1998 Nature America Inc. http://neurosci.nature.

com

contents

volume 1 no 3 july 1998

http://neurosci.nature.com

Timm staining of the hippocam-

pus, showing zinc localization. editorial
On page 185, Zheng et al.
provide evidence that Src mod-
Making sense of channel diversity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
1998 Nature America Inc. http://neurosci.nature.com

ulates NMDA receptor function

by reducing its inhibition by
zinc. Photo courtesy of Dr.
Jacqueline McGinty, Dept. of
news and views
Anatomy, East Carolina With color in mind . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
University.
Charles Heywood and Alan Cowey SEE ARTICLE, PAGE 235

Zinc, Src and NMDA receptorsa transmembrane connection . . . . . . . . . . . . . 173

Philippe Ascher SEE ARTICLE, PAGE 185

Probing a complex question: when are SNARE

proteins ensnared? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Timothy A Ryan SEE ARTICLE, PAGE 192

Getting a line on pain: is it mediated by

dedicated pathways? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Edward R Perl SEE ARTICLE, PAGE 218

Cortical control of the thalamus: top-down

processing and plasticity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Artists prefer light from Josef P Rauschecker SEE ARTICLE, PAGE 226
the upper left.
Page 183

book review
Brain, Vision and Memory: Tales of the History
of Neuroscience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
by C G Gross
REVIEWED BY D PURVES

Structure/function correlations
scientific correspondence
in neurons mediating pain. Where is the sun? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Pages 177 and 218 J Sun and P Perona

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, 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 Neuroscience
Subscription Department, PO Box 5054, Brentwood, TN 37024-5054. Telephone (800) 524-0328, Direct Dial (615) 377 3322, Fax (615) 377 0525. Outside North America: Nature
Neuroscience, Macmillan Magazines Ltd, Brunel Road, Basingstoke, Hants RG212XS, U.K. Annual subscription rates: U.S./Canada: U.S. $595, Canada add 7% for GST (institution-
al/corporate), U.S. $195, Canada add 7% for GST (individual making personal payment BN: 14091 1595 RT); U.K./Europe:395 (institutional/corporate), 175 (individual making
personal payment); Rest of world (excluding Japan): 450 (institutional/corporate), 195 (individual making personal payment); 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 home-
page at http://neurosci.nature.com. POSTMASTER: Send address changes to Nature Neuroscience Subscription Department, P.O. Box 5054, Brentwood, TN 37024-5054. Executive Offi-
cers of Nature America Inc: Nicholas Byam Shaw, Chairman of the Board; Mary Waltham, President; Edward Valis, Secretary-Treasurer. Printed by Publishers Press, Shepherdsville, KY,
USA. Copyright 1998 Nature America Inc.

nature neuroscience volume 1 no 3 july 1998 i

 1998 Nature America Inc. http://neurosci.nature.com

contents

***
*** *
articles
*
Tyrosine kinase potentiates NMDA receptor currents
* by reducing tonic zinc inhibition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
V1
V2 VP V4V
V8 F Zheng, M B Gingrich, S F Traynelis and P J Conn SEE NEWS AND VIEWS, PAGE 173

Multiple kinetic components of exocytosis distinguished

by neurotoxin sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
T Xu, T Binz, H Niemann and E Neher SEE NEWS AND VIEWS, PAGE 175
Mapping a new human
color center.
Pages 171 and 235 Presynaptic modulation of CA3 network activity . . . . . . . . . . . . . . . . . . . . . . . . . 201
K J Staley, M Longacher, J S Bains and A Yee

Input synchrony and the irregular firing of cortical neurons . . . . . . . . . . . . . . . . 210

100 C F Stevens and A M Zador
75
50 Nociceptive and thermoreceptive lamina I neurons
25 are anatomically distinct . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Z-S Han, E-T Zhang and A D Craig SEE NEWS AND VIEWS, PAGE 177
1998 Nature America Inc. http://neurosci.nature.com

0
25
50 Cortically induced thalamic plasticity in the
75 primate somatosensory system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
100 E R Ergenzinger, M M Glasier, J O Hahm and T P Pons SEE NEWS AND VIEWS, PAGE 179
Motion perception in
cortical color blindness. Strengthening of horizontal cortical connections
Page 242 following skill learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
M-S Rioult-Pedotti, D Friedman, G Hess and J P Donoghue

Retinotopy and color sensitivity in human visual

cortical area V8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
N Hadjikhani, A K Liu, A M Dale, P Cavanagh
and R B H Tootell SEE NEWS AND VIEWS, PAGE 171

Complete sparing of high-contrast color input to

motion perception in cortical color blindness . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
P Cavanagh, M-A Hnaff, F Michel, T Landis,
T Troscianko and J Intriligator

The effects of frontal eye field and dorsomedial

frontal cortex lesions on visually guided eye movements. . . . . . . . . . . . . . . . . . . 248
P H Schiller and I Chou

Top-down influences on stereoscopic depth-perception . . . . . . . . . . . . . . . . . . . 254

