an
RECEIVED 6 AUGUST; ACCEPTED 16 OCTOBER 2001 integration of behavioural, electrophysiological and anatomical data. Prog.
Neurobiol. 42, 719761 (1994).
31. Carlezon, W. A. Jr. & Wise, R. A. Rewarding actions of phencyclidine and
1. Nestler, E. J. Molecular basis of long-term plasticity underlying addiction. related drugs in nucleus accumbens shell and frontal cortex. J. Neurosci. 16,
Nat. Rev. Neurosci. 2, 119128 (2001). 31123122 (1996).
2. Robinson, T. E. & Berridge, K. C. The neural basis of drug craving: an 32. McGinty, J. F., ed. Advancing from the Ventral Striatum to the Extended
incentive-sensitization theory of addiction. Brain. Res. Brain. Res. Rev. 18, Amygdala: Implications for Neuropsychiatry and Drug Abuse Vol. 877 (New
247291 (1993). York Academy of Sciences, New York, New York, 1999).
3. Wolf, M. E. The role of excitatory amino acids in behavioral sensitization to 33. Carlezon, W. A. Jr., Devine, D. P. & Wise, R. A. Habit-forming actions of
psychomotor stimulants. Prog. Neurobiol. 54, 679720 (1998). nomifensine in nucleus accumbens. Psychopharmacology (Berl.) 122,
4. Vanderschuren, L. J. & Kalivas, P. W. Alterations in dopaminergic and 194197 (1995).
glutamatergic transmission in the induction and expression of behavioral 34. McKinzie, D. L., Rodd-Henricks, Z. A., Dagon, C. T., Murphy, J. M. &
sensitization: a critical review of preclinical studies. Psychopharmacology McBride, W. J. Cocaine is self-administered into the shell region of the
(Berl.) 151, 99120 (2000). nucleus accumbens in Wistar rats. Ann. NY Acad. Sci. 877, 788791 (1999).
5. Schenk, S. & Snow, S. Sensitization to cocaines motor activating properties 35. Pontieri, F. E. et al. Psychostimulant drugs increase glucose utilization in the
produced by electrical kindling of the medial prefrontal cortex but not of the shell of the rat nucleus accumbens. Neuroreport 5, 25612564 (1994).
hippocampus. Brain Res. 659, 1722 (1994). 36. Pontieri, F. E., Tanda, G. & Di Chiara, G. Intravenous cocaine, morphine, and
6. Ungless, M. A., Whisler, J. L., Malenka, R. C. & Bonci, A. Single cocaine amphetamine preferentially increase extracellular dopamine in the shell as
exposure in vivo induces long-term potentiation in dopamine neurons. compared with the core of the rat nucleus accumbens. Proc. Natl. Acad. Sci.
Nature 411, 583587 (2001). USA 92, 1230412308 (1995).
7. Pierce, R. C., Bell, K., Duffy, P. & Kalivas, P. W. Repeated cocaine augments 37. Caine, S. B., Heinrichs, S. C., Coffin, V. L. & Koob, G. F. Effects of the
excitatory amino acid transmission in the nucleus accumbens only in rats dopamine D-1 antagonist SCH 23390 microinjected into the accumbens,
having developed behavioral sensitization. J. Neurosci. 16, 15501560 (1996). amygdala or striatum on cocaine self-administration in the rat. Brain Res.
8. Pierce, R. C., Reeder, D. C., Hicks, J., Morgan, Z. R. & Kalivas, P. W. Ibotenic 692, 4756 (1995).
acid lesions of the dorsal prefrontal cortex disrupt the expression of 38. Pierce, R. C. & Kalivas, P. W. Amphetamine produces sensitized increases in
behavioral sensitization to cocaine. Neuroscience 82, 11031114 (1998). locomotion and extracellular dopamine preferentially in the nucleus
9. Li, Y. & Wolf, M. E. Ibotenic acid lesions of prefrontal cortex do not prevent accumbens shell of rats administered repeated cocaine. J. Pharmacol. Exp.
expression of behavioral sensitization to amphetamine. Behav. Brain. Res. 84, Ther. 275, 10191029 (1995).
285289 (1997). 39. Parkinson, J. A., Olmstead, M. C., Burns, L. H., Robbins, T. W. & Everitt, B. J.
10. Nicola, S. M., Kombian, S. B. & Malenka, R. C. Psychostimulants depress Dissociation in effects of lesions of the nucleus accumbens core and shell on
excitatory synaptic transmission in the nucleus accumbens via presynaptic appetitive pavlovian approach behavior and the potentiation of conditioned
D1-like dopamine receptors. J. Neurosci. 16, 15911604 (1996). reinforcement and locomotor activity by D-amphetamine. J. Neurosci. 19,
11. Zahm, D. S. Functional-anatomical implications of the nucleus accumbens 24012411 (1999).
core and shell subterritories. Ann. NY Acad. Sci. 877, 113128 (1999). 40. Robinson, T. E. & Kolb, B. Persistent structural modifications in nucleus
12. Zucker, R. S. Short-term synaptic plasticity. Annu. Rev. Neurosci. 12, 1331 accumbens and prefrontal cortex neurons produced by previous experience
(1989). with amphetamine. J. Neurosci. 17, 84918497 (1997).
13. Sah, P., Hestrin, S. & Nicoll, R. A. Tonic activation of NMDA receptors by 41. Robinson, T. E. & Kolb, B. Alterations in the morphology of dendrites and
ambient glutamate enhances excitability of neurons. Science 246, 815818 dendritic spines in the nucleus accumbens and prefrontal cortex following
(1989). repeated treatment with amphetamine or cocaine. Eur. J. Neurosci. 11,
14. Malenka, R. C. & Nicoll, R. A. Silent synapses speak up. Neuron 19, 473476 15981604 (1999).
(1997). 42. Rogan, M. T., Staubli, U. V. & LeDoux, J. E. Fear conditioning induces
15. Kullmann, D. M. Amplitude fluctuations of dual-component EPSCs in associative long-term potentiation in the amygdala. Nature 390, 604607
hippocampal pyramidal cells: implications for long-term potentiation. (1997).
Neuron 12, 11111120 (1994). 43. McKernan, M. G. & Shinnick-Gallagher, P. Fear conditioning induces a lasting
16. Goda, Y. & Stevens, C. F. Two components of transmitter release at a central potentiation of synaptic currents in vitro. Nature 390, 607611 (1997).
synapse. Proc. Natl. Acad. Sci. USA 91, 1294212946 (1994). 44. Moser, E. I., Krobert, K. A., Moser, M. B. & Morris, R. G. Impaired spatial
17. Oliet, S. H., Malenka, R. C. & Nicoll, R. A. Bidirectional control of quantal learning after saturation of long-term potentiation. Science 281, 20382042
size by synaptic activity in the hippocampus. Science 271, 12941297 (1996). (1998).
18. Thomas, M. J., Malenka, R. C. & Bonci, A. Modulation of long-term 45. Andersen, P., Moser, E., Moser, M. B. & Trommald, M. Cellular correlates to
depression by dopamine in the mesolimbic system. J. Neurosci. 20, spatial learning in the rat hippocampus. J. Physiol. (Paris) 90, 349 (1996).
55815586 (2000). 46. Rioult-Pedotti, M. S., Friedman, D. & Donoghue, J. P. Learning-induced LTP
19. Carroll, R. C., Beattie, E. C., von Zastrow, M. & Malenka, R. C. Role of AMPA in neocortex. Science 290, 533536 (2000).
receptor endocytosis in synaptic plasticity. Nat. Rev. Neurosci. 2, 315324 47. Hyman, S. E. & Malenka, R. C. Addiction and the brain: the neurobiology of
(2001). compulsion and its persistence. Nat. Rev. Neurosci. 2, 695703 (2001).

articles
Endogenous nicotinic cholinergic

activity regulates dopamine release
in the striatum
Fu-Ming Zhou, Yong Liang and John A. Dani
Division of Neuroscience, Baylor College of Medicine, Houston, Texas 77030-3498, USA

Correspondence should be addressed to J.A.D. (jdani@bcm.tmc.edu)
Dopamine is vital for coordinated motion and for association learning linked to behavioral reinforce-
ment. Here we show that the precise overlap of striatal dopaminergic and cholinergic fibers
underlies potent control of dopamine release by ongoing nicotinic receptor activity. In mouse
striatal slices, nicotinic antagonists or depletion of endogenous acetylcholine decreased evoked
dopamine release by 90%. Nicotine at the concentration experienced by smokers also regulated
dopamine release. In mutant mice lacking the 2 nicotinic subunit, evoked dopamine release was
dramatically suppressed, and those mice did not show cholinergic regulation of dopamine release.
The results offer new perspectives when considering nicotine addiction and the high prevalence of
smoking in schizophrenics.
Dopaminergic mechanisms of the striatum are intimately distribution of TH, ChAT and AChE was determined in coro-
involved in motor coordination, complex issues of behavioral nal sections (Fig. 1). The highest density of each enzyme was in
reinforcement, and disorders such as schizophrenia and Parkin- the striatum, extending dorsally into the olfactory tubercle.
sons disease17. The striatum receives the densest dopaminergic There is precise overlap in the distribution of the enzymes, and
innervation in the mammalian brain, which arises from neurons at the highest magnification individual cholinergic interneu-
located in the substantia nigra (SN) and ventral tegmental area rons are seen embedded within the intertwined dopamine and
(VTA) of the midbrain8,9. In addition, the striatum is densely ACh fibers (Fig. 1b).
innervated by local cholinergic interneurons10,11 that are toni- The precise overlap and close proximity of dopamine fibers
cally active and release acetylcholine (ACh)12,13. Histochemical and cholinergic enzymes led us to the following hypothesis. The
studies have indicated that nicotinic acetylcholine receptors high ChAT activity suggests ACh must be released often, and the
(nAChRs) are expressed on dopaminergic nerve terminals in the high AChE indicates that ACh must be broken down extremely
striatum14-17. Exogenous nicotinic agonists affect dopamine rapidly. The two enzymes together suggest fast nicotinic mech-
release in the striatum18,19, but the action of endogenous ACh anisms. The nAChRs are very susceptible to desensitization21,
release is not well understood. which can be avoided if the repeatedly released ACh is removed
To investigate cholinergic influence over dopamine release, rapidly by the abundant AChE.
carbon-fiber microelectrodes were placed into mice striatal brain
slices, and fast cyclic voltammetry was used to monitor the con- Depletion of ACh reduces dopamine release
centration of action-potential-dependent dopamine release in Fast-scan cyclic voltammetry with carbon-fiber microelectrodes
real time. We found that cholinergic interneurons acting via was used to monitor dopamine release on the subsecond time
nAChRs containing the 2 subunit potently regulate dopamine scale. A bipolar simulating electrode was placed in the striatum
release. Furthermore, in the concentration range experienced by about 150 m from the carbon-fiber microelectrode. Normally,
smokers, nicotine acts within the striatum and influences evoked dopamine release was electrically evoked every 2.5 min at 60%
dopamine release. Because the dopaminergic and nicotinic mech- of the maximal response. Under those conditions, the dopamine
anisms of the striatum are so intimately linked, the results have signal was stable for over two hours. The evoked dopamine con-
broad implications for understanding nicotine addiction and the centration at the tip of the carbon-fiber microelectrode was esti-
predilection for smoking by schizophrenic patients. mated to be 1.62 0.12 M (mean s.e.m., n = 29) in the dorsal
striatum, 1.46 0.13 M (n = 18) in the nucleus accumbens
RESULTS (NAc) core, and 1.44 0.13 M (n = 17) in the NAc shell. In
Dopamine and ACh fibers overlap in the striatum addition to electrically evoked dopamine release, we also mea-
There is dense expression of tyrosine hydroxylase (TH, indi- sured spontaneous dopamine release in the slice. Both the spon-
cating dopamine synthesis), choline acetyltransferase (ChAT, taneous and evoked dopamine release was prevented by 0.5 M
indicating ACh synthesis), and acetylcholinesterase (AChE, tetrodotoxin (n = 4) and by removing Ca2+ (n = 3), indicating
indicating ACh degradation) in the striatum20. The anatomical the release was Ca2+ dependent and action potential dependent.

articles
a (n = 5). Maximal inhibition of dopamine release was by

92 3.7% (n = 16) in 5 to 20 M mecamylamine (Fig. 3b). Fur-
thermore, mecamylamines inhibition of dopamine release was
not altered by atropine (n = 3).
In the absence of electrical stimulation, spontaneous,
action-potential-dependent dopamine release could be mon-
itored (Fig. 3c). When nAChRs were inhibited by mecamy-
lamine (1 or 5 M), the spontaneous dopamine release was
inhibited below our resolution (n = 6).

Microdialysis studies and measures of radioactive dopamine
release from minced striatal brain slices indicated that nicotine
b increases basal or ambient dopamine levels, and that increase was
resistant to dihydro--erythroidine (DHE) and to blockade of
action potentials by tetrodotoxin23. Fast cyclic voltammetry mea-
sures dopamine release driven by action potentials on a fast time
scale, which is different from the measures of basal dopamine
obtained from samples taken over much longer times24. Because
of the importance for addiction25, we bath-applied nicotine and
measured the effect on evoked dopamine. Under our experi-
Fig. 1. Dense and overlapping distribution of ACh and dopamine in the mental conditions, 50 nM nicotine decreased evoked dopamine
striatum. (a) Bright-field photomicrographs show TH, ChAT and AChE release by 73 4.8% (n = 4; Fig. 4a and b). The maximum inhi-
in the NAc and dorsal striatum as demonstrated by TH and ChAT anti- bition was 90.5 3.8% in 100 nM (n = 6) and 92.2 3.5% in
body staining and AChE histochemical staining. Arrows, anterior com- 500 nM (n = 5). The IC50 was estimated to be 30 nM (Fig. 4c).
missure. CC, corpus callosum, CPu, caudate putamen, NAc, nucleus In the absence of electrical stimulation, the same concentrations
accumbens, S, septum. (b) Pictures (left) taken from one slice show of nicotine inhibited spontaneous, action-potential-dependent
immunofluorescence double labeling for TH (green) and ChAT (red) at dopamine release (n = 4; Fig. 4d).
low magnification. The area in the white boxes is expanded to high mag-
nification (right), revealing dense fiber tracts and two cholinergic
interneurons. Scale bars, 50 m.
AChE inhibition reduces dopamine release
Based upon the distribution and the density of AChE in the stria-
tum (Fig. 1), we reasoned that AChE might be important for the
ongoing nAChR activity that enhances dopamine release. To test
To test for the importance of cholinergic activity, we pre- this idea, we used a potent AChE inhibitor, ambenonium26. Bath
vented endogenous ACh release by depleting ACh stores with application of 0.1 M (n = 6), 0.5 M (n = 3) or 1 M (n = 3)
()-vesamicol, a well characterized inhibitor of vesicular ACh ambenonium gradually decreased dopamine release by
transport that consequently exhausts ACh release 22 . The 90.5 2.6% (Fig. 5). The effect of ambenonium was reversed
()-vesamicol (2 M) presumably diminished ACh release and upon prolonged wash. This result demonstrates that AChE activ-
inhibited evoked dopamine release by 81 4.5% (n = 6; ity is essential for the ongoing nicotinic facilitation of dopamine
Fig. 2a and b). The maximum inhibition was 90.5 4.3% in release. By increasing extracellular ACh27, AChE inhibition may
5 M or 10 M ()-vesamicol (n = 12). The median inhibito- increase nAChR desensitization, as suggested by the results with
ry concentration (IC50) was estimated to be 1 M (Fig. 2c). bath-applied nicotine. This interpretation was supported by
These results show that intact cholinergic activity is necessary results obtained with a puffer pipette containing 1 mM ACh that
for normal dopamine release induced by action potentials. was positioned near to the carbon-fiber voltammetry electrode.
Under control conditions (without inhibition of AChE), puffs of
Nicotinic activity facilitates dopamine release ACh applied just before and during the electrical stimulation
When muscarinic acetylcholine receptors were inhibited by decreased dopamine release by 16 2.4% (n = 6). This result
1 or 2 M atropine, the evoked dopamine release slight-
ly increased by 4.8 3.2% (n = 6; Fig. 3a). In contrast,
the non-specific nicotinic antagonist mecamylamine a
(1 M) decreased evoked dopamine release by 83 4.8%
Fig. 2. Inhibition of ACh vesicular transport by vesamicol

reduces dopamine release in the striatum. (a) The three
dopamine responses were electrically evoked under control
conditions, during 2 M ()-vesamicol application, and after b c
recovery following 2 h of wash. The dopamine responses were
constructed from voltammograms that were obtained at the
rate of 10 Hz. The voltammogram (right) was obtained at the
peak of evoked dopamine response under control conditions.
The same results were obtained with or without atropine (12
M, data not shown). (b) Summary of the effect of 2 M ()-
vesamicol (n = 6). (c) The concentration dependence for inhibi-
tion of dopamine release produced by ()-vesamicol was fitted
by an Hill equation with IC50 of 1 M.

articles
Fig. 3. Nicotinic but not muscarinic ACh receptors regulate

dopamine release in the striatum. (a) The dopamine responses a
were evoked under control conditions, during 1 M atropine
application, and after recovery following a prolonged wash. The
voltammogram (right) was obtained at the dopamine peak of the
control trace. (b) The dopamine responses were evoked under
control conditions, during 1 M mecamylamine application, and
after recovery following a prolonged wash. The voltammogram
(right) is from the dopamine peak of the control trace.
b
(c) Without electrical stimulation, spontaneous dopamine

release is shown under control conditions and during application
of 5 M mecamylamine. The voltammogram was obtained at the
peak of the event indicated by the arrow.
again is consistent with desensitization of nAChRs caused by

the excess ACh. c
Mecamylamine is a non-specific nAChR inhibitor
(Fig. 3), but DHE is a specific inhibitor of 2* nAChRs28,29.
DHE decreased dopamine release by 47 5.3% (20 nM,
n = 4) and 90 5.2% (100 nM, n = 4; Fig. 6a and b), and
maximum inhibition was 91.6 4.3% (1,000 nM, n = 4). The
IC50 was estimated to be 20 nM (Fig. 6c).
The results were verified using mutant mice lacking the
2 subunit28,30. It was much more difficult to electrically evoke DISCUSSION
dopamine release from 2-null mice (Fig. 7a). The evoked Striatal dopamine and ACh fibers form an intertwined mesh-
dopamine concentration was only 0.31 0.07 M (n = 14), work that is the densest in the brain, and these fibers are associ-
which is an 80% decrease relative to wild-type littermates. Fur- ated with the densest expression of AChE8,9,11. Tonic activity of
thermore, 100 nM and 1,000 nM DHE (n = 4), 10 M mecamy- the cholinergic interneurons releases ACh12,13, and the AChE
lamine (n = 3, data not shown), and 1,000 nM nicotine (n = 6) no rapidly terminates the ACh signal. This situation optimizes ongo-
longer had any effect (Fig. 7b). Although electrically evoked ing nAChR activity by avoiding desensitization. Our results show
dopamine release was depressed by elimination of the nAChR 2 that dopamine release caused by action potentials is potently reg-
subunit, dopamine release induced by a depolarizing solution of ulated by 2* nAChR activity.
30 mM KCl produced the same dopamine signal in slices from Nicotine, at the concentrations achieved by smokers31, also
2-null mice (37.6 3.8 M, n = 4) and wild-type mice decreases action-potential-dependent dopamine release in the
(38.2 3.4 M, n = 6). This result indicates that the dopamine NAc, suggesting that nicotine is acting like an antagonist by caus-
content in 2-null mice was normal. ing desensitization32. This hypothesis is reasonable because the
The effect of nAChRs is specific to the 2 subunit because inhi- 2* nAChRs that regulate dopamine release have a high affinity
bition of 7* nAChRs had no detectable effect. In the presence of for nicotine and are readily desensitized by those concentra-
the 7*-specific inhibitor, 20 nM methyllycaconitine29, evoked tions33. The finding was unexpected, however, because older stud-
dopamine release was 98.6 2.3% of control level (n = 3). ies using in vivo microdialysis or loading of slices with
radiolabeled dopamine have shown that nicotine increases
the basal level of dopamine18,34. That increase, however, is
a at a very low concentration of dopamine and is often action
potential independent. Microdialysis is a slow process, often
taking 10 minutes per sample. The sample is collected over
a relatively large volume with probes of about 250500 m
diameter. The radiolabeled and microdialysis samples report
dopamine released from multiple sources and provide an
b c average baseline dopamine concentration that escapes re-
uptake or breakdown, giving estimates near 4 nM (ref. 36).
Fig. 4. Bath-applied nicotine reduces action-potential-dependent

dopamine release. (a) The dopamine responses were electrically
evoked under control conditions, during 50 nM nicotine applica-
tion, and after recovery following a prolonged wash. The voltam-
mogram (right) was obtained at the dopamine peak of the control
d trace. (b) Summary of the effect of 50 nM nicotine (n = 4). (c) The
concentration dependence for inhibition of dopamine release pro-
duced by nicotine was fitted by an Hill equation with IC50 of
30 nM. (d) Without electrical stimulation, spontaneous dopamine
release is shown under control conditions and during application
of 50 nM nicotine. The voltammogram was obtained at the peak of
the event indicated by the arrow.

articles
a
a
b b c
Fig. 6. A specific inhibitor of 2* nAChRs, DHE, potently reduces

Fig. 5. An AChE inhibitor, ambenonium, reduces dopamine release in evoked dopamine release. (a) The dopamine responses were evoked
the striatum. (a) The dopamine responses were evoked under control under control conditions, during 0.1 M DHE application, and after
conditions, during 100 nM ambenonium application, and after recovery recovery following a prolonged wash. The voltammogram (right) was
following a prolonged wash. The voltammogram (right) was obtained at obtained at the dopamine peak of the control trace. (b) Summary of the
the dopamine peak of the control trace. (b) Summary of the effect of effect of 0.1 M DHE (n = 5). (c) The concentration dependence for
100 nM ambenonium in 6 experiments, 3 of which lasted long enough to inhibition of dopamine release produced by DHE was fitted by an Hill
recover upon washing. equation with IC50 of 20 nM.
Fast voltammetry in our present study detects dopamine that is saliency1,7,32,34,37. The complexity of the dopamine signal is exem-
released by action potentials, and the measurement estimates plified by a recent study38. Rats learned to press a lever causing
dopamine concentration before uptake and diffusion have much intracranial self-stimulation of the midbrain dopamine areas.
effect. The voltammetry measurement is from a smaller volume, After learning the task, the rats continued self-simulations as if
with an electrode of 10 m diameter, and it detects dopamine it were pleasurable, but the self-simulations no longer increased
signals at speeds and concentrations that are indicative of direct the dopamine concentration at the target (NAc). Dopamine was
neuronal activity. These two very different measurements of released only during the initial phase while the rats were learn-
extracellular dopamine dynamics suggest that nicotine has mul- ing. Thus, even when the dopamine neurons are stimulated, other
tiple actions when analyzed in the dopamine target area of the regulatory processes can ultimately control dopamine release.
NAc. The unexpected result that nicotine strongly inhibits action- Our results identify a nicotinic cholinergic mechanism that reg-
potential-dependent dopamine release in the target has broad ulates dopamine release at the target.
implications, particularly when considering nicotine addiction. Ongoing 2* nAChRs activity in the NAc seems important
A simplification of a commonly held view of nicotine addic- for dopamine release driven by afferent action potentials, but
tion is the following: nicotine elevates dopamine in the NAc, and nicotine desensitizes those 2* nAChRs35. Although this result
that elevation reinforces continued use32. However, our under- is contrary to the simplest view of nicotine addiction, it offers a
standing of dopamine participation in reinforcement processes new clue to understand the high prevalence of smoking by schiz-
is far from complete. It is clear that dopamine in the NAc is not a ophrenic patients. Schizophrenics have impaired voluntary or
direct indication of reward. More sophisticated theories suggest sustained attention, and the positive symptoms of schizophre-
that the dopamine signal conveys novelty and/or prediction error nia, such as delusions and disorganized behavior, are associated
during the ongoing process of learning adaptive behaviors as an with an excess of dopamine in the striatum39. Schizophrenic
animal continually updates a construct of environmental patients are usually treated with neuroleptics, which inhibit spe-
cific dopamine receptors and ultimately decrease dopamine
signaling. Nicotine can transiently improve attention and
a some aspects of the positive symptoms in schizophrenic
patients 40,41. Such nicotine-induced improvements are
thought to account, at least partially, for the extraordinarily
high rate of smoking observed in schizophrenics42,43. How-
ever, it was difficult to explain why nicotine could help schiz-
ophrenics if it were increasing dopamine levels. Our finding
b
Fig. 7. 2-null mice have decreased dopamine release, and the
release is not regulated by DHE or nicotine. (a) The dopamine
responses were evoked under control conditions, during 0.1 M
DHE application, and during 1 M nicotine application. The
detected dopamine concentration was depressed in the 2-null
mice to about 0.3 M compared to about 1.5 M in WT mice. The
voltammogram (right) was obtained at the dopamine peak of the
control trace. (b) Summary data showing the lack of effect by 0.1
or 1 M DHE and by 1 M nicotine (n = 4).