I Blthoff, H Blthoff and P Sinha

Do depth cues contribute

to figure recognition? classified advertising
Page 254
see back pages

nature neuroscience volume 1 no 3 july 1998 ii

 1998 Nature America Inc. http://neurosci.nature.com
Making sense of channel diversity
When the eminent British geneticist J.B.S. Haldane was asked what
God had revealed about himself through his works, Haldane is said
to have replied an inordinate fondness for beetles. Were he alive
1998 Nature America Inc. http://neurosci.nature.com
today, Haldane might instead have cited ion channels; although their
diversity may no longer be absorbing the creative energies of the
Almighty, they did at least attract several hundred people to a three-
day meeting in New York last month. The event was organized by
the New York Academy of Sciences, and it provided an excellent
overview of recent progress in understanding the molecular basis of
ionic conductances, including ionotropic receptors as well as volt-
age-gated and other channels. Although the advances have been
impressive, it was also clear that the field faces a formidable chal-
lenge in making sense of what has already been discovered.
An unprecedentedly comprehensive picture of channel and
receptor diversity is now emerging from large-scale genome sequenc-
ing projects, notably of the nematode worm Caenorhabditis elegans
(the sequence of which is now around 80% complete). As discussed
by Larry Salkoff (St Louis), the worm sequencing project has already
led to the identification of large numbers of new channel and recep-
tor subunits. In particular, at least 80 potassium channel subunits
have been found, a remarkable number considering that the ner-
vous system of C. elegans contains only 302 neurons, which have
been classified into 118 types. About 50 of these genes belong to a
new class, distinguished both by their four transmembrane domains
and by the lack of knowledge about their function (the best guess is
that they are leak channels that regulate cell excitability). By tagging
the coding sequences with green fluorescent protein, it is possible
to visualize their expression patterns; many of the subunits are
restricted to single cell types, and at least one is expressed only in a
single interneuron. If this can be extrapolated to the mammalian
brain (which is not yet clear), not only does this imply a very large
number of channels, but some may be so restricted in expression
that they are unlikely to be discovered except by genomic sequencing.
The mammalian sequencing projects are far less advanced, the avail-
able evidence suggests that any given channel family will contain
many more members in mammals than in worms. For example, C.
elegans has a single voltage-gated K+ channel of the Shaker class,
whereas at least eight have already been identified in humans.
In addition to the large number of genes encoding channel and
receptor subunits, there are several other levels at which diversity
can arise. One is alternative splicing; some of the new K+ channel
genes discovered in C. elegans, for instance, can give rise to six or
seven different isoforms. Another is RNA editing, a remarkable
process by which single base changes (and hence changes in the
encoded protein) can be introduced into an already-transcribed
mRNA. Editing has been described in the mammalian brain for both
AMPA- and kainate-type glutamate receptors, where it is known to
regulate ion selectivity and channel kinetics (Rolf Sprengel, Heidel-
nature neuroscience volume 1 no 3 july 1998
editorial
berg; Steve Heinemann, Salk Institute). Robert Reenen (University of
Connecticut) has now found that the Drosophila Na+ channel encod-
ed by the paralytic gene is also edited at several different sites, and
that like the glutamate receptors, para editing is under tight devel-
opmental regulation. Why RNA editing has been exploited by the
nervous system in this way, and in such diverse species, is still unclear;
one possibility is that it may allow the expression of two almost iden-
tical sequences without the risk of gene conversion.
The greatest source of diversity, however, arises from the fact that
most channels and receptors are composed of multiple subunits,
which can be assembled in different combinations. In many cases,
a given channel can show profoundly different behavior depending
on which modulatory subunits are present. Many examples were
presented, including K+ and Ca2+ channels as well as all the major
classes of ionotropic receptors; to cite just one, Terry Snutch (Van-
couver) showed how P- and Q-type calcium currents, long thought
to be distinct (both are voltage-gated but they differ in their ability to
undergo spontaneous inactivation), can both arise from the same
pore-encoding 1a subunit. Moreover, this difference can be caused
not only by differential association with regulatory subunits, but
also by alternative splicing of the 1a subunit itself. The properties
of cloned channels must generally be studied in heterologous expres-
sion systems, but just because a particular subunit combination can
form in vitro does not necessarily mean that it occurs in the brain.
The process of determining which subunits associate with which
others in vivo is long and laborious, yet essential if the lessons from
recombinant channels in vitro are to be extended to real neurons.
The number of channels and receptors that can be encoded by
the genome is thus enormous, and although this may be good news
for pharmaceutical companies, it presents a daunting prospect for
any attempt to understand the underlying principles of brain orga-
nization. The challenge, of course, is to determine what this prodi-
gious molecular diversity might signify in functional terms. The easy
answer is that the brain is very complex and that it needs a corre-
spondingly vast number of molecules to perform its diverse func-
tions; but although this may be true, it is hardly satisfying.
It is possible, of course, that not all the observed diversity has any
adaptive significance. Gould and Lewontin have warned against the
uncritical acceptance of adaptive explanations in biology, and it is
at least possible that some of the diversity that has arisen among dif-
ferent families of channels and receptors has no purpose genes
may duplicate and diverge in evolution simply because they can, in
other words because there is no selective disadvantage to doing so
and because the process is not easily reversed once it has occurred. To
take one simple scenario, imagine that a gene encoding a particular
channel undergoes duplication, and that the two sequences drift
apart and acquire differences in their promoters; although they may
at first be mutually redundant, if certain populations of neurons lose
169
 1998 Nature America Inc. http://neurosci.nature.com
editorial
the ability to express one or the other gene, they both become essen-
tial even if they differ very little in their functional properties.
One way to address the question of adaptive significance is to
ask whether individual members of gene families show conserva-
tion in evolution. Broadly, the answer so far seems to be that they
do. For instance, Salkoff noted that the major classes of potassium
channels that exist in humans all have recognizable homologs in
C. elegans, suggesting that the evolution of more complex nervous
systems has not been accompanied by the appearance of new types
of channels, but rather by diversification of pre-existing types.
Moreover, even the individual family members often show high
conservation; the human Slo1 gene (which encodes a high-con-
ductance calcium-gated K+ channel), for example, is much closer
to nematode Slo1 than to human (or nematode) Slo2, and the
same principle holds for other classes of K channels. Perhaps the
most striking demonstration of functional conservation was the
worm homolog of one of the Long QT-type K+ channels, muta-
tions of which lead to abnormal heart rhythms in humans. The
1998 Nature America Inc. http://neurosci.nature.com
worm KQT homolog is expressed in the pharynx, which like the
heart generates a rhythmic pumping action. Moreover, when the
human mutation is introduced into the worm coding sequence,
the mutant animals show a defect analogous to the human condi-
tion, a long pharyngeal pump syndrome, as it were. Examples like
this offer hope that comparing model systems will reveal some
general principles of how different patterns of channel expression
determine the properties of different classes of neurons.
The current favorite method for determining the function of a
channel is to knock it out genetically, but it was clear from many of
the presentations that this approach has serious limitations. Although
gene knockouts avoid the problems associated with lack of speci-
ficity in pharmacological blocking agents, they raise interpretation-
al problems of their own. Often, mutant phenotypes are either
nonexistent or too subtle to be recognized using the available tech-
niques. Even in cases where a mutant phenotype is found, it is often
difficult to rule out the possibility that the absence of the gene dur-
ing development has led to compensatory changes that complicate
the interpretation of the result. The ideal gene-knockout method
would be cell-type specific and under tight temporal control. But
despite several apparently encouraging reports in the literature, such
techniques are far from robust, as Peter Seeberg (Heidelberg) empha-
sized. At present, the technology does not exist to inactivate specif-
ic genes in specific parts of the mammalian nervous system with
high efficiency and specificity, let alone in a rapidly inducible or
reversible manner that would eliminate concerns about develop-
mental compensation. A reliable method for doing this would be
invaluable, but it does not yet seem to be close to realization.
The challenge in understanding ion channel function may be
reduced to two broad questions: what do specific channels contribute
to the behavior of the cells in which they are expressed, and how
does the behavior of these cells contribute to the working of the sys-
tem as a whole? Although certain mutations have given interpretable
and interesting phenotypes, there are major obstacles to be over-
come before this can be achieved on a routine basis. For one thing,
the sheer effort of descriptive analysis will be considerable. It will be
essential to correlate ion channel expression with single-cell prop-
erties, and this is far from trivial in vivo. In situ hybridization or anti-
body staining can provide information about expression patterns,
but it is also necessary to correlate this with cellular physiology. One
powerful approach, discussed by Hannah Monyer (Heidelberg), is
to record from single neurons via a whole-cell patch pipette and then
to aspirate the contents of the cell into the pipette, so that mRNA
expression can be analyzed by PCR. Monyer has used this technique
to show that principal neurons and interneurons in the primary
170
visual cortex show different patterns of AMPA receptor expression,
and she hopes to determine how specific patterns of channel and
receptor expression can be related to cortical information processing.
Ultimately, it seems clear that to understand how channels and
receptors determine neuronal behavior, the field will have to go
beyond the level of molecular description and adopt a more quan-
titative and biophysical approach. This is perhaps the greatest chal-
lenge for the years ahead. To explain the electrical properties of a
neuron, it is not sufficient merely to specify the types of channels it
expresses; one must also know their densities and distributions, as
they relate to the fine structure and cable properties of axons and
dendrites. Such an analysis would seem essential for any serious
attempt at understanding channel function at the cellular level, but
surprisingly the question hardly came up in the meeting.
Consider, for instance, how a neuron might achieve the appro-
priate number and distribution of each of the channels it express-
es. To obtain the desired pattern of excitability, there must
presumably be some form of feedback from activity to channel
expression. Yet how this might occur is almost entirely mysterious.
To what extent is channel density regulated by activity, and if so by
what feedback pathways? At what level is control exerted? It could
be transcription, or at post-transcriptional levels such as protein
synthesis, degradation, trafficking or association with modulatory
proteins. In muscle fibers, the distribution and turnover times for
different types of acetylcholine receptors are regulated with great
precision, both during development and in response to changing
patterns of electrical activity, but whether this is also true for neu-
ronal receptors and ion channels is still very unclear.
Not only the number but also the precise localization of different
molecules must in some cases be specified. Efforts to understand
this were exemplified by the presentation from Ole Ottersen (Oslo),
who has used immunogold labeling to study the fine structure of
cerebellar Purkinje cells. He has shown that different molecules are
targeted to different sites; the 2 glutamate-like receptor, for instance,
is present at the postsynaptic sites formed with parallel fibers but
not with climbing fibers. Another molecule that is precisely local-
ized in these cells is the glutamate transporter EAAT4, which is
known to play a role in clearing glutamate and shortening the EPSC.
How it does so is unclear; Ottersen has shown that the main site of
EAAT4 expression is at the base of dendritic spines, close to the site
of contact with glial cells and several microns away from the post-
synaptic membrane where ionotropic glutamate receptors are con-
centrated. How EAAT4 can affect synaptic activation given its
exclusion from the site of transduction remains to be determined,
but the results highlight the importance of precise molecular descrip-
tions of synaptic structure if the details of synaptic transmission are
to be understood in quantitative terms.
The molecular basis of this structural specificity is even less clear,
but some details are starting to emerge. Morgan Sheng (Massachu-
setts General Hospital), Mary Kennedy (Caltech) Heinrich Betz
(Frankfurt) and Nat Heintz (Rockefeller University) each discussed
molecular components of postsynaptic sites, and have identified
various molecules that may govern how receptors and channels
become localized. An important goal now is to determine how these
various components interact, and how the appropriate density and
distribution of synaptic signaling components is achieved and main-
tained. In the longer term, it will also be important to find out
whether similar mechanisms regulate channel distribution elsewhere
on the membrane, and thus whether they play a more general role in
regulating neuronal excitability.
New York Academy of Sciences conference: Molecular and functional diversity of ion chan-
nels and receptors. New York, May 14-17, 1998. See http://www.nyas.org/brochion.html for
program details.
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com
With color in mind
Charles Heywood and Alan Cowey
Which human brain area produces color blindness when damaged? High-resolution
functional neuroimaging suggests that it is area V8, not the favorite candidate V4.
Dorothys whirlwind departure from a
monochrome Kansas into the vividly chro-
matic world of Oz, in the 1939 classic film
The Wizard of Oz, highlights the substan-
tial contribution that color makes to our
visual world. Yet equally revealing is the
ease with which we view monochrome film
or television, where the absence of color
1998 Nature America Inc. http://neurosci.nature.com
does not compromise our enjoyment and
can pass unnoticed. This is not so in the
clinical condition of cerebral achromatop-
sia where patients, following characteristi-
cally ventral occipitotemporal brain
damage (see Fig. 1a) inhabit a drab world,
devoid of color, and may be painfully aware
of their complete loss of chromatic vision1.
Attempts at understanding the nature of
cerebral achromatopsia and its neural basis
have spawned controversy ever since Louis
Verreys description2 of such a case in 1888
(see ref. 3 for review). The ensuing debate
lasted for more than a century, and not sur-
prisingly the protagonists reflected oppos-
ing views about whether any cognitive or
perceptual function was regionally local-
ized. At issue was whether achromatopsia
results from the deletion of a cortical
region specialized for the processing of
color. The demonstration4, using positron
emission tomography, of increased cere-
bral blood flow in an area of cortex when
observers view chromatic scenes was cer-
tainly consistent with this notion, because
the activated region, dubbed the human
color center, is invariably damaged in cases
of cortical color blindness. By then, how-
ever, the cluster of 2030 visual areas occu-
pying almost half of the neocortex of
monkeys had been identified, and the
debate turned to whether the human color
center was homologous to the fourth visu-
al area of the monkey, cortical area V4 (see
Fig. 1b). This correspondence had been
Charles Heywood is the Sir Derman
Christopherson Research Fellow at the
Department of Psychology, Science Laboratories,
South Road, Durham DH1 3LE, UK
(C.A.Heywood@Durham.ac.uk)
Alan Cowey is at the Department of
Experimental Psychology, South Parks Road,
Oxford OX1 3UD, UK
(alan.cowey@psy.ox.ac.uk)
nature neuroscience volume 1 no 3 july 1998
news and views
proposed chiefly on the a
A
(not uncontested) view that
V4 contains a comparative-
ly high proportion of cells
that respond selectively to
wavelength and color5. The
Calcarine
results of a functional imag- sulcus
ing study by Hadjikhani
and colleagues, reported in
this issue of Nature Neuro- Striate
science (pp 235241), sug- cortex
Lingual Fusiform Collateral
gest however that the gyrus gyrus sulcus
human color center is dis-
tinct from area V4. The
newlydefined color area b
B Lunate
sulcus
contains a complete retino-
topic map of the contralat-
eral visual half field, V4
responds more robustly to
color than neighboring
regions and, unlike V4, is
activated by the induction
of color aftereffects. The Superior
temporal
fMRI signal elicited by an sulcus
aftereffect thus mimics the
response to a real colored Inferior V4 TEO IT
occipital
stimulus, providing sup- sulcus
Bob Crimi
porting evidence that V8 is
implicated in processes Fig. 1. Visual areas in the human and monkey brains. (a)
involved in perceiving color. The medial view of the left hemisphere of the human brain.
These properties, the Striate cortex, area V1, is shown in red, partly buried in the
authors suggest, make it a calcarine sulcus. The region along the collateral sulcus,
ready candidate for a region whose destruction leads to cerebral achromatopsia, is
responsible for our con- shown in blue. Area V8 lies in the middle of the collateral
scious perception of a col- sulcus, whereas V4 lies slightly more posterior and medial.
The damage indicated by blue would therefore include V4v
ored world.
and V8. (b) A lateral schematic view of the right hemisphere
Hadjikhani and col-
of the macaque monkey, in which the labeled sulci have been
leagues used existing, but opened up. Areas V1 (red) and V4 (purple) extend onto the
improved, techniques of ventral and medial surfaces, respectively. Areas in the tem-
functional neuroimaging poral lobe (TEO, orange and IT, dark green), have been
to reveal, with increased implicated in color vision of the macaque monkey, and
senstivity, brain areas Hadjikhani and colleagues raise the possibility that TEO may
involved in the processing correspond to human V8.
of color. Functional mag-
netic resonance imaging
(fMRI) relies on endoge-
nous changes in magnetic susceptibility, the human brain6. These areas are the
which result from changes in local cere- presumed homologues of those identi-
bral blood flow and oxygenation. Such fied in the monkey brain using a variety
changes are activity dependent, and their of invasive techniques, such as cellular
measurement in response to visually pre- recording and experimental neuroanato-
sented stimuli have already established my. Delineation of an area relies on the
the borders of a number of visual areas in presence of an orderly retinotopic map
171
 1998 Nature America Inc. http://neurosci.nature.com
news and views
of the visual world, which is a feature of
many visual areas. The authors have
exploited the fMRI technique to identify
such maps, using stimuli consisting of
slow-moving patterns of luminance mod-
ulation. As the light and dark areas pass
across the visual field, they elicit period-
ic excitation at the associated cortical
location. Moreover, the phase of the
response specifies the polar angle or
eccentricity, for rotation or radial move-
ment around the fixation point respec-
tively, of the visual field region
represented at that location. Thus a
Fourier analysis on the response profile
of a single voxel of the image, along with
a consideration of the sign of the response
to identify mirror- versus non-mirror-
1998 Nature America Inc. http://neurosci.nature.com
image representations, will yield the
retinotopy of a visual area, which can dis-
played, by cortical flattening, as a two-
dimensional map. Using such techniques,
Hadjikhani and colleagues compared the
effect of luminance-defined visual pat-
terns with that of identical patterns that
were defined by equiluminant color vari-
ation, that is, variations in color but not
in luminance. In addition to finding
stronger activation to color than to lumi-
nance in cortical areas V1, V2, V3/VP and
the ventral subdivision of V4 (V4v), a
region in the middle of the collateral sul-
cus was identified that responded prefer-
entially and especially effectively to color.
Its location corresponded to that previ-
ously (and in the absence of identifica-
tion of retinotopic boundaries,
prematurely) described as human V4. By
adopting improved techniques, including
a high-field scanner, signal averaging and
improved visual displays, the authors
established that the true color center lies
beyond the anterior border of the previ-
ously reported area V4v (i.e. outside v4
altogether). Furthermore the retinotopy
of the color center differs from its neigh-
bors. Areas V4v, VP and inferior V2 con-
tain quarter-field representations of the
upper visual field and share a contiguous
representation of the fovea. In contrast,
the color area contains a map of the
upper and lower half-fields with a foveal
representation located at its anterior bor-
der. It now seems clear that the color cen-
ter is distinct from area V4, and
accordingly, the authors refer to this new,
previously unreported region as area V8.
Area V8 poses as many new questions
as its identification sought to resolve.
That it is distinct from area V4 is cer-
tainly consistent with hitherto puzzling
demonstrations that ablation of V4 in the
macaque monkey does not result in the
172
severe deficits in the discrimination of
hue that is the hallmark of cerebral
achromatopsia 7 . Conversely, impair-
ments in the discrimination of visually
present form and pattern vision that
invariably follow damage to V4 in the
monkey are not an invariable feature of
the vision of achromatopsic people. The
location of V8 within the region
destroyed in achromatopsic patients,
although the latter is always more exten-
sive and includes white matter damage,
is strong but not conclusive evidence that
V8 is the critical area whose removal can
result in the complete loss of the con-
scious representation of color. As Had-
jikhani and colleagues themselves point
out, the question naturally arises as to
where area V8 is concealed in the
macaque monkeys brain. Localization of
macaque V8 would lead to confirmato-
ry evidence that its removal results in
cortical color blindness. They speculate
that area TEO8,9 (Fig. 1b), a region lying
anterior to V4, may be the culprit.
Indeed this region, along with more
anterior temporal lobe areas (Fig. 1b),
has been implicated in color vision (Van-
duffel, W. et al., Soc. Neurosci. Abstr. 23,
334.7, 1997; Katsuyama, N. et al., Soc.
Neurosci. Abstr. 23, 803.11, 1997). Large
lesions to anterior and inferior portions
of the temporal lobe do render monkeys
achromatopsic10, and a high proportion
of color-selective cells, revealed by meta-
bolic labeling and electrophysiological
recording, reside in this area11.
Although the precise role of area V8
has yet to be clarified, determining what
is spared, as opposed to lost, in cerebral
achromatopsia may provide some sign-
posts. Color makes a ubiquitous contri-
bution to vision. Color differences can,
among other things, provide information
about form, texture and motion. This is
presumably reflected in the number, and
wide variation, of activated areas reported
in neuroimaging studies when subjects
perform a wide variety of color-related
tasks. Cerebral blood flow can be modu-
lated by the nature and difficulty of the
behavioral tasks (whether they entail pas-
sive viewing, active discrimination or
directed attention) and the properties of
the visual display (containing equilumi-
nant color, with or without form, with or
without associated brightness differences).
The physical basis of color is the wave-
length composition of light. Many corti-
cal areas prior to V8 contain cells that are
sensitive to, but not selective for, wave-
length differences. For example, cells may
respond vigorously to equiluminant
red/green borders without signaling the
nature of the colors of which the border
is composed, i.e. which is red and which is
green. Form can thus be derived from the
processing of wavelength differences and
yield information about the visual scene
without encoding its chromatic content.
It should perhaps come as no surprise
that achromatopsic patients, lacking area
V8, can show a preserved capacity to use
wavelength variation to detect motion
and form12, presumably mediated by ear-
lier and intact extrastriate areas. Detec-
tion of equiluminant chromatic form can
even be achieved when it is disguised by
accompanying rapid random luminance
variation. The latter implies that patients
must retain color-opponent processing
mediated by the well known P-channel of
primate vision, which unlike its partner
the M-channel is blind to the introduc-
tion of rapid flicker. Moreover, another
paper in this issue 13 reports that pro-
foundly achromatopsic patients lacking
evidence for residual color-opponent
processes are still able to extract motion
from high-contrast color cues.
Achromatopsic patients can therefore
process wavelength differences to extract
information about form and motion,
but their brain damage nevertheless ren-
ders them blind to color differences.
Such brain damage is, of course, likely
to encroach on territory other than area
V8, including the adjacent V4v.
Although loss of conscious representa-
tion of hue characterises achromatop-
sia, another very different explanation
has been offered3, namely that it is a fail-
ure of color constancya loss of the
invariance of an objects perceived color
despite wide variation in the wavelength
composition of the illuminating light.
Although it has been suggested that the
chromatic responses of neurons in area
V4 show color constancy14, the effects
of ablating this in the monkey have yet
to convincingly demonstrate a corre-
sponding deficit. Might V8 be assigned
such a role? An explanation of cerebral
achromatopsia as resulting from the
destruction of V8 with a concomitant
deficit in color constancy does not read-
ily explain why two very different equi-
luminant hues are indistinguishable to
an achromatopsic patient, nor why the
world should be described in shades of
gray. A direct test of color constancy in
one such patient does not lend unequiv-
ocal support for this view15. When two
patches of different spectral composi-
tion were presented against two differ-
ent backgounds, one to each eye, the
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com
patient reported them to be indistin-
guishable only when the ratios of the
retinal cone response of path/back-
ground were preserved. When the ratios
were different for each eye, the patches
no longer appeared identical. These
responses are akin to those of the nor-
mal observer and are presumed to be
mediated by retinal mechanisms. How-
ever, the achromatopsic patient departed
from normal performance when pre-
sented with more complex scenes requir-
ing multiple cone contrast comparisons.
Achromatopsic patients may therefore
retain rudimentary color constancy
mechanisms, and it remains an open
question as to whether V8 is essential for
consolidation of information across
1998 Nature America Inc. http://neurosci.nature.com
large regions of complex scenes.
However, what these patients do lack
is the conscious representation of color. If
it transpires that a single cortical area,
area V8, is indispensable for our con-
scious percept of color, it will indeed be a
rare triumph for the view that regional
specialization underlies the cluster of
visual areas that occupy a substantial
proportion of the neocortex in primates.
On a note of caution, however, Had-
jikhani and colleagues hint that area V8
responds to a wide variety of visual stim-
uli. The challenge will then be to estab-
lish the precise role of this, and other
areas, in the cortical processing of color.
Hadjikhani and colleagues have pointed
us in the correct direction.
1. Cowey, A. & Heywood, C.A. Trends Cog.
Neurosci. 1, 133139 (1997).
2. Verrey, L. Archs. Ophtalmol. (Paris) 8, 289301
(1888).
3. Zeki, S.A Vision of the Brain. (Blackwell
Scientific Publications: Oxford Univ. Press,
1993).
4. Zeki S. et al. J.Neurosci. 11, 641649 (1991).
5. Zeki, S. Nature 284, 412418 (1980).
6. Sereno, M.I. et al. Science 268, 998893
(1995).
7. Heywood, C.A., Gadotti, A. & Cowey, A. J.
Neurosci. 12, 40564065 (1992).
8. Boussaoud, D., Desimone, R. & Ungerleider,
L.G. J. Comp. Neurol. 306, 554575 (1991).
9. Zeki, S. Proc. R. Soc. Lond. B 263, 15391544
(1996).
10. Heywood, C.A., Gaffan, D. & Cowey, A. Eur. J.
Neurosci. 7, 10641073 (1995).
11. Komatsu, H., Ideura, Y., Kaji, S. & Yamane, S.
J. Neurosci. 12, 408424 (1992).
12. Heywood, C.A., Kentridge, R.W. & Cowey, A.
Exp. Brain Res. (in press).
13. Cavanagh, P. et al. Nature Neuroscience 1,
242247 (1998).
14. Zeki, S. Neuroscience 9, 741765 (1983).
15. Hurlbert, A.C., Bramwell, D.I., Heywood,
C.A. & Cowey, A. Exp Brain Res. (in press).
nature neuroscience volume 1 no 3 july 1998
Zinc, Src and NMDA
receptorsa transmembrane
connection
Philippe Ascher
Zinc and tyrosine kinases produce opposite effects on the
NMDA receptor; new evidence suggests that Src-induced
potentiation is due to the relief of zinc inhibition.
Recently, scientists have become increas-
ingly aware of the potent actions of zinc
and tyrosine kinases on NMDA receptors.
NMDA receptors are a group of ionotrop-
ic glutamate receptors that are permeable
to both calcium and sodium, and have
been implicated in many forms of synap-
tic transmission, synaptic plasticity and in
cell death. Zinc had long been known to
interact with many neurotransmitter
receptors, but its inhibition of NMDA
receptors has attracted particular interest
because of a possible functional role. Zinc
is stored in synaptic vesicles in a number
of glutamatergic terminals of the fore-
brain, is released during synaptic activity,
and its concentration in the synaptic cleft
has been suggested to rise to the micro-
molar range1. Early studies of the interac-
tion of zinc with NMDA receptors 2,3
identified two effects, a voltage-dependent
inhibition which resembles that of mag-
nesium, and a voltage-independent inhi-
bition, which occurs at a different site. The
two effects were initially described for zinc
concentrations in the micromolar range,
and this fitted nicely with estimates of zinc
concentration in the synaptic cleft fol-
lowing glutamate release. Recently, after
the cloning of NMDA receptors subunits
allowed the expression of recombinant
receptors, the two effects of zinc were fur-
ther analyzed using NMDA receptor sub-
types built from specific combinations of
subunits. Using these recombinant recep-
tors, additional evidence was obtained to
support the hypothesis that the voltage-
dependent effect of zinc occurs via the
same site as the voltage-dependent mag-
nesium block within the channel pore, but
that zinc permeates the channel better4.
The binding site involved in the voltage
independent block is less well identified,
Philippe Ascher is at the Laboratoire de
Neurobiologie, Ecole Normale Superieure, 46
rue dUlm, Paris 75005, France
e-mail: pascher@wotan.ens.fr
news and views
but in some recombinant NMDA recep-
tors (those assembled from NR1 and
NR2A subunits) the IC50 of the voltage-
independent inhibition is exceptionally
low (10 nM-100nM) 4-6. This is the range
of concentration at which zinc is present
as a contaminant in most experimental
solutions, but also in the cerebrospinal
fluid. Thus, even without adding any
additional zinc, a large fraction of the
high-affinity, voltage-independent,
inhibitory sites on NR1-NR2A receptors
are already occupied, so that the addition
of a zinc-chelating agent to the bath is
usually sufficient to double the amplitude
of the baseline response4.
The interaction of tyrosine kinases
with NMDA receptors a priori seemed to
have little relation with that of zinc. The
early observations of Salter and col-
leagues 7,8 showed that tyrosine kinase
inhibitors inhibit some NMDA respons-
es, which conversely can be potentiated
by intracellular injection of either a con-
stitutively-active form of the intracellular
tyrosine kinase Src or peptide fragments
that activate Src 8. Similar effects were
observed with another tyrosine kinase,
fyn9. These observations were reinforced
by reports that tyrosine kinase inhibitors
interfere with some forms of long-term
potentiation (a cellular model for learn-
ing and memory), and that some tyrosine
kinase deficient mice had perturbed LTP
and behavioral abnormalities (for refer-
ences see 10). Once again, recombinant
NMDA receptors were used to reveal a
subunit specificity: Src only acted on
receptors assembled from NR2A subunits
and a restricted group of splice variants
of NR19.
Zheng and colleagues on page 185 of
this issue of Nature Neuroscience11 have
linked these two sets of apparently inde-
pendent observations by experiments that
strongly suggest that the potentiating
effect of Src on NR1-NR2A receptors
results from the suppression of the ambi-
173
 1998 Nature America Inc. http://neurosci.nature.com
news and views
NR1 NR2A NR1 NR2A signaling (see Fig. 1).
Previous studies of
receptor tyrosine
kinases have popular-
ized an orthograde
model, where the
binding of an extra-
cellular ligand to a
receptor transmits a
signal through the
Src
membrane to activate
P Src a tyrosine kinase on
the cytoplasmic side.
Zinc
The data of Zheng
Gly
Glu
and colleagues sug-
Magnesium gest that the NMDA
Bob Crimi
receptor mediates a
Fig. 1. Left: the NMDA receptor has been activated by the simul- trans-membrane
1998 Nature America Inc. http://neurosci.nature.com
taneous binding of glycine to the NR1 subunit and of glutamate to message in the oppo-
the NR2A subunit, but the binding of zinc reduces the time spent
site direction; activa-
in the open state. Right: the interaction of Src with the cytoplas-
tion of a tyrosine
mic tail of the NR2 subunit induces a conformational change
which, by preventing the binding of zinc on the extracellular side
kinase on the cyto-
(arrow), potentiates the NMDA response. plasmic side alters the
binding of extracellu-
lar ligands. The ligand
that is most clearly
ent zinc inhibition. Src mimics the effect affected is zinc, which seems to be dis-
of adding an extracellular zinc chelator, placed by the process (its apparent bind-
and the potentiation induced by remov- ing affinity decreases), but the kinetics of
ing zinc from the extracellular medium glutamate dissociation from the receptor
and the potentiation induced by the addi- is also slowed. (Glycine affinity may also
tion of Src occlude each other, suggesting be affected but this has not yet been stud-
a shared mechanism. Dose response ied.) However, the retrograde signaling
curves confirm that the EC50 of the volt- analyzed by Zheng and colleagues does
age-independent, high-affinity zinc inhi- not exclude a mechanism working in the
bition is increased by Src. orthograde direction; if zinc and Src act
The picture of what exactly is happen- on allosteric sites of the NMDA receptor,
ing at the molecular level, however, allosteric theory predicts that if Src alters
remains fragmentary. From the work of zinc binding, zinc binding can alter Src
Salters group 9 we know that Src binds to action.
the NMDA receptor through a region that From a physiological point of view,
is distinct from the catalytic site, but we the observations suggest that Src and zinc
do not know whether Src directly phos- play opposite roles in synaptic transmis-
phorylates the NMDA receptor or sion. If there are glutamatergic synapses
whether it acts via a third protein. Zheng in vivo where the postsynaptic NMDA
and colleagues11 have now identified three receptors have the very high zinc sensi-
specific tyrosines in the NMDA receptor tivity observed in recombinant NR1-
subunit NR2A that appear to be involved NR2A receptors, extracellular ambient
in NMDA receptor modulation; it is zinc will set the NMDA response to about
tempting to speculate that they are targets half of its maximal value, which would
for Src-mediated phosphorylation, but place the receptors in an optimum posi-
this has not been directly demonstrated. tion to be regulated bidirectionally. The
The explanation of the high zinc senstivi- response could be either increased (by
ty of NR1NR2A remains uncertain. The activation of a tyrosine kinase) or inhib-
simple explanation was that either zinc or ited (by zinc released from glutamatergic
Src would specifically interact with NR2A. terminals). This hypothesis remains
However, two of the tyrosines identified untested, and in particular no study has
by Zheng and colleagues are common to yet investigated the modulation of
NR2A and NR2B. Furthermore, recent NMDA synaptic currents by zinc. In such
evidence12 suggests that NR1 is involved a study, one experimental obstacle will be
in the zinc modulation. Nevertheless, the difficulty of chelating synaptically
whatever the details of the picture, the released zinc. Zinc chelators like EDTA,
data strongly suggest a kind of retrograde TPEN or tricine allow one to control the
174
zinc concentration in experiments in
which glutamate or NMDA are bath
applied, and in which the concentration
of zinc is fixed. None of the available
chelators, however, will prevent the rise
of zinc concentration if zinc is released
in the synaptic cleft in the millisecond
range, especially if the local concentra-
tion in the cleft can reach the micromolar
range. This is mainly because calcium
also competes with zinc to bind to these
heavy metal chelators. The problem may
be resolved by controlling the presynap-
tic zinc content, either with chelators
having selective access to the synaptic
vesicles (for example see ref. 13) or from
genetically modified animals in which the
zinc transporters and buffers (and in par-
ticular the storage of zinc in synaptic vesi-
cles14) will have been modified.
Zinc is also strongly suspected to play
a role in various pathologic conditions,
such as neuronal death following global
ischemia or massive seizures 15. Zinc is
known to be cytotoxic, and there is evi-
dence that this toxicity results from its
entry into the cell, even though its intra-
cellular targets are not rigorously identi-
fied. The cell death induced by exposure
to zinc is increased by glutamate and can
be reduced by NMDA antagonists15, sug-
gesting that a major route of zinc entry
into neurons could be through NMDA
receptors, which appear to be both
blocked by and permeable to zinc. The
NR1-NR2A receptors, which contain a
high affinity, voltage independent, zinc
inhibitory site, could play a particular role
in the zinc toxicity because they have an
additional surprising property, namely
that the zinc inhibition remains incom-
plete even when this site is saturated57.
This means that at zinc concentrations in
the micromolar range, zinc retains access
to the channel and therefore to the cell
interior. Therefore, the NMDA receptors
most sensitive to zinc may be, paradoxi-
cally, the most dangerous route of zinc
entry into neurons. If we consider that Src
induces a rightward shift of the zinc con-
centration response curve, we can predict
that Src activation will potentiate the
entry of zinc. Thus in pathologic disease
states, Src and zinc may combine their
effects instead of neutralizing each other.
Although Zheng and colleagues have
not directly proven that Src can phos-
phorylate NMDA receptors and have not
identified the zinc binding site, they have
raised a number of intriguing questions
about the interaction between zinc bind-
ing and tyrosine phosphorylation. More
work is needed to determine whether
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com
these observations made on recombinant
NMDA receptors in vitro will apply to
real NMDA receptors in vivo (which may
have different combinations of subunits).
What zinc does in vivo, and why NMDA
receptors should have evolved zinc sensi-
tivity at all, still remains a mystery. How-
ever, Zheng and colleagues have put us
one step closer to understanding the intri-
cacies of this modulatory mechanism.
1. Frederickson, C.J. Intl. Rev. Neurobiol. 131,
145238 (1989).
2. Legendre, P., & Westbrook, G.L. J. Physiol.
Probing a complex question:
1998 Nature America Inc. http://neurosci.nature.com
when are SNARE proteins
ensnared?
Timothy A. Ryan
A recent study uses elegant microphysiological and
molecular tools to investigate the molecular basis and
kinetics of vesicle exocytosis.
Perhaps one of the most sensitive sub-
strates for modulating information flow
in the brain is the molecular machinery
of the synaptic terminal. A detailed mol-
ecular understanding of presynaptic
function is important because it is the
target of many therapeutic reagents for
neurological disease, as well as the site
of action of most stimulants and drugs
of abuse. The molecular description of
postsynaptic events during synaptic
transmission has advanced rapidly over
the last twenty years, thanks to the fruit-
ful combination of electrophysiology
and molecular biology 1 . Our under-
standing of molecular events in the
presynaptic terminal, however, has only
recently begun to emerge. The charac-
terization of presynaptic processes is dif-
ficult because of the inherent
complexity of the underlying organelle
cell biology and because, unlike with ion
channels, the study of any single mole-
cule reveals little about the system as a
whole. An example of the combined
functional and molecular approach
Timothy Ryan is at the Department of
Biochemistry, Cornell University Medical
College,1300 York Avenue,
New York, NY 10021, USA
email: taryan@mail.med.cornell.edu
nature neuroscience volume 1 no 3 july 1998
(Lond.) 429, 429449 (1990).
3. Christine, C.W. & Choi, D.W. J. Neurosci. 10,
108116 (1990).
4. Paoletti, P., Ascher, P. & Neyton, J. J.Neurosci. 17,
57115725 (1997).
5. Williams, K., Neurosci. Lett. 215, 912 (1996).
6. Chen, N., Moshaver, A. & Raymond, L.A. Mol.
Pharmacol. 51, 10151023 (1997).
7. Wang, Y.T. & Salter, M.W. Nature 369, 233235
(1994).
8. Yu, X.M., Askalan, R., Keil, G.J. & Salter, M.W.
Science 275, 674678 (1997).
9. Khr, G. & Seeburg, P.H. J. Physiol. (Lond.) 492,
445452 (1996).
required to tackle these issues is given
by Xu and colleagues in this issue of
Nature Neuroscience 2 , a study that
probes the molecular basis of neuro-
transmitter secretion using modern
microphysiological and molecular tools.
Several different approaches to
understanding membrane transport
have converged over the last five years,
revealing a cast of molecular players
that lie at the heart of vesicle trafficking
and presynaptic neurotransmitter
release 3,4. Considerable evidence now
implies that SNARE proteins5, a family
of compartmentally specific integral
membrane proteins with cytoplasmic
tails, provide a core interaction that
determines the specificity of membrane
pairing. Recently, the assembly of
appropriately paired SNAREs has been
shown to provide the minimal machin-
ery required for the successful mixing
of artificial membrane bilayers 6 . The
best-characterized SNAREs are involved
in neuronal exocytosis. They include the
synaptic vesicle associated protein
synaptobrevin, and the plasma mem-
brane associated proteins SNAP (synap-
tosome-associated protein)-25 and
syntaxin. In vitro, these three proteins
spontaneously assemble into a very sta-
ble ternary complex, whose disassem-
news and views
10. Lu, Y.M., Roder, J.C., Davidow, J. & Salter,
M.W. Science 279, 13631368 (1998).
11. Zheng. F., Gingrich, M.B., Traynelis, S.F. &
Conn, P.J. Nature Neurosci. 3, 185191
(1998).
12. Traynelis, S.F., Burgess, M.F., Zheng, F.,
Lyuboslavsky, P. & Powers, J.L. J. Neurosci.
(in press).
13. Budde, T., Minta, A., White, J.A. & Kay, A.
R. Neuroscience 79, 347358 (1997).
14. Wenzel, H.J., Cole, T.B., Born, D.E.,
Schwartzkroin, P.A. & Palmiter, R.D. Proc.
Natl. Acad. Sci. USA 4, 1267612681 (1997).
15. Choi D.W. & Koh, J.Y Annu. Rev. Neurosci.
21, 347375 (1998).
bly is carried out by the ATPase NSF (N-
ethyl-maleimide-sensitive fusion pro-
tein) together with SNAPs (soluble
NSF-attachment proteins).
The questions of where, when and
how SNARE assembly occurs within
secretory terminals and what specific
state the stable NSF-SNAP-sensitive
ternary complex corresponds to in the
progression of vesicle traffic are now
central to forming an accurate descrip-
tion of presynaptic function. Several
recent reports7,8 and a technical tour-de-
force in the current issue of Nature Neu-
roscience 2 have begun probing these
questions. The approach of Xu and col-
leagues was to apply a combination of
elegant techniques to precisely stimu-
late and measure catecholamine secre-
tion from neuroendocrine cells while
simultaneously interfering with SNARE
function. To accomplish this, they used
whole-cell capacitance measurements to
monitor cell-surface area, carbon-fiber
amperometry to detect catecholamine
release, and photo-uncaging of chelat-
ed calcium to deliver step changes in
intracellular calcium, triggering secre-
tion. The key to these studies is that
synaptobrevin, syntaxin and SNAP-25
are all specific substrates for digestion
by various botulinum and tetanus tox-
ins, and these proteins are only vulner-
able to these toxins when not assembled
in the tight ternary complex. Thus a
measurement of secretion after intra-
cellular toxin application provides an
estimate of the fraction of secretory-
competent vesicles that depend on
SNAREs in a toxin-sensitive, or
unassembled state.
Secretory terminals are comprised of
vesicles in at least two different states of
readiness with respect to the speed at
which they can be caused to fuse with
the plasma membrane. In typical synap-
tic terminals, there is a distinct subset
175
 1998 Nature America Inc. http://neurosci.nature.com
news and views
Fig. 1. A model of the role of SNAREs
in mediating membrane fusion in neu-
ronal exocytosis. Prior to docking with
the plasma membrane, SNARE proteins
on the vesicle membrane are subject to
cleavage by botulinum or tetanus toxin.
SNARE proteins on docked vesicles are
coiled tightly in a ternary complex, pro- NSF-ATP?
tecting them from toxin attack.
Following the elevation of intracellular
Ca++
calcium, the two membranes fuse. NSF
acts to dissamble the SNAREs, freeing X X X X
them and allowing the vesicle to be
recycled. In this model, the membrane
binding step is reversible, allowing
docked vesicles to cycle continuously
through a toxin-sensitive state, without Syntaxin SNAP-25
proceeding to fusion.
Synaptobrevin Tetanus or
Botulinum toxin
1998 Nature America Inc. http://neurosci.nature.com
of vesicles, representing roughly 10% of intracellular calcium elicit an exocytot-
the total population, that can be seen in ic burst proceeding very rapidly for
very close proximity to the plasma about one second, followed by a pro-
membrane, the so-called morphologi- longed slower phase of secretion. In
cally docked pool. Given the very short both types of secretory systems, it is
time delay between the arrival of a believed that release-ready vesicles are
presynaptic action potential and the the first to go when a stimulus arrives.
release of neurotransmitter (less than Given the importance of SNAREs in
one millisecond), it is believed that vesicleplasma membrane interactions,
secretion must draw upon a pool in a what is the arrangement of SNAREs in
release-ready state, probably a subset of the docked or release-ready vesicles? In
the morphologically docked vesicles. In its original form, the SNARE hypothe-
chromaffin cells, brief elevations in sis proposed that vesicles would fuse
with target membranes via
the following sequence of
events 3: first, SNAREs on
the vesicle and target
membranes assemble with
each other, forming the
tight ternary complex and
docking the vesicle to the
target membrane; second,
cytosolic factors, SNAP
NSF-ATP
and the ATPase NSF
Ca++ sequentially bind. The
energy derived from ATP
X X hydrolysis then serves to
fuse the two membranes
and to disassemble the
SNAREs. Thus, according
Bob Crimi to this scenario, one would
Syntaxin SNAP-25 predict that docked vesi-
Synaptobrevin Tetanus or cles would be invulnerable
Botulinum toxin
to toxin attack, and that
Fig. 2. Alternate model for SNARE-mediated exocytosis. measurements of secretion
As in Fig. 1, the SNARES on vesicles prior to docking are in the presence of toxins
sensitive to toxin attack. After vesicle docking, the SNAREs would reveal a small
are partially assembled and still sensitive to toxins. This amount of secretion corre-
state probably requires stabilization by other SNARE-inter- sponding to that derived
acting proteins. Vesicle fusion and recycling as in Fig. 1. The only from the docked
membrane binding step need not be reversible, but a new pool.
stable but toxin-sensitive SNARE arrangement that docks The experiments of Xu
vesicles is proposed. and colleagues indicate
176
that application of all but one
of the appropriate neurotoxins
leads to a complete abolition of
secretion, which is to say that
both the fast and slow phases
of membrane fusion are
blocked. Therefore, all of the
NSF-ATP SNAREs relevant for secretion
in this assay must have been in
a toxin-sensitive state. What
about the SNAREs on docked
vesicles? The authors propose
that the docked state be a
reversible one, so that during
Bob Crimi the toxin incubation period, all
of the readily releasable vesicles
cycle through a state where the
SNAREs are vulnerable to toxin
attack (Fig. 1). This preserves
the original idea that release-
ready vesicles have their SNAREs assem-
bled in a tight complex, but it predicts
that over a five-minute period (the time
to dialyze in the toxin), this state cycles
continuously through an assembly-dis-
assembly sequence. A second possibili-
ty consistent with these results is
depicted in Fig. 2. Here, the docked
state would consist of a partial assem-
bly of SNAREs that is primed and ready
to proceed promptly to fusion with the
elevation of intracellular calcium. Only
after fusion would SNAREs be trapped
in a tight ternary complex as was pro-
posed by Hanson and colleagues9. This
model does not require a reversible
docking process. At present there is lit-
tle in vitro biochemical evidence that a
partially assembled SNARE complex
sensitive to botulinum or tetanus toxin
exists. However, such partial assembly
could be stabilized by one of a number
of SNARE interacting proteins such as
Rab, Sec1, synaptotagmin, complexins
or the recently discovered tomosyn10.
The proposal that the stable ternary
complex represents a postfusion state is
also consistent with the ability of appro-
priately paired unassembled SNAREs to
mediate membrane fusion6. The ener-
gy released during the formation of the
ternary complex could potentially serve
to drive the fusion event, leaving a sta-
ble low-energy complex that requires
disassembly and energy input prior to
vesicle recycling.
These experiments thus constrain
the debate about the placement of the
stable SNARE complex in the secreto-
ry-vesicle life cycle. The use of the cel-
lular milieu as the modern cuvette for
studying native molecular interactions
promises to become increasingly impor-
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com
tant in determining the inner workings
of secretory machinery. In particular, it
should now be possible to pinpoint the
roles of NSF and SNAP as well as that of
the growing family of SNARE-interact-
ing proteins in the synaptic-vesicle life
cycle, using methods like those of Xu
and colleagues.
Getting a line on pain: is it
mediated by dedicated
pathways?
1998 Nature America Inc. http://neurosci.nature.com
Edward R. Perl
Are painful stimuli signaled to the thalamus by distinct
pathways? A new finding of structure/function correlations
in spinal neurons suggests that the answer is yes.
Most people accept that the sensations of
vision, hearing, taste and vibration are
subserved by dedicated neural pathways,
in which the physical stimulus is trans-
duced by specialized sensory cells and
processed by certain neurons and regions
of the central nervous system that are
dominated by their particular sensory
input. Is this also true of pain? Aristotle
held that pain was an emotion rather
than a sensation, and although this view
had been largely abandoned by the 20th
century 1 , the question of whether the
neural basis of pain resembles that of
other sensations remains highly contro-
versial to this day2,3. One concept is that
pain is processed by dedicated pain
receptors that provide input to specific
central pain pathways. An alternate
view3,4 is that pain is not signaled by a
specific labeled-line system of neurons
but instead by a special form of activa-
tion in neurons that are also concerned
with other somatic sensations. In this
view, painful events would be distin-
guished from innocuous stimuli by a
combination of functional characteris-
tics, including the frequency and pattern
of firing in neurons whose lesser or dif-
ferent activities result in other sensory
experiences. A new study in this issue
(pp. 218225) by Han, Zhang and Craig
Edward Perl is at the Department of Physiology,
University of North Carolina-Chapel Hill, CB#
7545, Chapel Hill, North Carolina 27599, USA
email: erp@med.unc.edu
nature neuroscience volume 1 no 3 july 1998
news and views
1. Coloquhoun, D. & Sakmann, B. Neuron 20, 6. Weber, T. et al. Cell 92, 759772 (1998).
381387 (1998).
7. Banerjee, A., Barry, V.A., DasGupta, B.R. &
2. Xu, T., Binz, T., Niemann, H. & Neher, E. Martin, T.F.J. J. Biol. Chem. 271,
Nature Neurosci. 1, 192200 (1998). 2022320226 (1996).
3. Rothman, J.E. Nature 372, 5563 (1994). 8. Schweizer, F.E. et al. Science 279, 12031206
(1998).
4. Scheller, R.H. Neuron 14, 893897 (1995).
9. Hanson, P.I., Roth, R., Morisaki, H., Jahn, R.
5. Sllner, T., Bennett, M.K., Whiteheart, S.W.,
Scheller, R.H. & Rothman, J.E. Cell 75, & Heuser, J.E. Cell 90, 523535 (1997).
409418 (1993). 10. Fujita,Y. et al. Neuron 20, 905915 (1998).
Midline
Central
sulcus
Area 3a
Area 24c
Cerebral
cortex
Thalamus Dorsal
anterior
insula
MDvc
VPI
VMpo
provides strong evidence favoring
at least a partially dedicated system.
Pain in normal animals and COLD
human beings is usually caused by NS
stimuli that are strong enough to HPC
threaten the integrity of the tissues
involved. Detection of such events Lamina I
WDR
is called nociception; sense organs
with the appropriate characteristics Spinothalamic
for nociception have been known Spinal cord tract
for over thirty years and are highly
conserved between different mam- Bob Crimi
malian species6. However, pain can Fig. 1. Diagram representing the relationships
arise from stimuli or events that are between functional category, cellular morphology
not noxious. This has been used to and thalamic projection of spinal dorsal horn lamina I
argue that nociception is not neurons proposed by Han, Zheng and Craig
directly associated with pain 3 , (adapted from ref. 2). MDvc, medial dorsal nucleus
although there is substantial evi- ventral caudal portion; VMpo, ventral medial
dence that noxious stimuli are nucleus, posterior part; VPI, ventral posterior infe-
closely related to indications of rior nucleus; NS, nociceptive specific neuron; COLD,
aversive behavior in animals or of innocuous cooling neuron; HPC, heat, pinch, cooling
pain in human beings. neuron; WDR, multireceptive neuron (wide
Han and colleagues have used a dynamic range).
combination of electrophysiology
and histology to characterize the
neurons of lamina I in the dorsal
horn of the cat spinal cord, one of the ear- thinly myelinated or unmyelinated fibers
liest stages in the nociception pathway. of specific thermoreceptors and various
The dorsal-horn gray matter is divided nociceptors7, suggesting a role in sensing
into zones, or laminae, based on varia- tissue damage and temperature. Consis-
tions in the size, shape and density of its tent with this, some neurons of lamina I
neurons. Primary sensory fibers from the are selective either for noxious stimuli or
somatic and visceral tissues terminate for innocuous changes in temperature8,9.
throughout the dorsal horn, and the ter- Moreover, lamina I neurons project to the
minals from different functional classes thalamus and several other brain regions
are segregated in different laminae. Lam- via the contralateral spinothalamic tract,
ina I receives its input largely from the which is important for the perception of
177
 1998 Nature America Inc. http://neurosci.nature.com
news and views
pain from the opposite side of the body1.
An important question is whether the
neurons of lamina I are specialized for
different types of painful stimuli, and
whether these different classes send their
outputs to particular brain regions. Such
an arrangement would have significant
implications for understanding how
afferent signals combine to give rise to
pain. Previous papers from Craigs labo-
ratory and others have described, in cat
and monkey, at least three functional
classes of neurons in lamina I that pro-
ject to the thalamus2,9. One class is noci-
ceptive-specific (NS) with inputs from
one or more types of nociceptor, anoth-
er responds to innocuous cooling
(COLD) and a third class is multimodal,
1998 Nature America Inc. http://neurosci.nature.com
responding to heat, pinch and cooling
(HPC). The authors have reported pre-
viously that these different functional
classes project to different thalamic
nuclei, which in turn project to different
cortical regions (Fig. 1). The present
findings greatly strengthen their pro-
posed functional classification, by show-
ing that the three classes of neurons
defined by their physiological selectivity
correspond to particular morphological
types, distinguished by the shapes of
their cell bodies and dendritic processes.
This is consistent with a particular
defined pathway for nociception and
pain and seems difficult to reconcile with
a model in which pain arises from func-
tional interactions between neurons that
are not themselves specified for particu-
lar sensory experience.
The new findings leave unresolved the
relative importance for pain of the NS and
HPC neurons of lamina I and the so-
called wide-dynamic-range (WDR) neu-
rons described by other researchers. WDR
neurons are excited only modestly by
innocuous mechanical stimuli and fire
most strongly in response to noxious
mechanical or heat stimuli. They are
found mainly in laminae V and VII and
project to the thalamus via the spinothal-
amic tract10. They are relatively rare in
lamina I (especially in cats), and so the
authors were not able to characterize them
in the present study. However, they have
been proposed to convey pain signals to
the brain in humans and to be important
for aversive behavior in animals6,11. WDR
neurons are multipolar cells12, similar to
HPC neurons, and project to at least one
thalamic locus2. Is it possible that HPC
and WDR neurons are of the same type
and serve similar functions?
Why has this structure/function cor-
relation not been found in previous stud-
178
ies? In most experiments, Han and col-
leagues used a more selective method than
other researchers to choose their neurons,
testing for antidromic activation of pro-
jecting neurons by finding and stimulat-
ing a restricted thalamic region that
activates lamina I neurons. Lamina I neu-
rons are difficult to record, because they
are flattened at the interface between the
outermost part of the gray matter and the
largely myelinated fibers of the spinal cord
tracts. Moreover, their form is not easily
appreciated unless viewed in horizontal
histological preparations, which were not
used in previous studies 1315. Another
consideration is that most previous stud-
ies did not concentrate exclusively on lam-
ina I neurons, but included observations
in the subjacent lamina II, possibly blunt-
ing distinctions. A further difficulty is the
variation between individual neurons
within the categories (e.g., fusiform,
pyramidal) reported by Han and col-
leagues; although this need not invalidate
their classification, it would have made
matters difficult for previous workers
lacking the advantage of the horizontal
plane in seeking structure/function rela-
tionships within this population.
The degree of specificity implied by
the new results does not necessarily indi-
cate that the system is genetically hard-
wired. The shape of neurons could
equally reflect their environments,
including the connections they receive
from other neurons, rather than being
an inherent property of each neuron
itself. But whichever mechanism is
responsible for the observed
structure/function relationship, the exis-
tence of neurons partially dedicated to
conveying information about noxious or
thermal stimuli fails to explain in itself
either the sensation of pain or its aber-
rations, just as our present understand-
ing of the auditory system fails to
explain hearing. Even if the functional
and structural specialization of spinal
neurons projecting pain-relevant infor-
mation is reminiscent of other sensory
systems, this does not indicate a tele-
phonic relay arrangement of the sort
implied by the term labeled line. Mam-
malian sensory systems are generally
plastic and reflect their past history. It is
not surprising that the same should be
true of a system associated with pain,
and as Han and colleagues acknowledge,
the specificity they describe may not be
absolute, so that transmission by the dif-
ferent pathways might be modulated not
only by past experience but also by
ongoing activity from other brain areas.
We now know from brain imaging stud-
ies that a given sensory stimulus acti-
vates multiple cerebral cortical zones.
Somatic sensations, including pain,
seem to involve an as yet unknown com-
bination of ascending afferent informa-
tion with an intrinsic background of
neural activity modified by past history
or disease.
It seems likely that there is merit in
the views of each of the most diametri-
cally opposed camps in the controversy
about the nature of pain mechanisms.
Accepting the observations of Han and
colleagues, there is a system of central
neurons, morphologically and func-
tionally distinct beginning in lamina I
of the spinal cord, whose usual function
is to inform the organism about tissue
damage or the threat of such damage. It
seems to be situated so that activity in
some of its constituents at several levels
of the central nervous system normally
leads to pain or equivalent reactions.
However, once these messages reach the
brain, they are processed in parallel by
more than one set of central neurons.
The eventual sensory experience almost
surely represents the interplay in a mix-
ture of serial and parallel processing
integrated into the activity of other
neural systems related to elaboration of
consciousness, emotion and memory.
1. Keele, K.D. Anatomies of Pain (Thomas,
Springfield, Illinois, 1957).
2. Craig, A.D. Pain Forum 7, 114 (1998).
3. Wall, P.D. Pain 62, 389391 (1995).
4. Melzack, R. & Wall, P.D. Science 150, 971979
(1965).
5. Belmonte, C. & Cervero, F. Neurobiology of
Nociceptors (Oxford Univ. Press, New York,
1996).
6. Dubner, R., Kenshalo, D.R., Maixner, W.,
Bushnell, M.C. & Oliveras, J.L. J.
Neurophysiology 62, 450457 (1989).
7. Light, A.R. The Initial Processing of Pain and
Its Descending Control: Spinal and Trigeminal
Systems (Karger, Basel, 1993).
8. Christensen, B.N. & Perl, E.R. J. Neurophysiol.
33, 293307 (1970).
9. Dostrovsky, J.O. & Craig, A.D. J. Neurophysiol.
76, 36563665 (1996).
10. Willis, W.D. The Pain System (Karger, Basel,
1985).
11. Mayer, D.J., Price, D.D. & Becker, D.B. Pain 1,
5158 (1975).
12. Ritz, L.A. & Greenspan, J.D. J. Comp. Neurol.
238, 440452 (1985).
13. Ferrington, D.G., Sorkin, L.S. & Willis, W.D. J.
Physiol. (Lond.) 388, 681703 (1987).
14. Light, A.R., Trevino, D.L. & Perl, E.R. J. Comp.
Neurol. 186, 151171 (1979).
15. Woolf, C.J. & Fitzgerald, M. J. Comp. Neurol.
221, 313328 (1983).
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com
Cortical control of the
thalamus: top-down
processing and plasticity
Josef P. Rauschecker
Blocking cortical NMDA receptors leads to a dramatic expansion
of thalamic receptive fields. The results help to illuminate the
role of corticofugal connections in sensory processing.
Sensory information is relayed to the
cortex via the thalamus, yet there are
nearly ten times as many fibers project-
1998 Nature America Inc. http://neurosci.nature.com
ing back from the cortex to the thala-
mus as there are in the forward
direction from thalamus to cortex 1,2 .
The function of this massive feedback
projection has puzzled neuroscientists
for decades. Initial attempts to address
this question were largely confined to
the visual system, but more recently,
studies on the corticofugal modulation
of subcortical responses in the auditory
system 3,4 have provided additional
insight. The most recent contribution5,
in this issue of Nature Neuroscience
(p 226), extends the story to the
somatosensory system and adds a new
twist. From the combined results in dif-
ferent sensory systems, we can see the
emergence of a more generalized under-
standing of the role of the corticothala-
mic loop in sensory information
processing.
Ergenzinger and colleagues 5 deliv-
ered an NMDA receptor antagonist
(APV) into the primary somatosensory
cortex of monkeys, at the site of the
hand representation. After several
months of treatment, the tactile recep-
tive fields within the somatosensory part
of the thalamus (the ventroposterior
(VP) nucleus) showed an enormous
enlargement in the hand region. Simi-
lar, albeit less dramatic, enlargements
were also found after acute treatment.
The results are intriguing because they
challenge the traditional view of senso-
ry plasticity as a bottom-up process,
whereby changes on one level are sim-
ply passed on to the next higher level.
By contrast, the new study shows that
Josef Rauschecker is at the Georgetown Institute
for Cognitive and Computational Sciences,
Georgetown University, Washington, DC 20007,
USA
email: rauscheckerj@giccs.georgetown.edu
nature neuroscience volume 1 no 3 july 1998
news and views
over, simultaneous recordings of cortical
and LGN activity revealed that cortical
firing was associated with an increased
probability of firing in LGN neurons
with overlapping receptive fields but not
more distally, suggesting that corticofu-
gal connections contribute to thalamic
excitation in a very specific manner
dependent on their respective locations
on the retinotopic grid.
The latter result ties in well with
another study8 on the visual system, in
which the authors recorded simultane-
ously from pairs of LGN neurons that
were stimulated by a moving bar and
plasticity also is a top-down process and then removed cortical feedback. With
that plastic changes can be propagated cortical feedback, the LGN neurons
in the reverse direction. fired synchronously, whereas without it,
These plastic changes seem likely to this synchrony was lost. In this case, the
be related to the normal role of the cor- corticothalamic loop would be expect-
ticofugal system in the transmission of ed to amplify the response to the mov-
sensory information. To
understand how this
might be, it is useful to Layer IV
review some of the
findings on corticofugal Cortex
connections in other Layer VI
systems. In one early
study6, the cat primary
visual cortex was
reversibly cooled to
inactivate the corticofu-
gal feedback to the lat-
eral geniculate nucleus
of the thalamus. This Thalamus
inactivation reduced
both the light-evoked
responses and the level
of spontaneous activity Sensory
in most thalamic relay periphery
neurons. Thus, corti-
cofugal fibers seem to
directly but weakly +
excite relay cells, as well
as disinhibit them by
Bob Crimi
reducing the strength of
intrinsic inhibitory Fig. 1. Thalamocortical network in which the corticofugal
mechanisms. A later projection provides narrow, highly specific positive feedback
study 7 provided more (dark orange) to a thalamic neuron with a classical center-sur-
specific evidence about round receptive field, as well as negative feedback (via
the wiring of the corti- inhibitory interneurons, shown in green) to a large number of
cothalamic loop; gluta- surrounding thalamic neurons (pale orange), creating a wide
inhibitory suppression field (ref. 2). During normal sensory
mate, applied
processing, this circuitry helps to sharpen the contrast of the
iontophoretically to a sensory input in time and space (or frequency in case of the
small patch of the pri- auditory system) and suppresses irrelevant information (ego-
mary visual cortex, centric selection, ref. 3, 4, 9). When the cortex is inactivated,
excited LGN neurons e.g. by means of NMDA receptor antagonists (ref. 5), this
that shared receptive leads to an unmasking of the suppressed input from recurrent
fields with cortical neu- collaterals of other thalamic neurons within the suppression
rons at the iontophore- field, which causes an enlargement of thalamic receptive
sis site but inhibited fields. As the cortical activity block continues for several
LGN neurons that were months, this effect on thalamic neurons can be surprisingly
out of register. More- large, perhaps magnified by secondary plastic changes.
179
 1998 Nature America Inc. http://neurosci.nature.com
news and views
ing bar stimulus, because synchroniza-
tion of thalamic relay neurons means
that they will elicit temporally overlap-
ping excitatory postsynaptic potentials
(EPSPs) in any common cortical target
cells, thereby increasing the chance that
such cortical cells will fire.
The early work of Tsumoto and col-
leagues in vision is complemented in an
intriguing fashion by recent work of
Suga and colleagues on the corticofugal
projections of the bats auditory system.
When cortical neurons tuned to specif-
ic frequencies are inactivated by local
injections of lidocaine 4 , the auditory
responses of neurons in the thalamic
medial geniculate nucleus (MGN) tuned
to the same frequency are reduced,
1998 Nature America Inc. http://neurosci.nature.com
whereas responses of neurons tuned to
different frequencies are enhanced. The
same is true, incidentally, for neurons
in another subcortical auditory nucle-
us, the inferior colliculus, suggesting
that the role of the corticofugal projec-
tion may be quite general9. Moreover, if
the cortical site is (electrically) stimu-
lated rather than blocked, the effect is
to increase and sharpen the responses of
a particular subclass of neurons impor-
tant in the bats echolocation behavior,
the delay-tuned neurons, in both thala-
mus and colliculus3. Like frequency and
retinotopic location, delay tuning is
mapped in a topographic fashion corti-
cally and subcortically and again, the
effect of cortical stimulation is highly
specific for neurons with the same delay
tuning on both levels.
These findings suggest a common
role for the corticothalamic loop in sen-
sory information processing in different
systems: the corticofugal projection pro-
vides positive feedback to the correct
input while at the same time suppress-
ing irrelevant information. It seems that
the positive feedback is very precisely
matched between cortex and thalamus,
such that a given thalamic neuron
receives excitation only from a small area
of cortex that shares the same stimulus
selectivity (for instance, a particular
location in visual space or a particular
sound frequency). Based on their work
in the auditory system, Suga and col-
leagues have termed this principle ego-
centric selection. This process leads to a
temporal and spatial contrast sharpen-
ing of the sensory input and, as suggest-
ed by Singer10, may play a crucial role in
figure-ground discrimination. Although
180
Singers classical paper was based solely
on data then available from the visual
system, the concept can now be extend-
ed to the auditory (and perhaps most
other) sensory systems, where corti-
cofugal feedback may play a role in anal-
ogous processes such as auditory stream
separation and scene analysis11.
The circuitry required for the mixed
excitatory-inhibitory feedback provid-
ed by the cortico-thalamic loop is
depicted in Fig. 1. The highly repetitive
structure seemingly does not leave any
room for selectivity, but as pointed out
by Koch 12 , the answer may lie in the
gating properties of the thalamic
NMDA receptors. Thalamic relay neu-
rons receive excitatory glutamatergic
input both from peripheral sensory
pathways and from corticofugal con-
nections. Assuming that these inputs act
partly through NMDA receptors, the
relay neurons would be activated only
if they receive sensory input and cortical
feedback at the same time. Given the
results of Sillito and colleagues8, only
those inputs that occur within a narrow
temporal window would be amplified.
Similarly, only stimuli that fall within a
defined region of the sensory receptor
surface (for instance the retina or
cochlea) would be amplified. Anything
that is part of the irrelevant din would
be suppressed, because without cortical
feedback (and hence NMDA receptor
activation), the inputs from the periph-
ery are not strong enough to overcome
the effects of thalamic inhibitory
interneurons, which are themselves dri-
ven by both sensory and corticofugal
inputs (see Fig. 1). Thus, although the
corticofugal input to the thalamus helps
to amplify certain inputs (similar to a
winner-take-all mechanism), it also
generates a large suppression field 2 ,
which exists outside the classical recep-
tive field.
This circuit model for the selective
modulation of sensory information can
also account for the results of Pons and
colleagues. Applying an NMDA recep-
tor antagonist to the cortical neurons
essentially silences this part of the cor-
tex (an effect that complicates the inter-
pretation of the early results on the
effects of NMDA receptor antagonists in
visual cortex; see ref. 13). The antago-
nist has no direct effect on the thalamus
(in contrast to cortical lesions, which
lead to retrograde thalamic degenera-
nature neuroscience volume 1 no 3 july 1998
tion). Thus, immediately after applying
APV to the cortex, receptive fields in the
thalamus are enlarged, which can be
explained as a disinhibition of the sup-
pression effects of the corticofugal fibers.
Applying the APV chronically, as in the
main finding of the study, renders the
thalamic receptive field enlargements
enormous, suggesting that the original
disinhibition effect leads to further plas-
tic changes, perhaps in the thalamus
itself. This may also yield a late explana-
tion of the earlier results by Pons and
colleagues of a massive expansion of the
cortical hand representation after chron-
ic deafferentation14.
As Yan and Suga point out in their
latest paper, published earlier this year
also in Nature Neuroscience9, more sub-
tle forms of plasticity could also be
mediated by the corticofugal loop. It
could be that thalamic maps are con-
stantly adjusted by sensory experience,
as are cortical maps. In conclusion,
function and plasticity have often been
thought of as highly interwoven aspects
of cortical structure, neural activity
being the factor that ties them together.
The study by Ergenzinger et al.5 makes
it clear that these dual aspects of high-
er perceptual and cognitive systems
need to be extended to the thalamus, at
least insofar as the latter is controlled by
its cortical input.
1. Guillery, R. W. J. Comp. Neurol. 130,
197222 (1967).
2. Jones E. G. The Thalamus (Plenum, New
York, 1985).
3. Yan, J. & Suga, N. Science 273, 11001103
(1996).
4. Zhang, Y. Suga, N. & Yan, J. Nature 387,
900903 (1997).
5. Ergenzinger, E. R., Glasier, M. M., Hahm, J.
O. & Pons, T. P. Nature Neurosci. 1, 226229
(1998).
6. Kalil, R. E. & Chase, R. J. Neurophysiol. 33,
459474 (1970).
7. Tsumoto, T., Creutzfeldt, O. D. & Legendy,
C. R. Exp. Brain Res. 32, 345364 (1978).
8. Sillito, A. M., Jones, H. E., Gerstein, G. L. &
West, D. C. Nature 369, 479482 (1994).
9. Yan, J. & Suga, N. Nature Neurosci. 1, 5458
(1998).
10. Singer, W. Physiol. Rev. 57, 386420 (1977).
11. Bregman, A.S. Auditory Scene Analysis (MIT
Press, Cambridge, Massachusetts, 1990)
12. Koch, C. Neuroscience 23, 399406 (1987).
13. Rauschecker, J.P. Physiol. Rev. 71, 587615
(1991).
14. Pons, T. P. et al. Science 237, 18571860
(1991).
 1998 Nature America Inc. http://neurosci.nature.com

scientific correspondence

Where is the sun?
Jennifer Sun1 and Pietro Perona1,2
1
California Institute of Technology 136-93, Pasadena, California 91125, USA
2 Universita di Padova, Via Ognissanti 72, 35131 Padova, Italy
Correspondence should be addressed to P.P. (perona@vision.caltech.edu)
When we interpret a shaded picture as a three-dimensional (3D)
scene, our visual system often needs to guess the position of the light
source in order to resolve a convex-concave ambiguity. For more
than a century, psychologists have known that the visual system
assumes that light comes from above and have argued that this
assumption is ecologically justified because our everyday light source
(the sun) is overhead. Our experiments reveal that peoples preferred
lighting direction is not directly overhead, but rather shifted to the
left, and this preference is reflected in art spanning two millennia.
1998 Nature America Inc. http://neurosci.nature.com
We addressed these questions by measuring the time it takes
to detect, within a group of distractor bubbles, a single target
bubble that is lit differently (Fig. 1b). Recent studies suggests that
the light-from-above assumption is used by the visual system for
interpreting quickly and in parallel some basic aspects of 3D
scenes48; the target pattern may be detected quickly (pop-out)
only when the distractors, but not the target, can be interpreted
as convex and lit from above59. We simulated different directions
of lighting by varying the shading gradient of the distractor bub-
bles. The target bubble was shaded to simulate illumination from
the opposite direction (Fig. 1b).
Data from twelve naive subjects shows that the visual system
does not respond uniformly to all lighting directions that are above
the horizon (Fig. 1c). There is clearly a preferred direction of light-
ing where detection requires the shortest display time. Surpris-
ingly, this preferred direction is not directly overhead (zero
degrees). Subject PG, for instance, performs best with a lighting
direction that is between 30 and 60 degrees left of the vertical