articles
that nicotine decreases action-potential-evoked dopamine release 2. Breiter, H. C. et al. Acute effects of cocaine on human brain activity and
may underlie nicotines transient positive influence for schizo- emotion. Neuron 19, 591611 (1997).
3. Grace, A. A. Gating of information flow within the limbic system and the
phrenic patients. pathophysiology of schizophrenia. Brain Res. Rev. 31, 330341 (2000).
4. Kalivas, P. W. & Nakamura, M. Neural systems for behavioral activation and
METHODS reward. Curr. Opin. Neurobiol. 9, 223227 (1999).
5. Graybiel, A. M., Aosaki, T., Flaherty, A. W. & Kimura, M. The basal ganglia
Wild-type C57BL/6J (Jackson Laboratory, Bar Harbor, Maine) and and adaptive motor control. Science 265, 18261831 (1994).
2-null28 mice were used at 3 to 6 months of age. Horizontal or coronal 6. Lang, A. E. & Lozano, A. M. Parkinsons disease. N. Engl. J. Med. 339,
striatal slices (400 m) were cut on a vibratome, held at room tempera- 11301143 (1998).
ture, and studied at 30C in 125 mM NaCl, 2.5 mM KCl, 1.3 mM MgCl2, 7. Schultz, W., Dayan, P. & Montague, P. R. A neural substrate of prediction and
2.5 mM CaCl2, 26 mM Na2HPO3, 1.25 mM NaHCO3 and 10 mM glu- reward. Science 275, 15931599 (1997).
8. Anden, N. E., Fuxe, K., Hamberger, B. & Hokfelt, T. A quantitative study on
cose saturated with 95% O2 and 5% CO2. Horizontal slices appeared the nigro-neostriatal dopamine neuron system in the rat. Acta. Physiol. Scand.
more robust than coronal sections, such that spontaneous dopamine 67, 306312 (1966).
events were recorded in about 10% of horizontal slices but none in the 9. Bjrklund, A. & Lindvall, O. in Classical Transmitters in the CNS Part I (eds.
coronal slices. Animal care was in accord with Baylor College of Medi- Bjrklund, A. & Hkfelt, T.) 55122 (Elsevier, Amsterdam, 1984).
10. Butcher, L. L. & Woolf, N. J. in Classical Transmitters and Transmitter
cines animal care committee. Receptors in the CNS (eds. Bjrklund, A., Hkfelt, T. & Kuhar, M. J.) 150
Homemade carbon-fiber (P55S, Amoco Polymers, Greenville, South (Elsevier, Amsterdam, 1984).
Carolina) microelectrodes were used for fast-scan cyclic voltammetry44. 11. Woolf, N. J. Cholinergic systems in mammalian brain and spinal cord. Prog.
Stable, low-noise and sensitive electrodes were selected, especially for Neurobiol. 37, 475524 (1991).
recording spontaneous dopamine release. An Axopatch 200B amplifier 12. Aosaki, T., Kimura, M. & Graybiel, A. M. Temporal and spatial characteristics
of tonically active neurons of the primates striatum. J. Neurophysiol. 73,
and pClamp 8 (Axon Instruments, Foster City, California) were used to 12341252 (1995).
acquire and analyze data. Scans of the microelectrode potential (12 ms 13. Bennett, B. D. & Wilson, C. J. Spontaneous activity of neostriatal cholinergic
duration, 10 Hz) were from 0 mV to 400 mV to 1,000 mV to 400 mV interneurons in vitro. J. Neurosci. 19, 55865596 (1999).
to 0 mV against an Ag/AgCl reference electrode at a rate of 300 mV/ms. 14. Colquhoun, L. M. & Patrick, J. W. Pharmacology of neuronal nicotinic
acetylcholine receptor subtypes. Adv. Pharmacol. 39, 191220 (1997).
The signal that formed the voltammogram was sampled at 50 kHz, and 15. Hill, J. A. Jr., Zoli, M., Bourgeois, J. P. & Changeux, J. P. Immunocytochemical
the net current due to the electrochemical reaction was obtained by dig- localization of a neuronal nicotinic receptor: the 2-subunit. J. Neurosci. 13,
ital subtraction before and after a stimulation. The voltammograms are 15511568 (1993).
of the proper form, having oxidation current peaks between 500 to 16. Schwartz, R. D., Lehmann, J. & Kellar, K. J. Presynaptic nicotinic cholinergic
700 mV and the reduction current peaks between 230 to 330 mV. receptors labeled by [3H]acetylcholine on catecholamine and serotonin
axons in brain. J. Neurochem. 42, 14951498 (1984).
Dopamine is more than 90% of striatal monoamines45, and the voltam- 17. Jones, I. W., Bolam, J. P. & Wonnacott, S. Presynaptic localisation of the
metry signal was calibrated against fresh solutions of 110 M dopamine. nicotinic acetylcholine receptor beta2 subunit immunoreactivity in rat
Intrastriatal stimuli were delivered using bipolar tungsten electrodes nigrostriatal dopaminergic neurones. J. Comp. Neurol. 439, 235247
with resistances of about 0.5 M. The two tips of the stimulating elec- (2001).
18. Marshall, D. L., Redfern, P. H. & Wonnacott, S. Presynaptic nicotinic
trode were about 100200 m away from each other. The tip of the
modulation of dopamine release in the three ascending pathways studied by
recording electrode was about 150 m away from the two tips of the stim- in vivo microdialysis: comparison of naive and chronic nicotine-treated rats.
ulating electrode. For wild-type striatal slices, stimuli were 1 ms in dura- J. Neurochem. 68, 15111519 (1997).
tion and 26 V to achieve about 60% maximum amplitude, but the same 19. Johnson, J. H., Zhao, C., James, J. R. & Rosecrans, J. A. Individual variability
results and conclusions were obtained with higher and lower stimulus of dopamine release from nucleus accumbens induced by nicotine. Brain Res.
Bull. 51, 249253 (2000).
intensities. For 2-null striatal slices, stimuli were 1 ms in duration and 20. Descarries, L., Gisiger, V. & Steriade, M. Diffuse transmission by
612 V, and stimulation and recording sites were adjusted to obtain acetylcholine in the CNS. Prog. Neurobiol. 53, 603625 (1997).
80% maximum amplitude. 21. McGehee, D. S. & Role, L. W. Physiological diversity of nicotinic
ChAT and TH immunohistochemistry and AChE histochemistry were acetylcholine receptors expressed by vertebrate neurons. Annu. Rev. Physiol.
adapted from published methods4648. For immunohistochemistry, brains 57, 521546 (1995).
22. Prior, C., Marshall, I. G. & Parsons, S. M. The pharmacology of vesamicol: an
were fixed in 0.1 M phosphate buffer containing 4% (w/v) paraformalde- inhibitor of the vesicular acetylcholine transporter. Gen. Pharmacol. 23,
hyde and 14% (v/v) picric acid. Coronal sections (30 m) were cut on a 10171022 (1992).
cryostat, incubated with either goat anti-ChAT or rabbit anti-TH anti- 23. Sacaan, A. I., Dunlop, J. L. & Lloyd, G. K. Pharmacological characterization
bodies, followed by peroxidae-conjugated antibodies, and were finally of neuronal acetylcholine gated ion channel receptor-mediated hippocampal
norepinephrine and striatal dopamine release from rat brain slices.
treated with chromogen diaminobenzidine. For ChAT and TH double J. Pharmacol. Exp. Ther. 274, 224230 (1995).
immunofluorescence, after treatment with the primary antibody mix- 24. Wu, Y., Pearl, S. M., Zigmond, M. J. & Michael, A. C. Inhibitory glutamatergic
ture, the secondary antibody mixture containing Cy2-conjugated don- regulation of evoked dopamine release in striatum. Neuroscience 96, 6572
key anti-rabbit and rhodamine-conjugated donkey anti-goat IgG antisera (2000).
was used. For AChE staining, sections were incubated for 15 h in 100 ml 25. Dani, J. A. & Heinemann, S. Molecular and cellular aspects of nicotine abuse.
Neuron 16, 905908 (1996).
solution containing 50 mM sodium acetate, 4 mM copper sulfate, 16 mM 26. Hodge, A. S., Humphrey, D. R. & Rosenberry, T. L. Ambenonium is a rapidly
glycine, 116 mg S-acetylthiocholine iodide and 3 mg ethopropazine. The reversible noncovalent inhibitor of acetylcholinesterase, with one of the
sections were then rinsed and developed with 1% sodium sulphide. All highest known affinities. Mol. Pharmacol. 41, 937942 (1992).
images were captured with a digital camera and processed in Adobe Pho- 27. Vinson, P. N. & Justice, J. B. Jr. Effect of neostigmine on concentration and
extraction fraction of acetylcholine using quantitative microdialysis.
toShop. J. Neurosci. Methods 73, 6167 (1997).
28. Xu, W. et al. Multiorgan autonomic dysfunction in mice lacking the beta2
ACKNOWLEDGEMENTS and the beta4 subunits of neuronal nicotinic acetylcholine receptors.
J. Neurosci. 19, 92989305 (1999).
We thank A. Beaudet and A. Orr-Urtreger for providing the mutant mice, and
29. Alkondon, M. & Albuquerque, E. X. Diversity of nicotinic acetylcholine
Y. Schmitz and D. Sulzer for advice with the techniques. The work was supported receptors in rat hippocampal neurons. I. Pharmacological and functional
by the National Institute on Drug Abuse (DA09411 and DA12661), the evidence for distinct structural subtypes. J. Pharmacol. Exp. Ther. 265,
14551473 (1993).
National Institute of Neurological Disorders and Stroke (NS21229) and the 30. Picciotto, M. R. et al. Acetylcholine receptors containing the 2 subunit are
NARSAD (to F.M.Z.). involved in the reinforcing properties of nicotine. Nature 391, 173177
(1998).
31. Henningfield, J. E., Stapleton, J. M., Benowitz, N. L., Grayson, R. F. &
RECEIVED 5 SEPTEMBER; ACCEPTED 29 OCTOBER 2001 London, E. D. Higher levels of nicotine in arterial than in venous blood after
cigarette smoking. Drug Alcohol Depend. 33, 2329 (1993).
32. Dani, J. A., Ji, D. & Zhou, F. M. Synaptic plasticity and nicotine addiction.
1. Berke, J. D. & Hyman, S. E. Addiction, dopamine, and the molecular Neuron 31, 349352 (2001).
mechanisms of memory. Neuron 25, 515532 (2000). 33. Fenster, C. P., Rains, M. F., Noerager, B., Quick, M. W. & Lester, R. A.

articles
Influence of subunit composition on desensitization of neuronal 41. Rezvani, A. H. & Levin, E. D. Cognitive effects of nicotine. Biol. Psychiatry 49,
acetylcholine receptors at low concentrations of nicotine. J. Neurosci. 17, 258267 (2001).
57475759 (1997). 42. Adler, L. E. et al. Schizophrenia, sensory gating, and nicotinic receptors.
34. Di Chiara, G. Role of dopamine in the behavioural actions of nicotine related Schizophr. Bull. 24, 189202 (1998).
to addiction. Eur. J. Pharmacol. 393, 295314 (2000). 43. Dalack, G. W., Healy, D. J. & Meador-Woodruff, J. H. Nicotine dependence in
35. Mansvelder, H. D. & McGehee, D. S. Long-term potentiation of excitatory schizophrenia: clinical phenomena and laboratory findings. Am. J. Psychiatry
inputs to brain reward areas by nicotine. Neuron 27, 349357 (2000). 155, 14901501 (1998).
36. Parsons, L. H. & Justice, J. B. Jr. Extracellular concentration and in vivo 44. Kawagoe, K., Zimmerman, J. B. & Wightman, R. M. Principles of
recovery of dopamine in the nucleus accumbens using microdialysis. voltammetry and the microelectrode surface states. J. Neurosci. Methods 48,
J. Neurochem. 58, 212218 (1992). 225240 (1993).
37. Balfour, D. J., Wright, A. E., Benwell, M. E. & Birrell, C. E. The putative role of 45. Schenk, J. O., Miller, E., Rice, M. E. & Adams, R. N. Chronoamperometry in
extra-synaptic mesolimbic dopamine in the neurobiology of nicotine brain slices: quantitative evaluations of in vivo electrochemistry. Brain Res.
dependence. Behav. Brain Res. 113, 7383 (2000). 277, 18 (1982).
38. Garris, P. A. et al. Dissociation of dopamine release in the nucleus accumbens 46. Chang, H. T. Dopamine-acetylcholine interaction in the rat striatum: a dual-
from intracranial self-stimulation. Nature 398, 6769 (1999). labeling immunocytochemical study. Brain Res. Bull. 21, 295304 (1988).
39. Abi-Dargham, A. et al. Increased baseline occupancy of D2 receptors by 47. Koelle, G. G. & Friedenwald, J. S. A histochemical method for localizing
dopamine in schizophrenia. Proc. Natl. Acad. Sci. USA 97, 81048109 (2000). cholinesterase activity. Proc. Soc. Exp. Biol. Med. 70, 617622 (1949).
40. Adler, L. E., Hoffer, L. D., Wiser, A. & Freedman, R. Normalization of 48. Schwartz, M. L. & Mrzljak, L. Cholinergic innervation of the mediodorsal
auditory physiology by cigarette smoking in schizophrenic patients. Am. thalamic nucleus in the monkey: ultrastructural evidence supportive of
J. Psychiatry 150, 18561861 (1993). functional diversity. J. Comp. Neurol. 327, 4862 (1993).

articles
Differential synaptic processing

separates stationary from transient
inputs to the auditory cortex
Marco Atzori1, Saobo Lei1, D. Ieuan P. Evans1, Patrick O. Kanold2, Emily Phillips-Tansey1,
Orinthal McIntyre1 and Chris J. McBain1
1 LCMN/NICHD/NIH, Rm 5A72, Bldg 49, Convent Drive, Bethesda, Maryland 20892-4495, USA
2 Department of Biomedical Engineering, Johns Hopkins University, Baltimore, Maryland 21205, USA
Correspondence should be addressed to C.J.McB. (chrismcb@codon.nih.gov)
Sound features are blended together en route to the central nervous system before being discriminated
for further processing by the cortical synaptic network. The mechanisms underlying this synaptic
processing, however, are largely unexplored. Intracortical processing of the auditory signal was investi-
gated by simultaneously recording from pairs of connected principal neurons in layer II/III in slices
from A1 auditory cortex. Physiological patterns of stimulation in the presynaptic cell revealed two pop-
ulations of postsynaptic events that differed in mean amplitude, failure rate, kinetics and short-term
plasticity. In contrast, transmission between layer II/III pyramidal neurons in barrel cortex were
uniformly of large amplitude and high success (release) probability (Pr). These unique features of audi-
tory cortical transmission may provide two distinct mechanisms for discerning and separating transient
from stationary features of the auditory signal at an early stage of cortical processing.
Sensory cortices perform sophisticated pattern recognition tasks, kinetics and short-term plasticity. In contrast, identical experi-
elaborating the thalamic input via a variety of functionally dif- ments done between layer II/III pyramidal cells in barrel cortex
ferent synapses across the six layers. The role of the auditory cor- reveal a uniformly homogeneous population of connections with
tex in processing and modifying auditory signals, however, is large amplitude EPSCs and high release probability, indicating
largely unknown. Thalamic inputs to primary auditory cortex that the properties of layer II/III auditory cortex connections do
(A1) show several types of responses to acoustic stimuli, which not represent two populations of connections common to other
fall into two broad classes of transient and sustained firing pat- sensory cortical areas. We propose that these two synaptic net-
terns13. Sustained responses saturate and can even be inhibited works implement two spectrally different tasks: the low proba-
at increasing sound pressure level (SPL), whereas transient bility synapses support ensemble-encoded narrow-band,
responses do not saturate or decrease at increasing SPL, suggest- long-lasting input, and the high probability synapses serve as reli-
ing differential mechanisms of synaptic processing. How these able event detectors and wide-band signal analyzers.
afferent signals are processed by downstream elements is unknown
and may rely on distinct intrinsic conductances, synaptic inputs, RESULTS
ensemble coding or a combination of these or other factors. We recorded from over 100 pairs of layer II/III pyramidal neu-
Neurons within layer II/III of the auditory cortex receive direct rons, which resulted in 37 synaptic connections in 35 pairs (two
thalamic input in addition to inputs from both layer IV spiny stel- were reciprocally connected). All recordings were made at room
late neurons and other layer II/III neurons46. Output from layer temperature (22C) with the presynaptic cell held under current
II/III pyramidal neurons project to local layer V and layer II/III clamp and postsynaptic cell held under voltage clamp. Synaptic
in neighboring and contralateral cortical areas, feeding all subse- currents evoked by single action potentials in the presynaptic cell
quent stages of auditory signal analysis. Anatomical studies have (stimulus frequency, 0.06 Hz) were blocked by the AMPA recep-
identified pyramidal cells in layer II/III of the auditory cortex with tor antagonist DNQX (10 M, n = 6), confirming that connec-
different types of axonal recurrent collateral patterns4,5,7 raising tions were glutamatergic in nature. Synaptic events were insensitive
the possibility of differential functional interactions within audi- to exogenous application of philanthotoxin (2 M, n = 11, data
tory columns8. We wanted to test the hypothesis that such intrin- not shown), indicating that synapses contained predominantly
sic connections had the potential for functionally differentiating GluR2-containing, Ca2+-impermeable AMPA receptors914.
between either transient or sustained firing patterns. Here we In 24 of 37 connections, transmission occurred with a low
demonstrate the existence of two modes of transmission between probability (weak connections), with most presynaptic action
pyramidal neurons of layer II/III of the auditory cortex using potentials failing to produce detectable postsynaptic events
paired whole-cell recordings. These two modes of transmission (Figs. 1a and 2a). The remaining 13 connections demonstrat-
differed with respect to their success probability, EPSC amplitude, ed synaptic transmission with a higher success probability with

articles
Fig. 1. Two layer II/III corticocorti-

a b c cal excitatory connection pheno-
types. (a) Representative example of
a weak connection synapse. Top,
single traces showing EPSCs evoked
by paired-pulse stimulation (stimuli
separated by 50 ms). Average
(n = 50) EPSCs and presynaptic
spikes are in red. Bottom, amplitude
histograms from 50 trials for the

first and second synaptic responses.
Synaptic failures and events are
reported in black and red, respec-
tively. (b) Representative example of
a strong connection synapse. Setup
for figure is the same as for (a).
Note scale differences between
(a) and (b). Bottom, EPSC ampli-
tude histograms from 70 trials.
(c) Camera lucida drawing of the
corresponding neurons used in
panel (b). Cell bodies were located
within layer III, whereas dendrites
(black) projected deep into layers
IIV. The axon (indicated in red)
made many local collaterals before
projecting to layer V.
only a small number of failures of transmission (strong dataset (Fig. 2a) represented a single population with a broad Pr
connections; Figs. 1b and 2a). distribution, data were fit with homogeneous, single Gaussian
A plot of the probability of transmission success (Pr = 1 fail- and single Poissonian distributions. None of these fits gave accept-
ures) between all connections revealed a histogram composed of able 2 values. In contrast, when the dataset were fit by the sum
two distinct clusters (Fig. 2a). One cluster, which we defined as of two Gaussians or the sum of two first-order Poissonian dis-
low-probability connections (LPCs), had Pr values
less than 0.4 (mean s.e.m., 0.13 0.02). We defined
the second cluster, which had Pr values greater than a
0.5 (mean s.e.m., 0.68 0.02), as high probability
connections (HPCs). To determine whether the Pr
Fig. 2. Differing properties of HPCs and LPCs. (a) Left, his-

togram plots of release probability (Pr1) for synaptic
responses in all connections (n = 37). The discontinuous
distribution reveals two modes of transmission, one with
low probability and one with high probability (24/37 and
13/37 connections, respectively). The largest population
b
comprised those with a high initial release probability
(HPC, open columns), with mean success probability (Pr) of
0.68 (right). The second class, low probability connections
(LPC, black columns) had a mean probability across con-
nections of 0.13. The Pr distribution was best fit by the sum
of two Poissonian distributions (continuous line). Right
three panels, mean values for the release probability, EPSC
amplitudes measured in the absence of event failures (Anf)
and total EPSC amplitude (A, measurement including fail-
ures) of the two groups. (b) 1090% rise time (left) and
decay time constant (, right) versus Pr1. Individual experi-
c
ments are shown by small symbols. The mean values across
all data are indicated by large symbols. LPC, triangles; HPC,
circles. (c) Left, paired-pulse ratio A2/A1 versus success
probability of the first EPSC. All but two HPCs demon-
strated paired-pulse depression, whereas LPCs could show
either depression or facilitation. Right, CV2 analysis sug-
gests that presynaptic mechanisms (indicated by hatched
areas) largely determine the response to paired-pulse stim-
uli. In this figure, error bars represent standard deviation to
allow better resolution of the data scatter.
articles
Fig. 3. Short-term plasticity at LFCs

and HFC connections. High- and low-P
a b
connections demonstrate significantly
different modes of short-term plasticity
in response to trains of action poten-
tials. Eight representative individual
records in response to 20-Hz trains of
presynaptic action potentials (lower
traces) in LPCs (a) and HPCs (b). The
mean response is shown below in bold

for both types of connections.
Amplitude histograms for each of the
four synaptic responses in the spike
train (n = 100 events for each histogram
for both the LPC and the HPC) are
shown below the individual traces.
LPCs had smaller but stationary ampli-
tudes during the course of the stimulus
train. In contrast, HPCs possessed a
larger amplitude response whose mean
c d e
amplitude shifted to lower values as the
train progressed. The number of fail-
ures increased during the stimulus train.
(ce) Average EPSC amplitude, proba-
bility and non-failure amplitude for all
connections within the two groups
(n = 13 for HPCs and n = 24 for LPCs).
Whereas the HPCs showed a reduction
in mean amplitude, probability and non-
failure amplitude during the spike
train, LPC parameters were constant
throughout the stimulus train. Solid lines are single exponential fits of strated either facilitation or depression in response to paired-
the experimental data. The numeric values for are 31 4 ms, pulse stimuli. The paired-pulse ratio (PPR) from all connec-
56 28 ms, and 80 11 ms for (c), (d) and (e), respectively. tions ranged from 0.18 to 5.3 (Fig. 2c). However, across all
recordings, the mean amplitude of the second EPSC was not
significantly different from the first EPSC amplitude
tributions, the 2 values (5.7 and 0.8, respectively) were well with- (EPSC1 = 1.1 0.5 pA; EPSC2 = 1.1 0.2 pA). In contrast, all
in the 20.05 range (<11.1 and <14.1, respectively). This suggests but two of the HPCs displayed strong depression
that the data do not represent an evenly distributed single pop- (EPSC1 = 6.9 1.9 pA; EPSC2 = 3.5 0.8 pA; Fig. 2c), with a
ulation of synaptic connections, but rather, indicate that HPC mean PPR in the range of 0.231.13. The ratio Pr2/Pr1 (Pr1,2,
and LPCs reflect synapses with two distinct modes of transmis- success probability of the first and second EPSC) also differed
sion between layer II/III pyramidal neurons. Throughout the between the two connection types: Pr2/Pr1 was 1.28 0.17 for
remainder of this report, all data were obtained from 24 cell pairs LPCs, and 0.78 0.06 for HPCs. Evidence that presynaptic
for LPCs and 13 cell pairs for HPCs, unless specified otherwise. mechanisms determine the observed response to paired-pulse
stimuli in both LPCs and HPCs was suggested by CV2 (ampli-
Biophysical and paired-pulse properties tude mean/variance) analysis16 (Fig. 2c).
We next wanted to determine whether LPCs and HPCs differed We next determined whether any of the measured parame-
in properties other than their release probability. We observed ters, Pr, EPSC amplitude, EPSCnf or the paired pulse ratio, were
no obvious differences in the resting membrane potentials correlated with the developmental age range used throughout
(LPCs, V r = 59.6 1.5 mV; HPCs, 61.3 1.6 mV), input our experiments (postnatal day 1828). None of these parameters
resistance of the postsynaptic cells (LPCs, Rinput = 228 38 M, were correlated with the postnatal age of animals (Pr versus age,
n = 11; HPCs, 243 53 M, n = 6), or latency of EPSC onset r = 0.16; mean amplitude, r = 0.31; EPSCnf, r = 0.3; paired-pulse
(LPCs, mean latency, 2.9 0.3 ms, n = 19; HPCs, 2.3 0.2, ratio, r = 0.09; data not shown). Similarly, no correlation exist-
n = 11). Analysis of EPSC kinetics revealed that the 1090% rise ed between the intersomatic distances between the pairs of layer
time differed between the two populations (LPCs, 1.5 0.2 ms, II/III pyramidal cells chosen for study (intersomatic distances
n = 19; HPCs, 2.0 0.3 ms, n = 13, p < 0.01) but not the decay measured from post hoc anatomical analysis) and the Pr (r = 0.04).
time constant (LPCs, 8.4 0.7 ms, n = 19; HPCs, 10.3 1.1 ms,
n = 13, p = 0.16; Fig. 2b). We also observed significant differ- Short-term plasticity in response to spike trains
ences between the mean EPSC amplitudes including failures Units in A1 recorded in vivo have an upper firing frequency limit
(LPCs, A = 1.1 0.5 pA; HPCs, A = 6.9 1.9 pA) and exclud- between 40 and 60 Hz17, and cortical integration times are of the
ing failures (LPCs, EPSC nf = 4.7 0.9 pA, n = 22; HPCs, order of 20150 ms18. Moreover, in vivo cortical units respond to
9.3 2.6 pA, n = 13, p = 0.046; Fig. 2a). repeated auditory stimuli up to frequencies of about 20 Hz19. There-
We next determined whether the paired-pulse ratio (PPR) fore, trains of spikes 150 ms long at 20 Hz (repeated 30150 times
of synaptic transmission in response to two closely time action at 15-second intervals, to allow complete recovery from depression
potentials (50 ms) in the presynaptic cell were similar for LPCs or facilitation) were selected to simulate physiological stimuli. The
and HPCs. At LPCs, the amplitude of the second EPSC demon- postsynaptic response to the presynaptic spike train markedly dif-

articles
Fig. 4. Time course of recovery from short-term depression at HPCs.

a To determine the time course of recovery from depression in HPCs,
experiments similar to those shown in Fig. 4 were repeated, but the ini-
tial 20-Hz train was followed by stimuli 1 and 2 s after the termination
of the train. (a) Example of HPC activity at 3 time points during the
train and 2 s following termination of the train. Each panel represents 10
traces evoked by single action potentials (bottom). Left and middle
traces represent the first and fourth events in the train and illustrate the
significant depression of the event amplitude. Right, recovery of the
event amplitude at a 2-s time point after termination of the train.

(b, c) The mean time course of the recovery for the EPSC amplitude
and probability, respectively (n = 6 connections).
b c
Time course of recovery from short-term depression

To determine the time course of recovery from short-term depres-
sion in HPCs, experiments similar to those described above were
repeated, but the initial 20 Hz train was then followed by stimuli
12 seconds after the train. In 6 HPCs (Fig. 4), the depression of
the EPSC mean amplitude occurring during the 20-Hz
conditioning train did not fully recover until about 2 seconds
after termination of the 20-Hz spike train (EPSC1 = 6.9 0.8 pA,
EPSC 4 = 2.9 1.3 pA, EPSC 1second = 2.5 0.5 pA,
EPSC2seconds = 5.5 1.3 pA; Fig. 4b and c). The time course for
amplitude recovery did not parallel that of the recovery of Pr
fered between LPCs and HPCs. On average, LPCs displayed no (Fig. 4c) (mean PrEPSC1 = 0.70 0.05, PrEPSC4 = 0.41 0.04,
change in mean amplitude, Pr or EPSCnf of any EPSCs occurring PrEPSC1second = 0.42 0.07, PrEPSC2seconds = 0.53 0.10). These
during the train (Fig. 3a, ce). In contrast, HPCs displayed depres- data suggest that two distinct mechanisms with distinct kinetics
sion of all three parameters when comparing the fourth to the first may underlie recovery from short-term depression.
event in the train (Fig. 3be). In all cases, the observed amplitude
(or probability) depression was best described by a single expo- [Ca2+]o-dependence of short-term plasticity
nential function (Fig. 3ce). Despite the observation of significant Several features of synaptic transmission, such as release proba-
short-term depression of transmission during trains of stimuli at bility and the degree of depression or facilitation in response to
HFCs, the mean EPSC amplitude remained significantly greater
than that of the corresponding LFCs (Fig. 3c). The mean value of
the fourth EPSC was 1.1 0.2 pA for LPCs, and 2.7 0.7 pA for
HPCs. Similarly, despite the decrease in Pr during successive events
in the train, Pr4 (success probability of the fourth response) in HPCs
remained significantly higher than that of the LPCs (HPCs,
Pr4 = 0.5 0.1; LPCs, 0.1 0.2; Fig. 3d). The stable response of
a b
LPCs is qualitatively different from the depressing response observed
in HPCs, and suggests that HPCs are not simply the numerical sum-
mation of multiple LPCs and that each represents a distinct mode
of transmission between auditory cortex pyramidal cells.
Fig. 5. Differential dependence of short-term plasticity on [Ca2+]o in

HPCs. (ad) Representative experiment demonstrating the effect of
reducing [Ca2+]o on short-term plastic properties. (e, f) Mean data
from five experiments at high-probability connections. (ad) Lowering
[Ca2+]o from 1.5 to 0.5 mM decreased both the mean amplitude and the
success probability of EPSCs occurring early in the train. In contrast, the c d
amplitudes of synaptic responses occurring later in the train were not
significantly different from those recorded in 1.5 mM [Ca2+]o (mean
change in EPSC4 amplitude, 5 2%). (c, e) This redistribution of EPSC
amplitude acts to minimize the short-term depression observed in con-
trol [Ca2+]o. The influence of [Ca2+]o on short-term plasticity is more
clearly observed when the two data sets are normalized to their
respective first EPSC amplitudes (e, inset), which clearly illustrates that e f
the depressing response of HPCs in control [Ca2+]o is converted to a
stationary response in low [Ca2+]o. (d) A comparison of the Pr in the
two [Ca2+]o conditions suggests that a reduction of Pr1 and Pr2 con-
tributes to the reduction of EPSC1 and EPSC2 amplitude observed in
panel (c). (f) Examination of the mean Pr data reveals that in general, the
largest reduction in Pr was associated with Pr1.
articles
Fig. 6. Differential dependence of short term plasticity on

[Ca2+]o in LPCs. Greater than twofold elevation in [Ca2+]o fails
to convert the stationary response of LPC short-term plastic- a b
ity into a depressing one. Figure setup is identical to Fig. 5. At
LPC connections, an elevation of [Ca2+]o from 1.5 to 3.5 mM
results in an increase in the mean amplitude and Pr of EPSCs
occurring early in the train (cf). However, this increase in
EPSC amplitude or Pr was not accompanied by any alteration of
the short-term plastic properties of the synapse (e, inset).
trains of stimuli, are tightly regulated by the extracellu-

lar calcium ion concentration ([Ca2+]o)20,21. For exam-
ple, at connections between hippocampal pyramidal
neurons, the degree of paired-pulse depression increas-
es with elevation of [Ca2+]o22, indicating that short-
term plasticity is strongly influenced by the initial
presynaptic release probability. To determine whether c d
the differential response to brief trains of stimuli of
HPCs and LPCs resulted from differences in the initial
Pr, we next examined the role of [Ca2+]o in shaping the
synaptic response to short trains of presynaptic activity.
Specifically, we wanted to determine whether the pat-
tern of short-term plasticity at HPCs could be convert-
ed to activity characteristic of LPCs by simply decreasing
[Ca2+]o, and vice versa. e f
In 5 HPC recordings, [Ca2+]o was decreased from 1.5
to 0.5 mM (whereas [Mg2+]o was raised from 1.5 to 2.5
mM). Such a reduction in Ca2+ and elevation of Mg2+
allowed us to manipulate the amplitude and P r of
EPSC1 to approximate LPCs. The reduction in [Ca2+]o
reduced the mean EPSC 1 amplitude by 40 8%
(Fig. 5), but had little effect on events occurring later in
the train (mean change in EPSC4 amplitude, 5 2%;
Fig. 5c and e). This redistribution of EPSC amplitude
reduced the short-term depression observed in control [Ca2+]o: in like short-term plasticity in elevated [Ca2+]o conditions. These
low-[Ca2+]o, EPSC1 and EPSC4 amplitudes were not significant- data suggest that although transmission in LPCs is sensitive to
ly different (EPSC1 = 7.3 0.8 pA versus EPSC4 = 5.4 1.2 pA, changes in [Ca2+]o, LPCs could not be converted to HPCs by
p = 0.28). The short-term plastic properties of HPCs in low [Ca2+]o simply changing their Pr.
were similar to those observed at LPCs in 1.5 mM [Ca2+]o (com-
pare with Fig. 3c). This was more clearly illustrated when the two Equivalent transmission is not observed in barrel cortex
data sets were normalized to their first EPSC amplitudes (Fig. 5e, To determine whether LPCs and HPCs were a general property
inset), which showed that a depressing response at HPCs is con- of synapses between layer II/III pyramidal neurons in other sen-
verted to a stationary response in lowered [Ca2+]o. sory cortical areas, similar recordings were made from pyramidal
Lowering [Ca2+]o significantly reduces the Pr of EPSC1 and neurons in barrel cortex slices23 across an identical age range.
EPSC2 (Fig. 5a and b). This reduction of Pr1 and Pr2 likely con- Synaptic transmission between connected pairs of layer II/III pyra-
tributed to the smaller EPSC1 and EPSC2 amplitude (Fig. 5c). midal neurons from barrel cortex possessed properties funda-
Across all experiments, the largest reduction in Pr of all events in mentally different from equivalent neurons in auditory cortex
the train was associated with P r1 (mean P r1low[Ca2+]o/mean (Fig. 7). From a total of 18 connected pairs, the mean EPSC ampli-
Pr1ctrl[Ca2+]o = 0.46 0.15, Fig. 5f). This reduction in Pr of events tude in response to single presynaptic stimulation was
early in the train endows HPC synapses with transmission prop- 19.6 1.8 pA (n = 18), almost 3 times greater than the mean
erties reminiscent of LPCs. amplitude of connections between HPC in auditory cortex and
In contrast, the pattern of short-term plasticity at LPCs was about 20 times greater than LPC amplitudes (compare Figs. 1
not altered by elevating [Ca2+]o. In 5 experiments, [Ca2+]o was and 2). The probability of synaptic transmission was also signifi-
elevated from 1.5 to 3.5 mM (and [Mg2+]o was decreased to cantly higher at barrel cortex connections (mean Pr = 0.93 0.01,
0.5 mM). Under these conditions, we observed an increase in n = 18; Fig. 7e) with few failures of synaptic transmission occur-
the mean amplitude and Pr of EPSCs occurring early in the train ring (Fig. 7a, b, d; compare with HPC mean P r = 0.68).
(Fig. 6; mean increase in EPSC1 amplitude, 160 98% of con- Consequently, the mean EPSC amplitude, excluding failures
trol). However, this increase in EPSC amplitude did not alter (mean, 20.9 1.9 pA) was significantly greater than observed in
the pattern of short-term plasticity at these synapses (that is, either auditory cortex HPC (9.3 pA) or LPC (4.7 pA).
we were unable to convert a stationary response to a depress- Responses to short trains of presynaptic stimuli were invari-
ing one; Fig. 6c and e and inset, compare Fig. 5c and e). Indeed, ably depressing (Fig. 7) with the mean EPSC amplitude, mean
on no occasion did we observe a LPC that demonstrated HPC- EPSC amplitude excluding failures and Pr all showing signifi-

articles
Fig. 7. Equivalent connections were not observed in barrel

a cortex. Experiments identical to those illustrated in Fig. 3
were made between pairs of connected neurons recorded
from layer II/III pyramidal neurons in barrel cortex.
(a) Representative single experiment and panels (bd) illus-
trate the mean pooled data from 18 connected pairs. (ac) In
all cases, connections between layer II/III pyramidal neurons
possessed large EPSC amplitudes far in excess of those
observed in the layer II/III auditory connection. (a) Seven
sample traces of synaptic activity in response to trains of