Furthermore, we find a strong correlation between peoples hand- (Fig. 1c). Furthermore, there is a consistent preference for left light-
edness and their preferred lighting. We suggest that what counts is ing over right lighting. This same marked leftright asymmetry is
not so much where the sun is, but where you like the sun to be. evident in the averaged data of all twelve subjects. As the angle of
The shaded shapes in Fig. 1a are typically perceived as convex illumination increases, the preference for left lighting becomes
bubbles surrounding concave indentations, all lit from above. Note increasingly pronounced (Fig. 1d).
that this image is also consistent with a different physical scene: This asymmetry in our data may explain a qualitative observa-
indentations surrounding bubbles, all lit from below. This second tion made by Gestalt psychologist Metzger, who noted that left-lit
perception, however, is difficult to achieve. Such an asymmetry scenes have a superior perceptual value over right-lit ones10. He
between the perceptual saliency of equally valid 3D interpretations ascribed this asymmetry to the convention of setting up desk lamps
demonstrates that our visual system prefers the assumption that on the left, presumably so that the writing hand does not cast a
light is coming from above13. Does this preference apply uniform- shadow on the page. Over time, he hypothesized, one learns to per-
ly to all lighting directions that are above the horizon? Perhaps there ceive left-lit scenes as being more natural. Metzgers explanation
is instead a preferred direction? If so, one might reason it to be direct- may be somewhat restrictive: our visual environment extends
ly overhead. Is this intuitive guess correct? beyond our writing desk. We tend to position a movable light
Performance
Performance versus
vs Lighting lighting
Direction direction
Subject PG
a b c 500 (subject PG)
Left Lighting Right Lighting
450 Left lighting Right lighting
(ms)

400
duration msec

350
MFV Duration

300
MFV

250

200

150
180 150 120 90 60 30 0 30 60 90 120 150
Lighting direction
Lighting Direction of distratorsdegrees
of Distractors (degrees)
Fig. 1. Shaded displays that may be interpreted as 3D shapes and measurement of preferred
lighting direction. (a) Rotate the page to invert the shapes. (b) Images were generated on a
Silicon Graphics Indigo2. Each `bubble spanned approximately one degree. One target pattern d Perfomance differences
g g
at
various lighting directions
was present at random among 23 distractor patterns in 50% of the trials. The remaining trials 140

contained 24 distractors and no target. The lighting direction is determined by the shading gra-
Durations msec

12 Subjects
Difference in MFV durations

120 pp<<0.001
0.001
dient of the distractors. Target-present test screens are shown for 2 of the 12 lighting direc-
tions used in our experiment. We denote a lighting direction by its deviation from the vertical 100
in degrees. Positive degrees indicate lighting from the left, and negative degrees indicate lighting
from the right. Accordingly, lighting from directly overhead is designated as 0 degrees, and light- 80
(ms)

pp<< 0.007
0.007
Difference in MFV

ing from directly below as 180 degrees. (c) We used a two-alternative forced-choice stimulus
60
onset asynchrony (SOA) design with masking. Data was collected using a staircase method that
converged at 67% accuracy performance. The most frequently visited (MFV) duration within 40 pp << 0.08
0.03
each block was used to estimate 67% accuracy performance. The duration necessary for 67%
accuracy for each lighting direction is shown for subject PG. (d) We computed the mean dif- 20
ference in necessary display duration between pairs of corresponding left-right lighting direc-
0
tions over all 12 naive subjects. The necessary duration for a left-lighting condition is 30, 30 60, 60 90, 90
subtracted from the necessary duration for the corresponding right-lighting condition. Lighting
Lighting Directionscompared
directions Compared (rightleft)
(right-left)

nature neuroscience volume 1 no 3 july 1998 183

 1998 Nature America Inc. http://neurosci.nature.com

scientific correspondence

Preferred angles of Preferred lighting Fig. 2. The preferred light direction corre-
a 35 lighting direction b 60
direction versus handedness lates with handedness. (a) Subjects were

Direction degrees
lefthander
50 categorized based on which hand they used
p << 0.0001
0.0001 righthander

Preferred lighting direction

30
40 for writing. The mean preferred lighting
Angle left of the vertical

25 direction was compared between the six

30
right-handed and the six left-handed sub-

(degrees)
20
20 jects. Although both groups show a prefer-
(degrees)

Preferred Lighting
15 p << 0.03
0.03 10 ence for left-lighting, right-handers, as a
Left Preference group, prefer a larger angle left of the verti-
10 0 cal (23.3 degrees) than do left-handers (7.9
Right Preference
5 -10 degrees). (b) Preferred lighting direction is
-20 correlated with handedness (r = 0.83,
0 - 10 -5 0 5 10 p<0.01). * lefthander, p righthander.
Lefthanders
Lefthanders Righthanders
Righthanders Handedness
Handedness

source, or position ourselves in relation to a fixed light source, such the effect mirror symmetric, with left-handers having a preference for
that our hand does not cast a shadow upon the object of our manip- right-lighting instead of left? This may be explained because left-
ulation. Right handers would then develop a preference for left- handers live in a right-handers world and are often forced to func-
1998 Nature America Inc. http://neurosci.nature.com

lighting, and, as suggested by van Fieandt11, left-handers may well tion in environments that are designed for right-handers. If lighting
show the opposite lighting preference. preference is not determined by a biological trait correlated with
To investigate this possibility, we derived the preferred light- handedness itself, but rather by handedness-related experience, then
ing direction and handedness for our subjects. We defined each one would expect that right-handers would prefer left-lighting,
subjects preferred lighting direction to be the lighting direction whereas left-handers, because of their mixed experience, would
for which target detection performance was best. This was esti- exhibit a weaker preference for either left or right lighting.
mated by fitting a parabola to the central portion of each subjects Quantitative measurements of the perception of shape from
time versus direction curve (e.g. Fig. 1c). and calculating the direc- shading using static displays have revealed no asymmetry between
tion for which the parabola reached a minimum. When prefer- left- and right-lighting conditions14,15, thereby failing to confirm
ence for lighting direction is considered for left- and right-handers Metzgers observation. One might, therefore, suspect that the pref-
separately, we find that both groups have a preference for left- erence for left lighting we observe with our fast-presentation para-
lighting. However, the right-handers as a group prefer a lighting digm is confined to the earlier stages of visual processing, and that
direction that is significantly more toward the left than the left- this effect may become negligible under ecological viewing condi-
handers preference (Fig. 2a). tions. We have reason to think otherwise. We asked two naive sub-
Handedness, however, is not a strictly binary trait; rather it varies jects, one right-hander and one left-hander, to survey 225 master
in a continuum12. We used a standard ten-item questionnaire that paintings and determine the predominant angle of lighting for each.
evaluated the relative strengths of our subjects handedness13. The The histogram of the measurements (Fig. 3) shows that the artists
resulting score ranges from -10 to 10, with positive values indicat- most often chose a lighting direction that is left of the vertical. This
ing a bias for the right hand, and negative values for the left. When preference for top-left lighting may have resulted from an accidental
each subjects preferred lighting direction is plotted against this hand- artistic convention. However, this is unlikely, as preference left-light-
edness score, a strong correlation is found (Fig. 2b). ing is found across schools and periods: from Roman mosaics,
If lighting preference is indeed related to handedness, why isnt through Renaissance, baroque, and impressionist art. It is therefore
possible that top-left lighting may actually have a higher perceptual
g g
Evaluation of lighting direction in paintings value than top-right lighting in natural viewing conditions, which
70 involve frequent saccades over the entire scene. Perhaps when sub-
Left Right jects are required to make their shape judgments under prolonged
60
scrutiny of a localized portion of the test stimulus14,15, the effect
50 drops below a measurable level.

40 Acknowledgements
The authors are grateful to Dr. Marianne Tauber for providing the references to the
30
Gestalt literature. Support was provided by the NSF Engineering Reasearch Center
20 for Neuromorphic Systems at Caltech.

10 1. Rittenhouse, D. Trans. Am. Phil. Soc. 2, 3742 (1786).

2. Brewster, D. Edinburgh Phil. Trans. 15, 657 (1847).
3. Ramachandran, V.S. Nature 331, 163166 (1988).
0
90 60 30 0 30 60 90 4. Enns, J.T. & Rensink, R.A. Science 247, 721723 (1990).
Lighting direction
Lighting of of
Direction painting (degrees)
Painting degrees 5. Braun, J. Perception 19, A112 (1990).
6. Braun, J. Spatial Vision 7, 311322 (1993).
Fig. 3. Painters tend to light scenes from top-left. 100, 100 and 25 7. Kleffner, D.A. & Ramachandran, V.S. Percept. Psychophys. 52, 1836 (1992).
8. Sun, J.Y. & Perona, P. Vision Res. 36, 25152529 (1996).
paintings were randomly selected from catalogues of the Louvre, 9. Sun, J.Y. & Perona, P. Nature 379, 165168 (1996).
the Prado, and the Norton Simon Museum. Two naive subjects eval- 10. Metzger, W., Gesetze des Sehens. Frankfurt am Main (1936).
uated these paintings for lighting direction. Using a protractor, 11. van Fieandt, K. Psychologischen Institut Monog. (Helsinki University, 1938).
12. Durost, W.N. Genet. Psychol. Monog. 16, 225335 (1934).
77 0.55% of the paintings were classified as being lit from the left 13. Oldfield, R.C. Neuropsychologia 9, 97113 (1971).
(p<0.05). The artists most often selected lighting directions that are 14. Blthoff, H.H. & Mallot, H.A. J. Opt. Soc. Am. A 5, 17491758 (1988).
between 30 and 60 degrees to the left of the vertical. 15. Todd, J.T. et al. Invest. Ophthal. Vis. Sci. 37, 4282 (1996).

184 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

Tyrosine kinase potentiates NMDA

receptor currents by reducing tonic
zinc inhibition
F. Zheng, M.B. Gingrich, S.F. Traynelis and P.J. Conn

Department of Pharmacology, Emory University School of Medicine, 1510 Clifton Road, Atlanta, Georgia 30322, USA
Correspondence should be addressed to F.Z. (fzheng@emory.edu)

Activation of the tyrosine kinase Src potentiates NMDA-receptor currents, which is thought to be
necessary for induction of hippocampal long-term potentiation. Although the carboxy(C)-terminal
1998 Nature America Inc. http://neurosci.nature.com

domain of the NR2A subunit contains potential tyrosine phosphorylation sites, the mechanism by
which Src modulates synaptic plasticity and NMDA receptor currents is not fully understood. Here
we present evidence from NR1 mutants and splice variants that Src potentiates NMDA-receptor cur-
rents by reducing the tonic inhibition of receptors composed of NR1 and NR2A subunits by extracel-
lular zinc. Using site-directed mutagenesis, we have identified three C-terminal tyrosine residues of
NR2A that are required for Srcs modulation of the zinc sensitivity of NMDA receptors. Our data link
two modulatory sites of NMDA receptors that were previously thought to be independent.

Recent studies suggest that tyrosine kinases are important for
synaptic plasticity 13. The requirement for Src activation in
long-term potentiation (LTP) induction has been attributed
to its enhancement of NMDA-receptor currents46 because an
NMDA-receptor antagonist blocks this effect3. The non-recep-
tor tyrosine kinases Src and Fyn potentiate receptors composed
of NR1 and NR2A subunits, but have no apparent effect on
receptors composed of NR1 and other NR2 subunits 5. Fur-
thermore, deletion of the C-terminal domain of NR2A elimi-
nates the potentiation of NR1/NR2A-receptor currents by Src5
and impairs synaptic plasticity and contextual memory7. Thus,
several lines of evidence suggest that a unique tyrosine phos-
phorylation site on the C-terminal of NR2A may confer sen-
sitivity to Src and be critical for the induction of LTP. However,
the mechanisms by which tyrosine kinase potentiates NMDA-
receptor function are unknown.
In addition, NR2B subunits are phosphorylated by one of
the Src family of tyrosine kinases8, but no functional effect of
this phosphorylation has been identified. This discrepancy
between biochemical and physiological data raises the possibil-
ity that either tyrosine phosphorylation of NR2B affects recep-
tor properties other than channel activity, or another property of
NR2A may be responsible for the apparent selective potentia-
tion by tyrosine kinase of receptors containing this subunit.
NMDA receptors are allosterically modulated by a variety of
endogenous extracellular ions913. Zinc, one of these modulators,
is accumulated in some nerve terminals in specific brain regions
and released during neuronal activity14. Zinc inhibits NMDA
receptors at two independent sites1517. A low-affinity zinc site is
probably located inside the channel pore, and binding of zinc to
this site causes a voltage-dependent inhibition of NMDA-receptor
channels. A high-affinity site is likely located outside the channel
pore and causes a voltage-independent inhibition. Receptors con-
taining NR2A subunits exhibit far greater affinity for zinc at the
extracellular, voltage-independent site than receptors containing
other NR2 subunits1821. This sensitivity to zinc is high enough
to allow ambient zinc (either in vivo22 or as a contaminant of
experimental solutions23) to tonically inhibit NR1/NR2A recep-
tors. Here we show that Src reduces zinc sensitivity of recombi-
nant NR1/NR2A and NR1/NR2B receptors, thereby relieving
tonic inhibition by zinc and potentiating the NMDA-receptor
response. Splice variants and mutants of NR1 subunits that have
a low apparent affinity for zinc are, as predicted, potentiated to a
lesser degree by Src. Conversely, NR1 mutants with higher appar-
ent zinc affinity are potentiated to a greater degree. Using site-
directed mutagenesis, we have identified three C-terminal tyrosine
residues of NR2A that are required for Srcs modulation of
NMDA-receptor zinc sensitivity. By showing that Src potentiates
NMDA receptors by specifically reducing tonic zinc inhibition,
our data link two modulatory sites of NMDA receptors that were
previously thought to be independent.
Results
SRC REDUCES TONIC INHIBITION OF NR1/NR2A RECEPTORS
Our hypothesis was that potentiation of NMDA receptor cur-
rents by Src may be due to its relief of tonic inhibition by zinc
acting at the voltage-independent site, in a manner analogous
to polyamine relief of tonic proton inhibition of NMDA recep-
tors 13 . To test this hypothesis, we examined the effects of
EDTA, a chelator of transition metals such as zinc, on Srcs
potentiation of NR1/NR2A receptors. The NMDA receptor
currents were recorded with a rapid perfusion system in
HEK293 cells transiently transfected with NR1-1a/NR2A
receptors. In control experiments, the amplitudes of evoked
NMDA-receptor currents did not change significantly over
time (Fig. 1a and b). When purified recombinant human c-
Src was added to the internal solution, the amplitude of
NMDA-receptor currents was gradually increased over a five-
minute period (Fig. 1a). On average, the peak of NMDA
receptor currents evoked by 100 M glutamate and 30 M

nature neuroscience volume 1 no 3 july 1998 185

 1998 Nature America Inc. http://neurosci.nature.com
articles
Fig. 1. Src potentiates NR1/NR2A receptors by
reducing tonic zinc inhibition. (a) Removal of tonic
a b
zinc inhibition of NR1-1a/NR2A receptors occludes
the potentiation of these receptors by Src. NMDA-
receptor current traces at left (labeled 1) were the
first of a series of current traces recorded 6090 sec-
onds after obtaining the whole-cell configuration.
Traces at right (labeled 30) were the last of 30 con-
secutive traces recorded every ten seconds over a
five-minute period. EDTA is added into extracellular
solutions to chelate zinc in some experiments,
whereas recombinant human c-Src is added into the
internal solution of patch pipettes. Under control
conditions, no significant rundown of the NMDA-
c Control
receptor current was observed. Inclusion of c-Src
(30 units per ml) resulted in a 50% potentiation.
However, this potentiation was not observed in the
presence of 10 M EDTA. (b) Average changes of
the NMDA-receptor current amplitudes over the
1998 Nature America Inc. http://neurosci.nature.com
same periods. Error bars are standard error for all
panels. Number of cells is indicated in parentheses.
P control (5), L Src (10), G Src and EDTA (5).
(c) The effects of EDTA (10 M) on NMDA currents
evoked by agonists (100 M glutamate, 30 M glycine) from representative HEK 293 cells with or without inclusion of Src in the pipette solu-
tion (Vh = -50 mV). (d) The potentiation of whole-cell NR1-1a/NR2A-receptor currents by EDTA in HEK cells is reduced by inclusion of Src
in the patch pipette (*p<0.01). Circles are control cells; squares are cells perfused with Src.
glycine was potentiated by 53 9 % (n = 10), whereas the peak
of currents evoked by 300 M glutamate and 100 M glycine
was potentiated by 53 10% (n = 4). Our data agree with the
40% potentiation of recombinant NR1-3a/NR2A receptor by
Src reported previously 5. As others5 and we have used satu-
rating concentrations of glutamate and glycine, it is unlikely
that the potentiation of NMDA receptors by Src is due to a
change in apparent affinity for glutamate or glycine.
If our hypothesis is correct, complete removal of the tonic inhi-
bition should prevent the potentiation of NR1/NR2A receptor cur-
rents by Src. To remove transition-metal contamination, we added
a trace amount of EDTA (10 M) to all recording solutions. Assum-
ing 100 nM contaminant zinc and 1 mM extracellular Ca2+, the free
zinc is less than 0.03 nM, whereas the free Ca2+ remains at 0.99 mM.
In the presence of EDTA, Src failed to potentiate NR1/NR2A recep-
tors (Fig. 1a and b). In the presence of 10 mM tricine, another metal
chelator, Src also failed to potentiate NR1/NR2A receptors (n = 10,
data not shown). Thus, removal of tonic inhibition of NR1/NR2A
receptors by transition metals occludes potentiation by Src.
We used the ratio of the peak NMDA-receptor current in the
presence of EDTA (IEDTA) to the peak current in the absence of
EDTA (INo EDTA) as a measure of the amount of tonic inhibition
of NR1/NR2A by transition metals. A smaller ratio indicates less
tonic inhibition. Without Src added to the patch pipettes,
IEDTA/INo EDTA ranged from 1.6 to 2.5 (mean, 2.2 0.1, n = 6,
Fig. 1c and d). With Src added to the patch pipettes, IEDTA/INo
EDTA ranged from 1.2 to 1.7 (mean = 1.4 0.1, n = 7). Thus, Src
significantly reduced the tonic inhibition of NR1/NR2A recep-
tors by transition metals (p<0.005). Inclusion of the phosphatase
inhibitor orthovanadate (500 M) with Src results in an
IEDTA/INo EDTA value of 1.18 0.03 (n = 3), which is not signif-
icantly different from the value with Src alone.
SRC SHIFTS IC50 FOR THE HIGH-AFFINITY, ZINC SITE
Zinc is probably the transition metal responsible for the EDTA-
sensitive tonic inhibition of NR1/NR2A receptors, although
186
No Src
% of initial current
Src
Src and EDTA
Time (min)
Src
d
IEDTA/INO EDTA
EDTA EDTA
Control Src
there are several other candidate ions, such as cadmium or cop-
per18. To directly assess the effect of Src on zinc-induced inhi-
bition of NR1/NR2A-receptor currents, we used
tricine-buffered zinc to determine the IC50 value of zinc act-
ing at its high-affinity, voltage-independent site on NR1/NR2A
receptors 18 . In Xenopus oocytes expressing NR1-1a/NR2A
receptors, tricine-buffered zinc inhibited NMDA-receptor cur-
rents with an IC50 value of 15 nM for the high-affinity zinc site
(data not shown), in agreement with a previous report18. In
HEK293 cells, low concentrations of zinc caused a similar con-
centration-dependent inhibition (Fig. 2a). With zinc concen-
trations up to 2.2 M, the NR1-1a/NR2A-receptor current still
exhibits a linear IV relationship (Fig. 2b), indicating that up
to this concentration, zinc acts exclusively on the voltage-inde-
pendent, high-affinity site (n = 3). The IC50 for the high-affin-
ity zinc site of NR1-1a/NR2A in HEK cells is 86 nM (Fig. 2c),
which is fourfold higher than the IC50 obtained in oocytes18.
We are uncertain about the cause of this fourfold difference.
Based on the fitted doseresponse curves and I EDTA/INo EDTA
values (2.86 for oocytes, n = 10; 2.13 for HEK cells, n = 6), we
estimate the zinc contamination to be 275 nM for both record-
ing systems, which is in line with a previously reported value
(300 nM)23 and our own measurements (338 nM by VG Plas-
ma Quad 3 ICP-MS mass spectrometer, 330 nM by Jarrell-
Ash965 ICP plasma emission spectrometer). Addition of Src
increased the IC50 for NR1-1a/NR2A receptors in HEK cells to
392 nM, approximately fivefold (Fig. 2c). Based on this IC50
value and the estimated zinc contamination level (275 nM),
we predict that the potentiation caused by EDTA in the pres-
ence of Src should be 42%, which is very close to the actual
potentiation observed experimentally (44 9%).
SRC POTENTIATION CORRELATES WITH ZINC INHIBITION
If Src potentiates NMDA receptors by reducing tonic inhibi-
tion by zinc, NMDA receptors that are less sensitive to zinc
should also be potentiated to a lesser degree by Src. Recent
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com

articles

NR1-1a/NR2A, Fig. 2. Src shifts IC50 for the

studies have identified some splice vari-
ants and mutant NR1 receptors with
reduced zinc sensitivity21. These recom-
a HEK293 b voltage-independent zinc site
of NR1/NR2A receptors. (a)
Typical current traces showing

binant receptors were tested in HEK cells
to determine the amount of tonic inhi-
bition and Src potentiation for each. The
IC50 value for NR1-1a(C744A,C798A)24,
co-expressed with NR2A, is 168 nM in
oocytes, about eightfold higher than the
wild-type NR1-1a/NR2A receptor (Fig.
3a). In HEK cells, this mutant is only
slightly inhibited by ambient zinc, as
indicated by the IEDTA/INo EDTA ratio of
1.11 0.03 (Fig. 3b). As predicted, NR1-
1a(C744A,C798A) is not potentiated by
Src in HEK cells (Fig. 3c). NR1-1b,
which has a 21-amino-acid insertion in
the N-terminal region encoded by alter-
1998 Nature America Inc. http://neurosci.nature.com
I (pA)
dose-dependent inhibition of
NR1-1a/NR2A receptors by
zinc in HEK 293 cells (Vh, -50
mV). Tricine was used to buffer
zinc. (b) The IV relationship
[Zn2+] (nM)
V (mV) of the peak NMDA-receptor
currents shows no voltage-
c dependent inhibition at up to 2
M buffered zinc. P 2230 nM,
L 223 nM, G 0 nM. (c) Src
shifts the IC50 for the high-
I/Imax
affinity, voltage-independent
zinc site of NR1-1a/NR2A
receptors. Tricine was used to
buffer zinc. Each point on the

natively spliced exon 5, also shows curve is averaged data from

reduced zinc sensitivity, with an IC 50
value of 198 nM for heteromeric recep-
tors containing NR2A in oocytes (see
also refs. 18 and 21). As expected,
recombinant NR1-1b/NR2A receptors,
with an IEDTA/INo EDTA ratio of 1.08 0.04 in HEK cells, were
not potentiated by Src (Fig. 3b and c).
With the NR1-1b and the double cysteine mutant of NR1-
1a, we demonstrated that reduction of tonic inhibition indeed
reduces the Src potentiation of NMDA receptors. By the same
reasoning, if a mutation within NR1 exon 5 could restore zinc
sensitivity of the heteromeric receptors to levels comparable
to wild-type receptors lacking exon 5, it should also restore the
potentiation by Src. Several point mutations within exon 5 are
reported to restore zinc sensitivity to various degrees21. NR1-
1bm207-211, a triple point mutant (K207G, R208G, K211G)
in the exon 5 region13, exhibits the highest level of zinc sensi-
tivity among all mutant receptors tested, with an IC50 value of
40 nM in oocytes (Fig. 3a). Recombinant NR1-1bm207-
211/NR2A receptors are under a significant amount of tonic
49 cells. The IC50 is 86 nM
without Src and 392 nM with
[Zn2+] (nM) Src. P control, L Src.
inhibition as indicated by an IEDTA/INo EDTA ratio of 1.77 0.09
(Fig. 3b). Again, consistent with our hypothesis, the peak cur-
rent of NR1-1bm207-211/NR2A receptors is potentiated by Src
(Fig. 3c). Overall, the amount of tonic inhibition of the NMDA
receptors (I EDTA /I No EDTA ) is highly correlated (R = 0.997,
p<0.005) with the amount of potentiation by Src of NMDA
receptors (Fig. 3d).
C-TERMINAL TYROSINES ARE CRITICAL FOR SRCS ACTION
To locate the tyrosine residues critical for modulation of zinc
sensitivity by Src, we made a series of point mutations in the C-
terminal region of NR2A (Fig. 4a). The amino-acid sequence of
the C-terminal region contains a consensus phosphorylation
site25 for Src around tyrosine residue 1267 (Fig. 4a). Thus, a
point mutation of Y1267 to phenylalanine was constructed.

Fig. 3. Potentiation of NMDA-receptor currents by Src corre- a b

lates with the amount of tonic inhibition of NR1/NR2A recep-
IEDTA/INO EDTA

tors by zinc. (a) Doseresponse curves for the high-affinity zinc

I/Imax

inhibition of heteromeric NR1/NR2A receptors in oocytes

(NR1-1a, n = 14; NR1-1b, n = 12; NR1-1a(C744A,C798A), n =
8; NR1-1bm207-211, n = 19). As previously reported21, a splice
variant with exon 5 (such as NR1-1b) exhibits an IC50 nearly
tenfold higher than a splice variant without exon 5 (such as
NR1-1a). Mutation of two cysteine residues of NR1-
1a(C744A,C798A)21 also shifts the IC50 for zinc by approxi- [Zn2+] (nM)
mately tenfold. A triple mutation in exon 5 (NR1-1bm207-211: d
K207G, R208G, K211G)21 nullifies the effects of exon 5 on zinc c
inhibition. L 1bm209211, g 1b, p 1a, H 1aC744,798A
Src potentiation
% of initial current

(b) The ratio of the current in the presence of 10 M EDTA to

that in the absence of EDTA is a measurement of tonic inhibi-
(%)

tion of NMDA-receptor currents by trace amounts of zinc in

HEK 293 cells. The amount of tonic inhibition by zinc is consis-
tent with the IC50 values obtained in oocytes. (c) Effects of Src
on current amplitudes of NR1/NR2A in HEK293 cells. Note
that NR1-1a and NR1-1bm207-211 are potentiated by Src, Time (min) IEDTA/INO EDTA
whereas NR1-1b and NR1-1aC744,798A are not. p 1a/2A
(10), P 1aC744,798A/2A (5), l 1b/2A (9), L 1bm207211/2A (6). (d) Correlation between the tonic inhibition by zinc and potentiation by Src
in HEK cells measured six minutes after obtaining the whole-cell configuration (correlation factor, R = 0.997, p<0.005).

nature neuroscience volume 1 no 3 july 1998 187

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 4. Point mutations of tyrosine

residues in NR2A C-terminal block Srcs
a d
potentiation of NR1/NR2A currents and Y1267F
relief of zinc inhibition. (a) A diagram
showing mutations of C-terminal tyrosine
residues of NR2A subunit. Note that
Consensus sequence for Src
Y1267 is part of a putative consensus
phosphorylation site for Src. Other tyro-
sine residues with high surface probabil- b
ity 37 were also mutated as indicated. Y1105F

% of initial
(b) Summary bar graph showing the

current
effects of nine point mutations of tyrosine
residues to phenylalanine on the Srcs
potentiation of whole-cell NR1/NR2A-
receptor currents measured six minutes
after obtaining the whole-cell configura- c
tion (*p<0.01 when compared to the wild-
type NR2A by ANOVA and Y1387F
% of initial
current

Newman-Keuls post hoc test). For all pan-

1998 Nature America Inc. http://neurosci.nature.com

els, number of cells is shown in parenthe-

ses. (c) Point mutation of Y1105F, Y1267F
or Y1387F blocks Srcs potentiation of
whole-cell NR1/NR2A-receptor currents.
p wt (10), L Y1105F (5), G Y1267F (8), Time (min) [Zn2+](nM)
H Y1387F (7). (d) Zinc sensitivity of het-
eromeric receptors containing three mutant NR2A subunits. Each data point is the average from 47 cells. Solid lines are fitted zinc curves
for wild-type NR1/NR2A receptor without Src; dashed lines are the curves with Src. Note that Y1267F has lower affinity for zinc and that
Src failed to relieve the zinc inhibition further. Although Y1105F and Y1387F have normal zinc affinity, Src failed to relieve zinc inhibition. p
control, l Src.

Additional tyrosine residues with high surface probability in the EDTA AND SRC INCREASE THE CURRENT DECAY TIME CONSTANT
neighboring region were also mutated. Point mutation Y1267F Published evidence conflicts about the mechanism by which
blocked the potentiation by Src (Fig. 4b and c). In addition, Src potentiates NMDA receptors. It has been reported that Src
mutants Y1105F and Y1387F also abolished Src potentiation increases the open probability, open time, burst length, clus-
(Fig. 4b and c). In contrast, mutations of several other tyrosine ter length and supercluster length of NMDA channels 6 .
residues in the same region had no detectable effects on Src- Although the relaxation time constant of macroscopic current
induced potentiation (Fig. 4b). Consistent with our hypothesis, may be influenced by the first latency of NMDA channels26, an
Src failed to change the zinc sensitivity of receptors with any one increase in the duration of individual channel activation by Src
of the three tyrosine mutations that eliminated Src potentiation should increase the relaxation time constant. However, no
of NMDA receptor currents (Fig. 4d). The zinc-inhibition curves change of the relaxation time constant is reported when the
for receptors with two of these mutations (Y1105F and Y1387F) macroscopic current amplitude is potentiated5. In our whole-
were identical to the zinc-inhibition curve for the wild-type cell recording, the decay of NR1/NR2A current could not be
receptor. In these cases, the only effect of the tyrosine mutation fitted with a single exponential component. To circumvent the
was to prevent the Src-induced reduction in zinc sensitivity. The possibility that the slower, second component reflects slow
third mutation (Y1267F) reduced zinc sensitivity of the recep- removal of the agonists due to turbulence around the whole
tor in the absence of Src, which precluded further reduction of cell, we investigated the decay time constant on NMDA-recep-
zinc sensitivity by Src. The IC 50 for zinc of NR1-
1a/NR2A(Y1267F) in oocytes was 92 nM (n = 13, data not
shown), which is approximately fivefold higher than the IC50 Table 1. Inhibition of wild-type versus mutant NR1/NR2A
for the wild-type NR1/NR2A. The amount of maximal inhibi- receptors by tricine-buffered zinc (223 nM) in HEK293
tion was not altered (data not shown). A similar shift of IC50 for cells
zinc was also observed in HEK cells (Fig. 4d). In contrast, the
Internal Solution No Src Src
EC50 values for glutamate (n = 7) and glycine (n = 7) and the
IC 50 value for proton inhibition (n = 8) for NR1- Wild-Type NR2A 0.48 0.04 (5) 0.69 0.05 (6)*
1a/NR2A(Y1267F) were not significantly different from those Y1105F 0.54 0.05 (7) 0.50 0.03 (5)
of the wild-type receptors (data not shown). These data suggest Y1267F 0.74 0.04 (4)* 0.75 0.05 (4)*
that the Y1267F mutation results in a receptor that behaves like Y1387F 0.49 0.04 (5) 0.49 0.05 (6)
one that is fully potentiated by Src. We have also constructed Y1105F,1387F 0.44 0.09 (4) 0.45 0.07 (5)
double tyrosine mutations NR2A(Y1105F,Y1387F) and
Y1267F,1387F 0.79 0.03 (3)* 0.82 0.04 (4)*
NR2A(Y1267F,1387F). There was no detectable difference
between NR2A(Y1105F,1387F) and the corresponding single Data presented in the table are I/Imax. Imax is the current recorded under
tyrosine mutants (Y1105F and Y3187F) (Table 1). nominally zinc-free conditions (in the presence of 10 mM tricine without
added zinc). Number of observations is indicated in parentheses. *significant
NR2A(Y1267F,Y1387F) behaved like Y1267F, with a reduced difference (p<0.05) compared to wild type by ANOVA and Newman-Keuls
apparent zinc affinity and lack of modulation by Src (Table 1). post-hoc test.

188 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 5. EDTA and Src increase the time constant of the

slower component of NMDA currents. (a) Typical cur-
a b Control EDTA

rent responses from an outside-out patch expressing

NR1/NR2A receptors. A brief glutamate-concentration
jump (10 ms, to 100 M) was applied to the patches.
Saturating glycine (1030 M) was present in all solu-
tions. Traces shown are averaged from 25 consecutive
responses for control and 20 responses for EDTA. The
top traces in (a) and (b) show the glutamate application
time. (b) The decay of NMDA-receptor-mediated cur-
rents was fitted with two exponential components
c
(smooth lines). The raw current trace is scaled to the
same size to illustrate the change of relaxation time con-
stant. (c) EDTA and Src increased the time constant of

1 (ms)

2 (ms)
the slower exponential component significantly (p<0.05,
paired t-test for EDTA, unpaired t-test for Src), whereas
the time constant for the faster exponential component
is not changed. p control, l EDTA, g Src.
1998 Nature America Inc. http://neurosci.nature.com

Control EDTA Src Control EDTA Src

tor currents in outside-out patches exposed to brief pulses (10
ms) of glutamate (with submillisecond solution-exchange rates
at the open tip 27). Under such conditions, two exponential
components were still needed to fit the decay of the macro-
scopic NMDA receptor currents. The two averaged time con-
stants were 36.6 5.3 and 79.8 8.9 ms without EDTA and
Src (n = 9). Inclusion of EDTA increased the peak of macro-
scopic NR1/NR2A currents recorded from outside-out patch-
es by 63 30% (n = 9, p<0.05, paired t-test), which is
consistent with our whole-cell data as well as a previous report
that Src increases open probability6. In addition, EDTA signif-
icantly increased the slower relaxation time constant to 130
19 ms, whereas the faster relaxation time constant remained
unchanged with a mean of 38.9 6.0 (Fig. 5). Interestingly, Src
induced a similar selective change in the slower relaxation time
constant. The averaged time constants in the presence of Src
were 42.0 4.1 and 117.7 9.5 ms (n = 7). Because the
amount of calcium influx during an EPSC is determined by
the charge integral (Q = iA i i) of the NMDA current, the
resulting increase of synaptic calcium influx could be ampli-
fied by as much as 2.3-fold by Src. These observations indicate
that Src and EDTA change the NMDA-receptor channel kinet-
ics in similar ways, consistent with our hypothesis that Src
potentiates NMDA receptors by relieving tonic zinc inhibition.
SRC REDUCES ZINC SENSITIVITY OF NR1/NR2B RECEPTORS
Is the reduction of zinc sensitivity by Src unique for receptors con-
taining NR2A subunits, or is it a more generalized regulatory
mechanism for other NMDA-receptor subunits? To address this
question, we examined the effects of Src on zinc sensitivity of
recombinant NR1-1a/NR2B receptors. Zinc inhibits NR1/NR2B
receptors with an IC50 value of 4.6 M under control conditions
in HEK293 cells (Fig. 6a and b). In the presence of c-Src (30 units
per ml), the IC50 was shifted to 17.6 M. Thus, NMDA receptors
comprised of NR2A or NR2B subunits are modulated by tyrosine
kinase Src in a similar way, through a reduction of zinc sensitivity.
Discussion
Taken together, these data suggest that Src potentiates recombi-
nant NR1/NR2A-receptor function by reducing tonic inhibition
of these receptors by extracellular zinc. Furthermore, Src also
reduces zinc sensitivity of NR1/NR2B receptors. Thus, reduction of
zinc inhibition by Src could be a widely utilized mechanism for
the modulation of NMDA receptor function in the brain. Our data
are consistent with biochemical observations that both NR2A and
NR2B are phosphorylated by the Src family of tyrosine kinases8,28
because they show that receptors comprised of either the NR2A
or NR2B subunit are modulated functionally by Src. The selective
potentiation of NR1/NR2A receptors by Src in HEK293 cells is
due to the high zinc sensitivity of NR1/NR2A receptors and the
ambient zinc contamination in the recording solutions that caus-
es them to be tonically inhibited. By contrast, receptors comprised
of NR1/NR2B are not potentiated by Src in HEK cells because their
relatively lower sensitivity to zinc prevents tonic inhibition.
Zinc is actively accumulated in certain nerve terminals in a
region-specific manner and can be released into synaptic clefts at

Fig. 6. Src reduces zinc sensitivity a Control Src b

of NR1/NR2B receptors. (a) The
effects of zinc (3 and 10 M) on
NR1/NR2B-receptor currents
I/Imax

evoked by agonists (100 M gluta-

mate, 30 M glycine) from repre-
[Zn2+](M) [Zn2+](M)
sentative HEK 293 cells with or
without inclusion of Src in the
patch pipette solution (V h, -50 mV). (b) Src shifts the IC 50 for the high-affinity, voltage-
independent zinc site of NR1/NR2B receptors. Each point on the curve is averaged data
from 35 cells. The IC 50 is 4.6 without Src and 17.6 M with Src. P control, L Src. [Zn2+] (M)

nature neuroscience volume 1 no 3 july 1998 189

 1998 Nature America Inc. http://neurosci.nature.com
articles
concentrations up to 0.1 mM14. The reported ambient zinc con-
centration in the brain varies from 150 nM to more than 2 M22.
Assuming ambient zinc levels of 1 M, more than 70% of
NR1/NR2A receptors would be inhibited, compared to 25% of
NR1/NR2B receptors. The native NMDA receptors in the brain
most likely include both NR2A and NR2B subunits. If, as seems
likely, ambient zinc concentrations are sufficient (in excess of 1
M22) to tonically inhibit NMDA receptors containing NR2A, then
such receptors might serve as as marginally functional receptors
whose full function can be restored through activation of tyrosine
kinases such as Src29. If ambient zinc concentrations are insuffi-
cient to cause significant tonic inhibition, Src could still alter the
responsiveness of NMDA receptors to synaptically released zinc.
The C-terminal region of NR2A is critical for modulation by
Src. A C-terminal deletion mutant of NR2A is not potentiated
by Src in HEK293 cells. Mutant mice expressing NR2A with the
C-terminal truncation have impaired synaptic plasticity. We have
identified three tyrosine residues in the C-terminal region of
1998 Nature America Inc. http://neurosci.nature.com
NR2A that are critical for the modulation of NR1/NR2A recep-
tors by Src. The Y1105 and Y1387 residues function as predict-
ed for potential tyrosine phosphorylation sites, which is that their
mutation to phenylalanine causes no apparent changes by itself
but blocks the reduction of zinc affinity by Src. Because both
these tyrosine residues are conserved between NR2A and NR2B,
it is tempting to speculate that phosphorylation of the corre-
sponding tyrosine residues in NR2B is responsible for reducing
the apparent zinc affinity of NR1/NR2B receptors. The function
of the third tyrosine residue, Y1267, is peculiar. Mutation of
Y1267 to phenylalanine results in a receptor with reduced zinc
affinity. This tyrosine residue and the small stretch of peptide
surrounding it are unique to NR2A. We do not know the mech-
anism behind this change and can only speculate at this point.
One possibility is that this residue is phosphorylated constitu-
tively and may be involved in interactions with the SH2 domain
of Src or other proteins. The C-terminal domains of NMDA
receptor subunits are larger in size than those of other receptor
subunits and may be involved in complex interactions with other
proteins in the postsynaptic density and cytoskeleton. One exam-
ple of such an interaction is the reduction of NMDA-receptor
activity by direct binding of calmodulin to NR1 subunits, which
inactivates NMDA receptors30. Little is known about the possible
proteinprotein interactions of the C-terminal regions of NR2
subunits other than their interaction with the PSD95 family of
proteins. Although the mechanism by which mutation of Y1267
reduces zinc sensitivity of NR1/NR2A receptors needs to be
explored, our data do show that its lack of tonic inhibition due
to reduced zinc sensitivity occludes Src potentiation, supporting
our hypothesis that Src potentiates NMDA-receptor function by
reducing tonic zinc inhibition.
From a structural perspective, our results suggest that tyro-
sine phosphorylation and/or Src binding on the cytosolic C-ter-
minal domain of NR2A alter the apparent affinity of an
extracellular zinc binding site. Although structural information
will ultimately be required to understand the interaction between
Src and the zinc site, several possible explanations deserve men-
tion. Conformational constraints imposed by Src binding and/or
tyrosine phosphorylation (the two events may occur separately)
may specifically reduce the affinity of zinc by altering the geom-
etry of the coordination site. Alternatively, Src may change other
aspects of the NMDA-channel properties that perturb the bind-
ing of zinc. Several modulators or mutations21 that modify zinc
sensitivity have previously been described. For example, expres-
sion of NR1 alternative exon 5, which encodes a highly charged
190
surface loop, seems to reduce zinc sensitivity in a manner that
can be mimicked by extracellular polyamines. In addition, muta-
tions of acidic residues in the NR1 subunit that reduce proton
sensitivity similarly reduce zinc sensitivity. One possible expla-
nation for these (and by analogy Srcs) effects is that they may be
due to reduced electronegativity of the electron-donating residues
that form the zinc coordination site, which would reduce the abil-
ity of these residues to coordinate zinc. Such an effect could occur
by inductive electron withdrawal through portions of the
polypeptide chain secondary to Src binding or phosphorylation.
Alternatively, Src could change the channel properties in such a
way that occupation of the zinc site is less likely to force closure of
an open channel, requiring a greater probability of zinc occu-
pancy to inhibit the channel, which might appear as a lower IC50.
In either case, the effects of Src and tyrosine phosphorylation
need to be transmitted from intracellularly accessible residues to
extracellular residues. Clearly additional functional as well as
structural information will be required to ultimately understand
Srcs effects on zinc modulation.
Finally, by demonstrating that potentiation of the NMDA-
receptor current by Src reduces the apparent zinc affinity of
NMDA receptors, we link two modulatory sites of NMDA
receptors that have been previously thought to be indepen-
dent. Our data provide further support for the convergence
of different NMDA receptor modulatory systems that has
become evident recently13,21,24,31.
Methods
SITE-DIRECTED MUTAGENESIS. All of the cDNAs used in this study except a
mutant of NR1-1b and NR2B (pCDNAI/amp, Invitrogen) were sub-
cloned into pCIneo vector (Promega). Site-directed mutagenesis was
done with Pfu DNA polymerase (Stratagene) to linearly replicate the
parental strand with desired mismatch incorporated into the primer.
Methylated parental DNA template was then degraded with Dpn I. The
nicked double-stranded mutant DNA was transformed into E. coli. The
nicks in the plasmid were repaired after transformation. Colonies were
selected by screening for a silent mutation that introduces a new restric-
tion site. The mutations were verified by didioxy sequencing through
both strands in the region of the mutation.
TRANSFECTION OF HEK CELLS. HEK293 cells (CRL 1573; American Type
Culture Collection, Rockville, Maryland) were maintained at 37C and
5% CO2 in DMEM supplemented with L-glutamine (200 M), sodium
pyruvate (100 M), penicillin/streptomycin (100 units per ml), and 10%
fetal bovine serum. Low-confluence cells were transfected by calcium-
phosphate precipitation32. Cells were cotransfected with a mixture con-
taining NR1, NR2 and green fluorescent protein33 plasmids (1, 2, and
0.3 g per 12 mm diameter coverslip, respectively). After transfection,
200 M D-AP5 was added to the culture medium.
EXPRESSION OF NMDA RECEPTORS IN XENOPUS OOCYTES. cRNA was synthe-
sized from linearized template cDNA according to manufacturers spec-
ification (Ambion). The quality of synthesized cRNA was assessed by gel
electrophoresis and quantified by spectroscopy and gel electrophoresis.
Preparation of oocytes and injection of cDNAs coding for wild-type and
mutant NMDA receptors were performed as described21.
BUFFERED ZINC SOLUTIONS. The tricine-buffered zinc solutions used to
obtain the zinc doseresponse curves were prepared according to the
empirically established binding constant18 105 M, by adding into 10 mM
tricine (pKa, 8.15) the following concentrations of zinc (in M): 0.26,
0.78, 2.6, 7.8, 26, 77.5, and 254. The corresponding estimated concen-
trations of free zinc for HEK recording at pH 7.4 were calculated with
WINMAXC34 and BAD35, and were (in nM) 2.3, 6.88, 22.3, 68.8, 223,
688, 2230, respectively.
VOLTAGE-CLAMP RECORDINGS FROM XENOPUS OOCYTES. Two-electrode volt-
age-clamp recordings were made as described21. Briefly, oocytes were
perfused with a solution comprised of (in mM) 90 NaCl, 1 KCl, 10
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com
HEPES and 0.5 BaCl2 at pH7.4, and held under voltage clamp at -30 to -
40 mV. Electrodes were filled with 300 mM KCl and had resistance of
110 M. Solution exchange was computer controlled through an eight-
valve manifold.
WHOLE-CELL AND OUTSIDE-OUT RECORDINGS FROM HEK293 CELLS. Patch-
clamp recording in the whole-cell configuration and outside-out patch
recording were done as described36 with Axopatch 200A or 200B ampli-
fier. Recording electrodes (512 M) were filled with (in mM) 140 Cs-
gluconate, 5 HEPES, 4 NaCl, 2 MgCl2, 0.5 CaCl2, 1 ATP, 0.3 GTP and 5
BAPTA (pH 7.4). The recording chamber was continually perfused with
recording solution composed of (in mM) 150 NaCl, 10 HEPES, 1.0 CaCl2,
3 KCl and 20 mM mannitol. Glutamate (20100 M) and glycine (1030
M) were applied using a multibarrel pipette driven by a piezo-electric
bimorph (exchange time under 0.5 ms) as described27. In some experi-
ments, recombinant human c-Src (Upstate Biotech.) was added into
internal solution to a final concentration of 530 units per ml.
DATA ANALYSIS AND STATISTICS. All pooled data were expressed as mean
standard error. Statistical comparisons were done with unpaired Stu-
dents t-test unless otherwise stated. To obtain IC50 values, pooled zinc
1998 Nature America Inc. http://neurosci.nature.com
inhibition data were fitted with least-squares criterion (SigmaPlot) to
equation I/Imax = (1-a)/(1+([Zn2+]/IC50)n)+a where n is the Hill slope
and a is the residual response. (n and a were not constrained.) If the fit-
ting algorithm returned a value less than 0.05 for a (our estimated limit
for detection), we fixed a to 0 and refit the zinc inhibition data to equa-
tion I/Imax = 1/(1+([Zn2+]/IC50)n) (as in Fig. 6).
Acknowledgements
We thank S.F.Heinemann for NR1 and NR2B cDNAs, N. Nakanishi for NR2A
cDNA and M. Chalfie, M. Mayer and P. Seeburg for green fluorescent protein
plasmid. We also thank P. Ascher, J. Neyton and I. Mody for critically reading
the manuscript.
RECEIVED 11 MARCH: ACCEPTED 22 MAY 1998
1. ODell, T.J., Kandel, E.R. & Grant, S.G.N. Long-term potentiation in the
hippocampus is blocked by tyrosine kinase inhibitors. Nature 353, 558560
(1991).
2. Grant, S.G.N. et al. Impaired long-term potentiation, spatial learning, and
hippocampal development in fyn mutant mice. Science 258, 19031910 (1992).
3. Lu, Y.M., Roder, J.C., Davidow, J. & Salter, M.W. Src activation in the induction
of long-term potentiation in CA1 hippocampal neurons. Science 279,
13631368 (1998).
4. Wang, Y.T. & Salter, M.W. Regulation of NMDA receptors by tyrosine kinases
and phosphatases. Nature 369, 233235 (1994).
5. Khr, G. & Seeburg, P.H. Subtype-specific regulation of recombinant NMDA
receptor-channels by protein tyrosine kinases of the Src family. J. Physiol.
(Lond.) 492, 445452 (1996).
6. Yu, X.M., Askalan, R., Keil, G.J. & Salter, M.W. NMDA channel regulation by
channel-associated protein tyrosine kinase Src. Science 275, 674678 (1997).
7. Sprengel, R. et al. Importance of the intracellular domain of NR2 subunits for
NMDA receptor function in vivo. Cell 92, 279289 (1998).
8. Miyakawa, T. et al. Fyn-kinase as a determinant of ethanol sensitivity:
Relation to NMDA-receptor function. Science 278, 698701 (1997).
9. Nowak, L., Bregestovski, P., Ascher, P., Herbert, A. & Prochiantz, A.
Magnesium gates glutamate-activated channels in mouse central neurones.
Nature 307, 462465 (1984).
10. Mayer, M.L., Westbrook, G.L. & Guthrie, P.B. Voltage-dependent block by Mg2+
of NMDA responses in spinal cord neurones. Nature 309, 261263 (1984).
11. Westbrook, G.L. & Mayer, M.L. Micromolar concentrations of Zn2+
antagonize NMDA and GABA responses of hippocampal neurons. Nature
328, 640643 (1987).
nature neuroscience volume 1 no 3 july 1998
articles
12. Peters, S., Koh, J. & Choi, D.W. Zinc selectively blocks the action of N-
methyl-D-aspartate on cortical neurons. Science 236, 589593 (1987).
13. Traynelis, S.T., Hartley, M. & Heinemann, S.F. Control of proton sensitivity
of the NMDA receptor by RNA splicing and polyamines. Science 268,
873876 (1995).
14. Smart, T.G., Xie, X. & Krishek, B.J. Modulation of inhibitory and excitatory
amino acid receptor ion channels by zinc. Prog. Neurobiol. 42, 393441
(1994).
15. Mayer, M.L., Vyklicky, L. Jr & Westbrook, G.L. Modulation of excitatory
amino acid receptors by group IIB metal cations in cultured mouse
hippocampal neurones. J. Physiol. (Lond.) 415, 329350 (1989).
16. Christine, C.W. & Choi, D.W. Effect of zinc on NMDA receptor-mediated
channel currents in cortical neurons. J. Neurosci. 10, 108116 (1990).
17. Legendre, P. & Westbrook, G.L. The inhibition of single N-methyl-D-
aspartate-activated channels by zinc ions on cultured rat neurones. J. Physiol.
(Lond.) 429, 429449 (1990).
18. Paoletti, P., Ascher, P. & Neyton, J. High-affinity zinc inhibition of NMDA
NR1-NR2A receptors. J. Neurosci. 17, 57115725 (1997).
19. Williams, K. Separating dual effects of zinc at recombinant N-methyl-D-
aspartate receptors. Neurosci. Lett. 215, 912 (1996).
20. Chen, N., Moshaver, A. & Raymond, L.A. Differential sensitivity of
recombinant N-methyl-D-aspartate receptor subtypes to zinc inhibition.
Mol. Pharmacol. 51, 10151023 (1997).
21. Traynelis, S.F., Burgess, M.F., Zheng, F., Lyuboslavsky, P. & Powers, J.L.
Control of voltage-independent zinc inhibition of NMDA receptors by the
NR1 subunit. J. Neurosci. (in press).
22. Bogden, J.D., Troiano, R.A. & Joselow, M.M. Copper, zinc, magnesium, and
calcium in plasma and cerebrospinal fluid of patients with neurological
diseases. Clin. Chem. 23, 485489 (1977).
23. Li, C., Peoples, R.W. & Weight, F.F. Proton potentiation of ATP-gated ion
channel responses to ATP and Zn2+ in rat nodose ganglion neurons. J.
Neurophysiol. 76, 30483058 (1996).
24. Sullivan, J.M. et al. Identification of two cysteine residues that are required
for redox modulation of the NMDA subtype of glutamate receptor. Neuron
13, 929936 (1994).
25. Kemp, B.E. & Pearson, R.B. Protein kinase recognition sequence motifs.
Trends Biochem. Sci. 15, 342346 (1990).
26. Edmonds, B. & Colquhoun, D. Rapid decay of averaged single-channel
NMDA receptor activations recorded at low agonist concentration. Proc. R.
Soc. Lond. B 250, 279286 (1992).
27. Traynelis, S.T. & Wahl, P. Control of rat GluR6 glutamate receptor open
probability by protein kinase A and calcineurin. J. Physiol. 503, 513531
(1997).
28. Lau, L.-F. & Huganir, R.L. Differential tyrosine phosphorylation of N-
methyl-D-aspartate receptor subunits. J. Biol. Chem. 270, 2003620041
(1995).
29. Dikic, I., Tokiwa, G., Lev, S., Courtneidge, S.A. & Schlessinger, J. A role for
Pyk2 and Src in linking G-protein-coupled receptors with MAP kinase
activation. Nature 383, 547550 (1996).
30. Ehlers, M.D., Zhang, S., Bernhardt, J.P. & Huganir, R.L. Inactivation of
NMDA receptors by direct interaction of calmodulin with the NR1 subunit.
Cell 84,745755 (1996).
31. Chen, L. & Huang, L.-Y.M. Protein kinase C reduces Mg2+ block of NMDA-
receptor channels as a mechanism of modulation. Nature 356, 521523
(1992).
32. Chen, C. & Okayama, H. High-efficiency transformation of mammalian cells
by plasmid DNA. Mol. Cell. Biol. 7, 27452752 (1987).
33. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W.W. & Prasher, D.C. Green
fluorescent protein as a marker for gene expression. Science 263, 802805
(1994).
34. Bers, D., Patton, C. & Nuccitelli, R. A practical guide to the preparation of Ca
buffers. Methods Cell Biol. 40, 329 (1994).
35. Brooks, S.P.J. & Storey, K.B. Bound and determined: a computer program for
making buffers of defined ion concentrations. Anal. Biochem. 201, 119126
(1992).
36. Hamill, O., Marty, A., Neher, E., Sakmann, B. & Sigworth, F. Improved patch-
clamp techniques for high-resolution current recording from cells and cell-
free membrane patches. Pflgers Arch. 391, 85100 (1981).
37. Emini, E.A., Hughes, J.V., Perlow, D.S. & Boger, J. Induction of hepatitis A
virus-neutralizing antibody by a virus-specific synthetic peptide. J. Virol. 55,
836839 (1985).
191
 1998 Nature America Inc. http://neurosci.nature.com