20-Hz presynaptic action potentials (bottom). An average of
30 traces (third from bottom) illustrates that this connection
possessed short-term depression in response to the 20-Hz
train. In the presence of the AMPA receptor antagonist,
DNQX, all transmission was blocked, indicating the gluta-
matergic nature of this connection. Right, amplitude his-
tograms constructed from cell shown on left. Failures are
indicated by open columns and successes by solid columns.
(be) Mean data from 18 connections. In all cases, the EPSC
amplitude (b) and EPSC amplitude excluding failures (c)
b d showed strong depression in response to the train of 20-Hz
stimuli. (d, e) The mean initial release probability of connec-
tions in barrel cortex was extremely high, with little evidence
for synaptic failures (compare with Figs. 13). The mean Pr
decreased during the train of presynaptic stimuli.
(e) Distribution of Pr1 from all 18 connections show that all
connections possess a uniformly high Pr1. Solid line is fit by
single Gaussian function.
c e
the incidence of synchronous events being similar in
LPCs (58%) and HPCs (64%). In contrast, only 18% of
non-connected pairs received synchronous spontaneous
synaptic activity (Fig. 8d). These data suggest the exis-
tence of synaptically connected ensembles whose com-
ponents may be activated in concert24.
cant depression during the 20-Hz train of stimuli (Fig. 7bd). DISCUSSION
Despite demonstrating significant reduction in the mean Pr dur- Here we provide functional evidence, based on differences in
ing the course of the stimulus train (Pr1 = 0.93 0.01 versus EPSC amplitude, success probability, kinetics and response to
Pr4 = 0.70 0.03, n = 18), Pr4 was still significantly greater in brief trains of stimuli, for two different modes of transmission
barrel cortex that Pr4 in HPC auditory cortex (0.5). Similarly, between excitatory connections of the primary auditory cortex7.
the mean value of the EPSC4 was 10.9 1.3 pA (n = 18) in bar- The most frequently observed transmission mode, the LPC, was
rel cortex compared to 1.1 pA and 2.7 pA for LPC and HPC in associated with small mean EPSC amplitude, a low event success
auditory cortex. Taken together, these data demonstrate that probability, and on average, a non-decremental response to 20-Hz
connections between layer II/III pyramidal neurons in barrel spike trains. Transmission via HPCs was characterized by a high
cortex possess properties distinct from those observed in audi- Pr, a larger mean EPSC amplitude, paired-pulse depression, and
tory cortex. Moreover, these data suggest that HPC and LPCs marked short-term depression in response to 20-Hz spike trains.
do not represent a homogeneous property of synaptic connec- The differential response of LPCs and HPCs to short trains of
tions common to other sensory cortical areas. presynaptic stimuli represented the most significant difference
between the two modes of transmission. At LPCs, the mean EPSC
A high level of connectivity within layer II/III amplitude and Pr remained constant at each successive event in
In recordings from auditory cortex, spontaneous excitatory the train. In contrast, both mean event amplitude and Pr decreased
synaptic activity often occurred synchronously on both the presy- during the 20-Hz train at HPCs. This suggests that whereas the
naptic and postsynaptic neuron, suggesting that common sources release machinery at LPCs can faithfully support brief trains of
may innervate layer II/III neurons. We next determined whether presynaptic stimuli, the decrease in both EPSC amplitude and Pr
the incidence of such synchronous synaptic activity preferential- at HPCs favors their involvement in transient episodes. It is uncer-
ly occurred on pairs of cells that were connected by LPC or HPCs tain at this time whether LPCs and HPCs represent two distinct
compared to pairs of cells that were not synaptically connected. classes of connection or represent the extreme ends of a contin-
Correlated spontaneous EPSCs were detected primarily in uum ranging from low- to high-probability connections. Because
recordings from connected cells (Fig. 8). The synchronous nature a uniform distribution for Pr between cortical pyramidal neurons
of spontaneous events detected in cell pairs was confirmed by the in layer V was reported previously15, we considered the possibil-
presence of a sharp peak in cross-correlograms (Fig. 8c). The per- ity that the clustering of LPCs and HPCs resulted from the finite
centage of cells displaying synchronous spontaneous currents was sample size (n = 37). We considered this unlikely; were the data
higher in connected cells compared to non-connected cells, with from a single population of connections, possessing a uniform

articles
Fig. 8. Synchronous spontaneous synaptic input onto connected layer II/III

cells. (a) An example of two connected cells (LPC-type). Lower traces a b
indicate presynaptic action potentials; upper traces represent four repre-
sentative trains of postsynaptic EPSCs. (b) The same two cells display
numerous synchronous spontaneous EPSPs (upward deflections)/EPSCs
(downward deflections) that originate from unidentified sources (indi-
cated by arrows). (c) A zero lag in the cross correlation analysis confirms
the synchronous nature of the co-occurring synaptic events. (d) Summary
histogram showing that the fraction of cells showing simultaneous sponta-
neous EPSC is high in both HPCs and LPC connected pairs but lower in
recordings of two cells showing no connectivity.
c d
distribution of Pr, the probability of observing such a gap in the
dataset within a single population would be extremely low
(p < 0.0001). In addition, Pr frequency histograms (Fig. 2a) were
well fit by the sum of two first-order Poissonian distributions but
poorly fit by either a single Poissonian or Gaussian distribution.
Thus, we consider it likely that LPCs and HPCs represent specif-
ic populations of synapse types whose transmission properties are
tuned for specific roles in auditory cortex. This is underscored by tory responses. Future experiments are aimed at determining
the experiments from equivalent connections in barrel cortex. the role of inhibitory neurons within the circuitry.
Transmission between layer II/III pyramidal neurons in barrel In conclusion, pyramidal cells within layer II/III of the audi-
cortex occurred via synapses that had high release probability tory cortex process incoming presynaptic signals using two
(mean Pr = 0.93) and large EPSC amplitudes, and that invariably synaptic networks with unique computational features. HPCs
demonstrated short-term depression of both EPSC amplitude and and LPCs networks in layer II/III provide a compelling synap-
Pr in response to brief trains of stimuli. tic substrate for performing spectrally heterogeneous demands
of auditory cortical analysis, such as event detection and sound
Physiological function of HPCs and LPCs fingerprinting26, possibly related to the what and where pro-
In vivo recordings from thalamus show that a limited number of cessing in the auditory pathway27.
firing patterns exist in response to sustained tones or repetitive
clicks. The response to tone stimulation, measured as mean firing METHODS
rate, is either sustained or transient in nature2. Similarly, the two Auditory cortex slices. C57/BL6 male mice (1828 days old) were anes-
most frequently encountered patterns, classified according to the thetized according to the NIH Animal Care and User Committee guide-
response to series of clicks, are the so-called lockers and special lines, and their brains were placed for 12 min in ice-cold saline solution
responders1. Lockers respond to clicks presented at a frequency of composed of 130 mM NaCl, 3.5 mM KCl, 24 mM NaHCO3, 1.25 mM
NaH2PO4, 0.5 mM CaCl2, 3.0 mM MgCl2 and 10 mM glucose. After
up to about 50 Hz, faithfully following the stimulus, whereas spe- removal of the cerebellum, 250-m-thick coronal slices from the first sixth
cial responders respond only to the onset of the stimulus train. of the caudal part of the brain, where the primary auditory area A1 lies,
How these two types of signal are integrated within auditory cor- were cut at 04C. Slices were then incubated at 32C in the same solu-
tex is largely unknown. We speculate that two synaptic channels tion until used for recordings. Slices were subsequently moved to a record-
composed of connections with temporal characteristics observed ing chamber and superfused with a solution as above except that [Ca2+]
in the present experiments could faithfully convey to cortical neu- and [Mg2+] were both 1.5 mM and maintained at room temperature.
rons (which fire up to approximately 20 Hz19) information encod-
ed in transient or sustained patterns of thalamic input. Barrel cortex slices. Coronal slices (400 m) containing the posteromedial
barrel subfield were prepared from C57/BL6 mice (P18P28) using previ-
Because of the distinct properties of short-term plasticity,
ously described methods23. Mice were anesthetized with forane and decap-
we propose that LPCs and HPCs may subserve fundamental- itated, and the brain was rapidly removed in ice-cold saline solution as
ly different functions in the auditory network. Input to a set described above. The posterior part of the left hemisphere was vertically
of LPCs could act to support locker responses, ensuring fideli- trimmed away at 50 degrees relative to the midsagittal plane. The trimmed
ty of the cortical output. A cortical unit receiving a sustained surface was then glued downward to the vibratome (Leica, Deerfield, Illi-
train of action potentials from a sufficient number of LPCs nois) stage and slices were cut from the rostral pole of the right hemisphere
could indeed produce a sustained noise-insensitive response parallel to the trimmed surface. The initial 4 slices (about 1600 m) were
discarded and the next 5 slices (about 2000 m) containing the whisker
by integration of ensemble-coded synaptic excitatory input. barrels were saved for recording. The slices were incubated at 32C in the
On the contrary, supra-threshold input through a set of HPCs recording solution for 30 min and then maintained at room temperature
would result in a transient, depressing response, largely inde- until use. The posteromedial barrel subfield was identified by the presence
pendent of the duration of the input signal. Thus, the brief of three to four large barrel-like structures in layer IV, visible under tran-
activation of a few HPCs may generate special responder pat- sillumination. Individual pyramidal neurons located within layer II/III were
terns in cortical units. Consequently, a unit that is postsynap- selected for recording based on somata size and position.
tic to HPCs may be particularly suited for detecting variability
Recording. For recording pairs of auditory cortical pyramidal neurons in
in the timing of signal location and frequency. An alternative or
layer II/III, cells were selected according to their position, dorsal to the
complementary role of depressing synapses is the detection of ectosylvian region, and shape of their cell bodies. Whole-cell patch-clamp
subtle differences of frequency-modulated signals 25. These recordings were performed using Axopatch-1D amplifiers. Signals were
interpretations do not rule out the involvement of additional filtered at 2 kHz and digitized at 10 kHz on a PC using a Digidata 1200
mechanisms, such as local inhibition or non-local cortical interface driven by pClamp8 (Axon Instruments, Foster City, California).
interactions, commonly invoked to explain shaping of excita- The intracellular electrode solution was composed of 90 mM KGluconate,

articles
12 mM KCl, 10 mM HEPES, 2 mM EGTA, 2.0 mM ATP.Na2, 0.3 mM RECEIVED 24 SEPTEMBER; ACCEPTED 16 OCTOBER 2001
GTP.Na, 1.0 mM MgCl2 and 0.5% biocytin. This solution was selected to
provide a reversal potential of 60 mV for Cl to minimize the contribu-
tion of GABAergic IPSCs. After obtaining cell pairs, suprathreshold elec- 1. Rouiller, E., de Ribaupierre, Y., Toros-Morel, A. & de Ribaupierre, F. Neural
coding of repetitive clicks in the medial geniculate body of the cat. Hear. Res.
trotonic current pulses (200300 ms) were delivered in current clamp to 5, 81100 (1981).
allow determination of the neuronal firing pattern to provide tentative 2. Phillips, D. P. & Kelly, J. B. Coding of tone-pulse amplitude by single neurons
identification that cells were pyramidal in nature. In all cases, the pres- in auditory cortex of albino rat. Hear. Res. 37, 269280 (1989).
ence of an accommodating action potential firing pattern in response to 3. Clarey, J. C., Barone, P. & Imig, T. J. in The Mammalian Auditory Pathway:
depolarizing current injection (postsynaptic cell now held under current- Neurophysiology (eds. Popper, A. N. & Fay, R. R.) 232334 (Springer, New
York, 1992).
clamp conditions) and post hoc cell identification following biocytin injec- 4. Winer, J. A. The pyramidal neurons in layer III of cat primary auditory cortex
tion (example in Fig. 1c) confirmed that all weakly and strongly connected (AI). J. Comp. Neurol. 229, 476496 (1984).
neuron pairs were between pyramidal neurons. 5. Winer, J. A. Structure of layer II in cat primary auditory cortex (AI). J. Comp.
Excitatory synaptic currents were recorded by holding the presynaptic Neurol. 238, 1037 (1985).
6. Metherate, R. & Aramakis, V. B. Intrinsic electrophysiology of neurons in
cell in current clamp and postsynaptic one in voltage clamp at a holding thalamorecipient layers of developing rat auditory cortex. Dev. Brain Res.
potential of 60 mV. Synaptic currents were evoked at 0.06 Hz by deliv- 115, 13144 (1999).
ering current pulses (13 ms, 0.32.0 nA) to the presynaptic cell held 7. Clarke, S., de Ribaupierre, F., Rouiller, E. M. & de Ribaupierre, Y. Several
under current clamp. In some experiments, trains of 4 action potentials neuronal and axonal types form long intrinsic connections in the cat primary
were generated at 20 Hz, repeated at 15-s intervals (40140 times; on auditory cortical field (AI). Anat. Embryol. (Berl.) 188, 117138 (1993).
8. Shen, J. X., Xu, Z. M. & Yao, Y. D. Evidence for columnar organization in the
average, trains were repeated 80 times). In experiments to determine the auditory cortex of the mouse. Hear. Res. 137, 174177 (1999).
recovery from short-term depression, additional EPSCs were elicited at 9. Blaschke, M. et al. A single amino-acid determines the subunit-specific spider
12-s intervals commencing after the spike train. toxin block of alpha-amino-3-hydroxy-5-methylisoxazole-4-propionate
kainate receptor channels. Proc. Natl. Acad. Sci. USA 90, 65286528 (1993).
Biocytin staining. Following all recordings, slices were immediately trans- 10. Brackley, P. T., Bell, D. R., Choi, D. K., Nakanishi, K. & Usherwood, P. N.
Selective antagonism of native and cloned kainate and NMDA receptors by
ferred to a 24-well plate and fixed in a solution containing 80 mM Na2HPO4, polyamine-containing toxins. J. Pharmacol. Exp. Ther. 266, 15731580
80 mM NaH2PO4 and 3.5% paraformaldehyde. Biocytin staining was then (1993).
processed using diaminobenzidine as chromogen, using a standard ABC 11. Herlitze, S. et al. Argiotoxin detects molecular differences in AMPA receptor
kit (Vector Labs, Burlingame, California). A light cresyl violet Nissl coun- channels. Neuron 10, 11311140 (1993).
12. Washburn, M. S. & Dingledine, R. Block of alpha-amino-3-hydroxy-5-
terstain was used to identify the cortical layers. methyl-4-isoxazolepropionic acid (AMPA) receptors by polyamines and
polyamine toxins. J. Pharmacol. Exp. Ther. 278, 669678 (1996).
Data analysis. The input and series resistance of the postsynaptic cell was 13. Toth, K., & McBain, C. J. Afferent specific innervation of two distinct AMPA
constantly monitored by delivering a 5-mV voltage command. Cells receptor subtypes on single hippocampal interneurons. Nat. Neurosci. 1,
whose input or series resistance changed by more than 30% of the ini- 572578 (1998).
14. Dingledine, R., Borges, K., Bowie, D. & Traynelis, S. The glutamate receptor
tial value were discarded. Lists of events and failures were created for each ion channels. Pharmacol. Rev. 51, 761 (1999).
recorded cell and for each peak amplitude in the train using a semiau- 15. Markram, H., Lubke, J., Frotscher, M., Roth, A. & Sakmann, B. Physiology
tomatic algorithm setting a threshold of two standard deviations of the and anatomy of synaptic connections between thick tufted pyramidal
noise amplitude. To measure peak amplitudes, the mean of the ampli- neurones in the developing rat neocortex. J. Physiol. (Lond.) 500, 409440
(1997).
tude of the failures, corresponding to the peak of the noise, was first sub- 16. Faber, D. & Korn, H. Applicability of the coefficient of variation method for
tracted from the event amplitude list. analyzing synaptic plasticity. Biophys. J. 60, 12881291 (1991).
Latencies were determined from the peak of the presynaptic spike to 17. DeCharms, R. C., Blake, D. T. & Merzenich, M. M. Optimizing sound
the zero crossing of the straight line that best fitted the rising phase of a features for cortical neurons. Science 280, 14391443 (1998).
sample of 1020 EPSCs. The same sample was used to calculate rise times 18. DeCharms, R. C. & Zador, A. Neural representation and the cortical code.
Annu. Rev. Neurosci. 23, 613647 (2000).
(1090%). Decay time constants were determined using single expo- 19. Kilgard, M. P. & Merzenich, M. M. Plasticity of temporal information
nential curve fitting routines. Mean and s.e.m. are reported throughout processing in the primary auditory cortex. Nat. Neurosci. 1, 727731 (1998).
unless otherwise indicated. Data were considered significant if p < 0.05. 20. Zucker, R. S. Short-term synaptic plasticity. Annu. Rev. Neurosci. 12, 1331
CV2 analysis was used to predict the locus of short term synaptic plas- (1989).
21. Zucker, R. S. Calcium- and activity-dependent synaptic plasticity. Curr. Opin.
ticity16. A change is represented as a point in the potentiationCV2 ratio
Neurobiol. 9, 305313 (1999).
() graph. 22. Debanne, D., Guerineau, N. C., Gahwiler, B. H. & Thompson, S. M. Paired
pulse facilitation and depression at unitary synapses in rat hippocampus:
CV2 /2, where mean amplitude and 2 amplitude variance quantal fluctuation affects subsequent release. J. Physiol. (Lond.) 491,
after/before and CV2 ratio CV2before/CV2after 163176 (1996).
23. Feldman, D. E. Timing-based LTP and LTD at vertical inputs to Layer II/III
pyramidal cells in rat barrel cortex. Neuron 27, 4556 (2000).
A data point in the regions between the = and the = 1 straight lines 24. Schreiner, C. E., Read, H. L. & Sutter, M. L. Modular organization of
(indicated by hatched area in Fig. 2d) corresponds to a presynaptic locus. frequency integration in primary auditory cortex. Annu. Rev. Neurosci. 23,
501529 (2000).
25. Abbott, L. F., Varela, J. A., Sen, K. & Nelson, S. B. Synaptic depression and
ACKNOWLEDGEMENTS control of cortical gain. Science 275, 220224 (1997).
We thank D. Feldman for his assistance in preparing Barrel cortex slices and 26. Giraud, A. et al. Representation of the temporal envelope sounds in the
human brain. J. Neurophysiol. 22, 15881598 (2000).
C. Trouth for cell reconstruction and camera lucida drawings of neurons. This 27. Kaas, J. & Hackett, T. What and where processing in the auditory cortex.
work was supported by an Intramural Research award to C.McB. Nat. Neurosci. 2, 10451047 (1999).

articles
Inducible, pharmacogenetic
approaches to the study of learning
and memory
Masuo Ohno, Paul W. Frankland, Adele P. Chen, Rui M. Costa and Alcino J. Silva
Departments of Neurobiology, Psychiatry and Psychology, Brain Research Institute, University of California, Los Angeles, California 90095-1761, USA
Correspondence should be addressed to A.J.S. (silvaa@ucla.edu)
Here we introduce a strategy in which pharmacology is used to induce the effects of recessive
mutations. For example, mice heterozygous for a null mutation of the K-ras gene (K-ras+/) show normal
hippocampal mitogen-activated protein kinase (MAPK) activation, long-term potentiation (LTP) and
contextual conditioning. However, a dose of a mitogen-activated/extracellular-signal-regulated kinase
(MEK) inhibitor, ineffective in wild-type controls, blocks MAPK activation, LTP and contextual learning in
K-ras+/ mutants. These indicate that K-Ras/MEK/MAPK signaling is critical in synaptic and behavioral
plasticity. A subthreshold dose of NMDA receptor antagonists triggered a contextual learning deficit in
mice heterozygous for a point mutation (T286A) in the CaMKII gene, but not in K-ras+/ mutants,
demonstrating the specificity of the synergistic interaction between the MEK inhibitor and the K-ras+/
mutation. This pharmacogenetic approach combines the high temporal specificity that pharmacological
manipulations offer, with the molecular specificity of genetic disruptions.
It is well established that inter-individual differences in the effi- presented here show the importance of K-Ras/MEK/MAPK sig-
cacy and toxicity of many medications are related to genetic diver- naling in synaptic and behavioral plasticity. Similarly, we also used
sity. Over the past 50 years, there has been significant progress this pharmacogenetic method to show that N-methyl-D-aspartate
in understanding the clinical relevance of genetic determinants (NMDA) receptor-dependent autophosphorylation at T286 of
of drug responses1,2. Here we used pharmacogenetic interactions -calcium/calmodulin protein-dependent kinase II (CaMKII) is
to induce the phenotypes of recessive mutations. required for contextual learning. These results indicate that this
Gene targeting in embryonic stem cells has been used to deter- pharmacogenetic procedure may be widely applicable.
mine the role of biochemical pathways, such as second-messenger
signaling cascades, in a variety of biological processes3,4. For exam- RESULTS
ple, gene knockout studies in mice have implicated a variety of MEK inhibition in K-ras+/ mice: biochemistry
neuronal proteins including neurotransmitter receptors, protein The Ras/MAPK cascade is important in many biological func-
kinases, phosphatases and transcription factors in learning and tions911. The K-ras gene is essential for mouse embryogenesis,
memory5,6. However, this methodology does not allow for tem- as K-Ras-deficient embryos die12,13. However, K-ras heterozy-
poral control of these mutations, thus limiting both the design and gous null mutants (K-ras+/) grow and develop normally12,13.
the interpretation of the experiments. Although inducible genetic MEK is downstream of Ras, and activates MAPK in hippocampal
systems7,8 promise to circumvent some of these problems, the use- neurons9,14. Therefore, we applied a pharmacogenetic approach
fulness of these systems remains limited in contrast to the great that combined K-ras +/ mutation and a MEK inhibitor
number of conventional gene targeting mutants available. It is esti- (SL327)15,16, to explore the function of K-Ras/MEK/MAPK sig-
mated that there are more than 4,000 targeted mutants already naling in the adult brain.
derived. In the heterozygote state, most of these mutations are Phorbol esters (for example, phorbol diacetate, or PDA) can
recessive, and therefore, could be used with the pharmacogenetic activate the MAPK cascade in hippocampal neurons14,1719. Expo-
approach introduced here. We used this approach to induce the sure to PDA (10 M) produced similar levels of p42 MAPK acti-
effects of two recessive mutations in mice. This approach takes vation in the hippocampus of K-ras+/ mutants and wild-type
advantage of synergism between pharmacological and genetic littermates (Fig. 1a and b). However, application of the MEK
manipulations. For example, the K-ras+/ mutation studied here inhibitor SL327 (1 M) attenuated MAPK activation produced
did not affect hippocampal MAPK activation, LTP or learning. by PDA in the hippocampus of K-ras+/ but not in wild-type lit-
Similarly, the dose of the MAPK kinase (MEK) inhibitor that we termates (F3,24 = 13.2, p < 0.05; Fig. 1a and b). Neither 1 M
used in the pharmacogenetic studies also did not affect these phe- SL327 nor the K-ras+/ mutation alone affected MAPK phospho-
nomena in wild-type mice. However, the combination of these rylation in response to PDA. The baseline levels of phosphorylat-
genetic and pharmacological manipulations had a profound effect ed p42 MAPK were also not altered by exposure to 1 M SL327
on hippocampal MAPK activation, LTP and learning. The studies in wild-type or K-ras+/ slices (F3,24 = 1.6, p > 0.05). Additional-

articles
Fig. 1. A subthreshold dose of a MEK inhibitor induces less MAPK acti-

vation in K-ras+/ mutants. (a) Representative western blots indicating
a phospho-MAPK
p44
protein bands visualized with antibodies to dually phosphorylated p42
p42/p44 MAPK and total p42/p44 MAPK. Hippocampal slices from total MAPK
K-ras+/ and wild-type mice were exposed to 10 M PDA or normal p44
p42
ACSF (control) under the presence or absence of 1 M SL327. The total
MAPK levels indicate equal protein loading. (b) Average results PDA + + + +
(s.e.m.) of phospho-p42 MAPK levels normalized with the values of SL327 + + + +
the wild-type control group. The SL327-treated K-ras+/ group showed +/ +/
wt K-ras wt K-ras
a significantly smaller increase in phospho-p42 MAPK levels induced by

PDA than each the other three groups (*p < 0.05), whereas the other
groups did not differ each other. b
400
ly, total MAPK levels were not different among the eight groups Control
p42 MAPK activation (%)

examined (Fig. 1a). These data show that a subthreshold dose of 300 PDA
the MEK inhibitor triggers a MAPK phosphorylation deficit in
K-ras+/ mutants. This suggests that K-Ras signaling is important
200
in regulating MAPK activation in the hippocampus. The syner-
gistic interaction between the subthreshold dose of the MEK *
inhibitor SL327 and the K-ras+/ mutation is not unlike previ- 100
ously described epistatic interactions between different mutations. *
MEK inhibition in K-ras+/ mice: LTP 0

wt K-ras
+/
wt +/
K-ras
MAPK signaling is important in synaptic plasticity; LTP-inducing + +
SL327 SL327
stimuli elicit hippocampal MAPK activation, and the inhibition of
MAPK activation blocks LTP induction17,20. Thus, we used the
pharmacogenetic approach outlined above to test whether K-Ras (F2,28 = 9.6, p < 0.05) without affecting baseline synaptic trans-
is involved in LTP, a cellular model of learning and memory. mission at Schaffer collateralCA1 synapses (Fig. 2b and c). Pre-
K-ras +/ mutants exhibited normal hippocampal LTP tetanic application of 1 M SL327 did not affect LTP in wild-type
(F1,19 = 0.02, p > 0.05; Fig. 2a), just as they showed normal MAPK slices, but significantly inhibited the induction of hippocampal LTP
activation. In wild-type slices, the MEK inhibitor SL327 prevented in slices from K-ras+/ mice (F3,38 = 8.5, p < 0.05; Fig. 2c and d).
the induction of LTP in a concentration-dependent manner Neither 1 M SL327 nor K-ras+/ mutation alone affected LTP. Just
as with the MAPK phosphorylation experiments, only the
a 250 combination of these two treatments was effective.
b 200
Post-tetanic application of 1 M SL327 did not affect
fEPSP (% of baseline)
Wild type the expression or maintenance of established LTP in

200 K-ras+/ 150 Wild type
K-ras+/ slices (F1,12 = 0.5, p > 0.05; Fig. 2e), indicating
10 M SL327/WT
150
a temporally specific role of K-Ras/MEK/MAPK signaling
100 in the induction but not in the expression or maintenance
100 of LTP. These results are consistent with the observation
50 that MAPK is activated 2 minutes after LTP induction
10 0 10 20 30 40 50 60 70 80 10 0 20 50 80 110 140
Time (min) Time (min)
and returns to baseline levels by 45 minutes17,20.
MEK inhibition in K-ras+/ mice: behavior

c d Contextual fear conditioning is a robust and long-lasting
form of learning, which is sensitive to hippocampal
250 Wild type 250 manipulations2124. In this test, mice learn to associate a
DMSO K-Ras+/ distinct context (conditioned stimulus; CS) with an aver-