articles

Multiple kinetic components of

exocytosis distinguished by
neurotoxin sensitivity
Tao Xu1, Thomas Binz2, Heiner Niemann2 and Erwin Neher1

1 Department of Membrane Biophysics, Max-Planck-Institute for Biophysical Chemistry, Am Fassberg 11, D-37077, Gttingen, Germany
2 Department of Biochemistry, Medizinische Hochschule, D-30625, Hannover, Germany
Correspondence should be addressed to T.X. (txu@gwdg.de)

The secretion of synaptic and other vesicles is a complex process involving multiple steps. Many mol-
1998 Nature America Inc. http://neurosci.nature.com

ecular components of the secretory apparatus have been identified, but how they relate to the
different stages of vesicle release is not clear. We examined this issue in adrenal chromaffin cells,
where capacitance measurements and amperometry allow us to measure vesicle fusion and
hormone release simultaneously. Using flash photolysis of caged intracellular calcium to induce exo-
cytosis, we observed three distinct kinetic components to vesicle fusion, of which only two are
related to catecholamine release. Intracellular dialysis with botulinum neurotoxin E, D or C1 or
tetanus-toxin light chains abolishes the catecholamine-related components, but leaves the third
component untouched. Botulinum neurotoxin A, which removes nine amino acids from the
carboxy(C)-terminal end of SNAP-25, does not eliminate catecholamine release completely, but
slows down both catecholamine-related components. Thus we assign a dual role to SNAP-25 and
suggest that its nine C-terminal amino acids are directly involved in coupling the calcium sensor to
the final step in exocytosis.

Functional studies on secretory cells have supplied ample evi-
dence that synaptic and other secretory vesicles can exist in
distinct functional states14. Typically, only a fraction of all
vesicles of a secretory cell can be released in a certain time
window, independent of the strength of a given stimulus.
These vesicles in a specific release-ready state are termed a
pool. The question arises whether pools defined in this or
other ways can be associated with well defined macromolecu-
lar complexes of synaptic proteins, which have recently been
characterized biochemically57. Some of the proteins of these
complexes, such as SNAP-25, synaptobrevin and syntaxin, are
the targets of very specific proteolytic neurotoxins, which
cleave the proteins at unique sites810. These proteins are high-
ly susceptible to toxin attack in monomeric form, whereas they
are protected to various degrees in macromolecular complex-
es11. This property and differences between the toxins actions
may allow us to link certain functional pools to some of the
biochemically defined complexes by their toxin susceptibili-
ty. Indeed, different kinetic components are impaired differ-
entially by toxins12,13. Also, botulinum neurotoxin type A may
be special among the clostridial neurotoxins, as its action in
some preparations can be overcome by strong stimuli14,15.
For a precise dissociation of functional steps and their cor-
relation with molecular data, it would be desirable to study
exocytosis at fast resolution, as the final steps in the neurose-
cretory pathway are known to be fast, and most likely neuro-
toxins act at later steps. We therefore studied catecholamine
secretion from bovine chromaffin cells using capacitance mea-
surement as a fast assay of exocytosis combined with flash
photolysis of caged calcium as a fast and strong stimulus. We
characterized two kinetic components that were differentially
affected by clostridial neurotoxins. We compared the capaci-
tance signal with simultaneously measured catecholamine
release, and thereby identified a third very prominent kinet-
ic component, which apparently was not related to cate-
cholamine release, and which was not affected by toxins.
Together, these data suggest possible relationships between
late steps in the release process and their biochemically defined
molecular counterparts.
Results
MULTIPLE COMPONENTS OF EXOCYTOSIS IN CHROMAFFIN CELLS
To study the kinetic components of secretion, we employed
fast-resolution capacitance measurements to estimate the secre-
tory response after spatially homogeneous elevation of inter-
nal calcium concentration ([Ca2+]i) by photorelease of caged
calcium, nitrophenyl-EGTA. Membrane capacitance (Cm) is
proportional to the surface area of the cell, and it increases
when secretory vesicles fuse with the plasma membrane 16,17.
In most of our measurements, we used the following protocol
(Fig. 1a). After ten minutes of whole-cell dialysis with either
toxin-free or toxin-containing pipette solutions, we first gave
low-intensity flashes (usually two) to generate small [Ca2+]i
jumps to about 20 M. This readily elicited secretory respons-
es (Fig. 1b). When there was a clear indication that secretion
was depressed, we then gave strong flashes to elevate [Ca 2+]i
to higher values (over 100 M). Intervals between the flashes
were 120 seconds. In response to the first flash, there was always
a robust Cm increase with two clearly distinct phases, which
we call the exocytic burst 18 and the slow component. Our

192 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 1. Multiple secretory

i(M)
a A 100

[Ca2+]]i(M)
components in response to
2+ 50

2+
different [Ca ]i levels.

Cm(pF) [Ca
0
(a) Overview of the experi-

Gs(nS) Cm(pF)
ment. Sample responses
from one cell are illustrated. 400

Gm(nS) Gs(nS)
The individual traces are 200
intracellular free calcium 0
1.0
concentration ([Ca2+]i) as

Imon(pA) Gm(nS)
0.5
measured by furaptra, mem-
0.0
brane capacitance (Cm),
Imon(pA)
0

series conductance of the -20

-40
equivalent circuit (Gs),
membrane conductance 0 200 400 600 800 1000
Time
Time(s)
(s)
(Gm) and membrane current
(Imon). Arrows indicate b
Cm(pF)

time points, where flashes

are given. They are followed
by gaps in the record, during
[Ca2+]i(M) IAMP(pA)
1998 Nature America Inc. http://neurosci.nature.com

which data sampling

switched to a high-resolu-
tion mode. During ten min-
utes of whole-cell dialysis,
no loading transient was
seen in NP-EGTA-containing
internal solution. The Cm
Time (s) Time (s) Time (s) Time (s)
traces remained flat until the
onset of flashes (arrows).
The flashes were given every 120 s. (b) Fast-resolution recordings of Cm, amperometric current and [Ca 2+]i in response to individual
flashes in (a) are displayed. The flashes were triggered at 200 ms in each trace.

working hypothesis is that the exocytic burst represents those
granules that are in a release-ready state and require only ele-
vation of [Ca2+]i for exocytosis. Subsequently, granules have
to undergo slower steps of recruitment to the release-ready
pool before they can exocytose. Therefore, further capacitance
increase occurs at a much slower rate (see Discussion for details
and a model). In 84 experiments, exocytic bursts were evoked
at [Ca2+]i averaging 29.3 18.9 M (mean SD). The time
course could be fitted by double exponentials with average time
constants of 62.6 5.9 ms (n = 44 of 84) and 295.6 23.1 ms
(n = 76 of 84), and average amplitudes of 368 40 fF and 302
27 fF, respectively. This is similar to previously reported val-
ues19 except for somewhat slower kinetics. The slow compo-
nent showed a time constant in the range of ten seconds, but
this is probably an underestimate, as capacitance at late times
may be reduced by endocytosis and may be curtailed by a slow
decline in [Ca2+]i. The amperometric current, which remained
on a plateau level for up to ten seconds in some cases, also sug-
gests that the time constant of this slow component may be
over ten seconds. The increase in Cm always was accompanied
by two phases of catecholamine release as monitored by car-
bon fibers (Fig. 2a). It should be noted that the faster compo-
nent of the exocytic burst was also accompanied by
catecholamine release (Fig. 2a, inset).
Following flashes, [Ca2+]i dropped back to basal values at
times longer than ten seconds. Second flashes in a given exper-
iment usually elicited very small Cm increases and only rarely
were accompanied by amperometric spikes, which might sug-
gest the depletion of a vesicle pool during first flashes. How-
ever, when [Ca2+]i was raised above 100 M in third flashes,
again an intense Cm increase was observed (Fig. 2b). This low-
calcium-affinity Cm component increased with two expo-
nentials. One had an amplitude of 404 76 fF and a time con-
stant of 1.1 0.2 s. The second exponential had a time con-
stant of 6.4 0.9 s and 361 62 fF in amplitude (n = 12). This
large membrane increase, which constituted roughly 10% of
the total cell membrane, was not associated with cate-
cholamine release (see Figs. 1b and 6). Also, the time constant
did not change with [Ca2+]i (Fig. 2b). It rather seemed that
Cm increases of this kind were elicited whenever [Ca 2+ ] i
exceeded a certain threshold.
ATP DEPENDENCE OF CALCIUM -TRIGGERED EXOCYTOSIS
To test the ATP dependence of different exocytic components,
we dialyzed the cytosol of chromaffin cells with 80% calcium-
loaded NP-EGTA internal solution containing
nonhydrolyzable ATP analogs, either adenosine 5-[,-meth-
ylene]triphosphate (AMP-PCP) or ,-Imidoadenosine 5-
triphosphate (AMP-PNP), via the patch pipette. While the
basal [Ca2+]i was kept at 100300 nM, too low to support exo-
cytosis, ATP was washed out of the cell and replaced by the
analogs during the first five minutes of perfusion. Subse-
quently no responses could be elicited by flash photolysis, nei-
ther when secretion was measured by Cm nor by amperometry
(Fig. 3a; n = 28 cells in paired experiments). These data com-
plement the finding that the release capability of permeabi-
lized cells can be increased (primed) by preincubation with
ATP and diminished (deprimed) by preincubation in the
absence of ATP1. Our results differ from a report that MgATP-
independent exocytosis remained undiminished for even six
minutes2, although that experiment was done at a lower tem-
perature than ours. The authors suggested that this form of
depriming may require magnesium, which was absent in their
experiments. We therefore performed additional experiments

nature neuroscience volume 1 no 3 july 1998 193

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 2. Kinetic analysis of multiple

using an internal solution without a Cm components. (a) An example
magnesium and DM-nitrophen to
of secretory kinetics at low
chelate magnesium. The free con-

Cm (pF)
[Ca2+]i (27 M). The Cm trace is
centration of magnesium in this reasonably fitted by two expo-
solution was estimated to be less nentials (superimposed thick
than 10 nM. Again no secretion was dashed line) with time constants
observed in the presence of 2 mM of 90 ms and 6.29 s. The corre-
AMP-PCP (data not shown). Thus sponding amperometric current
we conclude that magnesium does also shows a double exponential

not matter for this effect.
We also used AMP-PNP to
replace ATP. The effect of AMP-PNP
was weaker than that of AMP-PCP
(Fig. 3b). In some cells, AMP-PNP
only partially blocked exocytosis,
whereas in other cells AMP-PNP
showed strong inhibition of exocy-
tosis. However, AMP-PNP com-
1998 Nature America Inc. http://neurosci.nature.com
IAMP (pA)
decay. The onset of secretion
from 0.2 to 1 s (boxed area) is
expanded in the inset. (b) Kinetic
analysis of Cm response to high
[Ca2+]i. One example of Cm
Time (s) increase at [Ca2+]i of 287 M is
shown in the left panel.
b The Cm trace is reason-
ably fitted by two expo-

pletely blocked the secretion elicited nentials (superimposed

Cm (pF)

by a second flash, suggesting that thick dashed line) with

fast (s)
ATP is necessary for the refilling of time constants of 0.82 s
and 4.15 s. The right panel
vesicles (Fig. 3c).
displays a summary of the
Our results confirm the impor-
faster time constants of
tance of ATP in secretion. Although the Cm increases versus
replacement of ATP by the nonhy- the [Ca2+]i levels from 58
drolyzable analogs AMP-PCP and [Ca2+]i (M) experiments. The dashed
Time (s)
AMP-PNP completely blocked both line is the mean value.
the exocytic burst and the slow exocy-
tosis within five minutes, it is unlikely
that ATP hydrolysis is involved in the
exocytic burst for the following reasons. Experiments using the Complexes of synaptobrevin, syntaxin and SNAP-25
caged calcium DM-Nitrophen have shown that exocytosis can (termed SNARE complexes) are found in synaptic vesicle
readily be induced earlier in a whole-cell recording in the com- membrane in vitro 24,25 , and such ternary complexes are
plete absence of free magnesium and in a nominally ATP-free reversibly disassembled by treating synaptic vesicles with ATP-
solution2,4,18. This conclusion is in line with studies on perme- NSF and SNAPs25. It is therefore possible that, in living cells,
abilized cells2021 and with recent investigations showing that most of the SNARE complexes are continuously formed on
the action of N-ethylmaleimide fusion protein (NSF) is required the vesicle and continuously disassembled because of the activ-
before contact of vesicle and target membrane22,23. ity of ATP-NSF, which might be the last step of ATP hydroly-

a b c Fig. 3. ATP dependence of

secretion. (a) The effect of 56-
[Ca2+]i (M)
[Ca2+]i (M)

[Ca2+]i (M)

minute dialysis of AMP-PCP on

secretion. Averaged [Ca2+]i lev-
els, capacitance traces and
amperometric currents in
response to the first flashes
from paired experiments in the
Control Control presence of MgATP as control
Control
(solid lines; n = 11) and in the
Cm

Cm
Cm

AMP-PNP (2 mM)
presence of 2 mM AMP-PCP
substituting for MgATP (dashed
AMP-PCP (2 mM)
AMP-PNP (2 mM) lines; n = 17) are displayed.
Flashes were triggered at 200
ms in the graph. (b) The effect
of 56 minute dialysis of AMP-
PNP on secretion. Similar to (a),
IAmp (pA)

but 2 mM AMP-PNP was used to

IAmp (pA)
IAmp (pA)

substitute for MgATP (n = 12).

Control responses from paired
experiments are from 11 cells.
(c) Similar to (b) but for the
successive second flashes. Cells
Time (s) Time (s) Time (s) are the same as those in (b).

194 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com
a b
[Ca2+]i (M)
[Ca2+]i (M)
Control
Cm
Cm
BoNT/E
IAmp (pA)
IAmp (pA)
1998 Nature America Inc. http://neurosci.nature.com
Time (s)
c release process in the capacitance trace at later
d time. (c) BoNT/A slows down the exocytic burst. The upper
(Cm/t)max (pF/s)
Cm
Control
Time (s)
sis. We have shown that replacement of ATP by AMP-PCP
after a period of five minutes completely blocked both com-
ponents of secretion. The in vitro results mentioned above
offer an explanation for these findings: when ATP is replaced
by a nonhydrolyzable analogue, the continuous disassembly
of SNARE complex stops. Therefore the SNAREs on a given
vesicle tend to form ternary complexes and are no longer avail-
able for complexes linking vesicle and plasma membrane.
However, our data do not exclude additional requirements for
ATP, such as in maintaining PtdIns-4,5P2 levels2628. In fact,
our finding that the exocytic burst is lost within five minutes
in the absence of MgATP implies either that one of the last
steps before fusion is affected by the prolonged absence of
MgATP, or else that transmembrane SNARE complexes can
be disintegrated without the action of NSF.
CLOSTRIDIAL NEUROTOXINS AND CATECHOLAMINE SECRETION
Clostridial neurotoxins are potent inhibitors of synaptic-vesi-
cle exocytosis in nerve terminals. It is now well established
that the light chains (LC) of clostridial neurotoxins act as zinc-
dependent metallopreoteases, which specifically cleave SNARE
proteins. In particular, tetanus neurotoxin (TeNT) and botu-
nature neuroscience volume 1 no 3 july 1998
articles
Fig. 4. The effect of BoNT/E and BoNT/A on
secretion. (a) BoNT/E blocks both components of
secretion at low [Ca2+]i. Averaged [Ca2+]i levels,
capacitance traces and amperometric currents in
response to the first flashes from paired experi-
ments in control (solid lines; n = 18) and in the
presence of 400 nM BoNT/E-LC (dashed lines;
n = 15) are displayed. Flashes were triggered after
Control 10 min of whole cell dialysis, corresponding to 200
ms in the graph. The capacitance was normalized
to its pre-flash values. (b) BoNT/A partially blocks
BoNT/A
secretion at low [Ca2+]i. Averaged [Ca2+]i levels,
capacitance traces and amperometric currents in
response to the first flashes from paired experi-
ments in control (solid lines; n = 24) and in the
presence of 800 nM BoNT/A-LC (dashed lines;
n = 37) are displayed. The capacitance was normal-
ized to its pre-flash values. Flashes were triggered
after 10 min of whole cell dialysis, corresponding
to 200 ms in the graph. The capacitance trace gives
the impression that the slow phase of secretion is
blocked completely, whereas the amperometry
trace reports some continuing release two sec-
onds following the flash. This discrepancy may be
Time (s)
due to slow endocytosis, which conceals a slow-
solid trace is the averaged Cm response from 55 cells without
BoNT/A. It can be fitted by two exponentials (superimposed
dashed line) with a fast time constant of 51.6 ms and a slow one
of 195.3 ms. The lower solid trace is the averaged Cm response
from 40 cells poisoned with 800 nM BoNT/A-LC. The onset of
capacitance increase only shows a slow exponential component
with a time constant of 207.8 ms. Superimposed dashed line is a
single exponential fit. The capacitance was normalized to its pre-
flash values. (d) BoNT/A reduced maximal rate of secretion.
Maximal rates of secretion Cm/t were measured from the
capacitance response. t was set to half the fastest time con-
BoNT/A stant analyzed in the exponential fit of the cell.
linum neurotoxin (BoNT) serotypes B, D, F and G specifical-
ly cleave synaptobrevin at unique peptide bonds 810. BoNT
serotypes A and E cleave SNAP-25 at two different sites locat-
ed close to the carboxyl terminals29,30, whereas the targets of
BoNT serotype C are syntaxin and SNAP-2531,32. To explore
the role of SNARE proteins in the secretory pathway, we asked
whether the kinetic components described above are differ-
entially influenced by different clostridial neurotoxins.
If we assume there are 22,000 vesicles per chromaffin cell33,
and each vesicle contains 100 copies of synaptobrevin, the final
concentration of synaptobrevin is calculated to be 3 M in a
single chromaffin cell. To test the time course of the cleavage
of our recombinant light chains of clostridial neurotoxins, we
first performed in vitro cleavage assays. The results show that,
in our internal solution at 32C, 330 nM TeNT-LC are capa-
ble of cleaving more than 95% of 5 M recombinant GST-
synaptobrevin2 within three minutes, and 40 nM BoNT/E-LC
can cleave more than 95% of 5 M recombinant SNAP-25His6
(data not shown). For a 10 M access-resistance pipette, a
protein with molecular weight of 50 kD, which is the size of
the light chains of clostridial neurotoxins, can diffuse into the
cytosol with a time constant of five minutes 34. Thus within
195
 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 5. The effect of

a b c BoNT/C1, BoNT/D and
TeNT on secretion.
[Ca2+]i (M)

[Ca2+]i (M)

[Ca2+]i (M)
BoNT/C1 (a), BoNT/D
(b) and TeNT (c) block
both components of secre-
tion at low [Ca2+]i. The fig-
ure displays the averaged
Control [Ca2+]i levels, capacitance
Control Control traces and amperometric
currents in response to the
Cm

Cm
Cm

Tetx_LC (4 M)
first flashes from paired
BoNT/D (400 nM) experiments in the control
BoNT/C1 LC(2M)
condition (solid lines) and
in the presence (dashed
lines) of 2 M BoNT/C1-
LC (n = 7; control n = 6),
400 nM BoNT/D (n = 11;
IAMP (pA)

IAMP (pA)
IAMP (pA)

control n = 12) and 4 M

1998 Nature America Inc. http://neurosci.nature.com

TeNT (n = 14; control

n = 17). Flashes were trig-
gered after 10 min of
whole-cell dialysis, corre-
sponding to 200 ms in the
Time (s) Time (s) Time (s) graph. Scaling bars, 200 fF.

five minutes, the cytosol concentration of the light chains of in BoNT/A-poisoned cells, which was that exocytic bursts were
neurotoxins would reach 36% of the pipette concentration. slowed down. The averaged Cm response from 50 control cells
Therefore, to test the role of SNAP-25 in exocytic processes, displayed an exocytic burst having two exponentials with time
we used 400 nM BoNT/E-LC in the patch pipette and waited constants of 51.6 ms and 195.3 ms. However, the response of
for ten minutes before elevation of [Ca2+]i to ensure complete 46 BoNT/A-treated cells could be well fitted by a single expo-
action of the toxin. As the action of clostridial neurotoxins are nential with a time constant of 207.8 ms, which is similar to
also steeply temperature dependent35, the experiments were the second component of the control cells (Fig. 4c). As a
conducted at 32C. result, the initial maximal rate of rise of Cm was reduced 4.7-
Under our experimental protocol, BoNT/E-LC not only fold (see Fig. 4d). When Cm responses from BoNT/A-treat-
blocked the slow phase of Cm increase, but also inhibited the ed cells were analyzed individually, 33 out of 38 flash responses
exocytic burst (Fig. 4a; see also Table 1). Next, we wanted to could be adequately fitted by a single exponential with an
assess the effects of BoNT/A, which cleaves the Q197R198 pep- overall average amplitude of 398 37 fF and a time constant
tide bond very close to the C-terminus of SNAP-25 and there- of 281 27.8 ms. This value is not different from that of the
by removes only nine amino acids, in contrast to BoNT/E, second component of exocytic bursts in control cells. More-
which cleaves the R 180 I 181 peptide bond, 17 residues over, the time constant of this slow component was more or
upstream. In vitro cleavage studies showed that BoNT/A has less constant in the [Ca2+]i range between 15 and 70 M. We
similar kinetics as BoNT/E (data not shown). However, even at therefore suggest that the truncated form of SNAP-25 cannot
800 nM, which is twice the concentration used for BoNT/E, mediate a particularly rapid interaction between calcium-sen-
BoNT/A only partially inhibited secretion (Fig. 4b). BoNT/A sor and release machinery for catecholamine release (see Dis-
reduced both the exocytic burst (Table 1) and the slow phase cussion). Alternatively, this suppression of the rapid phase of
of secretion. The capacitance trace (Fig. 4b) seems to indicate the exocytic burst by BoNT/A may represent the block of a
complete block of the slow phase. However this result cannot
be taken at face value, because endocytosis may compensate Table 1. Amplitudes of the exocytic burst in paired
for ongoing exocytosis. Indeed, the amperometric trace indi- experiments
cates a small slow exocytic component after toxin treatment.
Experiment Control (fF) Treatment (fF)
By analyzing the amperometric trace, we concluded that the
exocytic burst (approximately the integral response over the AMP-PCP/NP-EGTA 242 47 (n = 9) 21 3 (n = 16)
first two seconds) is reduced by a factor of 1.7, whereas the AMP-PCP/DM-nitrophen 340 82 (n = 4) 14 7 (n = 10)
slow component (two to ten seconds) is reduced by a factor AMP-PNP/DM-nitrophen 496 61 (n = 11) 79 27 (n = 12)
of 2.7. This strong reduction of the slow component suggests BoNT/A (800 nM) 511 38 (n = 50) 363 27 (n = 38)
that BoNT/A affects a relatively early step in the secretory BoNT/E (400 nM) 196 27 (n = 13) 37 9 (n = 14)
pathway by reducing the supply rate of release-ready granules.
BoNT/C1 (2 M) 260 37 (n = 6) 29 4 (n = 6)
The complete block by BoNT/E (as compared to the partial
one of BoNT/A) indicates that residues 181197 at the C-ter- BoNT/D (400 nM) 423 86 (n = 11) 43 11 (n = 13)
minus of SNAP-25 are essential for exocytosis, in accordance TeNT (4 M) 226 28 (n = 13) 44 8 (n = 12)
with previous studies13,36. The capacitance values were measured at 800 ms after flashes. Values given
Scrutinizing exocytic bursts revealed an additional change are mean SE.