200 1 M SL327 200 DMSO sive stimulus such as foot shock (unconditioned stimu-
10 M SL327 1 M SL327
lus; US). When placed back in the same training context,
150 150
the mice exhibit a range of conditioned fear responses,
100 100
including freezing25,26. Contextual conditioning specifi-
cally activates p42 MAPK signaling in the hippocampus15,
10 0 10 20 30 40 50 60 70 80 10 0 10 20 30 40 50 60 70 80 and MEK inhibitors disrupt contextual condition-
Time (min) Time (min)
e
Fig. 2. A subthreshold dose of a MEK inhibitor induces an LTP impairment in K-ras+/ mutants. Each point
250 K-Ras+/
indicates the field EPSP slope (mean s.e.m.) normalized to the average baseline response before the tetanus
DMSO
200 1 M SL327
(delivered at time 0). (a) K-ras+/ mice showed normal hippocampal LTP. (b) Application of SL327 (indicated
by the bar) had no effect on baseline synaptic transmission at CA1 synapses. (c, d) Hippocampal slices from
150 wild-type and K-ras+/ mice were exposed to pre-tetanic application of SL327. SL327 (1 M) triggered LTP
deficits in K-ras+/ slices without affecting LTP induction in wild-type slices. Traces in panel (c) (wild-type con-
100
trol, left; wild-type + 1 M SL327, right) and in panel (d) (K-ras+/ control, left; K-ras+/ + 1 M SL327, right)
10 0 10 20 30 40 50 60 70 80 are the average of fEPSPs recorded during baseline and 7080 min after tetanization. Scale bars, 10 ms,
Time (min) 0.5 mV. (e) Post-tetanic SL327 had no effect on hippocampal LTP in slices from K-ras+/ mutants.

articles
a 60 b 70 Fig. 3. A subthreshold dose of a MEK inhibitor induces a

60
contextual conditioning impairment in K-ras+/ mutants.
50 Each column represents the percentage of freezing (mean
50 s.e.m.) tested 24 h after training. (a) K-ras+/ mice
Percent freezing
Percent freezing
40 showed normal levels of contextual freezing. (b) Pre-

40
30 training administration of SL327 decreased contextual
30
freezing in wild-type mice in a dose-dependent manner.
20 20 *p < 0.05 versus DMSO-treated group. (c) K-ras+/ mice
10 10
* were administered DMSO or 20 mg/kg of SL327 1 h
before training, which had no effect on freezing in wild-

0 0 type littermates. SL327 (20 mg/kg) significantly decreased
3 10 20 30 100 mg/kg i.p.
contextual freezing in K-ras+/ mice (*p < 0.05). (d) SL327
DMSO SL327
(20 mg/kg) administered 2 h post-training had no effect
c on contextual freezing in K-ras+/ mice.
d
70 90
60 80 Wild-type
70
50 K-ras Specificity of pharmacogenetic interactions
+/
Percent freezing
Percent freezing
60
Previous studies suggested that NMDA receptor-
40 50
* dependent activation of hippocampal MAPK is crit-
30 40
ical for contextual fear conditioning15. However, it
30
20 is unknown how NMDA receptor function is cou-
20
10
pled to MAPK activation during learning. There-
10
fore, we next tested whether a subthreshold dose of
0 0
20 mg/kg i.p. 20 mg/kg i.p. an NMDA receptor antagonist could induce a con-
DMSO SL327 DMSO SL327 textual conditioning deficit in K-ras+/ mice. Pre-
training administration of 10 mg/kg but not
ing15,16,27. Therefore, we used the pharmacogenetic approach out- 5 mg/kg CPP disrupted contextual conditioning (24-h test) in
lined above to explore the role of K-Ras/MEK/MAPK signaling in both wild-type (F 2,34 = 8.4, p < 0.05) and K-ras +/ mice
contextual learning and memory. (F2,32 = 8.8, p < 0.05; Fig. 4). These two doses of CPP affected
K-ras+/ mice tested 24 hours after training showed normal wild-type and K-ras+/ mice in the same manner, demonstrating
contextual conditioning (F1,18 = 0.3, p > 0.05; Fig. 3a). Pre-train- that decreases in NMDA receptor signaling do not affect K-ras+/
ing administration of the MEK inhibitor SL327 impaired contex- mice more severely than their wild-type littermates.
tual conditioning in wild-type mice (24-h test) in a dose-dependent There is much evidence for a close structural and functional
manner (F5,63 = 4.4, p < 0.05; Fig. 3b). Administration of 20 mg/kg link between the NMDA receptor and CaMKII2832. NMDA-
of SL327 did not affect contextual conditioning in wild-type lit- receptor-dependent autophosphorylation of CaMKII at T286 is
termates, but significantly impaired it in K-ras +/ mutants required for the activation of this kinase3235. Our previous stud-
(F1,16 = 7.8, p < 0.05; Fig. 3c). Thus, only the combination of the ies demonstrated that mice homozygous for a point mutation at
MEK inhibitor SL327 (20 mg/kg) and the K-ras+/ mutation dis- T286 that blocks the autophosphorylation of CaMKII
rupted contextual conditioning; neither manipulation alone had (CaMKIIT286A/) have normal NMDA function, but have severe
an effect. The unconditioned response (UR) to shock was not impairments in hippocampal plasticity and hippocampus-depen-
affected by SL327 in either K-ras+/ mutants or wild-type mice dent learning and memory36. Similarly, this homozygous mutation
(data not shown), suggesting that this pharmacogenetic manipu- completely blocked contextual fear conditioning (data not shown).
lation did not disrupt the mouses ability to perceive the foot shock. In contrast, the heterozygous CaMKIIT286A+/ mutants showed
Post-training (2 h) administration of 20 mg/kg of SL327 did not normal levels of contextual conditioning tested 24 h after train-
affect contextual conditioning in K-ras+/ mutants (F1,18 = 1.4, ing (Fig. 5b and d). In wild-type mice, contextual fear condition-
p > 0.05; Fig. 3d), just as post-tetanic application of this compound ing was disrupted by pre-training administration of NMDA
failed to affect established LTP. These results indicate that receptor antagonists CPP (F 3,62 = 3.7, p < 0.05) or MK-801
K-Ras/MEK/MAPK signaling at the time of training, but not some (F4,48 = 3.8, p < 0.05) in a dose-dependent manner (Fig. 5a).
time afterwards, is critical for formation of contextual memory. Administration of doses of CPP (5 mg/kg) and MK-801
A weak training protocol (30-s placement-to-shock) was (0.01 mg/kg), which were behaviorally ineffective in wild-type
unable to trigger a contextual conditioning deficit in K-ras+/
mutants. K-ras +/ mice exhibited contextual freezing
(23.5 4.4%) comparable to that of wild-type littermates 60
Wild-type
(21.9 4.4%; F1,38 = 0.07, p > 0.05), suggesting that the K-ras+/
mutation alone does not weaken contextual conditioning. 50 +/
K-ras
Percent freezing
Whereas contextual learning is normal in K-ras+/ mutants, more 40

complex hippocampus-based learning such as spatial learning
in water maze is affected in these mice (data not shown). 30
20 * *
Fig. 4. A subthreshold dose of an NMDA receptor antagonist does not 10
induce a contextual conditioning impairment in K-ras+/ mutants. Each
column represents the percentage of freezing (mean s.e.m.) tested 0
5 10 5 10 mg/kg i.p.
24 h after training. Pre-training administration of CPP affected contex-
tual learning in K-ras+/ mutants and wild-type littermates similarly. Saline CPP Saline CPP

articles
a 40
b 40 Wild-type
+/
T286A
30 30
Percent freezing
Percent freezing
20 20 *
10 10
* *
0 0
5 10 20 0.01 0.03 0.1 0.3 mg/kg i.p. 5 5 mg/kg i.p.
Saline CPP MK-801 Saline CPP Saline CPP
c 50
d 50
40
40
Percent freezing
Percent freezing
30
30
20 *
20
10 10
0 0
0.01 0.01 mg/kg i.p. 5 5 mg/kg i.p.
Saline MK801 Saline K-801 Saline CPP Saline CPP
Fig. 5. Subthreshold doses of NMDA receptor antagonists induce a contextual conditioning impairment in CaMKIIT286A heterozygotes, but not in
K-ras+/ mutants. Each column represents the percentage of freezing (mean s.e.m.) tested 24 h after training. (a) Pre-training administration of CPP
and MK-801 decreased contextual freezing in wild-type mice in a dose-dependent manner. *p < 0.05 versus saline-injected group.
(b, c) CaMKIIT286A+/ mice and wild-type littermates were administered saline, 5 mg/kg CPP or 0.01 mg/kg MK-801 30 min before training. The CPP-
or MK-801-injected CaMKIIT286A+/ group showed significantly smaller freezing responses than each the other three groups (*p < 0.05), whereas the
other groups did not differ from each other. (d) CPP (5 mg/kg) administered 2 h post-training had no effect on contextual freezing in CaMKIIT286A+/
mice or wild-type littermates.
mice, disrupted contextual fear conditioning in CaMKIIT286A+/ plete block of hippocampal MAPK activation in the K-ras+/
mutants (CPP, F3,115 = 3.9, p < 0.05, Fig. 5b; MK-801, F3,56 = 6.4, mice. The same dose of this inhibitor did not affect MAPK acti-
p < 0.05, Fig. 5c). Therefore, 5 mg/kg of CPP triggered a contex- vation in wild-type littermate controls.
tual conditioning deficit in CaMKIIT286A+/ mutants, but not in MEK and MAPK have previously been shown to be involved
K-ras+/ mice. Therefore, the effects of this dose of CPP are specific in LTP17,20 and learning15,16,27,37. Our pharmacogenetic approach
to the CaMKIIT286A+/ mutants, reflecting the close interaction indicated that K-Ras signaling is critical for the activation of the
between NMDA receptors and CaMKII. Similarly, our data also MAPK cascade during synaptic and behavioral plasticity.
suggest that K-Ras is critical in MEK-dependent activation of Although the K-ras+/ mutants showed normal hippocampal LTP
MAPK. These findings attest to the specificity of the synergistic and contextual conditioning, a dose of the MEK inhibitor inef-
interactions at the heart of this pharmacogenetic approach. fective in wild types induced both LTP and learning deficits in
The conditioning deficits in CaMKIIT286A+/ mutants treat- these mutants. Previous studies indicated that a H-ras mutation
ed with either CPP or MK-801 were not due to changes in noci- enhanced hippocampal NMDA currents and LTP38. However, it
ception, as neither of these pharmacogenetic manipulations is unclear whether this enhancement in LTP is a compensatory
affected unconditioned responses to foot shock (data not shown). reaction to the homozygous loss of H-ras during development.
Post-training injection of 5 mg/kg CPP also had no effect on con- The advantage of our pharmacogenetic approach is that it cir-
textual learning in CaMKII T286A+/ mutants (F 3,36 = 0.6, cumvents developmental confounds. Unlike H-ras homozygous
p > 0.05; Fig. 5d), indicating that contextual conditioning requires null mutants, the K-ras+/ mice show normal LTP; their deficits
NMDA-receptor-dependent autophosphorylation of CaMKII, are observed only after drug treatment.
specifically around the time of training. The pharmacological induction of the K-ras+/ phenotype
showed temporal specificity, as pre-training but not post-
DISCUSSION training (2 h) administration of the MEK inhibitor triggered
Here we have introduced a way to manipulate molecular func- the contextual learning deficit in this mutant. Similarly, the
tion in the central nervous system. This approach takes advan- MEK inhibitor is only effective when applied before the induc-
tage of synergistic interactions between pharmacological and tion of LTP. Indeed, p42 MAPK activation is significantly
genetic manipulations to alter the function of specific signaling increased in the hippocampus one hour after training, but
pathways in a temporally controlled manner. For example, the returns to baseline levels within two hours of contextual con-
K-ras+/ mutation alone did not affect hippocampal MAPK acti- ditioning 15 . Taken together, these findings indicate that
vation. However, a subthreshold dose of an inhibitor of MEK, a K-Ras/MEK/MAPK signaling is critical in both the induction of
component of the Ras/MEK/MAPK pathway9, triggered a com- LTP andhippocampus-dependent learning.

articles
Although we did not find any evidence linking NMDA recep- reaction analysis of tail DNA samples. All experiments were done with mice
tor signaling and K-Ras-dependent plasticity and learning, a phar- 37 months old, and a similar number of males and females were used. The
macogenetic approach like the one described here could be used mice were housed in groups and kept on a 12 h light/dark cycle, and the
to explore the proposed connection between NMDA signaling experiments were always conducted during the light phase of the cycle. With
the exception of testing times, the mice had ad lib access to food and water.
and different Ras proteins. For example, H-Ras is thought to reg-
All the procedures used were approved by UCLAs Animal Research Com-
ulate NMDA currents38. Additionally, NMDA signals may acti- mittee. Animals were maintained in accordance with the applicable por-
vate specific Ras proteins through Ras-GRF (a Ras guanine tions of the Animal Welfare Act and the Department of Health and Human
nucleotide releasing factor) thought to have a role in plasticity Services Guide to the Care and Use of Laboratory Animals.
and learning3941. NMDA receptor activation may also amplify
certain Ras signals by inhibiting SynGAP, a synaptic Ras-GTPase Drugs. SL327 (provided by DuPont Pharmaceuticals, Wilmington,
activating protein42,43. Our present findings are not inconsistent Delaware) and phorbol 12,13-diacetate (PDA; Sigma, St. Louis, Missouri)
with a possible link between NMDA signals and other Ras iso- were dissolved in 100% DMSO. []-3-[2-Carboxypiperazin-
forms (such as H-Ras and N-Ras) signaling to the MAPK cascade. 4-yl]propanephosphonic acid (CPP; Sigma) and (+)-MK-801 hydrogen
maleate (RBI, Natick, Massachusetts) were dissolved in saline or artifi-
Although a subthreshold dose of an NMDA receptor block-
cial cerebrospinal fluid (ACSF). All biochemical, electrophysiological and
er did not trigger deficits in the K-ras+/ mutants, it induced behavioral experiments were conducted with the experimenter blind to
learning and LTP deficits (data not shown) in heterozygous the drug treatments as well as the genotype of mice.
CaMK-IIT286A mice. Mice homozygous for this point mutation
show profound deficits in spatial36 and contextual conditioning Electrophysiology. Transverse hippocampal slices (400 m thick) were
learning that could be conceivably due to developmental deficits. maintained in a submerged recording chamber perfused with ACSF equi-
Our present data discount this possibility. We found that although librated with 95% O 2 and 5% CO 2 at 30C. The ACSF contained
heterozygous CaMKIIT286A mice did not show a contextual 120 mM NaCl, 3.5 mM KCl, 2.5 mM CaCl2, 1.3 mM MgSO4, 1.25 mM
deficit, a dose of NMDA receptor antagonists ineffective in wild- NaH2PO4, 26 mM NaHCO3 and 10 mM D-glucose. Extracellular field
type mice induced a contextual conditioning deficit in these EPSPs were recorded with a metallic electrode from the stratum radia-
tum layer of the area CA1, and the Schaffer collaterals were stimulated
mutants. Importantly, this deficit was only induced when the
with a bipolar electrode. The intensity of stimulation (100-s duration)
drug was administered before training. Therefore, these data was adjusted to give field EPSP approximately 33% of maximum. LTP
demonstrate that the NMDA receptor-dependent autophospho- was induced by a tetanic stimulation (100 Hz, 1 s) delivered at the test
rylation of CaMKII specifically during training is critical for intensity. After the responses were monitored at least for 20 min to ensure
learning32. The subsequent phosphorylation of key cellular sub- a stable baseline, drugs were applied to hippocampal slices. SL327 (max-
strates, such as glutamate receptors44, is thought to be required imal final DMSO concentration, 0.1%) was applied for 1 h before
for early stages of memory formation45,46. Because the pharma- tetanization, and was maintained in the bath throughout the recording
cogenetic approach introduced here uses drugs at concentrations period. In some slices, SL327 was applied post-tetanically (20 min after
that are ineffective in wild types, the nonspecific effects of these tetanization). To determine whether the magnitude of LTP differed sig-
drugs should be reduced. nificantly between the groups, responses from the last 10-min block of
recordings (7080 min) were compared statistically.
The results presented here demonstrate that pharmacologi-
cal manipulations can be used to induce the phenotype of reces- Western blotting. Hippocampal slices prepared from K-ras+/ mice and
sive mutations in mice. Although neither the CaMKIIT286A+/ wild-type littermates were transferred to the same chamber as used for
nor K-ras+/ mice showed contextual learning deficits, partial electrophysiology, and then exposed to PDA (10 M) or normal ACSF
pharmacological disruption of specific signaling components for 10 min. The slices were preincubated with SL327 for 70 min. In the
upstream or downstream from those genetically targeted mole- last 10 min of preincubation, slices were exposed to PDA or ACSF. After
cules triggered learning deficits in the mutants. Importantly, these pharmacological activation with PDA, the CA1 subregions of hip-
partial pharmacological disruptions only affected contextual con- pocampal slices were dissected out on ice and homogenized in an
ice-cold lysis buffer (containing 25 mM HEPES (pH 7.5), 150 mM NaCl,
ditioning in the presence of the mutations. Thus, epistatic-like
2 mM EDTA, 10% glycerol, 1% Triton X-100, 100 g/ml phenylmethyl-
interactions between pharmacological and genetic manipulations sulfonyl fluoride (PMSF), 1 g/ml leupeptin, 1 mM sodium orthovana-
can be used to induce the effects of mutations in mice in a tem- date (Na3VO4), 25 mM -Na glycerophosphate and 10 mM NaF). After
porally controlled manner, and to identify novel functional rela- insoluble material was removed by centrifugation (13,000 g for 5 min),
tions between signaling pathways. This approach combines the protein concentration of the supernatant was determined by using a
high temporal specificity that pharmacological manipulations PIERCE BCA Protein Assay Kit (Pierce, Rockford, Illinois). Samples con-
offer, with the molecular specificity of genetic disruptions. Spa- taining equivalent amount of protein (20 g) were separated by 12%
tial control of these effects could also be accomplished by neu- SDS-PAGE gels and transferred onto nitrocellulose membranes (Bio-
roanatomically guided injections of compounds of interest. Rad, Hercules, California). Membranes were blocked in TBS with 0.1%
Tween-20 and 5% dry milk overnight at 4C. Then the blots were incu-
Importantly, this pharmacogenetic approach will be applicable
bated for 2 h at room temperature with the primary antibody which
not only to neuroscience questions, but also to other biological selectively recognizes p42/p44 MAPK dually phosphorylated at Thr202
problems and to other genetic systems (such as Drosophila, and Tyr204 (1:1,000; New England Biolabs, Beverly, Massachusetts).
Caenorhabditis elegans and yeast). Another antibody that detects both phosphorylated and unphosphory-
lated forms of p42/p44 MAPK (1:1,000; New England Biolabs) was used
METHODS to measure total MAPK levels. After membranes were washed, the blots
Mice. We used K-ras mutants (K-ras+/)13 that were F1 progeny derived were incubated with HRP-conjugated secondary antibody (1:2,000; Bio-
from a cross between mice heterozygous for a null mutation in the K-ras Rad). Protein signals were visualized by enhanced chemiluminescence
gene (129T2/SvEmsJ background) and C57Bl/6N mice. Starting with the (ECL Western Blotting Analysis system, Amersham, Arlington Heights,
CaMKIIT286A+/ chimeras (contributing to 129 background), this muta- Illinois). Densitometric analysis for quantification of the signals was per-
tion was backcrossed into the C57Bl/6 genetic background 36. The formed using a desktop scanner and ImageQuaNT software (Molecular
CaMKIIT286A+/ mice used in the experiments were heterozygotes derived Dynamics, Sunnyvale, California). For each experiment, both phospho-
after 89 backcrosses into C57Bl/6N. At 45 weeks postnatally, the mice rylated and total MAPK levels were normalized to those observed in the
were weaned and their genotypes were determined with polymerase chain control group of wild-type mice.

articles
Contextual fear conditioning. The contextual conditioning experiments 18. Ebinu, J. O. et al. RasGRP, a Ras guanyl nucleotide- releasing protein with
were performed as described previously47. During training, mice were placed calcium- and diacylglycerol-binding motifs. Science 280, 10821086
in the conditioning chamber for 2.5 min, and then CaMKIIT286A+/ and K- (1998).
19. Tognon, C. E. et al. Regulation of RasGRP via a phorbol ester-responsive C1
ras+/ mice were exposed to a 0.75 and 0.50 mA foot shock for 2 s, respec- domain. Mol. Cell Biol. 18, 69957008 (1998).
tively. In a separate experiment, we assessed contextual fear conditioning 20. English, J. D. & Sweatt, J. D. Activation of p42 mitogen-activated protein
in mice receiving a single foot shock 30 s following placement in the con- kinase in hippocampal long term potentiation. J. Biol. Chem. 271,
ditioning chamber during training. After shock delivery, mice were left in the 2432924332 (1996).
chamber for another 30 s, and then returned to their home cage. CPP and 21. Frankland, P. W., Cestari, V., Filipkowski, R. K., McDonald, R. J. & Silva, A. J.
The dorsal hippocampus is essential for context discrimination but not for
MK-801 were administered intraperitoneally (i.p.) in a volume of 10 ml/kg contextual conditioning. Behav. Neurosci. 112, 863874 (1998).
30 min before or 2 h after training. Mice received an i.p. injection of SL327
22. Fanselow, M. S. Contextual fear, gestalt memories, and the hippocampus.

in a volume of 2 ml/kg 1 h before or 2 h after training. The mice were test- Behav. Brain Res. 110, 7381 (2000).
ed for contextual conditioning 24 h after training. Conditioning was deter- 23. Rampon, C. et al. Enrichment induces structural changes and recovery from
mined by scoring freezing behavior (absence of all but respiratory nonspatial memory deficits in CA1 NMDAR1-knockout mice. Nat. Neurosci.
3, 238244 (2000).
movement) with automated procedures described previously47 when the 24. Shimizu, E., Tang, Y. P., Rampon, C. & Tsien, J. Z. NMDA receptor-dependent
mice were placed back into the conditioning chamber. synaptic reinforcement as a crucial process for memory consolidation.
Science 290, 11701174 (2000).
Data analysis. The significance of differences between the groups was 25. Fanselow, M. S. Factors governing one-trial contextual conditioning. Anim.
Learn. Behav. 18, 264270 (1990).
determined by a one-way analysis of variance followed by post-hoc New-
26. Phillips, R. G. & LeDoux, J. E. Differential contribution of amygdala and
manKeuls test when F ratios reached significance (p < 0.05). hippocampus to cued and contextual fear conditioning. Behav. Neurosci. 106,
274285 (1992).
27. Schafe, G. E., Nadel, N. V., Sullivan, G. M., Harris, A. & LeDoux, J. E.
ACKNOWLEDGEMENTS Memory consolidation for contextual and auditory fear conditioning is
We thank S.A. Josselyn, N.B. Fedorov and K.P. Giese for discussions, and R. Chen dependent on protein synthesis, PKA, and MAP kinase. Learn. Mem. 6,
and M. Lacuesta for help with genotyping. We also thank J.M. Trzaskos (DuPont 97110 (1999).
28. Husi, H., Ward, M. A., Choudhary, J. S., Blackstock, W. P. & Grant, S. G.
Pharmaceuticals Research Laboratories) and T. Jacks (Department of Biology, Proteomic analysis of NMDA receptor-adhesion protein signaling complexes.
MIT) for donating SL327 and K-ras+/ mutants, respectively. This work was Nat. Neurosci. 3, 661669 (2000).
funded by grants from the McKnight Foundation, Merck Foundation and the 29. Kennedy, M. B. Signal-processing machines at the postsynaptic density.
Science 290, 750754 (2000).
NIH (P01HD33098 and AG13622) to A.J.S. M.O. was partially supported by a 30. Leonard, A. S., Lim, I. A., Hemsworth, D. E., Horne, M. C. & Hell, J. W.
research fellowship from the Uehara Memorial Foundation for Life Sciences. Calcium/calmodulin-dependent protein kinase II is associated with the N-
methyl-D-aspartate receptor. Proc. Natl. Acad. Sci. USA 96, 32393244
RECEIVED 20 AUGUST; ACCEPTED 25 OCTOBER 2001 (1999).
31. Silva, A. J. & Giese, K. P. in Neurobiology of Learning and Memory (eds.
Martinez, J. & Kesner, R.) 89142 (Academic, San Diego, California, 1998).
32. Lisman, J. The CaM kinase II hypothesis for the storage of synaptic memory.
1. Steimer, W., Muller, B., Leucht, S. & Kissling, W. Pharmacogenetics: a new Trends Neurosci. 17, 406412 (1994).
diagnostic tool in the management of antidepressive drug therapy. Clin. 33. Fukunaga, K., Soderling, T. R. & Miyamoto, E. Activation of
Chim. Acta 308, 3341 (2001). Ca2+/calmodulin-dependent protein kinase II and protein kinase C by
2. McLeod, H. L. & Evans, W. E. Pharmacogenomics: unlocking the human glutamate in cultured rat hippocampal neurons. J. Biol. Chem. 267,
genome for better drug therapy. Annu. Rev. Pharmacol. Toxicol. 41, 101121 2252722533 (1992).
(2001). 34. Ouyang, Y., Kantor, D., Harris, K. M., Schuman, E. M. & Kennedy, M. B.
3. Brandon, E. P., Idzerda, R. L. & McKnight, G. S. Knockouts. Targeting the mouse Visualization of the distribution of autophosphorylated calcium/calmodulin-
genome: a compendium of knockouts (Part I). Curr. Biol. 5, 625634 (1995). dependent protein kinase II after tetanic stimulation in the CA1 area of the
4. Brandon, E. P., Idzerda, R. L. & McKnight, G. S. Targeting the mouse genome: hippocampus. J. Neurosci. 17, 54165427 (1997).
a compendium of knockouts (Part II). Curr. Biol. 5, 758765 (1995). 35. Hanson, P. I. & Schulman, H. Neuronal Ca2+/calmodulin-dependent protein
5. Silva, A. J., Smith, A. M. & Giese, K. P. Gene targeting and the biology of kinases. Annu. Rev. Biochem. 61, 559601 (1992).
learning and memory. Annu. Rev. Genet. 31, 527546 (1997). 36. Giese, K. P., Fedorov, N. B., Filipkowski, R. K. & Silva, A. J.
6. Chen, C. & Tonegawa, S. Molecular genetic analysis of synaptic plasticity, Autophosphorylation at Thr286 of the calcium-calmodulin kinase II in
activity-dependent neural development, learning, and memory in the LTP and learning. Science 279, 870873 (1998).
mammalian brain. Annu. Rev. Neurosci. 20, 157184 (1997).
37. Berman, D. E. & Dudai, Y. Memory extinction, learning anew, and learning
7. Furth, P. A. et al. Temporal control of gene expression in transgenic mice by a
the new: dissociations in the molecular machinery of learning in cortex.
tetracycline-responsive promoter. Proc. Natl. Acad. Sci. USA 91, 93029306
Science 291, 24172419 (2001).
(1994).
8. Ray, P. et al. Regulated overexpression of interleukin 11 in the lung. Use to 38. Manabe, T. et al. Regulation of long-term potentiation by H-Ras through
dissociate development-dependent and -independent phenotypes. J. Clin. NMDA receptor phosphorylation. J. Neurosci. 20, 25042511 (2000).
Invest. 100, 25012511 (1997). 39. Farnsworth, C. L. et al. Calcium activation of Ras mediated by neuronal
9. Derkinderen, P., Enslen, H. & Girault, J. A. The ERK/MAP-kinases cascade in exchange factor Ras-GRF. Nature 376, 524527 (1995).
the nervous system. Neuroreport 10, R2434 (1999). 40. Giese, K. et. al. Hippocampus-dependent learning and memory is impaired
10. Fukunaga, K. & Miyamoto, E. Role of MAP kinase in neurons. Mol. in mice lacking the Ras-guanine-nucleotide releasing factor 1 (RAS-GRF1).
Neurobiol. 16, 7995 (1998). Neuropharmacology 41, 791800 (2001).
11. Sweatt, J. D. The neuronal MAP kinase cascade: a biochemical signal 41. Brambilla, R. et al. A role for the Ras signalling pathway in synaptic
integration system subserving synaptic plasticity and memory. J. Neurochem. transmission and long- term memory. Nature 390, 281286 (1997).
76, 110 (2001). 42. Kim, J. H., Liao, D., Lau, L. F. & Huganir, R. L. SynGAP: a synaptic RasGAP that
12. Koera, K. et al. K-ras is essential for the development of the mouse embryo. associates with the PSD-95/SAP90 protein family. Neuron 20, 683691 (1998).
Oncogene 15, 11511159 (1997). 43. Chen, H. J., Rojas-Soto, M., Oguni, A. & Kennedy, M. B. A synaptic Ras-
13. Johnson, L. et al. K-ras is an essential gene in the mouse with partial GTPase activating protein (p135 SynGAP) inhibited by CaM kinase II.
functional overlap with N-ras. (published erratum, Genes Dev. 11, 3277, Neuron 20, 895904 (1998).
1997) Genes Dev. 11, 24682481 (1997). 44. Barria, A., Muller, D., Derkach, V., Griffith, L. C. & Soderling, T. R.
14. Roberson, E. D. et al. The mitogen-activated protein kinase cascade couples Regulatory phosphorylation of AMPA-type glutamate receptors by CaMK-II
PKA and PKC to cAMP response element binding protein phosphorylation during long-term potentiation. Science 276, 20422045 (1997).
in area CA1 of hippocampus. J. Neurosci. 19, 43374348 (1999). 45. Tan, S. E. & Liang, K. C. Spatial learning alters hippocampal
15. Atkins, C. M., Selcher, J. C., Petraitis, J., Trzaskos, J. & Sweatt, J. The MAPK calcium/calmodulin-dependent protein kinase II activity in rats. Brain Res.
cascade is required for mammalian associative learning. Nat. Neurosci. 1, 711, 234240 (1996).
602609 (1998). 46. Cammarota, M., Bernabeu, R., Levi De Stein, M., Izquierdo, I. & Medina, J.
16. Selcher, J. C., Atkins, C. M., Trzaskos, J. M., Paylor, R. & Sweatt, J. D. A H. Learning-specific, time-dependent increases in hippocampal
necessity for MAP kinase activation in mammalian spatial learning. Learn. Ca2+/calmodulin-dependent protein kinase II activity and AMPA GluR1
Mem. 6, 478490 (1999). subunit immunoreactivity. Eur. J. Neurosci. 10, 26692676 (1998).
17. English, J. D. & Sweatt, J. D. A requirement for the mitogen-activated protein 47. Anagnostaras, S. G., Josselyn, S. A., Frankland, P. W. & Silva, A. J. Computer-
kinase cascade in hippocampal long term potentiation. J. Biol. Chem. 272, assisted behavioral assessment of Pavlovian fear conditioning in mice. Learn.
1910319106 (1997). Mem. 7, 5872 (2000).

articles
Inferotemporal neurons represent

low-dimensional configurations of
parameterized shapes
Hans Op de Beeck1,2, Johan Wagemans2 and Rufin Vogels1
1 Laboratorium voor Neuro- en Psychofysiologie, K.U. Leuven, Campus Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium
2 Laboratory of Experimental Psychology, K.U. Leuven, Tiensetraat 102, B-3000 Leuven, Belgium
Correspondence should be addressed to R.V. (rufin.vogels@med.kuleuven.ac.be)
Behavioral studies with parameterized shapes have shown that the similarities among these complex
stimuli can be represented using a low number of dimensions. Using psychophysical measurements
and single-cell recordings in macaque inferotemporal (IT) cortex, we found an agreement between
low-dimensional parametric configurations of shapes and the representation of shape similarity at the
behavioral and neuronal level. The shape configurations, computed from both the perceived and neu-
ron-based similarities, revealed a low number of dimensions and contained the same stimulus order
as the parametric configurations. However, at a metric level, the behavioral and neural
representations deviated consistently from the parametric configurations. These findings suggest an
ordinally faithful but metrically biased representation of shape similarity in IT.
The capacity to categorize stimuli is fundamental to all living As the analysis of the visual input in the visual system is high-
organisms1,2. Theories of categorization agree upon the impor- ly nonlinear, the neuronal representation space could deviate
tance of the similarity between stimuli to account for many from the configurations in parameter space in several ways. Pre-
aspects of categorization performance 35. However, it is not vious psychophysical studies found an ordinal agreement between
straightforward to compute the degree of similarity between stim- parametric configurations and their perceptual representation
uli that can vary across a high number of dimensions, like com- (the same number of dimensions and the same stimulus
plex shapes. Fortunately, the similarities among a set of complex order)912, so we also expected this result at the neuronal level.
stimuli can often be described in a more compact way68. Indeed, However, configurations that fit perfectly on an ordinal level can
stimuli from many behaviorally relevant sets can be represented differ on a metric level. Thus, apart from investigating ordinal
in a low-dimensional representation space in which the proxim- relationships, we looked for consistent metric anomalies in per-
ity between stimuli is related to their similarity. For example, by ceptual and neuronal representation spaces with respect to the
presenting the randomly ordered shapes of Fig. 1d in a particular parametric similarities.
order (Fig. 1ac), the similarities can be easily described by a two- In this study, we varied complex shape dimensions, that is,
dimensional square-like configuration. Several behavioral stud- radial frequency components (RFCs), that have been used pre-
ies that have varied complex shape differences parametrically viously with human subjects10. These parametric variations
revealed that primates are able to represent the similarities affect a high number of perceptually salient dimensions10,18, but
between shapes in a low-dimensional representation space with- are not coded directly in the macaque visual system19. Never-
out ever seeing these stimuli in their parametric configuration912. theless, in agreement with computational work11, we found that
Here we aim to study directly the neural basis of these low- human and monkey subjects were able to represent the similar-
dimensional representation spaces. Object recognition and cate- ities between such shapes in low-dimensional representation
gorization in macaques is thought to depend on the inferotemporal spaces that agree well with pixel-based configurations (Fig. 2ae).
cortex (IT)13,14. Single IT neurons are selective for moderately Moreover, the perceived similarities of the monkeys cor-
complex object features15, but several studies have found little rela- responded better with the stimulus similarities when these were
tionship between the similarities between complex objects and the computed using IT neuron responses (Fig. 2f) than with the
responses of single IT neurons16,17. However, one needs to manip- pixel-based similarities. Further behavioral experiments in mon-
ulate shape similarity parametrically to investigate how the respons- keys showed that the low-dimensional stimulus configuration
es of IT neurons to complex stimuli are related to the proximity predicts categorization performance.
of these stimuli in a low-dimensional space. Thus, we investigat-
ed whether the response pattern across a population of IT neurons RESULTS
can reveal a low-dimensional and faithful representation of shape We investigated the underlying representation space of 24 shapes
similarity using parameterized shapes. that were divided into 3 groups (Fig. 1). The within-group con-

articles
Fig. 1. Visual stimuli. (ac) Three groups of eight

a b c shapes were used. Within each stimulus group, the
parameterspace configuration of the stimuli is rep-
resented by the square arrangement of the stimuli.
The top-left stimulus in each square has a low ampli-
tude value for both manipulated radial frequency
components. (d) Same 24 stimuli, but the 8 shapes
from each stimulus group are presented in a random
order in a row.
d points is 90 at corner points, and 180 at points

along the sides. Other configurations with the
same stimulus order, such as eight equidistant
points on a circle, do not possess this property.
A chi-square test was used to assess the depen-
dence of angle size (smaller or larger than 135)
and type of point (corner or side point in the
parametric configuration) for each subject.
Angles were greater at side points compared to
figurations formed two-dimensional squares in the parametric corner points for each subject (2 = 10.74, p < 0.01 for subject 1;
space, and similar but not identical configurations are present 2 = 20.17, p < 0.01 for subject 2). Thus, the configurations of
when using pixel-based similarities (Fig. 2a). each subject differentiate between configurations of the same
stimulus order, such as a square or a circle.
Perceptual representation space: human data To determine whether remaining metric deviations from the
We verified the perceptual representation of the similarities expected configurations were systematic rather than due to mea-
between shapes in two human subjects performing similarity surement noise, we determined the inter-subject consistency of
ratings. First, we applied hierarchical cluster analysis to deter- deviations of the perceived similarities with respect to the pixel-
mine whether these similarity ratings reflected the expected clus- based similarities. For group A and B stimuli, C was lower when
tering of stimuli into the three groups. As expected, all stimuli the perceived similarity ratings were compared between the two
from a given stimulus group were assigned to the same cluster subjects (Table 1, C(human 1, human 2)) than when each sub-
before they were clustered with stimuli from the other two jects data were compared with the pixel-based similarities
groups. A substantial proportion of variance (85%) in the sim- (C(pixel, human 1) and C(pixel, human 2)). This suggests that
ilarity judgments was accounted for when reducing the 24 stim- the deviations of the perceived similarities were likely due to mea-
uli to these 3 clusters. surement noise. However, there was a striking consistency in the
Next, we analyzed the similarities among the eight shapes way the perceived similarities of the two subjects deviated from
from each group. To determine whether the perceived similari- the pixel-based similarities for group C stimuli (compare length
ties converge with the pixel-based similarities, we computed the and direction of red lines in Fig. 2b and c).
congruence C between these two sets of similarities. For all groups
of stimuli and for both subjects, significantly high congruences Perceptual representation space: monkey data
were found (Table 1). Two monkeys (Y and E) were trained in a same-different task. Fol-
We used multidimensional scaling (MDS) to determine lowing previous studies in macaques12,20, the proportion of cor-
whether the shape similarities could be captured by the distance rect different responses for a particular stimulus pair was taken
between the stimuli in a low-dimensional space. Two-dimen- as measure of the dissimilarity of that pair. This measure can be
sional MDS-derived configurations were computed for each sub- subjected to the same analyses as performed on the human data.
ject and each group of stimuli separately (Fig. 2b and c). The First, hierarchical cluster analysis showed that the most sim-
more similar the two stimuli, the smaller the distance between ilar stimuli belong to the same stimulus group. All stimuli from
the points representing each stimulus. Even with as few as two a given stimulus group were assigned to the same cluster, before
dimensions, these configurations accounted for most of the vari- they were clustered with stimuli from the other two groups. How-
ance in the similarity ratings. All within-group similarities were ever, the clustering in monkeys was less than in human subjects.
represented in low-dimensional configurations that strongly The proportion of variance accounted for by reducing the 24
resembled the expected configurations. First, no qualitative devi- stimuli to 3 clusters was smaller for the monkey data (averaged
ations, such as the reversal of a pair of shapes, were seen. We across monkeys, 47% of variance explained; humans, 85%).
quantified this ordinal agreement between perceived and expect- Second, analysis of the similarity data for each stimulus group
ed configurations by computing the Spearman rank order cor- separately showed that the congruence with the pixel-based sim-
relation (r s) of the polar angles of the stimulus points with ilarities was always highly significant (Table 1). Two-dimension-
respect to the center of each configuration. If one takes stimu- al MDS-derived configurations (Fig. 2d and e) explained most
lus 1 as the reference with polar angle zero and proceeds clock- of the variance in the similarity ratings (averaged across mon-
wise, then one would expect stimulus 2 to have the next-smallest keys, 96%, 88% and 97% for group A, B and C, respectively).
polar angle, followed by stimulus 3. The expected order of the Overall, the within-group similarities were represented in low-
stimuli was preserved perfectly in all stimulus groups (rs = 1.0, dimensional configurations that captured most aspects of the
p < 0.01). Second, details of the expected configurations, such expected configurations. First, the stimulus order matched the
as their squareness, were maintained. In a square, the angle expected order (rs = 1, p < 0.01) for all stimulus groups in each
between two lines connecting a point with its two neighboring monkey, except for group B in monkey Y (reversal of stimuli 6

articles
Fig. 2. MDS-derived two-dimensional configurations of

stimuli. (a) Configurations based on the pixel-based simi- a
larity between stimuli. The numbers refer to the stimuli as
labeled in Fig. 1. The other panels show a comparison
between the configurations (blue squares) of stimuli in (a)
and the configurations (green diamonds) found for the
perceived similarities of the first (b) and second (c) human
subject, the perceived similarities of monkey Y (d) and
monkey E (e), and the neuron-based similarities (f). The
amount of variance (r2) of the similarity values that was

explained by the two-dimensional configuration is denoted b
below each panel, as is the range ([minimum; maximum])
of similarity values. These ranges are expressed in different
scales for each dataset and can only be compared among
stimulus groups within each dataset (pixels, Euclidean dis-
tance; human, mean similarity rankings; monkey, propor-
tion different responses; neuronal data, distance in
124-dimensional neuronal space). To aid a visual compari-
son of the pixel-based with the other configurations, the
latter were Procrustes transformed (combination of trans-
lation, scaling, rotation and reflection)11. Red lines connect c
corresponding points in both configurations. As the con-
figurations have arbitrary origin, scale and orientation, the
labeling and scale on both axes in each graph are omitted.
and 8 (rs = 0.93, p < 0.01)). The latter deviation was

not a consequence of a poor reliability, as it was repli-
cated. Second, there was a clear difference in the
d
MDS-derived configurations between corner and side
points as defined in the expected configurations, with
larger angles at side points compared to corner points
( 2 = 8.17, p < 0.01 for monkey Y; 2 = 10.74,
p < 0.01 for monkey E).
We determined whether the remaining deviations
of the perceived from the pixel-based similarities were
consistent between the two monkeys. There was a e
significantly high congruence between the data of
monkeys Y and E for stimulus groups A and C,
although no such inter-subject consistency was pre-
sent for group B (Table 1, C(monkey Y, monkey E)).
Each monkey tended to represent stimulus 3 of
group A more toward the center of the configuration
than expected from the pixel-based configurations
(Fig. 2d and e). For group C, the perceived shape
configurations showed a bias towards the vertical f
parametric dimension (Fig. 1c), a trend also present
in human subjects (Fig. 2b and c). This consistency
between humans and monkeys for group C was
assessed using the averaged human and monkey data,
and was significant (Table 1, C(humans, monkeys);
Supplementary Fig. 1, see supplementary informa-
tion page of Nature Neuroscience on line for the appli-
cation of MDS on these averaged data).
neuron in Fig. 3a responded differently to similar shapes within
Neuronal representation space in IT cortex each stimulus group (for example, shapes C1 and C3), whereas it
We recorded from 124 single neurons in area TE of monkeys F responded with similar strength to shapes that look dissimilar to
(n = 51) and M (n = 73) while they were learning to categorize human and monkey observers (for example, shapes C1 and A6).
the 24 stimuli in 2 classes. Recordings were performed in the Many neurons were very selective within stimulus groups, but
lower bank of the superior temporal sulcus and lateral to the ante- stimuli from different groups often elicited similar responses
rior middle temporal sulcus. Most neurons (82%) responded (Fig. 3b). Indeed, the median of the distribution of the selectiv-
twice as strongly to their preferred shape than to the least pre- ity index (SE, see Methods; Fig. 4a) was 2.4, indicating that most
ferred shape. Many of these neurons combined high within-group neurons do not respond more strongly to all stimuli of one group
selectivity with lower between-group selectivity. For example, the compared to the responses to stimuli of the other groups (as can

articles
Table 1. Congruence between configurations for each group of stimuli. shape C1 and was still responsive
for the neighboring shapes C2
Group A Group B Group C
and C8, but no comparable
Congruence between pixel-based similarities responses were elicited at any
and empirical similarities
point further in stimulus space.
C (pixels, human 1) 0.973+ 0.987+ 0.960+
As a consequence, polar plots for
C (pixels, human 2) 0.989+ 0.987+ 0.976+
each stimulus group revealed reg-
C (pixels, humans) 0.991 + 0.995 + 0.974 +
ular and unimodal tuning curves
0.979+ 0.984+ 0.979+
C (pixels, monkey Y) (Fig. 3b). The ratio index (RR)

C (pixels, monkey E) 0.988+ 0.983+ 0.976+ captures the deviation of this tun-
C (pixels, monkeys) 0.989+ 0.990+ 0.979+ ing curve with respect to an uni-
C (pixels, neurons) 0.996+ 0.993+ 0.991+ modal sinusoidal modulation
(see Methods). The three neurons
Congruence between two sets in Fig. 3b illustrate tuning curves
of empirical data across the whole distribution of
C (human 1, human 2) 0.967 0.976 0.981* RR values (Fig. 4b). For all cases
C (monkey Y, monkey E) 0.985* 0.972 0.990* with a good within-category
C (humans, monkeys) 0.987 0.988 0.992*
selectivity (that is, a response dif-
ference between preferred and
C (monkeys, neurons) 0.990* 0.991* 0.984*
least preferred shape above 50%,
+Probability of random configuration, p < 0.001. *Probability that deviations from pixel-based similarities are n = 121), the median RR index
random, p < 0.05. was 1.9. Thus, the responses of
most IT neurons were closely
be illustrated by sorting the stimuli according to response related to the parametric variation of shape similarity.
strength, Supplementary Figs. 1 and 2). We used the responses of the 124 neurons to compute the neu-
The within-category selectivity of these neurons was regular, ron-based similarity between each pair of stimuli. Notwithstand-
meaning that their responses decreased monotonically with ing the low between-group selectivity of many neurons, hierarchical
increasing parametric distance from the preferred shape of a stim- cluster analysis of the neuron-based similarities revealed the expect-
ulus group. For example, the neuron in Fig. 3a responded well to ed clustering at the population level: all stimuli from a group were
Fig. 3. Responses of single IT neurons. (a) Peristimulus time histograms of a single IT neuron from monkey M. The ordering and numbering of the pan-
els is the same as in Fig. 1. Stimulus presentation (300 ms) is indicated by the bar underneath each histogram. The histograms have a bin width of 25 ms
and the height of each bin is normalized to the maximum bin across all histograms (94 spikes/s). (b) Polar plots of the within-group response pattern,
making use of the radial position of each stimulus with respect to the center of the parametric square configuration (numbering of shapes is the same
as in a). The black tuning curves represent the responses of the neuron of (a), whereas the responses of two other single neurons (monkey F) are
shown in red and blue. Responses across all three panels are normalized to the maximum response for each neuron separately (maximum responses
were 27, 59 and 34 spikes/s for the black, blue and red curves, respectively). The standard error of the mean for the maximum of each tuning curve is
indicated. The RR indices for stimulus groups A, B and C were 1.8, 3.6 and 2.8, respectively, for the neuron indicated by black lines (SE = 27); 5.2, 3.0
and 4.8 for the neuron indicated by blue lines (SE = 2.3); and 1.6, 0.68 and 0.33 for the neuron indicated by red lines (SE = 3.8).
articles
a b ties for all stimulus groups (Table 1;

C(monkeys, neurons)). The neuron-based
similarities were a better predictor of the
averaged perceived similarities than were
the pixel-based similarities (the same result
was obtained with the behavioral data con-
sidered for each monkey separately). The
neural as well as the perceptual represen-
tation of stimulus A3 tended more toward

the center of the configuration than
expected from the pixel-based configura-
tions. For group B, both the averaged per-
ceived similarities and the neuron-based
Fig 4. Distribution of SE and RR indices. (a) The SE index is calculated for all neurons responding configurations tend to differentiate the
with at least twice as many spikes to their preferred shape than to the least preferred shape stimuli on the right of the parametric con-
(n = 102). (b) The RR index is computed for all cases with a good within-category selectivity (that figuration more strongly than those on the
is, a response difference between preferred and least preferred shape above 50%, n = 121). All val- left (Fig. 1b; although the behavior-based
ues above five are collapsed in one column (maximum was 38 and 14 for SE and RR, respectively). representation of the stimuli on the left is
The median of each distribution is indicated by an arrow. not consistent between monkeys, each ani-
mal was more sensitive for differences
between stimuli on the right compared to
assigned to the same cluster, before they were clustered with stim- stimuli on the left, as was the population of neurons). The results
uli from other groups. Because 45% of the variance was explained for group C indicate that the neurons tend to be more sensitive
by assigning the stimuli to 3 groups, the degree of clustering at the for differences along the vertical parametric dimension, as were
neuronal level is comparable to that found in the monkey behav- the animals and human subjects at the behavioral level.
ioral data, but less so than in human subjects.
For each stimulus group, the congruence between the neu- Effects of categorization on neuronal representation
ron-based similarities and the pixel-based similarities was high- During the recordings, the monkeys learned to categorize the
ly significant (Table 1). The proportion of variance explained by shapes from each stimulus group into two classes (Fig. 6). In
2-dimensional MDS-derived configurations (Fig. 2f) exceeded some stimulus groups, the division of the stimulus space into
99% for each group. As found for the perceived similarities, the specific regions followed a simple linear decision rule (group C
positions of the eight shapes in these configurations captured and A in monkeys F and M, respectively). In other groups, how-
most aspects of the expected configurations. First, we found the ever, the categorization rule required the stimulus space to be
same stimulus order in the neuron-based configurations com- parceled into specific quadrants (group B in monkey M) or into
pared to the expected configurations for stimulus groups A and arbitrary decision regions (groups A and B in monkey F and
C (r s = 1, p < 0.01), and only a small deviation for group B group C in monkey M). In this last category rule, highly similar
(rs = 0.98, p < 0.01). Second, the MDS-derived configurations shapes require a different response.
were square-like, with larger angles at side points compared to A possible effect of categorizing the images during the
corner points (2 = 8.17, p < 0.01). recordings on the neural responses was examined by comparing
Reasonable fits could already be obtained with smaller samples the neuronal data between monkeys for each stimulus group
of neurons. The mean congruence between neuron- and pixel- separately. Indeed, each monkey learned a different rule for a
based similarities was significant with as few as eight neurons particular group, allowing a comparison between the neural
(Fig. 5a). The congruence depends also on
the tuning properties of the neurons in
addition to the effect of sample size (Fig. a
b
5b). First, congruence was lower for a sam-
ple of neurons with an irregular tuning (RR
< median RR) compared to neurons with a
regular tuning (RR > median RR). Second,
restricting the range of preferred shapes
(tuning optima) reduced congruence when
combining at least four neurons. Thus, both
the regularity of the tuning of single neu-
rons and the range of their preferred shapes
contribute to the efficiency of the coding of
shape similarity at the population level. Fig. 5. Congruence (C) between pixel-based and neuron-based similarities as a function of number
The neuron-based similarities seem to and type of neurons. Each data point represents the mean congruence of 150 random samples
faithfully represent many metric aspects of (whiskers indicate the standard error of the mean). (a) Congruence as a function of sample size.
the perceived similarities of the monkey. The ordinate scale starts from the expected value when the pixel-based similarities are compared
Indeed, the congruence between the behav- with the similarities in random configurations (C = 0.86). The congruence of an individual sample is
significantly different from this expected value for all values above the dashed line (C = 0.96). (b)
ioral and neuronal data was greater than
Difference between the congruence from (a) and the congruence obtained when neurons are
would be expected based on their respective selected from subsets of the population. A positive difference corresponds to a higher congruence
congruences with the pixel-based similari- relative to that shown in (a). RR, ratio index.

articles
a the easier it will be to categorize any given stimulus. Second, the

more closely a stimulus is located to a boundary between two
categories, the more difficult it will be to assign it to one or the
other of those categories.
The errors the monkeys made while learning these catego-
rization problems confirmed both predictions. First, the more
tightly that shapes having to be categorized into the same class
are clustered, the more easily a particular categorization was
b learned (Fig. 7a). Linear rules were learned more easily than a
quadrant rule, but even the latter was more easily learned than
the arbitrary rules. Second, in the case of a linear rule, the cate-
gory assignments of shapes that were close to a category bound-
ary were learned more slowly compared to more distantly located
shapes (Fig. 7b). Nevertheless, performance on the stimuli close
to a linear category border was above 90% after one week of train-
ing. This shows that the monkeys were able to discriminate neigh-
Fig. 6. The division of each stimulus group into two response cate-
gories for each monkey. Filled and open squares refer to stimuli that boring stimuli in the parametric space. So, the difficulties with
were associated with a leftward or a rightward eye movement, respec- the arbitrary rules (performance after several weeks of training
tively. The stimulus order is the same as in Fig. 1. lower than 85%, Fig. 7a) were not merely due to stimulus dis-
criminability, but indeed reflect the clustering of the stimuli with-
in the representation space.
similarity spaces for different category rules. We observed no DISCUSSION

systematic effects of categorization. First, the within-category Our data reveal a rather faithful representation of the physi-
selectivity of single neurons was regular with respect to the para- cal similarities among high-dimensional stimuli in monkeys
metric similarity between stimuli and it did not follow the cat- at the neuronal level. The neuronal representation of a con-
egory rule. For example, when a monkey had learned an figuration of shapes preserves the dimensionality and stimu-
arbitrary category rule, no neurons responded stronger to all lus order of the configuration, and even some metric
stimuli from one category compared to the responses to the properties like the difference between a square and circle con-
other category. Second, arbitrary rules did not change the figuration. The response patterns of single neurons were
dimensionality or the order of shapes within the MDS-derived related to the distance between shapes within these low-
configurations. Third, even more subtle metric changes, like dimensional representation spaces, a property that contributes
expansions/contractions of dimensions or parts of the stimu- to the efficient coding of similarity at the population level.
lus space, were not induced by the categorization task. Notwithstanding the close agreement between physical and
neuron-based similarities, our results revealed that the neu-
The effect of similarity on categorization performance ron-based similarities were a better predictor of perceived sim-
If stimulus similarity determines categorization performance, ilarities than were the physical similarities. Finally, the
then several predictions can be made regarding the difficulty of low-dimensional representation of shape similarity determined
learning the different category rules. First, the more tightly stim- the difficulty of learning specific categorization rules, but the
uli from a given category are clustered (that is, the larger the reverse was not the case: applying a categorization rule did
inter-category distance relative to the intra-category distance), not change the representation of similarity in IT.
a b
Fig. 7. The performance of the two monkeys in the categorization task for each stimulus group. The stimulusresponse associations for each group
and monkey are shown in Fig. 4. (a) Performance averaged across all stimuli from each stimulus group for different sessions: day 1 and 2 (performance
at the end of the first and second session, respectively), and rec 1 and 2 (behavioral performance during the first and second half of the recording ses-
sions, respectively). Point style represents the type of category rule, whereas the line style refers to the different stimulus groups (group A, dotted
line; B, dashed line; C, solid line). Vertical bars indicate 95% confidence intervals of one line plot (the number of observations is equal for the other
stimulus groups). (b) Performance for the linear categorization problems in the first two sessions as a function of the distance between each stimulus
and the optimal decision boundary in parametric space. Pairwise close stimuli are C1, C2, and C5, C6 for monkey F, and A1, A2, and A5, A6 for mon-
key M. Vertical bars indicate 95% confidence intervals for performance with close stimuli in each monkey.
articles
Fig. 8. Possible responses of a hypothetical neuron. The stimuli to

which the neuron responds are indicated with open squares; black cir- a b
cles indicate no response. (a) A representation of three stimuli within
two dimensions. (b) Including a parametric variation of the similarity
between stimuli 1 and 3 requires the addition of a third dimension for all
similarities between stimuli to be captured.
Previous behavioral studies have already shown that

humans and monkeys represent parameterized two- and three-
dimensional shapes in a manner that preserves relative simi-
larities among the stimuli in parameter space (refs. 912 and However, by examining consistent metric deviations from the
Sigala et al., Soc. Neurosci. Abstr. 26, 448.10, 2000). Computa- expected representation spaces we found that the representation
tional work21 has revealed that such representation of similar- of similarity at the behavioral and neuronal level is not com-
ity can be simulated by a network composed of units that pletely faithful at the metric level. Indeed, the neuron-based sim-
encode shapes by computing their similarity to reference shapes ilarities were always a better predictor of the monkey-perceived
(radial basis functions22,23). Our data indicate that shape sim- similarities than were the pixel-based similarities. With the idea
ilarity is implemented at the neuronal level in a similar way. of a network using radial basis functions, one can account for
First, we noted responses of single neurons within a single stim- most aspects of our data, but cannot predict these metric biases
ulus group that follow the pattern expected of radial basis func- (although they could be explained ad hoc). We applied measures
tions: the responses to other shapes than the most optimal of physical similarity that make no assumptions about how the
shape of a group (the reference shape) decreased gradually image is analyzed in the visual system (in contrast to, for exam-
with increasing distance from the position of the reference ple, a wavelet analysis with oriented filters). Of course, physical
shape in parameter space. The responses of single neurons similarity can be quantified by an almost infinite number of mea-
across different stimulus groups seemed to contradict this con- sures beyond the limited set we have used, and some of these
clusion, insofar as many neurons responded with similar alternative measures could do better. But, even in the latter case,
strength to dissimilar shapes from different stimulus groups we need a model that can explain why some measures are better
(Fig. 3). The same observation has been made in previous stud- than others although they all agree in an ordinal sense. Many the-
ies that revealed that images capable of activating an IT neu- oretical models have tried to characterize the nonlinear process-
ron need not be similar to one another17,18,24. It is tempting to ing steps in the hierarchically organized visual system (for
relate this observation to the fact that the concept of radial example, see refs. 2528). They make different predictions about
basis function networks has been introduced to explain the the occurrence of systematic biases toward a higher sensitivity
representation of similar but not distant objects, and that for some stimulus differences than for others, but our study was
objects that are highly dissimilar need not be embedded into not designed to differentiate between these models.
the same low-dimensional space6. However, the previous dif- In agreement with categorization models using the concept
ficulties with finding a correlation between shape similarity of similarity35, we showed that the representation of similarity
and the responses of IT neurons could also be related to the has a profound influence on categorization performance. How-
lack of proper stimulus parameterization. Indeed, the high- ever, computing similarity is only a first step. Learning a partic-
dimensional nature of these complex stimuli makes it difficult ular categorization rule implies that the low-dimensional
to ascertain what relationship the shapes may have with regard representation space is parceled into regions that require the same
to one another as long as similarity is not controlled for in a response. Neuropsychological theories of categorization localize
parametric way. For instance, consider a realistic pattern of the representation of stimulus similarity within extrastriate cor-
responses to three dissimilar stimuli whose physical similari- tex, but the stimulusresponse mapping is presumed to be done
ties can be described within a two-dimensional space (data in other areas such as prefrontal cortex, hippocampus and basal
points 13 in Fig. 8a). Within this stimulus set, a neuron ganglia 2933 . In line with this strict segregation of
responding strongly to stimuli 1 and 3 but not to stimulus 2 stimulusresponse associations from the representation of stim-
would exemplify a response pattern bearing no relationship to ulus similarity, the representation spaces we found at the neu-
the similarity between the stimuli. However, including addi- ronal level in visual cortex were not altered by learning a specific
tional stimuli in the stimulus set (Fig. 8b) might reveal a more categorization rule. Further research is needed to compare the
regular response pattern with a systematic unimodal tuning visual responses in different brain regions using parameterized
for a third stimulus dimension. Thus, we need a parametric stimuli to see how these visual representation spaces are trans-
control of similarity to arrive at a more principled under- formed into category or response spaces.
standing of the tuning characteristics of these neurons. Indeed,
using parameterized stimuli, we found regular tuning curves METHODS
of IT neurons for combinations of complex shape dimensions. Subjects. Four monkeys (Macaca mulatta) and two naive humans were
A second point of agreement between our data and the com- subjects. All procedures34,35 were approved by the K.U. Leuven Ethical
putational efforts is the demonstration that a population of neu- Committee for animal experiments and followed NIH guidelines.
rons tuned to a set of reference shapes can underlie a Stimuli. The stimuli were 24 closed contours defined by 7 RFCs36, and
low-dimensional and ordinally faithful representation of shape were divided into 3 groups with distinct RFCs. Within each group, dif-
configurations. In addition, our results confirm that both ferences between shapes were induced by independently varying the
assumptions of the model, a regular tuning on one hand, and amplitude of two RFCs, creating eight shapes arranged in a square-like
different optimal shapes for different neurons on the other hand, manner in the two-dimensional amplitude space (Fig. 1). The shapes
contribute to the efficiency of the coding of shape similarity. (maximum size, 6) were presented on a gray background (luminance,

articles
17 candelas/m2 (cd/m2)) and filled with noise (pixel luminances, either the parametric configuration, and the unimodality of the selectivity was
0.05 or 34 cd/m2). The mean luminance of the shapes and background determined by a fast Fourier transform of these plots42. Unimodal selec-
was equal. Shape position was randomized within a square region rang- tivities will be reflected in a Fourier spectrum dominated by the first-
ing from 2 to 4 centered on the fixation spot, except during the single- order component. The ratio (RR) of the size of this first-order component
cell recordings where all shapes were presented foveally. to the largest of all other components was taken as a measure of the dom-
As a first physical dissimilarity measure, we computed the distance inance of the first-order component. A high RR value reflects an uni-
between two stimuli in a two-dimensional configuration equal to the para- modal tuning, but a low RR does not necessarily correspond to an
metric space. Second, the Euclidean and the city-block distance between irregular tuning. (Many factors contribute to a low RR, such as an asym-
two shapes, i and j, within the space of 256 256 pixels11,37, was computed: metrical but unimodal tuning.) So, the RR index underestimates the rela-
tionship between shape similarity and IT selectivity.
To analyze the representation of the shape similarities at the population