196 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 6. Secretion at high [Ca2+]i is sensitive to ATP but not to

clostridial neurotoxins. (a) Comparison of secretory
a Control AMP-PCP

responses to high [Ca2+]i after exhaustion of exocytosis at

[Ca2+]i (M)
low [Ca2+]i. Averaged [Ca2+]i levels, capacitance traces and
amperometric currents from paired experiments in the pres-
ence of MgATP as control (n = 7) and in the presence of
2 mM AMP-PCP substituting for MgATP (n = 17) are dis-
played. The flashes were usually the third ones in a given cell.
(b) The effect of TeNT on secretory responses to high
[Ca2+]i after exhaustion of exocytosis at low [Ca2+]i.
Averaged [Ca2+]i levels, capacitance traces and amperometric

Cm
currents from paired experiments in control (n = 7) and in
the presence of 4 M TeNT-LC (n = 6) are displayed. The
flashes were usually the third ones in a given cell.

IAMP (pA)
small, more rapid and more toxin-sensitive compo- Time (s) Time (s)
nent of exocytosis, not necessarily related to cate- b
1998 Nature America Inc. http://neurosci.nature.com

cholamine release. In conclusion, BoNT/A has a dual [Ca2+]i (M) Control Tetx LC
effect, one at a relatively early step and another one
modifying the exocytic burst, which is considered to
be the final step in exocytosis.
BoNT/C1 cleaves both syntaxin and SNAP-25 in
chromaffin cells 31. This toxin cleaves SNAP-25 at a
peptide bond (R 198A 199) adjacent to the BoNT/A
cleavage site (V.V. and H.N., unpublished results). To
test the role of syntaxin in different components of
Cm

secretion in chromaffin cells, we used 2 M

BoNT/C1-LC in the pipette solution. If an intact syn-
taxin were not essential for exocytosis, we would
IAMP (pA)

expect the same incomplete inhibition as seen with

BoNT/A. However, BoNT/C1 completely blocked the
secretion at low [Ca 2+]i, suggesting that syntaxin is
important in the final step of exocytosis (Fig. 5a).
Synaptobrevin is the substrate for TeNT and
BoNT/B, D, F and G. In the presence of 400 nM BoNT/D-LC,
capacitance and amperometric measurements (Fig. 5b) show
that both components of secretion are blocked at low [Ca2+]i.
Similar results were also obtained for 4 M TeNT-LC (Fig. 5c),
suggesting that synaptobrevin also is critical for the final steps
of exocytosis. Comparing effective toxin concentrations shows
that BoNT/D is much more potent in blocking secretion than
TeNT. The inactive mutants of TeNT-LC (E234Q) and
BoNT/C1-LC (E230A) had no effect on secretion, suggesting
that the blockage of clostridial neurotoxins that we observed
here was due to the toxins metalloprotease activities.
It may be considered surprising that exocytosis is completely
blocked within a time window of about ten minutes by many
toxins that do not cleave their targets as long as those proteins
are part of a SNARE complex, as many recent models envisage
an intact SNARE complex as part of a release-ready fusion
machinery. If this is the case, our findings imply that the
respective proteins must cycle through a nonprotected state
(which is the monomeric form) within a few minutes. Other-
wise, we would expect the exocytic burst to be toxin resistant.
ATP-DEPENDENT, TOXIN-INSENSITIVE CAPACITANCE COMPONENT
We have shown that there is an additional large Cm increase at
high [Ca2+]i that is not related to catecholamine release. This
Cm response is dependent on ATP. When we substituted ATP
with AMP-PCP, this component vanished completely (Fig. 6a).
This result gives us some confidence that the signal is not mere-
ly an artifact of Cm measurement.
Time (s) Time (s)
We further tested this Cm increase for its sensitivity to clostridi-
al neurotoxins. To do so, we first depleted the Cm responses by
repetitive stimulations at low [Ca2+]i. Then, we elevated [Ca2+]i
to high levels in order to elicit this component. This way we could
infer that this Cm increase does not represent the same pool as
that measured at low [Ca2+]i. This protocol also provided the
advantage that we could judge whether the neurotoxins were active
in a given experiment. Even when the neurotoxins succeeded in
blocking secretion at low [Ca2+]i, they failed to block the Cm
increases at high [Ca2+]i. Figure 6b shows that TeNT-LC did not
inhibit Cm responses at high [Ca2+]i. The same results were
obtained with BoNT/E, D, A and C1. These results suggest the
existence of an additional type of vesicles devoid of catecholamines,
which are fusion competent at [Ca2+]i over 100 M. Exocytosis of
these vesicles is not sensitive to clostridial toxins but is dependent
on ATP. Above, we concluded that this intermediate component
is elicited whenever [Ca2+]i exceeded a certain threshold. This point
is strengthened by the toxin results reported here, as no indication
of this component was observed after toxin treatment when [Ca2+]i
increases were restricted to below 70 M.
The existence of a Cm increase not related to catecholamine
release that is triggered at [Ca2+]i above 100 M is somewhat
puzzling. The presence of vesicles other than chromaffin gran-
ules may be an explanation. One possibility might be that low
[Ca2+]i mainly triggers the fusion of catecholamine-containing
chromaffin granules, whereas high [Ca2+]i preferentially triggers
the exocytosis of synaptic-like microvesicles, which contain
acetylcholine. This postulate agrees with the general view that

nature neuroscience volume 1 no 3 july 1998 197

 1998 Nature America Inc. http://neurosci.nature.com

articles

-SNAP

NSF
BoNT/B,D,F,G
BoNT/C
t-SNAREs
1 BoNT/A,E
TeNT
v-SNARE
Morphological

Ca2+-Sensor
Docking

-SNAP
NSF Pro
MgATP Rea ductive
ssem
2 3
al

bly
tin ic

Maturation Exocytosis
ac ort

NSF 10 s 10 - 100 ms
C

?
4 5 Ca2+ 6
1998 Nature America Inc. http://neurosci.nature.com

Fig. 7. Hypothetical model for the kinetic steps leading to exocytosis. Left, Morphological docking may require cortical actin, Munc18, Doc2
and similar proteins, but may not involve the interaction of v- and t-SNAREs. The 7S complex existing on the vesicle membrane prevents the
v-SNAREs from interacting with plasma membrane t-SNAREs. The 20S complex is formed after the 7S complex binds to -SNAP and NSF.
It is disassembled by ATP hydrolysis, which makes v-SNAREs available to interact with plasma membrane t-SNAREs. For simplicity, 7S com-
plexes are displayed as consisting only of one v-SNARE and one t-SNARE. In the productive reassembly of SNARE complexes between mem-
branes, the energy released by complex formation is used to put the membranes under tension and to make the vesicle release ready.
Calcium-sensor protein may interdigitate with the distal tail of the SNARE complexs coiled coil structure. A calcium-dependent conforma-
tional change might put additional torsion on the coiled-coil, forcing the two membrane anchors together and initiating fusion. The numbers
represent functionally discrete pools of vesicles.

the [Ca2+]i threshold for triggering small clear vesicles is higher
than that for large dense-core vesicles. However, our data are
inconsistent with this idea. Chromaffin cells express acetylcholine
receptor channels. Therefore one might expect to observe a tran-
sient conductance following a flash if acetylcholine were released.
This was not the case, however. Furthermore, the resistance of
this non-catecholamine Cm increase to clostridial neurotoxins
suggests that the involvement of SNAREs is not necessary for the
fusion of the underlying vesicles. By exclusion, then, the expla-
nation for this peculiar Cm increase seems to be that of an unspe-
cific membrane fusion triggered by high level of [Ca2+]i, which
the absence of ATP somehow prevents. Alternatively, this fusion
event could represent a previously undetected process that
involves a toxin-resistant set of SNARE proteins, as has been sug-
gested for the apical route in polarized cells37,38.
Discussion
In previous work, at least three late steps have been distinguished
in the secretory pathway of neuroendocrine cells: morphological
docking, priming, and exocytosis. Various properties have been
assigned to the individual stages; thus, docking is ATP-dependent28,
priming is dependent on both ATP and calcium20,39 and exocytosis
shows a steep calcuim dependence18,19. Evidence for these proper-
ties has been obtained from different laboratories by very different
techniques and by experiments on diverse time scales. It is a ques-
tion of intense current research how features defined in such dif-
ferent ways correspond to each other, how they relate to molecularly
defined states, and which of these states are susceptible to (or else
protected from) the proteolytic action of clostridial neurotoxins.
Electrophysiological experiments showed that sufficiently
strong stimuli elicit a so-called exocytic burst, which is inter-
preted to represent complete exocytosis of a pool of release-ready
granules (the readily releasable pool, RRP). This size of this pool
depends on [Ca2+]i preceding the stimulus and on activation of
PKC4,40. Following the exocytic burst, release proceeds at a much
slower rate, which is believed to represent the combined process
of recruiting vesicles to the RRP. Flash photolysis of caged calci-
um together with rapid assays of exocytosis, such as membrane-
capacitance measurement and amperometry, allow a detailed
kinetic analysis of secretory responses. Our goal was to establish
links between the distinct kinetic components of such [Ca2+]i
jump experiments and molecular processes.
Five of the findings reported in our study bear on this issue.
First, we eliminated one of the kinetic components as a candidate
for a step in the release of catecholamines. Second, we showed
that five minutes of neurotoxin action in the absence of stimula-
tion are sufficient to eliminate the exocytic burst. This implies all
of SNAP-25 or syntaxin or synaptobrevin must cycle through a
toxin-sensitive state within that time period. Thus, any toxin-pro-
tected SNARE complex as part of a mature fusion machinery must
have a lifetime shorter than five minutes. Third, cleavage of SNAP-
25 by BoNT/A suppresses exocytosis only partially and manifests
itself both at a slow (seconds to minutes) and at the fastest step
by retarding the reactions in which SNAP-25 is involved. Fourth,
the size of the exocytic burst is only partially affected by BoNT/A.
Fifth, the time constant of the exocytic burst after BoNT/A action
is not calcium dependent above 20 M.
In the following, we explain these findings in the framework
of a specific model of exocytosis control although we are aware
that alternative interpretations are possible. We start with the
recent proposal of a productive reassembly of SNARE complexes
between membranes, in which the energy released by complex
formation is used to put the membranes under tension41. We
incorporate this into a model similar to those previously pro-

198 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com
posed2,40, including the recent finding that v-SNARE and t-
SNAREs are in parallel orientation in the complex7 and arrive at
Fig. 7. In this model, we allow for morphological docking, for
NSF action to separate SNARE complexes42, for SNARE complex
formation (productive reassembly) and for fusion. For rapid
action of calcium, the last step should be steeply dependent on
calcium and fast, such that the readily releasable pool (state 5)
can be depleted rapidly within the exocytic burst. In addition, a
step of SNARE complex maturation is considered in the model
to allow for the finding that this pool recovers from depletion
(following strong stimulation) on the time scale of ten seconds43
and that its size depends on PKC and [Ca2+]i39,40. This step (from
state 4 to state 5) may correspond to the recruitment of other
synaptic proteins to the complex such as the calcium-dependent
activator protein for secretion (CAPS)28 and the calcium sensor
(possibly synaptotagmin). We also allow morphological undock-
ing on the time scale of minutes44. Reversal of the productive
reassembly step may involve NSF action. Such a model fits all of
1998 Nature America Inc. http://neurosci.nature.com
the findings summarized above. It explains why SNARE com-
plexes are not protected from clostridial toxin action because
NSF intermittently separates SNAREs within a time span of five
minutes, which permits proteolytic action of the toxins. Action of
most of the toxins, with the exception of BoNT/A, would com-
pletely prevent the productive reassembly and thereby prevent
the formation of a release-ready pool of vesicles.
The model also provides a hint about why BoNT/A fails to elim-
inate secretion completely and seems to act both at an early and at a
late step. BoNT/A cleaves off SNAP-25 the nine amino acids farthest
from the plasma membrane, at the C-terminus. Truncated SNAP-
25 still is able to form SNARE complexes that are disassembled by
NSF45,46, but the C-terminal end of the molecule is likely to be
involved in initial contact during productive reassembly41. Lack of
these nine amino acids may slow down rather than prevent this step.
In our experiments, such a change would be reflected in a decrease
of the rate of secretion at high [Ca2+]i, once the readily releasable
pool is consumed. In other words, it would slow down the late com-
ponent of secretion, as we observed. The same lack of nine amino
acids may also explain the slowdown of the exocytic burst itself, if
the calcium sensor for exocytosis (possibly synaptotagmin) inter-
acts with the C-terminal end of the SNARE complex. Such interac-
tion between the C-terminus of SNAP-25 and synaptotagmin was
actually shown to occur in vitro (R.R.L. Gerona & T.F.J. Martin, per-
sonal communication). A specific proposal of how this might hap-
pen is shown in Fig. 7, referring to the finding that the ternary
complex of SNAREs has a coiled-coil structure41. If the calcium sen-
sor interdigitates with the tail of this structure, a calcium-dependent
conformational change might put additional torsion on the coiled
coil, forcing the two membrane anchors together and initiating
fusion. Shortening of the SNAP-25 tail would most likely reduce the
efficacy of the calcium sensor in transmitting the calcium signal to
the SNARE complex. On the other hand, the affinity of the sensor
itself is not expected to change (apart from possible allosteric cou-
pling effects). In the model of Heinemann and colleagues19, such a
change would be most conveniently represented by a reduction of
the rate constant (the rate constant of exocytosis from a triply cal-
cium-bound state of the sensor) to a value of about five per second.
Given the intrinsic dissociation constant of the calcium sensor of 13
M19 and a calcium-binding rate to the sensor of 8106 per mole per
second47, a saturation of the calcium dependence of the release-rate
constant would be expected for [Ca2+]i in the range 20 to 70 M, as
observed after BoNT/A treatment.
Irrespective of the specific mechanism suggested here, the
finding that removal of nine amino acids from SNAP-25 slows
nature neuroscience volume 1 no 3 july 1998
articles
down the exocytic burst and does so more at high than at low
[Ca2+]i is strong evidence that this portion of the molecule is
involved in mediating the action of a calcium sensor in rapid cal-
cium-regulated exocytosis.
Methods
CELL PREPARATION AND SOLUTIONS. Chromaffin cells from bovine adrenal
glands were prepared and cultured as described48. Cells were used one
to four days after preparation. The external bathing solutions for exper-
iments contained (in mM) 150 NaCl, 2.8 KCl, 2 CaCl2, 1 MgCl2, 10
HEPES and 2 mg per ml glucose (pH 7.2, 320 mosm). For preparing
pipette solutions, we generally used 2 concentrated buffers, which con-
tained 250 mM Cs-glutamate, 80 mM HEPES (pH 7.2). We added dif-
ferent concentrations of NP-EGTA (gift from Dr. Ellis-Davies,
Philadelphia), fura-2 (Texas Fluorescence Labs, Austin, TX), furaptra
(Molecular Probes, Eugene, OR), CaCl2 or ATP for different purposes
as indicated in the text. The resulting mixtures were diluted with dou-
ble-distilled water for the appropriate osmolarity (310 mosm). The NP-
EGTA-containing internal solutions for control consisted of (in mM)
84.5 Cs-glutamate, 10 NP-EGTA, 8 CaCl2, 1 MgCl2, 2 MgATP, 0.3 GTP,
0.5 furaptra, 27 HEPES. For the DM-nitrophen experiment, internal
solutions contained (in mM) 110 Cs-glutamate, 5 DM-nitrophen, 4
CaCl2, 2 MgATP, 0.3 GTP, 0.5 furaptra, 35 HEPES. The basal [Ca2+]i was
measured to be 100300 nM by fura-2. The pipette solution was adjust-
ed to pH 7.2 by either HCl or CsOH. All experiments were performed
at 32-33C.
PHOTOLYSIS OF CAGED CALCIUM AND [CA2+]I MEASUREMENT. Flashes of ultra-
violet light and fluorescence excitation light were generated as
described48. To avoid any influence resulting from the loading tran-
sient4, we used the more calcium-selective caged compound Nitro-
phenyl-EGTA49. The flash photolysis efficiency was also measured as
described48 except for the following changes: we used 1 mM fura-2, 2
mM NP-EGTA and 2 mM CaCl2 during flash experiments. Small flash
intensities were applied by adding neutral-density filters. From mea-
surements of [Ca2+]i before and immediately after flashes, we calculated
the photolysis efficiency for NP-EGTA. The photolysis efficiency of a 375
V discharge flash for NP-EGTA was determined to be 52%. As the [Ca2+]i
should decay significantly during 10 s of Cm measurement after flash-
es48, we used the fluorescence excitation light to measure [Ca2+]i and to
simultaneously photorelease calcium after the flashes in order to keep
[Ca2+]i more or less constant. [Ca2+]i was calculated from the fluores-
cence ratio R according to ref. 50. The calibration constants for furap-
tra measurements before and after 375 V discharge flashes were measured
as described48. For 10 mM NP-EGTA, 0.5 mM furaptra, the changes of
the calibration constants for a ratiometric measurement at 340 and 380
nm were as follows: Rmin changed from 0.238 to 0.247, Rmax from 6.554
to 4.708, Keff from 2128 M to 2053 M.
WHOLE-CELL PATCH CLAMP AND CAPACITANCE MEASUREMENT. Convention-
al whole-cell recordings were done with sylgard-coated 2-3 M pipettes.
Series resistance ranged from 412 M. An EPC-9 patch-clamp ampli-
fier was used together with Pulse software (HEKA Electronics, Lam-
brecht, Germany). Capacitance measurements used the Lindau-Neher
technique17 implemented as the sine+dc mode of the software lock-in
extension of pulse software, which allowed long duration Cm measure-
ment in single sweeps. A 800 Hz, 50 mV peak-to-peak sinusoid voltage
stimulus was applied above a DC holding potential of -70 mV. Currents
were filtered at 2 kHz and sampled at 12 kHz. The capacitance traces
were imported to IGOR Pro (WaveMetrics, Inc., Lake Oswego, OR). The
analyses were conducted on a Macintosh computer using IGOR Pro.
Unless otherwise stated, the data are given as mean SE.
AMPEROMETRY. Carbon-fiber electrodes were prepared from 10 m
diameter carbon fibers (Amoco performance products, Greenville,
South Carolina) and were canulated through glass capillaries. A con-
stant voltage of 780 mV versus Ag/AgCl reference was applied to the
electrode. The tip of the carbon-fiber electrode was gently pressed
against the cell surface. The amperometric current was filtered at 3
kHz, sampled at 10 kHz and further digitally filtered at 1 kHz. Arti-
199
 1998 Nature America Inc. http://neurosci.nature.com
articles
facts of amperometry due to flash irradiation were subtracted using
the averaged trace for the same fiber at the end of the experiment
when there was no secretion.
Acknowledgments
We would like to thank Dr. Ellis-Davies for samples of NP-EGTA, Drs. Corey
Smith, Reinhard Jahn and Tobias Moser for feedback on the manuscript, and
Frauke Friedlein and Michael Pilot for cell preparation. This work was
supported by grants from the Deutsche Forschungsgemeinschaft (Nr. CHV-
113/65/0) and from the European Community (Nr. CHRX-CT940500 ) to E.N.
T.B. and H.N. were supported by the Fonds der chemischen Industrie and by the
Deutsche Forschungsgemeinschaft (Nr. IIB2-Bi660/1-1).
RECEIVED 29 APRIL: ACCEPTED 23 MAY 1998
1. Holz, R.W., Bittner, M.A., Peppers, S.C., Senter, R.A. & Eberhard, D.A.
MgATP-independent and MgATP-dependent exocytosis. J. Biol. Chem. 264,
54125419 (1989).
1998 Nature America Inc. http://neurosci.nature.com
2. Parsons, T.D., Coorssen, J.R., Horstmann, H. & Almers, W. Docked granules,
the exocytic burst, and the need for ATP hydrolysis in endocrine cells. Neuron
15, 10851096 (1995).
3. Rosenmund, C. & Stevens, C.F. Definition of the readily releasable pool of
vesicles at hippocampal synapses. Neuron 16, 11971207 (1996).
4. Neher, E. & Zucker, R.S. Multiple calcium-dependent processes related to
secretion in bovine chromaffin cells. Neuron 10, 2130 (1993).
5. Sllner, T. et al. SNAP receptors implicated in vesicle targeting and fusion.
Nature 362, 318324 (1993).
6. Sdhof, T.C. The synaptic vesicle cycle: a cascade of protein-protein
interactions. Nature 375, 645653 (1995).
7. Hanson, P.I. & Jahn, R. Structure and conformational changes in NSF and its
membrane receptor complexes visualized by quick-freeze/deep-etch electron
microscopy. Cell 90, 523535 (1997).
8. Foran, P., Lawrence, G. & Dolly, J.O. Blockade by botulinum neurotoxin B of
catecholamine release from adrenochromaffin cells correlates with its
cleavage of synaptobrevin and a homologue present on the granules.
Biochemistry 34, 54945503 (1995).
9. Niemann, H., Blasi, J. & Jahn, R. Clostridial neurotoxins: new tools for
dissecting exocytosis. Trends Cell Biol. 4, 179185 (1994).
10. Montecucco, C. & Schiavo, G. Mechanism of action of tetanus and botulinum
neurotoxins. Mol. Microbiol. 13, 18 (1994).
11. Hayashi, T., Yamasaki, S., Nauenburg, S., Binz, T. & Niemann, H. Synaptic
vesicle membrane fusion complex: action of clostridial neurotoxins on
assembly. EMBO J. 13, 50515061 (1994).
12. McMahon, H.T. et al. Tetanus and botulinum toxins type A and B inhibit
glutamate, GABA, asparate and metenkephalin release from synaptosomes:
clues to the locus of action. J. Biol. Chem. 267, 2133821343 (1992).
13. Lawrence, G.W., Foran, P., Mohammed, N., DasGupta, B.R. & Dolly, J.O.
Importance of two adjacent C-terminal sequences of SNAP-25 in exocytosis
from intact and permeabilized chromaffin cells revealed by inhibition with
Botulinum neurotoxins A and E. Biochemistry 36, 30613067 (1997).
14. Dreyer, F., Rosenberg, F., Becker, C., Bigalke, H. & Penner, R. Differential
effects of various secretagogues on quantal transmitter release from mouse
motor nerve terminals treated with botulinum A and tetanus toxin. Naunyn-
Schmiedebergs Arch. Pharmacol. 335, 17 (1987).
15. Capogna, M., McKinney, R.A., OConnor, V., Ghwiler, B.H. & Thompson,
S.M. Ca2+ or Sr2+ partially rescues synaptic transmission in hippocampal
cultures treated with botulinum toxin A and C, but not tetanus toxin. J.
Neurosci. 17, 71907202 (1997).
16. Neher, E. & Marty, A. Discrete changes of cell membrane capacitance
observed under conditions of enhanced secretion in bovine adrenal
chromaffin cells. Proc. Natl Acad. Sci. USA 79, 67126716 (1982).
17. Gillis, K. D. in Single-Channel Recording 2nd edn. (eds. Sakmann, B. & Neher,
E.) 155198 (Plenum, NewYork, 1995).
18. Thomas, P., Wong, J.G., Lee, A.K. & Almers, W. A low affinity Ca2+ receptor
controls the final steps in peptide secretion from pituitary melanotrophs.
Neuron 11, 93104 (1993).
19. Heinemann, C., Chow, R.H., Neher, E. & Zucker, R.S. Kinetics of the
secretory response in bovine chromaffin cells following flash photolysis of
caged Ca2+. Biophys. J. 67, 25462557 (1994).
20. Bittner, M.A. & Holz, R.W. Kinetic analysis of secretion from permeabilized
adrenal chromaffin cells reveals distinct components. J. Biol. Chem. 267,
1621916225 (1992).
21. Banerjee, A., Barry, V.A., DasGupta, B.R. & Martin, T.F.J. N-Ethylmaleimide-
sensitive factor acts at a prefusion ATP-dependent step in Ca2+-activated
exocytosis. J. Biol. Chem. 271, 2022320226 (1996).
200
22. Nichols, B.J., Ungermann, C., Pelham, H.R.B., Wickner, W.T. & Hass, A.
Homotypic vacuolar fusion mediated by t- and v-SNAREs. Nature 387,
199202 (1997).
23. Colombo, M.I., Taddese, M., Whiteheart, S.W. & Stahl, P.D. A possible
predocking attachment site for N-ethylmaleimide-sensitive fusion protein.
Insights from in vitro endosome fusion. J. Biol. Chem. 271, 1881018816
(1996).
24. Hhne-Zell, B. & Gratzl, M. Adrenal chromaffin cells contain functionally
different SNAP-25 monomers and SNAP-25/syntaxin heterodimers. FEBS
Lett. 394, 109116 (1996).
25. Otto, H., Hanson, P.I. & Jahn, R. Assembly and disassembly of a ternary
complex of synaptobrevin, syntaxin, and SNAP-25 in the membrane of
synaptic vesicles. Proc. Natl Acad. Sci. USA 94, 61976201 (1997).
26. Hay, J.C. & Martin, T.F.J. Phosphatidylinositol transfer protein required for
ATP-dependent priming of Ca2+-activated secretion. Nature 366, 572580
(1993).
27. Hay, J.C. et al. ATP-dependent inositide phosphorylation required for Ca2+-
activated secretion. Nature 374, 173177 (1995).
28. Martin, T.F.J. Stages of regulated exocytosis. Trends Cell Biol. 7, 271276
(1997).
29. Binz, T. et al. Proteolysis of SNAP-25 by types E and A botulinal neurotoxins.
J. Biol. Chem. 269, 16171620 (1994).
30. Blasi, J. et al. Botulinum neurotoxin A selectively cleaves the synaptic protein
SNAP-25. Nature 365, 160163 (1993).
31. Foran, P., Lawrence, G., Shone, C.C., Foster, K.A. & Dolly, J.O. Botulinum
neurotoxin C1 cleaves both syntaxin and SNAP-25 in intact and chromaffin
cells: Correlation with its blockade of catecholamine. Biochemistry 35,
26302636 (1996).
32. Blasi, J. et al. Botulinum neurotoxin C1 blocks neurotransmitter release by
means of cleaving HPC-1/syntaxin. EMBO J. 12, 48214828 (1993).
33. Plattner, H., Artalejo, A.R. & Neher, E. Ultrastructural organization of bovine
chromaffin cell cortexAnalysis by cryofixation and morphometry of
aspects pertinent to exocytosis. J. Cell Biol. 139, 17091717 (1997).
34. Pusch, M. & Neher, E. Rates of diffusional exchange between small cells and a
measuring patch pipette. Pflgers Arch. 411, 204211 (1988).
35. Poulain, B. et al. Differences in the multiple step process of inhibition by
tetanus toxin and botulinum neurotoxins type A and B at aplysia synapses.
Neuroscience 70, 567576 (1996).
36. Bittner, M.A. & Holz, R.W. Protein kinase C and clostridial neurotoxins
affect discrete and related steps in the secretory pathway. Cell. Mol. Neurobiol.
13, 649664 (1993).
37. Ikonen, E., Tagaya, M., Ullrich, O., Montecucco, C. & Simons, K. Different
requirements for NSF, SNAP, and rab proteins in apical and basolateral
transport in MDCK cells. Cell 81, 571580 (1995).
38. Weimbs, T., Low, S.H., Chapin, S.J. & Mostov, K.E. Apical targeting in
polarized epithelial cells: theres more afloat than rafts. Trends Cell Biol. 7,
393399 (1997).
39. von Rden, L. & Neher, E. A Ca-dependent step in the release of
catecholamines from adrenal chromaffin cells. Science 262, 10611065
(1993).
40. Gillis, K.D., Mner, R. & Neher, E. Protein kinase C enhances exocytosis
from chromaffin cells by increasing the size of the readily releasable pool of
secretory granules. Neuron 16, 12091220 (1996).
41. Hanson, P.I., Heuser, J.E. & Jahn, R. Neurotransmitter releasefour years of
SNARE complexes. Curr. Opin. Neurobiol. 7, 310315 (1997).
42. Barnard, R.J.O., Morgan, A. & Burgoyne, R.D. Stimulation of NSF ATPase
activity by alpha-SNAP is required for SNARE complex disassembly and
exocytosis. J. Cell Biol. 139, 875883 (1997).
43. Moser, T. & Neher, E. Rapid exocytosis in single chromaffin cells recorded
from mouse adrenal slices. J. Neurosci. 17, 23142323 (1997).
44. Steyer, J.A., Horstmann, H. & Almers, W. Transport, docking and exocytosis of
single secretory granules in live chromaffin cells. Nature 388, 474478 (1997).
45. Otto, H., Hanson, P.I., Chapman, E.R., Blasi, J. & Jahn, R. Poisoning by
botulinum neurotoxin A does not inhibit formation or disassembly of the
synaptosomal fusion complex. Biochem. Biophys. Res. Comm. 212, 945952
(1995).
46. Hayashi, T., Yamasaki, S., Nauenburg, S., Binz, T. & Niemann, H.
Disassembly of the reconstituted synaptic vesicle membrane fusion complex
in vitro. EMBO J. 14, 23172325 (1995).
47. Klingauf, J. & Neher, E. Modeling buffered Ca2+ diffusion near the
membrane: implications for secretion in neuroendocrine cells. Biophys. J. 72,
674690 (1997).
48. Xu, T., Naraghi, M., Kang, H. & Neher, E. Kinetic studies of Ca2+ binding and
Ca2+ clearance in the cytosol of adrenal chromaffin cells. Biophys. J. 73,
532545 (1997).
49. Ellis-Davies, G.C. & Kaplan, J.H. Nitrophenyl-EGTA, a photolabile chelator
that selectively binds Ca2+ with high affinity and releases it rapidly upon
photolysis. Proc. Natl Acad. Sci. USA 91, 187191 (1994).
50. Grynkiewiez, G., Poenie, M. & Tsien. R.Y. A new generation of Ca2+
indicators with greatly improved fluorescence properties. J. Biol. Chem. 260,
34403450 (1985).
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com