(x=1:256y=1:256 |pixi(x, y) pixj(x, y)|r)1/r (1)
level, we computed the distance between a pair of stimuli i and j in the
multidimensional space spanned by the responses of all neurons35,43,44:
For shape and background pixels, pix(x, y) is 1 and 0, respectively; for
city-block and Euclidean metrics, r is 1 and 2, respectively. Finally, we
[(n = 1:124 |Respi(n) Respj(n)|2)/124]1/2 (3)
calculated the average information that a pixel of one image provides
about the corresponding pixel in another image 38. The congruence
Here, n is the cell number. The inverse of these distances measures the
between these pixel-based dissimilarity measures was high. The Euclid-
similarity of the neural representations of two shapes.
ean distances provided the best fit with the behavioral and neuron-based
The effect of tuning properties on the coding of similarity at the pop-
similarity measures and are used in the Results.
ulation level was assessed as follows. First, we drew a random sample of
neurons (with replacement) from the population. The sample size ranged
Behavioral tasks. Monkeys Y and E were trained in a same-different task
from 2 to 64, and 150 samples of each size were analyzed (50 for each
with fixation control39. After 700-ms fixation (window, 2), two shapes
stimulus group). For each stimulus group, only neurons that responded
were shown successively for 300 ms (interstimulus interval,
to at least one shape were included, giving a population size of 105, 109
500 ms). A leftward and rightward saccade was the correct response in
and 117 for groups A, B and C, respectively. Because we found no con-
trials with identical (50% of the trials) or different shapes, respectively.
sistent differences among stimulus groups, the results were pooled. Sec-
Aborted trials (for example, responses before the end of the second stim-
ond, we selected neurons from four different subsets. For each subset,
ulus) were not included in the analyses. Correct responses were rewarded
we selected 150 samples for each of 3 sizes (2, 4 or 8). In the first and sec-
by juice. Each monkey was trained for several months in this task using
ond subset, we selected only neurons with RR lower and higher, respec-
a wide variety of images. Their performance level for highly dissimilar
tively, than the median RR of the population. In the third subset, the
novel stimuli was 95% correct. During the experiment, all pairs of shapes
neurons in a sample had similar preferred shapes. In each sample, the
were presented for 16 trials each, whereas additional testing (72 trials/pair)
first neuron was randomly drawn but the other neurons were required
was done for pairs of shapes within a stimulus group. One additional ses-
to have tuning curves with an optimum at the same or a neighboring
sion was run for monkey Y with group B stimuli (34 trials/pair). The per-
stimulus. In the fourth subset, the optima of the neurons were distrib-
formance level was 79% (Y) and 88% (E) correct when all shapes were
uted (optima were 180 apart for two neurons, 90 apart for four neu-
paired and 69% (Y) and 80% (E) correct when only within-group com-
rons, and were at all positions for eight neurons).
parisons were shown, which is as good as in other studies12,20.
The 2 other monkeys learned to categorize the 24 shapes into 2
Analysis of similarity data. The different sets of similarity data were ana-
response categories (Fig. 6). The procedure was the same as for monkey
lyzed using Statistica software (StatSoft, Tulsa, Oklahoma). First, we
Y, except that only one shape was shown, after which the monkeys had
to make either a leftward or a rightward eye movement. Each monkey checked whether the 24 shapes showed the expected clustering in three
had several months of experience in this task with other stimuli. groups by performing hierarchical cluster analysis45. This algorithm starts
Similarity judgments from the human subjects were obtained with a from a configuration with as many clusters as stimuli, and groups simi-
rating method. The rating and the same-different technique produce lar stimuli in several steps (starting with the most similar stimuli) until all
equivalent results11,40. The subjects were asked to fixate a spot while two stimuli form one cluster. We describe the results obtained by using a
stimuli were shown for 300 ms (interstimulus interval, 500 ms). They had weighted pair-group average, but other rules such as single and complete
to rate the similarity between the two stimuli on a scale from one (very linkage revealed the same pattern of results. We have summarized the
similar) to nine (very dissimilar). All possible pairs of shapes were pre- results by describing at which level stimuli are clustered with stimuli from
sented four times for each subject, and additional testing (10 trials/pair) the same and different stimulus groups. The proportion of variance in
was done for pairs of shapes within a stimulus group. the original data that could be accounted for by three clusters of stimuli
measures the degree of clustering. The dissimilarity between stimuli with-
Recordings. IT recordings34,35 were performed in monkeys F and M dur- in a cluster was set to zero, whereas the linkage distance at the level at
ing the categorization task. Recording sites were verified using CT images which two clusters were grouped was taken as the dissimilarity between
with the guiding tube in situ in monkey M and post mortem in monkey F. stimuli from different clusters. Second, we analyzed the similarity data
We searched for responsive neurons by presenting all 24 shapes. Respon- from each stimulus group separately with nonmetric MultiDimensional
sive neurons were investigated further by presenting all shapes for at least Scaling (MDS)43,45.
6 trials (median 12 trials). All further analyses were done using the mean As a statistical criterion to decide whether the perceived and neuron-
number of spikes in the 50350-ms interval after stimulus onset after based similarities converged with the similarities from the parametric or
normalization to the maximum response. All neurons were shape- pixel-based configurations, we computed the congruence coefficient C
selective (one-way ANOVA41). The selectivity index (SE) compared the between different sets A and B of similarity data46:
selectivity within the group to which the optimal stimulus belonged
(maxR and minR(within) being the responses to the best and the worst C = (n = 1:28 dA(n) dB(n))/(ndA2(n) n dB2(n))1/2 (4)
stimulus within that group), with the maximum response elicited by a
stimulus from the other groups (maxR(between)): The index n provides a summation across the 28 pairs of distances. The
closer C approximates 1, the better the fit. The significance of a C value
SE = (maxR minR(within))/(maxR maxR(between)) (2) was evaluated using Monte Carlo simulations (n = 1000) to determine
the expected congruence when a square configuration is compared with
SE ranges between 0 (no within-group selectivity) and infinity (no a random two-dimensional configuration of eight points. This distrib-
between-group selectivity). ution approximated a normal distribution with mean 0.865 and a stan-
The within-group response pattern was represented on a polar plot dard deviation of 0.029, meaning that a value above 0.960 is highly
using the radial position of each stimulus with respect to the center of significant (p < 0.001), which is similar to the result found for a range

articles
of configurations11. As this index is skewed to unity, small differences in 16. Desimone, R., Albright, T. D., Gross, C. G. & Bruce, C. Stimulus-selective
C in the range of 0.951 are meaningful. properties of inferior temporal neurons in the macaque. J. Neurosci. 4,
20512062 (1984).
Measurement noise would introduce deviations of a measured con- 17. Gochin, P. M., Colombo, M., Dorfman, G. A., Gerstein, G. L. & Gross, C. G.
figuration with respect to the actual configuration that are unrelated Neural ensemble coding in inferior temporal cortex. J. Neurophysiol. 71,
from one set of measurements to the other. However, if the deviations 23252337 (1994).
are consistent across datasets, then one can conclude that the actual con- 18. Wilkinson, F., Wilson, H. R. & Habak, C. Detection and recognition of radial
figuration is distorted with respect to the expected configuration. This frequency patterns. Vision Res. 38, 35553568 (1998).
19. Albright, T. D. & Gross, C. G. Do inferior temporal cortex neurons encode
is analogous to determining whether a correlation between two variables shape by acting as Fourier Descriptor filters? Proc. Int. Conf. Fuzzy Logic &
(the two measured configurations) can be explained by their correlations Neural Networks, Izuka, Japan, 375378 (1990).
with a third variable (the expected configuration).To simulate the effect 20. Sands, S. F., Lincoln, C. E. & Wright, A. A. Pictorial similarity judgements and
of measurement noise, we added uniform noise to the similarities expect- the organization of visual memory in the rhesus monkey. J. Exp. Psychol. 111,
369389 (1982).
ed in a square configuration. We used several noise magnitudes to cover 21. Edelman, S. Representation is representation of similarities. Behav. Brain Sci.
the range of congruences found in our study. The fit between two noisy 21, 449467 (1998).
datasets was almost never higher than the best of both fits found by com- 22. Poggio, T. & Girosi, F. Regularization algorithms for learning that are
paring each noisy dataset with the perfect square similarities (p < 0.001). equivalent to multilayer networks. Science 247, 978982 (1990).
Even a fit between two noisy sets that is larger than the worst but lower 23. Poggio, T. & Edelman, S. A network that learns to recognize three-
dimensional objects. Nature 343, 263266 (1990).
than the best of these is significant (p < 0.05). If a configuration is mea- 24. Vogels, R. Categorization of complex visual images by rhesus monkeys. Part
sured twice, then the difference between the measurements will be larg- 2: single-cell study. Eur. J. Neurosci. 11, 12391255 (1999).
er than the difference between one measurement and the configuration. 25. Marr, D. Vision: A Computational Investigation into the Human
If the difference between the two measurements is smaller, then proba- Representation and Processing of Visual Information (Freeman, San Francisco,
California, 1982).
bility is high that a systematic factor determines the way in which the 26. Riesenhuber, M. & Poggio, T. Models of object recognition. Nat. Neurosci. 3,
measurements deviate from the expected configuration (Table 1). 11991204 (2000).
Note: Supplementary information is available on the Nature Neuroscience web 27. Biederman, I. Recognition-by-components: a theory of human image
understanding. Psychol Rev. 94, 115147 (1987).
site (http://neuroscience.nature.com/web_specials). 28. Wallis, G. & Rolls, E. T. Invariant face and object recognition in the visual
system. Prog. Neurobiol. 51, 167194 (1997).
29. Ashby, F. G., Alfonso-Reese, L. A., Turken, A. U. & Waldron, E. M. A
neuropsychological theory of multiple systems in category learning. Psychol.
ACKNOWLEDGEMENTS Rev. 105, 441481 (1998).
This work was supported by the Geneeskundige Stichting Koningin Elizabeth, 30. Murray, E. A., Bussey, T. J. & Wise, S. P. Role of prefrontal cortex in a network
Interuniversitaire Attractiepolen (IUAP P4/22) and the University Research for arbitrary visuomotor mapping. Exp. Brain Res. 133, 114129 (2000).
31. Wise, S. P. & Murray, E. A. Role of the hippocampal system in conditional
Council (IDO/98/002). We thank M. De Paep, P. Kayenbergh, G. Meulemans, motor learning: mapping antecedents to action. Hippocampus 9, 101117
G. Vanparrys, S. Van Wetter and G. Kayaert for technical assistance, and S. (1999).
Raiguel and F. Wichmann for their comments on the manuscript. H.O.d.B. is a 32. Asaad, W. F., Rainer, G. & Miller, E. K. Neural activity in the primate
prefrontal cortex during associative learning. Neuron 21, 13991407 (2000).
research assistant of the Fund for Scientific Research (FWO) Flanders. 33. Freedman, D. J., Riesenhuber, M., Poggio, T. & Miller, E. K. Categorical
representation of visual stimuli in the primate prefrontal cortex. Science 291,
RECEIVED 6 SEPTEMBER; ACCEPTED 25 OCTOBER 2001 312316 (2001).
34. Op de Beeck, H. & Vogels, R. Spatial sensitivity of macaque inferior temporal
neurons. J. Comp. Neurol. 426, 505518 (2000).
1. Edelman, G. M. Neural Darwinism (Basic Books, New York, New York, 1987). 35. Vogels, R., Biederman, I., Bar, M. & Lorincz, A. Inferior temporal neurons
2. Smith, E. E. & Medin, D. L. Categories and Concepts (Harvard Univ. Press, show greater sensitivity to nonaccidental than to metric shape differences.
Cambridge, Massachusetts, 1981). J. Cogn. Neurosci. 13, 444453 (2001).
3. Nosofsky, R. M. Choice, similarity, and the context theory of classification. 36. Zahn, C. T. & Roskies, R. Z. Fourier descriptors for plane closed curves. IEEE
J. Exp. Psychol. Learn. Mem. Cogn. 10, 104114 (1984). Trans. Comp. 21, 269281 (1972).
4. Ashby, F. G. & Perrin, N. A. Toward a unified theory of similarity and 37. Grill-Spector, K. et al. Differential processing of objects under various
recognition. Psychol. Rev. 95, 124150 (1988). viewing conditions in the human lateral occipital complex. Neuron 24,
5. Shepard, R. N. Toward a universal law of generalization for psychological 187203 (1999).
science. Science 237, 13171323 (1987). 38. Rieke, F., Warland, D., de Ruyter van Steveninck, R. R. & Bialek, W. Spikes:
6. Edelman, S. Representation and Recognition in Vision (MIT Press, Cambridge, Exploring the Neural Code Vision (MIT Press, Cambridge, Massachusetts,
Massachusetts, 1999). 1997).
7. Roweis, S. T. & Saul, L. K. Nonlinear dimensionality reduction by locally 39. Robinson, D. A. A method of measuring eye movements using a scleral search
linear embedding. Science 290, 23232326 (2000). coil in a magnetic field. IEEE Trans. Biomed. Eng. 101, 131145 (1963).
8. Tenenbaum, J. B., de Silva, V. & Langford, J. C. A global geometric framework 40. Cortese, J. M. Perceptual Similarity of Closed Contours. Thesis, Univ. of Illinois
for nonlinear dimensionality reduction. Science 290, 23192323 (2000). (1992).
9. Shepard, R. N. & Cermak, G. W. Perceptual-cognitive explorations of a 41. Kirk, R. E. Experimental Design: Procedure for the Behavioral Sciences
toroidal set of free-form stimuli. Cogn. Psychol. 4, 351377 (1973). (Brooks-Cole, Belmont, California, 1968).
10. Cortese, J. M. & Dyre, B. P. Perceived similarity of shapes generated from 42. Worgotter, F. & Eysel, U. T. Quantitative determination of orientational and
fourier descriptors. J. Exp. Psychol. Hum. Percept. Perform. 22, 133143 (1996). directional components in the response of visual cortical cells to moving
11. Cutzu, F. & Edelman, S. Representation of object similarity in human vision: stimuli. Biol. Cybern. 57, 349355 (1987).
psychophysics and a computational model. Vision Res. 38, 22292257 (1998). 43. Young, M. P. & Yamane, S. Sparse population coding of faces in the
12. Sugihara, T., Edelman, S. & Tanaka, K. Representation of objective similarity inferotemporal cortex. Science 256, 13271331 (1992).
among three-dimensional shapes in the monkey. Biol. Cybern. 78, 17 (1998). 44. Kobatake, E., Wang, G. & Tanaka, K. Effects of shape-discrimination training
13. Dean, P. Effects of inferotemporal lesions on the behavior of monkeys. on the selectivity of inferotemporal cells in adult monkeys. J. Neurophysiol.
Psychol. Bull. 83, 4171 (1976). 80, 324330 (1998).
14. Logothetis, N. K. & Sheinberg, D. L. Visual object recognition. Annu. Rev. 45. Shepard, R. N. Multidimensional scaling, tree-fitting, and clustering. Science
Neurosci. 19, 577621 (1996). 210, 390398 (1980).
15. Tanaka, K. Inferotemporal cortex and object vision. Annu. Rev. Neurosci. 19, 46. Borg, I. & Leutner, D. Measuring the similarity of MDS configurations.
109139 (1996). Multivar. Behav. Res. 20, 325334 (1985).

articles
Genetic influences on brain structure

Paul M. Thompson1, Tyrone D. Cannon2, Katherine L. Narr1, Theo van Erp2, Veli-Pekka Poutanen3,
Matti Huttunen4, Jouko Lnnqvist4, Carl-Gustaf Standertskjld-Nordenstam3, Jaakko Kaprio5,
Mohammad Khaledy1, Rajneesh Dail1, Chris I. Zoumalan1 and Arthur W. Toga1
1 Laboratory of Neuro Imaging and Brain Mapping Division (Rm. 4238, Reed Neurological Research Center), Department of Neurology,
UCLA School of Medicine, 710 Westwood Plaza, Los Angeles, California 90095-1769, USA
2 Departments of Psychology, Psychiatry, and Human Genetics, UCLA School of Medicine, 1285 Franz Hall, 405 Hilgard Avenue,
Los Angeles, California 90095-1563
3 Department of Radiology, University of Helsinki Central Hospital, Meilahti Clinics, FIN-00290 Helsinki, Finland
4 Department of Mental Health and Alcohol Research, National Public Health Institute of Finland, Mannerheimintie 166, SF-00300 Helsinki, Finland
5 Department of Public Health, Universities of Helsinki and Oulu, P.O. Box 41, Mannerheimintie 172, University of Helsinki,
FIN-00014 Helsinki, Finland
Correspondence should be addressed to P.T. (thompson@loni.ucla.edu)
Here we report on detailed three-dimensional maps revealing how brain structure is influenced by
individual genetic differences. A genetic continuum was detected in which brain structure was
increasingly similar in subjects with increasing genetic affinity. Genetic factors significantly
influenced cortical structure in Brocas and Wernickes language areas, as well as frontal brain
regions (r2MZ > 0.8, p < 0.05). Preliminary correlations were performed suggesting that frontal gray
matter differences may be linked to Spearmans g, which measures successful test performance
across multiple cognitive domains (p < 0.05). These genetic brain maps reveal how genes determine
individual differences, and may shed light on the heritability of cognitive and linguistic skills, as well
as genetic liability for diseases that affect the human cortex.
The degree to which genes and environment determine brain terning, and brain asymmetry. These features each vary with
structure and function is of fundamental importance. Large- age, gender, handedness, hemispheric dominance and cogni-
scale neuroimaging and genetic studies are beginning to uncov- tive performance in both health and disease. Composite maps
er normal and disease-specific patterns of gene and brain of these features, generated for large populations, can reveal
function in large human populations1,2. Yet, little is known patterns not observable in an individual 12 . Such patterns
about the genetic control of human brain structure, and how include statistical maps that show whether heredity and non-
much individual genotype accounts for the wide variations genetic factors are involved in determining specific aspects of
among individual brains. Recent reports show that many cog- brain structure.
nitive skills are surprisingly heritable, with strong genetic influ- Among the structural features that are genetically regulated and
ences on IQ3,4, verbal and spatial abilities, perceptual speed5 have implications for cortical function is the distribution of gray
and even some personality qualities, including emotional reac- matter across the cortex. This varies widely across normal indi-
tions to stress6. These genetic relationships persist even after viduals, with developmental waves of gray matter gain and loss
statistical adjustments are made for shared family environ- subsiding by adulthood13, and complex deficit patterns observed in
ments, which tend to make members of the same family more Alzheimers disease, schizophrenia, and healthy subjects at genet-
similar. Given that genetic and environmental factors, in utero ic risk for these disorders. In this study, we began by comparing
and throughout lifetime, shape the physical development of the average differences in gray matter (Fig. 1) in groups of unre-
the brain, we aimed to map patterns of brain structure that are lated subjects, dizygotic (DZ) and monozygotic (MZ) twins (see
under significant genetic control, and determine whether these Methods). Although both types of twins share gestational and post-
structural features are linked with measurable differences in gestational rearing environments, DZ twins share, on average, half
cognitive function. The few existing studies of brain structure their segregating genes, whereas MZ twins are normally geneti-
in twins suggest that the overall volume of the brain itself7 and cally identical (with rare exceptions due to somatic mutations).
some brain structures, including the corpus callosum8,9 and We found that brain structure is under significant genetic con-
ventricles, are somewhat genetically influenced, whereas gyral trol, in a broad anatomical region that includes frontal and lan-
patterns, observed qualitatively10 or by comparing their two- guage-related cortices. The quantity of frontal gray matter, in
dimensional projections, are much less heritable11. To make particular, was most similar in individuals who were genetically
the transition from volumes of structures to detailed maps of alike; intriguingly, these individual differences in brain structure
genetic influences, advances in brain mapping technology have were tightly linked with individual differences in IQ (intelligence
allowed the detailed mapping of structural features of the quotient). The resulting genetic brain maps reveal a strong rela-
human cortex, including gray matter distribution, gyral pat- tionship between genes, brain structure and behavior, suggest-

2001 Nature Publishing Group http://neurosci.nature.com articles
Fig. 1. Genetic continuum of similarity in brain structure. Differences in the quantity of gray matter at each region of cortex were computed for iden-
tical and fraternal twins, averaged and compared with the average differences that would be found between pairs of randomly selected, unrelated indi-
viduals (blue, left). Color-coded maps show the percentage reduction in intra-pair variance for each cortical region. Fraternal twins exhibit only 30%
of the normal inter-subject differences (red, middle), and these affinities are largely restricted to perisylvian language and spatial association cortices.
Genetically identical twins display only 1030% of normal differences (red and pink) in a large anatomical band spanning frontal (F), sensorimotor
(S/M) and Wernickes (W) language cortices, suggesting strong genetic control of brain structure in these regions, but not others (blue; the signifi-
cance of these effects is shown on the same color scale).
ing that highly heritable aspects of brain structure may be fun- guage area (r2 = 0.70.8; p < 0.0001) and highly similar in pari-
damental in determining individual differences in cognition. eto-occipital association areas (6070% correlation; p < 0.001).
They also showed significantly less affinity (p < 0.05) in a
RESULTS sharply defined region that included the frontal cortices
MZ within-pair gray matter differences were almost zero (intra- (p > 0.05). The resulting pattern of twin correlations suggests
class r 0.9 and higher, p < 0.0001 corrected; Fig. 1, right col- substantial genetic influences in this region.
umn) in a broad anatomical band encompassing frontal,
sensorimotor and linguistic cortices, including Brocas speech Mapping genetic correlations
and Wernickes language comprehension areas. Because MZ twins With a sample size of only 40 twins, heritability coefficients can-
are genetically identical, any regional differences would be inter- not be estimated precisely, and limited statistical power precludes
preted as being attributable to environmental effects or the detection of differences in heritability between individual
geneenvironment interactions. Meanwhile, sensorimotor and regions of cortex. Preliminary comparisons of MZ and DZ cor-
parietal occipital but not frontal territory was significantly more relations suggested that frontal, sensorimotor and anterior tem-
similar in DZ twins than random pairs (Figs. 1 and 2). Affinity poral cortices were under significant genetic control (p < 0.05,
was greatest in the MZ pairs, suggesting a genetic continuum in rejecting the hypothesis that heritability (h2) = 0; one-tailed). Pre-
the determination of structure. liminary estimates suggested that discrete middle frontal regions,
near Brodmann areas 9 and 46 (ref. 27), displayed a 9095%
A genetic continuum genetic determination of structure (that is, h2 0.900.95). Many
In population genetics, a feature is heritable if it shows a genet- regions are under tight genetic control (bilateral frontal and sen-
ic cascade in which within-pair correlations (Fig. 2) are high- sorimotor regions, p < 0.0001; Fig. 3). Due to small sample sizes,
est for MZ twins, lower for DZ twin pairs and lowest of all for any provisional heritability estimates should be interpreted with
unrelated subjects. As we expected specific regions of cortex caution, but were comparable with twin-based estimates for the
to be more heritable than others, we plotted these correlations most highly genetically determined human traits, including fin-
across the cortex (Fig. 2) and assessed their statistical signifi- gerprint ridge count (h2 = 0.98), height (h2 = 0.66) and systolic
cance (see Methods). This uncovered a successively increasing blood pressure (h2 = 0.57)28. Genetic influences here are far high-
influence of common genetics. A 95100% correlation was er than for the most environmentally influenced characters (such
revealed between MZ twins in frontal, linguistic and parieto- as social maturity, for which h2 = 0.16; ref. 29).
occipital association cortices, suggesting individual differences
in these regions can be largely attributed to genetic factors. DZ Language asymmetry
twins, who share half their genes on average, were still near- Given the high heritability of reading skills and performance on
identical in the supramarginal component of Wernickes lan- linguistic tasks30, we were interested in whether the structure of

articles
Fig. 2. Correlation between twins in gray matter distri-

bution. Genetically identical twins are almost perfectly
correlated in their gray matter distribution, with near-
identity in frontal (F), sensorimotor (S/M) and perisyl-
vian language cortices. Fraternal twins are significantly
less alike in frontal cortices, but are 90100% correlated
for gray matter in perisylvian language-related cortex,
including supramarginal and angular territories and
Wernickes language area (W). The significance of these
increased similarities, visualized in color, is related to

the local intra-class correlation coefficents (r).
language cortices would also be heritable, and if so,

whether heritability would be higher in the left
hemisphere, which is dominant for language in
most (right-handed) subjects. The heritability of
brain size does not vary markedly by hemisphere31,
but one MZ twin study of cortical surface areas32
suggested the possibility of differing left/right
genetic influences. Intriguingly, when a three-
dimensional map was made subtracting the heri-
tability of the structures in one hemisphere from
their counterparts in the other, differences in Wer-
nickes language area were highly significant, even after hemi- h 2 = 0.48 (ref. 33); h 2 = 0.60.8 (ref. 34, compare with
spheric differences in gyral patterning were directly refs. 3538)). We found that differences in frontal gray matter
accommodated. Heritability was significantly greater on the left were significantly linked with differences in intellectual function
(p < 0.05, corrected). Although no other regions displayed this (Table 1; p < 0.0044; p < 0.0176 after correction for multiple tests)
lateralized effect, we cannot infer that there are no such asym- as quantified by g, which was itself also highly heritable
metries elsewhere, as we were underpowered, in a sample of 40, (h2 = 0.70 0.17 in this study). Although these preliminary cor-
to make general comparisons of heritability among cortical relations should be evaluated in a larger sample, a recent abstract
regions. Nonetheless, the asymmetry in language-related cortex also observed that differences in regional gray matter volume
was significant, and was corroborated by the genetic correlation were significantly correlated with differences in IQ, in a sample of
maps as well (Figs. 2 and 3), in that Wernickes and Brocas speech 28 pediatric MZ twin pairs (mean age, 12.1 years) studied volu-
area displayed highly significant heritability on the left metrically (E. Molloy et al., 7th Annual Meeting of the Organiza-
(p < 0.0001) but not on the right (p > 0.05). tion for Human Brain Mapping, 447, Brighton, England, 2001).
In frontal brain regions, a regionally specific linkage has
Cognitive linkages previously been found39 between g and metabolic activity mea-
To make a preliminary assessment of whether gray matter dif- sured by positron emission tomography (PET), suggesting that
ferences between subjects were significantly linked with differ- general cognitive ability may in part derive from a specific
ences in cognitive function, a cognitive measure termed frontal system important in controlling diverse forms of behav-
Spearmans g was assessed for all 40 twins. Like IQ, this widely ior. Frontal regions also show task-dependent activity in tests
used measure isolates a component of intellectual function com- involving working (short-term) memory, divided and sustained
mon to multiple cognitive tests, and has been shown to be high- attention, and response selection40. Genetic factors may there-
ly heritable across many studies, even more so than specific fore contribute to structural differences in the brain that are
cognitive abilities (h 2 = 0.62 (ref. 4, compare with ref. 24); statistically linked with cognitive differences. This is especial-
ly noteworthy, as cognitive performance seems
to be linked with brain structure in the very
regions where structure is under greatest genet-
ic control (Figs. 2 and 3). This emphasizes the
pronounced contribution of genetic factors to
structural and functional differences across indi-
viduals, as detected here in frontal brain regions.
Fig. 3. Significance of genetic control of gray matter dis-

tribution. Brain regions for which cortical gray matter
distribution is under significant genetic control are
shown in red. Frontal (F) and lateral temporal (T)
regions show significant heritability, consistent with
their near-identity in identical twins (Fig. 2) and the
weaker patterns of correlations observed in fraternal
twins, who have less similar genotypes. Wernickes area
shows significantly higher heritability in the left hemi-
sphere (Wleft), which is generally dominant for language
function (p < 0.05 for asymmetry).
articles
Table 1. Random effects analysis regressing individual regional gray matter measures on the IQ measure, Spearmans g
(n = 40 subjects).
(a) Random effects analysis Controlling for overall gray matter only After controlling
for other predictors
Measure Regression Effect size (t) Significance F1,33 Significance

coefficient ()
Whole brain gray matter volume 0.0037 1.73 0.046 3.92 0.0561
Frontal gray matter volume 0.072 1.95 0.029** 9.37 0.0044**
Temporal gray matter volume 0.039 0.23 0.411 3.77 0.0607
Parietal gray matter volume 0.055 0.41 0.343 0.54 0.4690
Occipital gray matter volume 0.033 0.15 0.439 0.04 0.8376
(b) Correlation analyses in independent samples
First sample using twin 1; n = 20 r (twin 1) Significance

Frontal gray matter volume 0.45343 0.0256**
Independent sample using twin 2; n = 20 r (twin 2) Significance
Frontal gray matter volume 0.37392 0.0574*
A highly significant relationship (**p < 0.0044) exists between gray matter volume in the frontal cortex and Spearmans g. The random effects analysis (a) is
most powerful4749. It uses all 40 subjects data and it explicitly models and controls for correlations between twins in both measures4749. The first columns
show regression coefficients, effect sizes (t) and significance values for each regional brain measure, in a step-wise regression where only the predictive effects
of overall gray matter volume (t = 1.73, p < 0.046) on IQ are factored out. In the final two columns, a Type III (simultaneous) regression model48 was used,
meaning that each predictor was tested controlling for all other model terms simultaneously. In this second analysis, correlations between brain regions are
accounted for in assessing the significance of each regional effect. An F statistic and a significance value are shown assessing the fit of each model parameter.
Although the power is substantially less, correlations were also significant if analyses were restricted to independent samples including one twin from each pair.
In (b), correlations are measured in 20 twins (arbitrarily termed twin 1) selected randomly, one from each of the 20 twin pairs (n = 20). Correlations are
repeated in twin 2 (n = 20; using the other subject from each of the 20 twin pairs). Pearson partial correlation coefficients (r), and their significance levels (one-
tailed) are shown, suggesting in each independent sample a positive relationship between (greater) frontal gray matter and better cognitive performance.
Although this analysis is slightly less powerful (L.A. Kurdek, Technical Report, Wright State University, Department of Psychology, 2001, available at
http://www.psych.wright.edu/lkurdek/analyze.htm) due to splitting the sample into halves, the cognitive relationship appears at trend level in one sample
(*p < 0.0574*) and significantly in the other (**p < 0.0256).
DISCUSSION cortex, including frontotemporal dementia and primary pro-

Genetic brain mapping gressive aphasia. The genetic cascades implicated in these dis-
Influences of nature and nurture in the determination of indi- eases may or may not overlap with those involved in cortical
vidual brain structure are not independent; genes necessarily determination, but the genetic coupling of brain structure we
operate through the environment, particularly if they concern report here may result in increased familial liability to cortical
susceptibilities to environmental stressors or hazards41. Nonethe- degenerative disease, specifically in highly genetically determined
less, twin designs can reveal the degree to which heredity is frontal regions. By controlling for nongenetic factors, twin stud-
involved, and the extent to which individual differences can be ies may offer unique advantages in isolating disease-specific dif-
attributed to genetic and environmental factors. Whereas genet- ferences in these highly heritable brain regions.
ic influences strongly determine aspects of intellect and its close- Genetic brain maps, such as those introduced in this study,
ly related traits, the extent to which genes shape brain structure is may reveal how genes determine individual differences in brain
heterogeneous. The gene control of brain structure displays asym- structure and function. Additional linkages were observed
metries that mirror asymmetries in the brains functional orga- between cortical differences and intellectual function, suggest-
nization, and genes strongly control a broad anatomical band ing that genetic brain mapping may shed light on the heritabili-
encompassing frontal, linguistic and sensorimotor cortex. As with ty of cognitive and linguistic skills, as well as familial liability for
any polygenic trait, multiple genes are likely to combine addi- diseases that affect the human cortex.
tively or interact at the same or different loci (dominance or epis-
tasis) to structure the adult brain. Future studies mapping METHODS
quantitative trait loci are likely to provide insight into the genes Subjects. 40 healthy normal subjects, consisting of 10 monozygotic (MZ)
that determine brain structure42, and neurocognitive skills that and 10 dizygotic (DZ) twin pairs, were drawn from a twin cohort con-
in some cases depend on it43. sisting of all the same-sex twins born in Finland between 1940 and 1957,
The tight coupling of brain structure and genetics, particu- inclusive, in which both members of each pair were alive and residing in
Finland as of 1967 (n = 9,562 pairs, 2,495 MZ; 5,378 DZ; 1,689 of
larly in frontal brain regions, may contribute to the genetic lia- unknown zygosity)14. Pairs were excluded if either member or any of
bility for diseases that affect the integrity of the cortex. Frontal their first-degree relatives had a history of hospitalization, medicine pre-
gray matter deficits are found in both schizophrenia patients and scriptions, or work disability due to a psychiatric indication from 1969 to
their healthy first-degree relatives4446, and there is a strong famil- 1991. MZ pairs were matched with the DZ pairs for age (48.2 3.4 years),
ial risk for many neurodegenerative diseases that affect the frontal gender, handedness, duration of cohabitation and parental social class.

articles
Each zygosity group included five male pairs and five female pairs. The anatomic variability in some cortical regions, high-dimensional elastic
study protocol was reviewed and approved by the institutional review matching of cortical patterns21, constrained by all 1520 (40 38) three-
boards of the University of California (Los Angeles) and the National dimensional sulcal models, associated measures of gray matter density
Public Health Institute of Finland, and all subjects signed IRB-approved from homologous cortical regions across subjects. Maps of intra-pair gray
informed-consent forms. matter differences, generated within each MZ and DZ pair, were subse-
quently elastically realigned for averaging across the 10 pairs within each
Cognitive testing. Each twin in a pair received a neuropsychological test group, before inter-group comparisons.
battery15 from a different examiner blind to zygosity. All subjects received
the same test battery in a fixed order. Seventeen different cognitive Mapping genetic correlations and asymmetry. Intra-class correlation
domains were assessed, including verbal and spatial working memory, between pairs of each zygosity was computed at each cortical point, after
selective and divided attention, verbal knowledge, motor speed and visu- testing for heteroscedastic variance across each group. First, to assess
ospatial ability. A measure of general cognitive ability, in the form of an whether it was significantly non-zero, broad-sense heritability was com-
overall IQ, was prorated from age-scaled scores on the Vocabulary, Sim- puted using Falconers method22 to determine all genic influences on
ilarities, Block Design, and Digit Symbol subtests of the Wechsler Adult the phenotype (with heritability, h2, defined as twice the difference
Intelligence ScaleRevised (WAIS-R; D. Wechsler, WAIS-R Manual, Psy- between MZ and DZ intra-class correlation coefficients). Because non-
chological Corporation, Cleveland). This measure exhibited 98% corre- genetic familial effects contribute to the resemblance between relatives,
lation with full-scale IQ based on all of the WAIS-R subtests. such effects were accommodated, if not entirely eliminated, by assuming
the same common environmental variance for MZ and DZ pairs (com-
Zygosity. For all pairs, zygosity was determined by DNA analysis using pare with ref. 4). For this study, random field models were preferred as
the following markers: DIS80 (20 alleles), DI7S30 (13 alleles), apoB (20 opposed to a full structural equation model (for example, ref. 23) given
alleles), COL2A1 (10 alleles), vWA (9 alleles) and HUMTH01 (6 alleles). the low degrees of freedom per point available to estimate dominance
Assuming an average heterozygosity rate of 70% per marker, this proce- and epistatic variance terms and reject simpler models based on the
dure will falsely classify a DZ pair as MZ in approximately 1/482 cases. available database of 40 scans. Interaction and geneenvironment covari-
ance terms, as well as unique and shared environment factors, may be
Magnetic resonance imaging. Three-dimensional maps of gray matter estimable with a more general familial design, an adoption design, or
and models of cortical surface anatomy were derived from high-resolu- by using sample sizes much larger than available in the present study23
tion three-dimensional (2562 124 resolution) T1-weighted (MPRAGE) (compare with ref. 24). The significance of genetic effects was comput-
magnetic resonance images acquired from all 40 subjects on a 1.5 T ed point-wise by reference to an analytical null distribution (F-test) and
Siemens scanner (New York, New York). was confirmed separately by assembling an empirical null distribution
using 1,000,000 random pairings to avoid assuming bivariate normali-
Image processing and analysis. A radio-frequency bias field correction ty. All map-based inferences were corrected for multiple comparisons
algorithm eliminated intensity drifts due to scanner field inhomogeneity. by permutation. We used permutations to make statistical inferences
A supervised tissue classifier generated detailed maps of gray matter, that were not based on any assumptions about the error covariances. To
white matter and cerebrospinal fluid (CSF). Briefly, 120 samples of each correct for the multiple comparisons implicit in our brain maps we
tissue class were interactively tagged to compute the parameters of a established the null distribution of the largest statistic over the voxels
Gaussian mixture distribution that reflects statistical variability in the analyzed. By adopting the critical threshold of this largest statistic, we
intensity of each tissue type16,17. A nearest-neighbor tissue classifier could then maintain both strong and weak control over false-positive
assigned each image voxel to a particular tissue class (gray, white or CSF), rates over the voxels analyzed.
or to a background class. The inter/intra-rater reliability of this proto- Asymmetric heritability was tested by computing a set of 40 flows
col, and its robustness to changes in image acquisition parameters, have driving each subjects left hemisphere model onto the right, matching
been described previously17. The error variance, that is, the variation gyral patterns, and computing a field of heritability differences. This
associated with map error and reproducibility, was further confirmed to field was compared with its standard error (pooled across contralateral
be small by the extremely high intra-class correlations in the MZ pairs cortical points, after testing equality of variance across hemispheres
(around 1.0), which would not otherwise be obtainable (Fig. 2). Gray (compare with ref. 26)). Regions of asymmetric heritability were detect-
matter maps were retained for subsequent analysis. ed and their significance was assessed by permuting the covariate vec-
tor coding for hemisphere. Maps of MZ intra-pair gray matter
Three-dimensional cortical maps. To facilitate comparison and pooling differences associated with intra-pair differences in the cognitive mea-
of cortical data across subjects, a high-resolution surface model of the sure, Spearmans g, were generated by elastically realigning three-dimen-
cortex was automatically extracted for each subject18. Thirty-eight gyral sional maps for averaging across all MZ twin pairs and modeling g as a
and sulcal boundaries, representing the primary gyral pattern of each continuous covariate for linkage with local gray matter distribution.
subject, were digitized on the highly magnified three-dimensional sur- Maps identifying these linkages were computed point-wise across the
face models. Gyral patterns and cortical models were used to compute a cortex and assessed statistically by permutation by computing the area of
three-dimensional vector deformation field, which reconfigures each the average cortex with statistics above a fixed threshold in the signifi-
subjects anatomy to the average configuration of the entire group cance maps (p < 0.01). Null distributions were assembled from random
(n = 40), matching landmark points, surfaces and curved anatomic inter- pairings of unrelated subjects. We preferred this to an analytical null
faces. Data were accordingly averaged or compared, to the maximum distribution to avoid assuming that the smoothness tensor of the resid-
possible degree, across corresponding cortical regions12. Additional three- uals of the statistical model was stationary across the cortical surface26.
dimensional vector deformation fields reconfigured one twins anatomy In each case, the covariate vector was permuted 1,000,000 times on an
into the shape of the other, matching landmark points, surfaces and SGI RealityMonster supercomputer with 32 internal R10000 processors.
curved anatomic interfaces on the pair of three-dimensional image sets. An algorithm was then developed to report the significance probability
Given that the deformation maps associate cortical locations with the for each map as a whole12, so the significance of intra-pair variance
same relation to the primary folding pattern across subjects, a local mea- reduction by zygosity, heritability, asymmetry and cognitively linked
surement of gray matter density was made in each subject and averaged patterns of gray matter distribution could be assessed after the appro-
across equivalent cortical locations. priate correction for multiple comparisons.
Gray matter mapping. To quantify local gray matter, we used a measure

termed gray matter density, which has been used in previous studies to
ACKNOWLEDGEMENTS
Grant support was provided by a P41 Resource Grant from the National Center
compare the spatial distribution of gray matter across subjects19,20,12. This
measures the proportion of tissue that segments as gray matter in a small for Research Resources (P.T., A.W.T.; RR13642) and by a National Institute of
region of fixed radius (15 mm) around each cortical point. Given the large Mental Health grant (T.D.C.). Additional support for algorithm development

articles
was provided by the National Library of Medicine, NINDS, the National Science 24. Feldman, M. W. & Otto, S. P. Twin studies, heritability, and intelligence.
Foundation, and a Human Brain Project grant to the International Consortium Science 278, 13831384 (1997).
25. Loehlin, J. C. & Nichols, R. C. Heredity, Environment and Personality (Univ. of
for Brain Mapping, funded jointly by NIMH and NIDA. Special thanks go to Texas Press, Austin, Texas, 1976).
U. Mustonen, A. Tanksanen, T. Pirkola, and A. Tuulio-Henriksson for their 26. Thompson, P. M. et al. in Handbook on Medical Image Analysis (ed.
contributions to subject recruitment and assessment. Fitzpatrick, M.) 10661131 (SPIE, Bellingham, Washington, 2000).
27. Rajkowska, G. & Goldman-Rakic, P. S. Cytoarchitectonic definition of
prefrontal areas in the normal human cortex: II. Variability in locations of
RECEIVED 4 JUNE; ACCEPTED 19 OCTOBER 2001 areas 9 and 46 and relationship to the Talairach Coordinate System. Cereb.
Cortex 5, 323337 (1995).
28. Smith, C. in Textbook of Human Genetics (eds. Fraser, G. & Mayo, O.
1. Collins, F. S. & McKusick, V. A. Implications of the Human Genome Project Blackwell, Oxford, 1975).
for medical science. JAMA 285, 540544 (2001). 29. Huntley, R. M. C. in Genetic and Environmental Factors in Human Ability
2. Huerta, M. F. & Koslow, S. H. Neuroinformatics: opportunities across (eds. Meade, J. E. & Parkes, A. S., Oliver and Boyd, Edinburgh, Scotland,
disciplinary and national borders. Neuroimage 4, 46 (1996). 1966).
3. Plomin, R. & Loehlin, J. C. Direct and indirect IQ heritability estimates: a 30. Gayan, J. & Olson, R. K. Reading disability: evidence for a genetic etiology.
puzzle. Behav. Genet. 19, 331342 (1989). Eur. Child Adolesc. Psychiatry 8, 5255 (1999).
4. McClearn, G. E. et al. Substantial genetic influence on cognitive abilities in 31. Pennington, B. F. et al. A twin MRI study of size variations in human brain.
twins 80 or more years old. Science 276, 15601563 (1997). J. Cogn. Neurosci. 12, 223232 (2000).
5. Alarcn, M., Plomin, R., Fulker, D. W., Corley, R. & DeFries, J. C. 32. Tramo, M. J. et al. Surface area of human cerebral cortex and its gross
Multivariate path analysis of specific cognitive abilities data at 12 years of age morphological subdivisions. J. Cogn. Neurosci. 7, 292301 (1995).
in the Colorado Adoption Project. Behav. Genet. 28, 255264 (1998). 33. Devlin, B., Daniels, M. & Roeder, K. The heritability of IQ. Nature 388,
6. Eley, T. C. & Plomin, R. Genetic analyses of emotionality. Curr. Opin. 468471 (1997).
Neurobiol. 7, 279284 (1997). 34. Finkel, D. et al. Longitudinal and cross-sectional twin data on cognitive
7. Tramo, M. J. et al. Brain size, head size, and intelligence quotient in abilities in adulthood: the Swedish adoption/twin study of aging. Devel.
monozygotic twins. Neurology 50, 12461252 (1998). Psychol. 34, 14001413 (1998).
8. Oppenheim, J. S., Skerry, J. E., Tramo, M. J. & Gazzaniga, M. S. Magnetic 35. Swan, G. E. et al. Heritability of cognitive performance in aging twins. The
resonance imaging morphology of the corpus callosum in monozygotic National Heart, Lung, and Blood Institute Twin Study. Arch. Neurol. 47,
twins. Ann. Neurol. 26, 100104 (1989). 259262 (1990).
9. Pfefferbaum, A., Sullivan, E. V., Swan, G. E. & Carmelli, D. Brain structure in 36. Loehlin, J. C. Partitioning environmental and genetic contributions to
men remains highly heritable in the seventh and eighth decades of life. behavioral development. Am. Psychol. 44, 12851292 (1989).
Neurobiol. Aging 21, 6374 (2000). 37. Chipuer, H. M., Rovine, M. J., Plomin, R. LISREL Modeling: genetic and
10. Biondi, A. et al. Are the brains of monozygotic twins similar? A three- environmental influences on IQ revisited. Intelligence 14, 1129 (1990).
dimensional MR study. Am. J. Neuroradiol. 19, 13611367 (1998). 38. Plomin, R. & Petrill, S.A. Genetics and intelligence: whats new? Intelligence
11. Bartley, A. J., Jones, D. W. & Weinberger, D. R. Genetic variability of human 24, 5378 (1997).
brain size and cortical gyral patterns. Brain 120, 257269 (1997). 39. Duncan, J. et al. A neural basis for general intelligence. Science 289, 457460
12. Thompson, P. M. et al. Cortical change in Alzheimers disease detected with a (2000).
disease-specific population-based brain atlas. Cereb. Cortex 11, 116 (2001). 40. Funahashi, S., Bruce, C. J. & Goldman-Rakic, P. S. Mnemonic coding of
13. Giedd, J. N. et al. Brain development during childhood and adolescence: a visual space in the monkeys dorsolateral prefrontal cortex. J. Neurophysiol.
longitudinal MRI study. Nat. Neurosci. 2, 861863 (1999). 61, 331349 (1989).
14. Kaprio, J., Koskenvuo, M. & Rose, R. J. Change in cohabitation and intrapair 41. Rutter, M., Silberg, J., OConnor, T. & Simonoff, E. Genetics and child
similarity of monozygotic (MZ) cotwins for alcohol use, extraversion, and psychiatry: I. Advances in quantitative and molecular genetics. J. Child
neuroticism. Behav. Genet. 20, 265276 (1990). Psychol. Psychiatry 40, 318 (1999).
15. Cannon, T. D. et al. The inheritance of neuropsychological dysfunction in 42. Williams, R. W., Strom, R. C., & Goldowitz, D. Natural variation in neuron
twins discordant for schizophrenia. Am. J. Hum. Gen. 67, 369382 (2000). number in mice is linked to a major quantitative trait locus on Chr 11.
16. Zijdenbos, A. P. & Dawant, B. M. Brain segmentation and white matter lesion J. Neurosci. 18, 138146 (1998).
detection in MR images. Crit. Rev. Biomed. Eng. 22, 401465 (1994). 43. Hill, L. et al. DNA pooling and dense marker maps: a systematic search for
17. Sowell, E. R. et al. In vivo evidence for post-adolescent brain maturation in genes for cognitive ability. Neuroreport 10, 843848 (1999).
frontal and striatal regions. Nat. Neurosci. 2, 859861 (1999). 44. Cannon, T. D. et al. Regional gray matter, white matter, and cerebrospinal
18. MacDonald, D., Avis, D. & Evans, A. C. Multiple surface identification and fluid distributions in schizophrenic patients, their siblings, and controls.
matching in magnetic resonance images. Proc. SPIE 2359, 160169 (1994). Arch. Gen. Psychiatry 51, 651661 (1998).
19. Wright, I. C. et al. A voxel-based method for the statistical analysis of gray and 45. Weinberger, D. R., DeLisi, L. E., Neophytides, A. N. & Wyatt, R. J. Familial
white matter density applied to schizophrenia. Neuroimage 2, 244252 aspects of CT scan abnormalities in chronic schizophrenic patients.
(1995). Psychiatry Res. 4, 6571 (1981).
20. Ashburner, J. & Friston, K. J. Voxel-based morphometrythe methods. 46. Suddath, R. L. et al. Anatomical abnormalities in the brains of monozygotic
Neuroimage 11, 805821 (2000). twins discordant for schizophrenia. N. Engl. J. Med. 322, 789794 (1990).
21. Thompson, P. M., Woods, R. P., Mega, M. S. & Toga, A. W. 47. Kreft, I. & de Leeuw, J. Introducing Multilevel Modeling (Sage, London,
Mathematical/computational challenges in creating deformable and 1998).
probabilistic atlases of the human brain. Hum. Brain Mapp. 9, 8192 (2000). 48. Littell, R. C., Milliken, G. A., Stroup, W. W. & Wolfinger, R. D. SAS System for
22. Falconer, D. S. Introduction to Quantitative Genetics Edn. 3 (Longman, Essex, Mixed Models (SAS Institute, Cary, North Carolina, 1996).
UK, 1989). 49. Hedeker, D., Gibbons, R. D. & Flay, B. R. Random-effects regression models
23. Neale, M. C. & Cardon, L. R. Methodology for Genetic Studies of Twins and for clustered data with an example from smoking prevention research.
Families (Kluwer Academic, Boston, 1992). J. Consult. Clin. Psychol. 62, 757765 (1994).

articles
Human memory formation is

accompanied by rhinalhippocampal
coupling and decoupling
Jrgen Fell1, Peter Klaver1, Klaus Lehnertz1, Thomas Grunwald1, Carlo Schaller2,
Christian E. Elger1 and Guilln Fernndez1
Departments of Epileptology1 and Neurosurgery2, University of Bonn, Sigmund-Freud Str. 25, D-53105 Bonn, Germany
Correspondence should be addressed to J.F. (juergen.fell@meb.uni-bonn.de)
In humans, distinct processes within the hippocampus and rhinal cortex support declarative memory
formation. But do these medial temporal lobe (MTL) substructures directly cooperate in encoding
new memories? Phase synchronization of gamma-band electroencephalogram (EEG) oscillations
(around 40 Hz) is a general mechanism of transiently connecting neural assemblies. We recorded
depth-EEG from within the MTL of epilepsy patients performing a memorization task. Successful as
opposed to unsuccessful memory formation was accompanied by an initial elevation of
rhinalhippocampal gamma synchronization followed by a later desynchronization, suggesting that
effective declarative memory formation is accompanied by a direct and temporarily limited coopera-
tion between both MTL substructures.
Lesions of certain MTL substructures, most notably of the hip- from rhinal and hippocampal electrodes in epilepsy patients dur-
pocampus and the parahippocampal region, disturb the declar- ing a word memorization task. We found successful declarative
ative memory system, the system that makes memory accessible memory formation to be associated with a transient reduction
to conscious recollection14. Neuroimaging studies suggest that of gamma power in rhinal and hippocampal recordings together
both these MTL substructures participate in memory forma- with an initial enhancement of gamma synchronization between
tion510. A process in the rhinal cortex, which consists of the his- both MTL substructures, followed by a later desynchronization.
tologically distinct entorhinal and perirhinal cortices and is part
of the parahippocampal region, precedes a hippocampal RESULTS
process11. Anatomically, the perirhinal and parahippocampal We took EEG recordings from nine patients with unilateral tem-
cortices provide most of the neocortical input to the entorhinal poral lobe epilepsy while they performed a single-trial word-list
cortex, which in turn provides the predominant cortical input to learning protocol with a free recall test after a distraction task.
the hippocampus via the perforant path1214. However, there is no We implanted multicontact depth electrodes bilaterally along the
stringent evidence for a direct interaction between rhinal cortex longitudinal axis of each MTL during presurgical evaluation
and hippocampus during declarative memory formation in because the zone of seizure onset could not be determined
humans15. Moreover, the exact time when such a presumed inter- unequivocally by noninvasive investigations23. Depth electrodes
action might take place is unknown. had a cylindrical surface area of 10 mm2. Sensitivity is maximal
Phase synchronization of gamma oscillations (electrical brain for field potentials generated within the adjacent brain tissue and,
activity) of around 40 Hz is a general mechanism underlying in general, decays with the inverse square of the distance24. For
transient functional coupling of neural assemblies16,17. This example, compared to adjacent brain tissue 0.1 mm away from
mechanism provides an explanation for the flexibility and speci- the recording electrode (0.1 mm is the order of magnitude of the
ficity of functional associations between brain modules. Evidence thickness of hippocampal layers), a source 1 cm away only con-
has accumulated that phase coupling of induced (that is, not tributes 0.01% to the recorded signal. The placement of electrode
stimulus-locked) gamma activity is essential in object represen- contacts within the hippocampus and the anterior parahip-
tation. Different object features processed by distinct neural pocampal gyrus, which is covered by rhinal cortex13, were ascer-
assemblies are bound together to one coherent percept by syn- tained by magnetic resonance images in each patient (Fig. 1).
chronized induced activity in the gamma range18,19. Long-range Only EEG recordings from the MTL contralateral to the zone of
association of cortical modules via induced gamma-band cou- seizure origin were analyzed to reduce poorly controllable effects
pling subserves the integration of cognitive processes2022. In introduced by the epileptic process25. Moreover, none of the
contrast, the evoked (stimulus-locked) gamma response seems MTLs investigated here showed any pathology, such as hip-
not to be involved in assembly coupling, and its functional role is pocampal atrophy, on clinical MR scans performed during the
still unclear19. Thus, we analyzed induced gamma synchroniza- presurgical work-up. Within these non-epileptic MTLs, we ana-
tion of depth-EEG activity, which was recorded simultaneously lyzed rhinal and hippocampal recordings with the best signal-to-

articles
Fig. 1. Localization of MTL depth electrodes. (a) Hippocampally adjusted

a axial MR image of a patient with bilateral depth electrodes in situ (Hi, hip-
pocampus; R, right; L, left). Due to MR artifacts, the electrodes appear
much larger than their actual size of 1 mm diameter. (b) Black dots (coro-
M nal) and bars (sagittal) in standardized drawings indicate the approximate
location of MTL electrode contacts, which were used to record the EEG
tracings analyzed in this study (Am, amygdala; Cs, collateral sulcus; Hi,
hippocampus; Pg, parahippocampal gyrus; Rc, rhinal cortex).
we investigated induced gamma activity, that is, gamma activi-

ty occurring in a non-time-locked fashion in response to the
stimuli, analysis was based on single-trial evaluations19. Phase
b synchronization values between electrode contacts within the
rhinal cortex and the hippocampus were calculated from the
Rhinal cortex
individual wavelet-transformed EEG segments. The higher the

synchronization value, the more constant is the phase differ-
ence between the two electrodes over all trials. Additionally,
averaged power values were determined separately for rhinal
and hippocampal recordings for subsequently recalled and
unrecalled words. Finally, synchronization and power values
were averaged for consecutive 100-ms time windows from
100 ms to 1500 ms relative to stimulus onset.
The time course of phase synchronization between rhinal cor-
Hippocampus
tex and hippocampus averaged over all frequencies between 32

and 48 Hz shows a dissociation between subsequently recalled
and unrecalled words starting within the first 100 ms after stim-
ulus onset (Fig. 2; main effect of time, F14,1008 = 3.47, p < 0.001,
= 0.60 and interaction of memory time, F 14,1008 = 3.20,
p < 0.01, = 0.50). Average gamma synchronization between rhi-
Coronal Sagittal
nal and hippocampal recordings was increased by up to 16% for
subsequently recalled as opposed to subsequently forgotten words
from 100 up to 300 ms (main effects of memory, 100200 ms,
noise ratio, assessed by amplitudes of event-related potentials F1,72 = 9.39, p < 0.005; 200300 ms, F1,72 = 3.94, p = 0.051). After
(ERPs) recorded during the same task11. The average distance the early enhancement in gamma synchronization, a second
between the selected rhinal and hippocampal locations was increase was detected from 500 to 600 ms (main effect of mem-
1.6 0.25 cm (mean s.e.m.; range, 0.82.6 cm). ory, F1,72 = 6.17, p < 0.02) and, finally, a decrease of synchro-
Each patient participated in 20 study test blocks plus 2 nization was observed from 1000 to 1100 ms (main effect of
training blocks immediately before the experiment, each con- memory, F1,72 = 6.62, p < 0.02; Fig. 2).
taining 12 semantically unrelated German nouns. Patients were The early synchronization effect is most pronounced in the
instructed to memorize each word presented sequentially on frequency range from 36 to 40 Hz and for these frequencies reach-
a computer monitor. To prevent ongoing rehearsal, a distrac- es up to 30% (Fig. 3). The second synchronization increase
tion task was conducted after each block. Thereafter, patients (500600 ms) and the later desynchronization are prominent in
were asked to recall freely the previously displayed words in the frequency range from 32 Hz to 40 Hz. Phase lag distributions
any order (mean recall rate, 29.7%; range, 20.054.6%). In for both conditions (subsequently recalled and unrecalled words)
accord with neuropsychological findings in large series of have a Gaussian shape and are centered around zero (Fig. 4). The
patients with temporal lobe epilepsy26, memory performance difference in synchronization for successful and unsuccessful
was poorer in the patients investigated here than in healthy encoding thus arises from a narrowing of the phase lag distrib-
subjects performing a similar task27. ution caused by an increased amount of phase differences close to
To compare successful and unsuccessful memory encoding,
EEG was separated offline into segments for subsequently
synchronization relative to baseline level (mean s.e.m.)
recalled and unrecalled items. To obtain an optimal time, as

well as frequency resolution within the gamma band, we then
Percentage change of rhinalhippocampal
applied a wavelet technique to the EEG instead of the tradi-

tional coherence method. EEG was wavelet-filtered in the
gamma frequency range from 32 to 48 Hz (2 Hz steps). Because
Fig. 2. Changes of phase synchronization between rhinal cortex and hip-

pocampus (%) relative to prestimulus baseline for subsequently recalled
versus unrecalled words. Synchronization values were averaged over all
analyzed gamma frequencies (3248 Hz); mean and s.e.m. are plotted.
The x-axis depicts the time with respect to stimulus onset (word presen-
tation). Time course of synchronization values was calculated from over-
lapping windows of 100 ms duration shifted in steps of 20 ms. Time (s)

2001 Nature Publishing Group http://neurosci.nature.com articles
Fig. 3. Differences of phase synchronization between rhinal cortex and hippocampus (%) relative to prestimulus baseline for subsequently recalled
versus unrecalled words. Color-coded plot of S(recalled) S(unrecalled), with S(recalled) = phase synchronization (%) for subsequently recalled
words and S(unrecalled) = phase synchronization (%) for subsequently unrecalled words. The different gamma range frequencies (3248 Hz) are rep-
resented in the y direction and time is depicted in the x direction. Synchronization/desynchronization is coded on a color scale: red areas show an
enhancement; blue areas, a reduction of synchronization for subsequently recalled versus unrecalled words.
zero. This finding indicates that rhinal and hippocampal neu- pocampal recordings, gamma power was significantly diminished
rons oscillate together in a more synchronous rhythm when in EEG segments related to successful as opposed to unsuccessful
encoding leads to successful memory formation. memory formation. This difference was detected between 100
To explore the possibility that the subsequent memory effects and 400 ms after stimulus onset.
identified here could be related to general, ubiquitous effects Finally, we compared absolute synchronization and power
found throughout the brain, we examined the synchronization values for subsequently recalled and unrecalled words during
between rhinal cortex and a temporolateral location (gyrus tem- the prestimulus baseline window (100 to 0 ms) to examine
poralis superior). For identification of the zone of seizure onset, whether EEG changes before word presentation are responsible
EEG recordings were made from this location with subdural strip for our findings, which would have suggested a slowly modu-
electrodes in seven of the nine patients28. The two remaining lated encoding state rather than transient processes8. However,
patients had no electrodes outside the MTL. The analysis of rhi- we did not find significant baseline differences between subse-
nal-temporolateral synchronization values revealed neither a quently recalled and unrecalled trials for either rhinalhip-
memory effect (F1,54 = 1.20, p = 0.28), nor a memory time pocampal synchronization values (F1,72 = 0.28, p = 0.60), or for
interaction (F14,756 = 1.15, p = 0.33, = 0.63). Moreover, none gamma power from rhinal (F 1,72 = 1.85, p = 0.18) or hip-
of the individual time windows showed a statistically significant pocampal (F1,72 = 1.05, p = 0.31) recordings.
memory effect (each p > 0.05).
Absolute gamma power values at hippocampal sites were DISCUSSION
about threefold larger than values from rhinal contacts (average Our results show EEG activity in the gamma frequency range in
over all frequencies, time windows and conditions, for hip- field recordings from within the human MTL during a memory
pocampus, 39.6 39.8 V2; for rhinal cortex, 14.2 12.2 V2; task. A previous study29 revealed a generally higher gamma power
paired two-tailed t-tests for each individual frequency, condition in parahippocampal than neocortical recordings in humans. We
and time window; all p < 0.05, T8 > 2.35). The time course of extend this knowledge by showing that gamma power in hip-
gamma power at rhinal cortex and hippocampus (Fig. 5) disso- pocampal recordings is even threefold higher than in the parahip-
ciated significantly between conditions (ANOVA effects, for rhi- pocampal region, suggesting that high-frequency oscillations of
nal cortex, time (F 14,1008 = 14.64, p < 10 12 , = 0.41) and around 40 Hz have a prominent involvement in medial temporal
memory time (F14,1008 = 4.64, p < 0.0001, = 0.54); for hip- and especially hippocampal information processing.
pocampus, time (F14,1008 = 4.68, p < 0.0001, = 0.62) and mem- Intracranial EEG recordings allow the reliable separation of
ory time (F14,1008 = 6.37, p < 107, = 0.59)). In the rhinal synchronization and power effects. In view of the anatomical
cortex, a gamma peak was observed for both conditions at around proximity of the inspected areas (mean distance, 1.6 cm), such
250 ms. However, gamma power was reduced for subsequently a separation would be impossible with surface EEG record-
recalled compared to unrecalled words (significant reductions ings 30,31 . Previous ERP data indicate 11,32 that there is no
from 600 to 800 ms and 1300 to 1400 ms). Similarly, in hip- detectable correlation between EEG recorded from within the

articles
450
Subsequently recalled words
Fig. 4. Distribution of phase differences between rhinal cortex
Subsequently unrecalled words and hippocampus in the gamma frequency range (3248 Hz) for
400 subsequently recalled versus unrecalled words. Phase differences
were evaluated from the time window between 100 and 200 ms
Number of occurences
350
after item presentation.
300
the hippocampus11 and terminate the communication

250
between both structures. Such a functional decoupling has

200 been found in a visual face perception task and has been
termed active desynchronization21.
150 The timing of rhinalhippocampal coupling and
-180 -90 0 90 180
decoupling fits well with the sequence of processes as
Phase differences (degrees)
monitored by ERPs recorded separately from the rhinal
cortex and hippocampus during the same task11. Rhi-
hippocampus and the rhinal cortex, even with electrode distances nal ERPs in response to subsequently recalled words start to
of less than 1 cm. The large anterior medial temporal lobe N400 differ from ERPs in response to subsequently forgotten words
component, for instance, which reflects word processing and can about 300 ms after stimulus onset. This subsequent memory
be recorded with an amplitude of up to 70 V from rhinal cortex, effect in the rhinal cortex is followed by a hippocampal effect
is usually not observable in recordings from within the hip- some 200 ms later that lasts until about 2000 ms after stimu-
pocampus11,27. On the other hand, hippocampal activity is shield- lus onset. Assuming that rhinalhippocampal information
ed toward the outside by the radial cylindrical arrangement of transfer occurs between the onset of the rhinal and the hip-
hippocampal pyramidal layers33. Thus, it is highly unlikely that pocampal ERP effects, the early beginning of gamma phase
our results are biased by correlated EEG recordings. We found coupling revealed here would allow the preparation for and the
successful memory formation to be accompanied by two factors: actual transfer of information. The decoupling observed fol-
an early increase and a later decrease in gamma synchronization lows the end of the rhinal subsequent memory effect at about
between rhinal and hippocampal recording sites, and a transient 900 ms after stimulus onset, the time point when information
reduction of gamma power at both locations partly within the transfer to the hippocampus might be accomplished.
same time window.
The enhancement of gamma-band synchronization observed
here between the rhinal cortex and the hippocampus occurs at
zero-phase lag. Such a synchronization requires highly accurate
timing of neural discharges within a time range of just a few mil-
liseconds34. By achieving this, gamma synchronization enables a
precise functional association between specific brain regions over
short as well as longer distances16,17. Thus, gamma-band cou-
Percentage changes of gamma band power relative to baseline (mean s.e.m.)
pling between rhinal cortex and hippocampus is likely to estab-

lish a transient connection between both MTL structures
initializing declarative memory formation.
The time course of modulation of gamma synchronization
found here is consistent with reports of altered firing rates of sin-
gle anterior rhinal neurons within 200 ms after visual object pre-
sentation 35. However, it remains unclear whether semantic
information provided by each stimulus is already available during
the initiation of rhinalhippocampal coupling or not. If not, direct-
ed attention might, in a first step, allocate specific connections nec- Time (s)
essary for memory formation before actual information transfer
takes place. Attention-driven enhancement of gamma-band phase
synchronization has been shown in several studies3638. In princi-
ple, cortical regions like the prefrontal cortex and the superior tem-
poral sulcus, which are anatomically connected with the MTL,
could influence rhinalhippocampal interaction. The early tim-
ing, however, suggests that the observed coupling might be initiated
by a top-down process mediated directly by the thalamus39. The
later decrease in synchronization (10001100 ms) may occur fol-
lowing information transfer from the parahippocampal region to
Fig. 5. Changes of EEG gamma power (%) averaged over all frequencies
(3248 Hz) relative to prestimulus baseline for subsequently recalled
versus unrecalled words. Mean s.e.m. is plotted. The x-axis depicts the
time with respect to stimulus onset (word presentation).
Time (s)

articles
Gamma oscillations induced in the CA1 field of hip- EEG recording. Depth electroencephalograms were referenced to linked
pocampal slices within the first second following stimulation mastoids, bandpass-filtered (0.03 to 85 Hz, 6 dB/octave), and recorded
are associated with a prolonged elevation of excitatory postsy- with a sampling rate of 173 Hz (12-bit analogdigital conversion). To
naptic potentials 40,41 , suggesting a crucial involvement in determine the anatomical positions of electrode contacts, MRI scans were
synaptic plasticity, the synaptic correlate of memory forma- acquired in sagittal and adjusted coronal planes, perpendicular to the
longitudinal axis of the hippocampus. Electrode contacts were mapped by
tion 42 . These in vitro data and our findings underline the
transferring their positions from MRI to standardized anatomical draw-
importance of medial temporal gamma activity in memory ings46 (Fig. 1). EEG trials and corresponding power spectra were visu-
formation. A direct line linking these two sets of data, howev- ally inspected for artifacts in the gamma frequency range and 4.9% of all
er, cannot be drawn, as we recorded macroscopic field poten- trials were excluded from analysis.
tials summing up hippocampal mass activity and were not able
to distinguish oscillations generated specifically by distinct hip- Measuring phase synchronization and power. EEG trials were filtered
pocampal subregions like the CA1 field. in the gamma frequency range from 32 Hz to 48 Hz (2-Hz steps) by
Our analysis of macroscopic field potentials revealed that wavelet transforms implementing Morlet wavelets of 7 cycles length. The
efficient medial temporal information processing, leading to filtered signals wj,k (j, time point within a trial; k, trial number) hereby
result from the time convolution of original signals and the complex
successful memory formation, correlates with reduced gamma
wavelet function47. From the wavelet transformed signals wj,k, the phas-
power at both recording sites. The transient reduction of gamma es j,k (j,k = arctan (Im(wj,k)/Re(wj,k))) and the power values Pj,k (Pj,k =
oscillations might be explained by the necessity to suppress Re(wj,k)2 + Im(wj,k)2) were extracted for each time point j of each trial k.
noise-like ambient gamma activity unrelated to specific study Power values were averaged separately for trials corresponding to subse-
items43. One might speculate that in an event of unsuccessful quently recalled and unrecalled words. For each time point of each trial,
encoding, ongoing background gamma activity interferes with phase differences j,k between hippocampal and rhinal electrode con-
item related activity and distorts the process of memory forma- tacts were determined. Phase synchronization values Sj were calculated
tion. Thus, reduced gamma power, as assessed here during suc- based on the definition of circular variance48.
cessful encoding, might be a correlate of a higher specificity of
local assembly activation. Another interpretation could be that
e
N
1 i j , k
gamma activity behaves like a hippocampal resting rhythm sim- Sj =
ilar to the occipital alpha rhythm. In this picture, certain com- N k=1
where N is the number of trials; Sj [0,1].
ponents of rhinalhippocampal circuits might be shut off or
reset during successful memory encoding. Different numbers of trials for subsequently recalled and unrecalled words
The results presented here suggest that gamma power reduc- would cause a bias in the absolute values of the synchronization mea-
tion and appropriate neural coupling and decoupling interact in sure. Therefore, trial numbers were adjusted between conditions using
declarative memory formation within the human MTL. Our randomized trial lists for the condition with the originally larger trial
data do not exclude a third pacemaker site driving phase-locked number. Finally, power and phase synchronization values were averaged
gamma activity in rhinal cortex and hippocampus indepen- for non-overlapping successive time windows of 100 ms duration from
100 to 1500 ms (16 windows in total).
dently from each other. The strong anatomical connection
between both structures13,14, however, supports the hypothe- Statistical analysis. Synchronization and power values were normalized
sis of a direct rhinalhippocampal interaction underlying with respect to prestimulus values (window 1) separately for each subject
gamma-band phase coupling and decoupling. Thus, our find- and each filter frequency. For statistical evaluation, we conducted three-
ings accord with models11,44, proposing that the formation of way ANOVAs with time (15 windows) and memory (subsequently recalled
new declarative memories requires a direct cooperation between versus unrecalled) as repeated measures, and frequency (9 levels) as inde-
rhinal cortex and hippocampus. pendent variable. p-values were HuynhFeldt corrected for inhomo-
geneities of covariance when necessary49. The notation of the F-values is
Fx,y with x being the model degrees of freedom, that is, the number of
METHODS
adjustable parameters in the model, and y being the residual degrees of
Subjects. All 9 patients (6 women, 3 men; mean age, 34.1 8.3 years)
freedom, that is, the number of degrees of freedom that are not taken up
had pharmacoresistant unilateral temporal lobe epilepsies (mean dura-
by the model. In a subsidiary analysis, each time window was tested sep-
tion of illness, 26.4 9.1 years). They were native German speakers. At the
arately by two-way ANOVAs. Effects for time windows with p-values less
time of the experiment, all patients received carbamazepine (plasma con-
than 0.05/15 (Bonferroni correction) and for doublets of neighboring
centration 8 to 12 g/ml) as the only centrally acting drug. During presur-
time windows each with a p-value less than 0.05 (combined probability
gical evaluation, at least three spontaneous seizures were recorded
less than 0.052 14 = 0.035) were regarded as statistically significant.
invasively using bilateral MTL depth electrodes in all patients and tem-
poro-lateral strip electrodes in seven patients. In six patients, seizures
originated exclusively from the right MTL; in three patients, exclusively ACKNOWLEDGEMENTS
from the left MTL. No seizure occurred within 24 hours before the inves- This research was supported by the Deutsche Forschungsgemeinschaft (DFG
tigation. After resection of the epileptic MTL, all patients remained free Fe479/4-1). We wish to thank H. Beck, W. Burr, A. Engel, C. Koch, M. Reuber
of seizures (follow-up, 6 to 15 months). In all patients, histopathologi- and I. Tendolkar for comments on earlier drafts of the manuscript. We also
cal examination of tissue resected at the time of epilepsy surgery revealed
thank H. Urbach for providing the MR images and I. Blmcke for providing the
hippocampal sclerosis. The EEG study was approved by the local med-
ical ethics committee. Each patient gave written informed consent. histopathological reports.
Word memorization protocol. Words were presented in uppercase letters RECEIVED 12 JULY; ACCEPTED 10 OCTOBER 2001
for 400 ms. Interstimulus intervals were randomized and ranged from
2.3 s to 2.7 s (mean 2.5 s). Word length ranged from 4 to 11 (mean 6) letters
1. Scoville, W. B. & Millner, B. J. Loss of recent memory after bilateral
and word frequency ranged from 15 to 175 per million (mean 75 per mil- hippocampal lesions. Neurol. Neurosurg. Psychiatry 20, 1121 (1957).
lion)45. Patients were asked to use a rote strategy to memorize each word, 2. Eichenbaum, H. A cortical-hippocampal system for declarative memory. Nat.
avoiding memory aids such as making rows, sentences, stories or pictures. Rev. Neurosci. 1, 4150 (2000).
During the distraction task, patients were instructed to count backward by 3. Gabrieli, J. D. E. Cognitive neuroscience of human memory. Annu. Rev.
threes, starting at a number between 81 and 99 displayed on screen. Neurosci. 49, 87115 (1998).

articles
4. Zola-Morgan, S. & Squire, L. R. Neuroanatomy of memory. Annu. Rev. memory: dissociation between the effects of subsequent memory
Neurosci. 16, 547563 (1993). performance and distinctiveness. Psychophysiology 35, 709720 (1998).
5. Otten, L. J., Henson, R. N. A. & Rugg, M. D. Depth of processing effects on 28. Elger, C. E. et al. Human temporal lobe potentials in verbal learning and
neural correlates of memory encoding. Brain 124, 399412 (2001). memory processes. Neuropsychologia 35, 657667 (1997).
6. Kirchhoff, B. A., Wagner, A. D., Maril, A. & Stern, C. E. Prefrontal-temporal 29. Hirai, N., Uchida, S., Maehara, T., Okubo, Y. & Shimizu, H. Enhanced
circuitry for episodic and subsequent memory. J. Neurosci. 20, 61736180 gamma (30150 Hz) frequency in the human medial temporal lobe.
(2000). Neuroscience 90, 11491155 (1999).
7. Henke, K., Weber, B., Kneifel, S., Wieser, H. G. & Buck, A. Human 30. Menon, V. et al. Spatio-temporal correlations in human gamma band
hippocampus associates information in memory. Proc. Natl. Acad. Sci. USA electrocorticograms. Electroencephalogr. Clin. Neurophysiol. 98, 89102
96, 58845889 (1999). (1996).
8. Fernndez, G., Brewer, J. B., Zhao, Z., Glover, G. H. & Gabrieli, J. D. Level of 31. Bullock, T. H. et al. EEG coherence has structure in the millimetre domain:
sustained entorhinal activity at study correlates with subsequent cued-recall subdural and hippocampal recordings from epileptic patients.
performance: a functional magnetic resonance imaging study with high Electroencephalogr. Clin. Neurophysiol. 95, 161177 (1995).
acquisition rate. Hippocampus 9, 3544 (1999). 32. McCarthy, G., Nobre, A. C., Bentin, S. & Spencer, D. D. Language-related
9. Brewer, J. B., Zhao, Z., Desmond, J. E., Glover, G. H. & Gabrieli, J. D. E. field potentials in the anterior-medial temporal lobe: I. Intracranial
Making memories: brain activity that predicts how well visual experience will distribution and neural generators. J. Neurosci. 15, 10801089 (1995).
be remembered. Science 281, 11851187 (1998). 33. Klee, M., & Rall, W. Computed potentials of cortically arranged populations
10. Wagner, A. D. et al. Building memories: remembering and forgetting of of neurons. J. Neurophysiol. 40, 647666 (1977).
verbal experiences as predicted by brain activity. Science 281, 11881191 34. Traub, R. D., Whittington, M. A., Stanford, I. M. & Jefferys, J. G. R. A
(1998). mechanism for generation of long-range synchronous fast oscillations in the
11. Fernndez, G. et al. Real-time tracking of memory formation in the human cortex. Nature 383, 621624 (1996).
rhinal cortex and hippocampus. Science 285, 15821585 (1999). 35. Kreiman, G., Koch, C. & Fried, I. Imagery neurons in the human brain.
12. Lavenex, P. & Amaral, D. G. Hippocampal-neocortical interaction: a Nature 408, 357361 (2000).
hierarchy of associativity. Hippocampus 10, 420430 (2000). 36. Steinmetz, P. N. et al. Attention modulates synchronized neuronal firing in
13. Amaral, D. G. & Insausti, R. in The Human Nervous System (ed. Paxinos, G.) primate somatosensory cortex. Nature 404, 187190 (2000).
711755 (Academic, San Diego, 1990). 37. Desmedt, J. E. & Tomberg, C. Transient phase-locking of 40 Hz electrical
14. Witter, M. P., Groenewegen, H. J., Lopes da Silva, F. H. & Lohman, A. H. M. oscillations in prefrontal and parietal human cortex reflects the process of
Functional organization of the extrinsic and intrinsic circuitry of the conscious somatic perception. Neurosci. Lett. 168, 126129 (1994).
parahippocampal region. Prog. Neurobiol. 33, 161253 (1989). 38. Mller, M. M., Gruber, T. & Keil, A. Modulation of induced gamma band
15. Brown, M. W. & Aggleton, J. P. Recognition memory: what are the roles of the activity in the human EEG by attention and visual information processing.
perirhinal cortex and hippocampus? Nat. Rev. Neurosci. 2, 5161 (2001). Int. J. Psychophysiol. 38, 283299 (2000).
16. Engel, A. K. & Singer, W. Temporal binding and the neural correlates of 39. LaBerge, D. Attention, awareness, and the triangular circuit. Conscious. Cogn.
sensory awareness. Trends Cogn. Sci. 5, 1625 (2001). 6, 149181 (1997).
17. Varela, F. J., Lachaux, J.-P., Rodriguez, E. & Martinerie, J. The brainweb: phase 40. Traub, R. D., Whittington, M. A., Buhl, E. H., Jefferys, J. G. & Faulkner, H. J.
synchronization and large-scale integration. Nat. Rev. Neurosci. 2, 229239 On the mechanism of the _ frequency shift in neuronal oscillations
(2000). induced in rat hippocampal slices by tetanic stimulation. J. Neurosci. 19,
18. Sauv, K. Gamma-band synchronous oscillations: recent evidence regarding 10881105 (1999).
their functional significance. Conscious. Cogn. 8, 213224 (1999). 41. Whittington, M. A., Traub, R. D., Faulkner, H. J., Stanford, I. M. & Jefferys,
19. Tallon-Baudry, C. & Bertrand, O. Oscillatory gamma activity in humans and J. G. Recurrent excitatory postsynaptic potentials induced by synchronized
its role in object representation. Trends Cogn. Sci. 3, 151162 (1999). fast cortical oscillations. Proc. Natl. Acad. Sci. USA 94, 1219812203
20. Singer, W. Striving for coherence. Nature 397, 391393 (1999). (1997).
21. Rodriguez, E. et al. Perceptions shadow: long-distance synchronization of 42. Bliss, T. V. P. & Lomo, T. Long-lasting potentiation of synaptic transmission
human brain activity. Nature 397, 430433 (1999). in the dentate area of the anaesthetized rabbit following stimulation of the
22. Miltner, W. H., Braun, C., Arnold, M., Witte, H. & Taub, E. Coherence of perforant path. J. Physiol. (Lond.) 232, 331356 (1973).
gamma-band EEG activity as a basis for associative learning. Nature 397, 43. Craik, F. I. M., Govoni, R., Naveh-Benjamin, M. & Anderson, N. D. The
434436 (1999). effects of divided attention on encoding and retrieval processes in human
23. Van Roost, D., Solymosi, L., Schramm, J., Van Oosterwyck, B. & Elger, C. E. memory. J. Exp. Psychol. Gen. 125, 159180 (1996).
Depth electrode implantation in the length axis of the hippocampus for the 44. Buzsaki, G. The hippocampo-neocortical dialogue. Cereb. Cortex 6, 8192
presurgical evaluation of medial temporal lobe epilepsy: a computed (1996).
tomography-based stereotactic insertion technique and its accuracy. 45. Baayen, R. H., Piepenbrock, R. & van Rijn, H. CELEX Lexical Database
Neurosurgery 43, 819826 (1998). (Linguistic Data Consortium, University of Pennsylvania, Philadelphia,
24. Morris, H. H. & Luders, H. Electrodes. Electroencephalogr. Clin. Neurophysiol. 1993).
Suppl. 37, 326 (1985). 46. Jackson, G. D. & Duncan, J. S. MRI Neuroanatomy (Churchill Livingstone,
25. Paller, K. A., McCarthy, G., Roessler, E., Allison, T. & Wood, C. C. Potentials New York, 1996).
evoked in human and monkey medial temporal lobe during auditory and 47. Daubechies, I. The wavelet transform, time-frequency localisation and signal
visual oddball paradigms. Electroencephalogr. Clin. Neurophysiol. 84, 269279 analysis. IEEE Trans. Inform. Theory 36, 9611005 (1990).
(1992). 48. Mardia, K. V. Probability and Mathematical Statistics: Statistics of Directional
26. Hermann, B. P., Seidenberg, M., Schoenfeld, J. & Davies, K. Data (Academic, London, 1972).
Neuropsychological characteristics of the syndrome of mesial temporal lobe 49. Huynh, H. & Feldt, L. S. Estimation of the box correction for degrees of
epilepsy. Arch. Neurol. 54, 369376 (1997). freedom from sample data in the randomized plot and split plot designs.
27. Fernndez, G. et al. Event-related potentials of verbal encoding into episodic J. Educ. Stat. 1, 6982 (1976).

errata
Reversal of subjective temporal order due to arm crossing

Shinya Yamamoto and Shigeru Kitazawa
Nat. Neurosci. 4, 759765 (2001)
Because of an error in proof corrections, a word was misprinted in the third line of the abstract. The correct abstract
appears below.
How does the brain order successive events? Here we studied whether temporal order of two stimuli
delivered in rapid succession, one to each hand, is determined before or after the stimuli are
localized in space. When their arms were uncrossed, subjects could accurately report the temporal
order, even when the interval between stimuli was as short as 70 ms. In most trials, subjects could
also judge temporal order when their arms were crossed, but only if given adequate time (>1 s). At
moderately short intervals (<300 ms), crossing the arms caused misreporting (that is, inverting) of
the temporal order. Thus, at these intervals, the determining factor of temporal order was the
spatial location of the hands. We suggest that it is not until the spatial locations of the hands are
taken into account that the cutaneous signals from the respective hands are ordered in time.

Nature Neuroscience December 2001

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nature Neuroscience December 2001

Uploaded by

Copyright:

Available Formats

2001 Nature Publishing Group http://neurosci.nature.

volume 4 no 12 december 2001

nature neuroscience volume 4 no 12 december 2001 i

Memory retrieval impairment induced by hippocampal CA3 lesions is

Distinct regulators control the expression of the mid-hindbrain

Protecting axons from

nature neuroscience volume 4 no 12 december 2001 ii

2001 Nature Publishing Group http://neurosci.nature.com editorial

nature neuroscience volume 4 no 12 december2001 1151

news and views

Genes, brain and cognition

nature neuroscience volume 4 no 12 december 2001 1153

news and views

mate of heritability is to ferent brain regions are

the demands for statisti- factor (first unrotated

1154 nature neuroscience volume 4 no 12 december 2001

news and views

give rise to individual differences in g. For tomography, single photon emission

Although it remains to be seen whether

nature neuroscience volume 4 no 12 december 2001 1155

news and views

1156 nature neuroscience volume 4 no 12 december 2001

news and views

gest that the two classes of synaptic con-

computation processing information about these two

nature neuroscience volume 4 no 12 december 2001 1157

news and views

quite large (n 104), whereas the dynamics in both vertical and

made it easier to study not

1158 nature neuroscience volume 4 no 12 december 2001

news and views

recalled words. These synchronization dif-

gone Im lost without a both rhinal and hippocampal regions dur-

later desynchronization, or decreased

nature neuroscience volume 4 no 12 december 2001 1159

news and views

1160 nature neuroscience volume 4 no 12 december 2001

Induction of ically expressed in the photoreceptors of the mature retina and

(EGFP)-transducing adenovirus did (Fig. 2c). Similar results were

nature neuroscience volume 4 no 12 december 2001 1163

Fig. 2. Induction of rod photoreceptor-specific antigens with Crx gene

1164 nature neuroscience volume 4 no 12 december 2001

Winterthurerstr. 190, 8057 Zurich, Switzerland

image of the pattern for spiny cells in other 40

cortical layers whose axons bear mainly 30

en passant boutons and form synapses 20

bouton types as being of any particular significance. Are they

Extraocular proprioceptive signals reach the SC8,9, but mod-

nature neuroscience volume 4 no 12 december 2001 1167

Fig. 1. Contraction of SRF resulting Control 252 spikes C296

the left of each SRF is a dot raster 30

response profile data. (Diameters of

the axes.) The SRFs in the center 30

averaging filter) and plotted using

cance, the mean-squared-difference 30

the interval (2.33, 2.33) indicate 0 10 20 30 40 50 0021-3

1168 nature neuroscience volume 4 no 12 december 2001

5 between control and eye-displaced conditions. (c) Azimuthal functions

0.2 0.5 1. Jay, M. F. & Sparks, D. L. J. Neurophysiol. 57, 2234 (1987).

Memory retrieval synthesis inhibitor metyrapone, administered shortly before

nature neuroscience volume 4 no 12 december 2001 1169

(Fishers, p < 0.01). Lesioned rats also had longer latencies to

(Fishers, p = 0.32; data not shown). These results indicate that

Fig. 2. Effect of metyrapone (35 mg/kg) injections