articles

Presynaptic modulation of CA3

network activity
Kevin J. Staley, Mark Longacher, Jaideep S. Bains and Audrey Yee

Departments of Neurology and Pediatrics, University of Colorado Health Sciences Center, Box B182, 4200 East 9th Avenue, Denver, Colorado 80262, USA
Correspondence should be addressed to K.J.S. (Kevin.Staley@UCHSC.edu)

The simultaneous discharge of hippocampal CA3 pyramidal cells is a widely studied in vitro model of
physiological and pathological network synchronization. This network is rapidly activated because of
extensive positive feedback mediated by recurrent axon collaterals. Here we show that population-
burst duration is limited by depletion of the releasable glutamate pool at these recurrent synapses.
Postsynaptic inhibitory conductances further limit burst duration but are not necessary for burst ter-
1998 Nature America Inc. http://neurosci.nature.com

mination. The interval between bursts in vitro depends on the rate of replenishment of releasable
glutamate vesicles and the probability of release of those vesicles at recurrent synapses. Therefore
presynaptic factors controlling glutamate release at recurrent synapses regulate the probability and
duration of synchronous discharges of the CA3 network.

Recurrent axon collaterals of CA3 pyramidal cells make extensive
excitatory synapses on the dendrites of neighboring pyramidal
cells1, so that when one pyramidal cell fires, its neighbors are pow-
erfully excited2,3. This facilitates the rapid synchronization of
action-potential firing in CA3 neurons that underlies the normal
electroencephalographic pattern known as hippocampal sharp
wave activity4,5 and periodic in vitro burst discharges6. This syn-
chrony is important for activity-dependent modification of synap-
tic strength4, but the positive feedback from recurrent collaterals
that underlies the synchronization should produce continuous
discharging of all CA3 neurons7, such as may occur during human
temporal lobe seizures8. Instead, in hippocampal slices from nor-
mal animals, experimental manipulations that decrease inhibi-
tion or increase excitation 6,8 produce only brief population
discharges that resemble sharp waves4 and the pathological elec-
troencephalographic interictal spike pattern that indicates a
propensity for temporal lobe seizures8. During these discharges,
CA3 neurons fire a burst of action potentials during a 50 to 200
millisecond depolarization that begins and ends within a few mil-
liseconds of the rest of the population4,6,8, followed in vitro by a
relatively silent period that lasts until the next burst6. What process
terminates burst activity? Calcium influx during the burst trig-
gers a potassium current that results in an afterhyperpolarization
that can terminate bursts of action potentials initiated by a depo-
larizing current injection9. Thus the afterhyperpolarization, in
conjunction with inhibitory postsynaptic conductances, is a log-
ical mechanism for burst termination6.
Three lines of evidence, however, raise the possibility that
inhibitory conductances are not the primary mechanism termi-
nating CA3 bursts. First, the inhibitory conductance necessary
to counteract the excitatory input to a bursting CA3 pyramidal
cell should produce a dramatic increase in membrane conduc-
tance during the course of the burst10, but studies to date do not
provide evidence for such a conductance increase11,12. Second,
CA3 bursts are only modestly prolonged by experimental manip-
ulations that block the afterhyperpolarization and inhibitory
postsynaptic -aminobutyric acid (GABA) type A13 and B14
receptor conductances, indicating that these conductances are
not necessary for burst termination. Finally, systems characterized
by the amount of positive feedback present in CA3 become
locked into the state favored by the feedback7,15. For example,
during an action potential, depolarization of the neuronal mem-
brane is accelerated by the positive feedback provided by volt-
age-dependent activation of depolarizing sodium conductances.
The positive feedback must be removed by inactivation of the
depolarizing sodium conductances before inhibitory potassium
conductances can repolarize the membrane16.
Analogous to the action potential, positive feedback in area
CA3 could be reduced during the burst by a use-dependent lim-
itation of synaptic strength at recurrent synapses. Modeling stud-
ies demonstrate that the bursts are not sustained below a critical
synaptic strength6. Synaptic strength could be decreased by recep-
tor desensitization or by inhibition of glutamate release via presy-
naptic metabotropic glutamate receptors17. However, these effects
may not produce the necessary rate of decrease in synaptic
strength1719, and use-dependent failure of synapses from CA3
to CA1 pyramidal cells is independent of these two processes20.
Only a few neurotransmitter-containing vesicles are imme-
diately available for release from hippocampal terminals.
Anatomical studies suggest there are 2 to 36 docked glutamate
vesicles per pyramidal cell synapse21,22, and electrophysiological
studies indicate that between 1 and 15 vesicles are available for
immediate release20,2325. When this supply is exhausted, it is
replenished from a larger reserve pool23 with a time constant (at
36 C) of two to ten seconds20,24,25. Here we demonstrate that
the number of releasable glutamate vesicles and the probability of
their release regulates the positive feedback mediated by recur-
rent collateral synapses, and thereby the synchronous activity of
the CA3 network.
Results
INHIBITORY CONDUCTANCES DURING CA3 POPULATION BURSTS
We performed three experiments to assess the role of inhibitory
conductances in terminating CA3 population bursts. First, we
reasoned that if CA3 population bursts are terminated by nega-
tive feedback in the form of inhibitory conductances, then the

nature neuroscience volume 1 no 3 july 1998 201

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 1. Membrane currents

in CA3 pyramidal cells dur-
a b
ing population bursts. (a)
Top panel, whole-cell
recording of the membrane
potential of a pyramidal cell
during a CA3 population
burst induced by 6 mM
extracellular potassium
([K+]o) (resting membrane 50 ms
50 ms
potential -55 mV). Middle
25 mV
panel, in the same cell, 1 nA
1 nA
membrane currents 20 nS
recorded when the mem- Voltage clamp
brane was clamped at the
indicated potentials during
a population burst. Inset, in
a different slice, simultane-
ous recording of CA3 EC field potential
1998 Nature America Inc. http://neurosci.nature.com

extracellular field potential

and the membrane current
in a pyramidal cell demon- Clamp current
strates that the initial rip-
ples in intraburst voltage
clamp currents reflect
spikes in the extracellular
field potential. Bottom
panel, estimate of input
conductance during a pop-
ulation bursts, obtained by subtracting the currents shown in the middle panel, and dividing by the difference in test potential. There
is no evidence of an inhibitory current or conductance of sufficient amplitude to terminate the burst10. Recording performed with a
potassium electrode solution. (b) Membrane currents during CA3 population bursts before and after blocking GABA A conductances
with picrotoxin. Currents were recorded using cesium and QX314 in the electrode solution. Currents at all time points during the
burst reverse at similar potentials (symbols, inset), so that there is no evidence for burst termination by a large, late inhibitory cur-
rent. Bottom panel, in the same cell, the burst is prolonged after blocking GABA A receptors, and the reversal potential of the currents
at all time points during the burst is shifted positive by 10 mV. The uniform shift in reversal potential at all time points during the burst
indicates that the GABA A conductance is already maximal at the earliest part of the burst.

input conductance of neurons participating in the burst should
increase as these conductances become large enough to negate
excitatory input from recurrent collaterals and terminate the
burst10. We induced CA3 population bursts by increasing the
extracellular potassium concentration to 8.5 mM, and then esti-
mated the membrane conductance of CA3 pyramidal cells by
measuring the currents required to clamp the membrane at var-
ious potentials during the bursts (Fig. 1). The membrane con-
ductance was maximal near the beginning rather than the end
of the burst, and the reversal potential of the membrane currents
did not change substantially during the burst (n = 3), suggesting
that there was no increase in inhibitory conductances near the
end of the burst. Further, blocking inhibitory GABAA conduc-
tances with picrotoxin had the same effect on the reversal poten-
tial of membrane currents at all time points during a burst (n =
3). This indicates that GABAA currents do not increase near the
end of a burst, and excludes the possibility that a progressive
increase in GABAA inhibitory conductance was masked by a pro-
portional decrease in excitatory conductances.
The second experiment was based on the effect of postsynaptic
inhibitory conductances on the probability of firing an action
potential in response to excitatory postsynaptic current10. If CA3
population bursts are limited by inhibitory conductances, then
the number of action potentials triggered by depolarizing cur-
rent injection should be substantially reduced at the end of a
burst compared to just prior to a burst. We induced spontaneous
CA3 population bursts using 8.5 mM extracellular potassium,
and recorded the number of action potentials that could be trig-
gered in pyramidal cells by a 500-millisecond depolarizing current
of the same magnitude as that recorded during spontaneous
bursts (1 nA; e.g. Fig. 1b). The number of action potentials elicit-
ed within one second after a population burst (11.53 0.3) was
not significantly different from the number elicited one second
prior to a burst (12.44 0.4; Fig. 2a and b; n = 4 cells, 35 cur-
rent injections). Thus, although the pyramidal cell firing rate
drops during prolonged depolarizing currents, the time scale over
which this occurs is poorly correlated with burst duration (Fig.
2a), and bursts produce only a small decrement in the probabil-
ity of action potential firing.
Finally, if CA3 population bursts are limited by negative
feedback in the form of inhibitory conductances, then phar-
macological blockade of these conductances should result in
continuous activation of the CA3 network. When the after-
hyperpolarization was blocked with norepinephrine26, popu-
lation bursts were only moderately prolonged, and there was
no significant impact on burst frequency (n = 6; Fig. 2b and c).
Burst length and frequency were also modestly affected by
enhancing or blocking inhibitory postsynaptic GABA A con-
ductances (n = 10; Fig. 2d). Even when the afterhyperpolariza-
tion, GABAA, and GABAB conductances were simultaneously

202 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 2. Role of postsynaptic inhibitory

conductances in the control of CA3
a b
burst discharges. (a) Effect of the after-

Number of spikes in burst

hyperpolarization on CA3 pyramidal
cell burst firing. Similar trains of action
potentials were elicited in a CA3 pyra-
midal cell by 500 ms, 1 nA current
injections at various times after bursts
induced by 8.5 mM [K+]o. Top traces
are membrane potential responses to
current injection; bottom traces are
schematics of the timing of current Time after spont. burst (ms)
injections. (b) Number of action
potentials in each train as a function of
c d
the delay between the end of a popula-
tion burst and the start of the current
injection for the cell shown in (a). The

Burst interval (sec) P

average interburst interval was 2.6 s.

Burst length(ms) J
For each delay, points represent means
1998 Nature America Inc. http://neurosci.nature.com

and bars represent standard error; n =

1 to 4 for each delay. (c) Effect of
blocking IAHP (afterhyperpolarization
current) on burst discharge.
Intracellular and extracellular record- 1s
ings are superimposed. Top two traces,
in control solution (8.5 mM [K+]o), the
duration of the afterhyperpolarization
does not predict the timing of the next Minutes
burst. Bottom trace, blocking IAHP with
40 M norepinephrine does not alter e f
burst timing; burst length increased by
23 7% (average standard error), and

Burst length (ms) j

interburst interval decreased by 1
Burst interval (sec) P

12% (n = 6). Scale, 20 mV IC, 2 mV EC.

(d) Effect of GABAA conductances on
burst length (large symbols) and the
interval between CA3 population
bursts (small symbols) was assessed by
bath application of 60 M pentobarbital
(pb) and 100 M picrotoxin (ptx).
Bursting was induced by 8.5 mM [K+]o.
Heavy line is a moving average.
Burst number
Increasing the GABAA conductance
with 60 M pentobarbital decreased
burst length by 12 11% and decreased the interburst interval by 20 14 % (n = 10). Blocking GABAA conductances with picrotoxin
increased burst length by 39 19 % and increased the interburst interval by 50 15 % compared to control (n = 10). (e) Bursts recorded
from two CA3 pyramidal cells in 6 mM [K+]o (control) and after blockade of the afterhyperpolarization, GABAA, GABAA, conductances and
glutamate receptor desensitization. Top panels are current-clamp recordings in a cell recorded with K+ electrode solution; bottom panels are
membrane currents in a cell voltage clamped at 55 mV with cesium and QX314 in the electrode solution. (f) Burst length (l ) and inter-
burst interval (P ) for the current-clamped pyramidal cell shown in (e), during perfusion with 6 mM [K +]o and after addition of NE/PTX/CGP
and cyclothiazide. Lines represent average values for each condition.

blocked by bath application of 40 M norepinephrine, 100 M
picrotoxin and 1 M CGP55845A, prolonged CA3 discharges
never occurred (Fig. 2e and f), indicating that inhibitory con-
ductances are not necessary for burst termination.
REGULATION OF BURSTS BY SUPPLY OF RELEASABLE GLUTAMATE
We then considered the idea that burst termination is due at
least in part to a decrease in the positive feedback mediated by
glutamatergic recurrent collateral synapses. In computer mod-
els, population bursting is a very sharp function of the strength
of recurrent excitatory synapses6. Synaptic strength is deter-
mined by several pre- and postsynaptic factors; we considered
the effect of the rate of glutamate release from the presynaptic
terminal (RGlu ) on synaptic strength. RGlu is the product of the
number of releasable vesicles (NR) in the presynaptic terminal
multiplied by the probability of release of a releasable vesicle
(PR) (reviewed in ref. 27). PR is considered here to be a contin-
uous function of time whose value is relatively low between
bursts and high during a burst due to the barrage of action
potentials reaching the presynaptic terminal27. As glutamate
vesicles are released at recurrent synapses during a burst, N R
decreases until exhaustion of releasable glutamate results in
synaptic failure (i.e. NR 0 causes RGlu 0). If this occurs at
enough synapses, the strength of recurrent synapses and their

nature neuroscience volume 1 no 3 july 1998 203

 1998 Nature America Inc. http://neurosci.nature.com

articles

positive feedback would be diminished to a

Normalized burst length

the point that bursts are terminated6.
The rate of increase in the ability of the
CA3 network to support a burst was mea-
sured by triggering a second burst at ran-
dom time intervals after a spontaneous
burst in 8.5 mM extracellular potassium
spont
(Fig. 3a and b). The length of the evoked
burst increased in proportion to the time
elapsed since the end of the last sponta-
neous burst with a time constant of 0.6 to Stimulus delay (ms)
2 seconds (n = 12). Burst length was not
zero at the shortest stimulation times at b
least in part because of evoked glutamate
release at synapses that did not participate

Burst length (ms)

in the burst, such as at terminals on per-
forant path axons, mossy fibers, and cut
recurrent collaterals. Although these slices
1998 Nature America Inc. http://neurosci.nature.com

exhibited a wide range of spontaneous spont

burst lengths (60420 ms; compare Fig. 3a
to Fig. 3b), the fractional increase in burst
length followed similar time courses.
Because the afterhyperpolarization and
postsynaptic inhibitory conductances that Stimulus delay (ms)
modulate burst length (Figs 1 and 2) were c
not blocked, the time constants obtained
in this experiment reflect the decay of

Normalized burst length

those conductances as well as the re-accu-
mulation of releasable glutamate. When
the afterhyperpolarization, GABAA, and
GABAB conductances were blocked, the
time constant for the rate of increase in
spont
burst length increased to 7.1 1.9 seconds
(Fig. 3c; n = 7). The time course for the
increase in burst length is similar to the
time course both for successfully inducing
a second population burst by triggering Stimulus delay (ms)
action potentials in a single CA3 pyrami-
dal cell after a spontaneous burst3, and for Fig. 3. Releasable glutamate +
and burst timing. (a) In a CA3 network bursting spontaneously
the replenishment of releasable glutamate in 8.5 mM extracellular K , the length of pyramidal cell bursts evoked by single electrical
at individual synapses in the hippocampal stimuli in s. moleculare is proportional to the time interval between stimulation and the ter-
mination of the preceding spontaneous burst. Resting membrane potential is -63 mV. Left
slice20. These findings support the idea
panel, examples of bursts evoked at the delays indicated (seconds) compared to a sponta-
that depletion of releasable glutamate is a
neous burst (Spont). Right panel, plot of evoked burst length versus post-burst stimulus delay
determinant of burst termination, and that (bars are standard error, n = 2 to 12 for each delay value). Solid line represents the least
the degree of replenishment modulates the squares fit exponential (time constant 600 ms). (b) Extracellular recording from the same
probability of a subsequent discharge. type of experiment as shown in (a). Although the CA3 burst length in this slice is only half as
Alternatively, the strength of recurrent long as in (a), the rate of increase in burst length shows a similar proportionality to the inter-
synapses could be limited by a postsynaptic val since the last burst (time constant of 900 ms). Average interval between spontaneous
mechanism such as desensitization. How- bursts was 5.4 seconds. (c) Extracellular CA3 recording of the same experiment in a slice in
ever, when glutamate receptor desensitiza- which GABAA and GABAB postsynaptic conductances and IAHP were blocked. Evoked burst
tion was inhibited by 100 M length increased with a time constant of seven seconds following a spontaneous burst.
cyclothiazide28, there was only a modest
increase in the length of CA3 bursts record-
ed during blockade of GABAA, GABAB,
and after-hyperpolarization conductances (17 12%; n = 4; Fig. eight cells; Fig. 4ac). However, as in the experiments utilizing
2e and f). To directly evaluate postsynaptic glutamate receptor electrical stimulation (Fig. 3), the later, polysynaptic response was
function, we measured the current induced by glutamate appli- substantially increased at longer post-burst application intervals,
cation to the dendrites of bursting CA3 pyramidal cells. There was indicating an increase in the strength of recurrent synapses.
no correlation between the amplitude of the initial glutamate- These results suggest that during a burst, the strength of recur-
induced current and the time elapsed since the end of the last rent synapses is limited by a presynaptic mechanism. Evaluation of
burst (basal dendrites, n = 4; distal apical dendrites, n = 22; 10 spontaneous excitatory postsynaptic currents recorded before
cells, 9 slices, 6 animals; bursts induced by 8.5 mM external potas- bursting began compared to events recorded between CA3 bursts
sium in two cells, and by tetanization of pyramidal cell layer in induced by 8.5 mM extracellular potassium revealed a much larg-

204 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 4. Glutamate receptor function is

not altered following a burst. (a)
a glutamate b glutamate

Currents evoked by four consecutive 0.6 s

applications of 100 M glutamate (arrow-
head) to the distal apical dendrites of a 100 ms 50 ms
20 pA
pyramidal cell voltage clamped at -50 mV. 10 ms 100 pA
Bursting induced by 8.5 mM extracellular 2s 20 pA
0.2 s
potassium. Intervals between the end of 1.5 s
the preceding burst and glutamate appli-
2.5 s
cation were randomly generated and are 9.4 s 2.6 s
8.0 s
noted next to the current recordings. The
average interval between bursts was 2.3 cC d
D
seconds; thus the glutamate application
mean (%)
100
2.5 seconds after the previous burst trig-
1.0 -
gers a burst, but the initial rate of rise of
from mean
+
50
the current is the same as for the current control
elicited by glutamate application 0.2 sec-

Probability
from

Probability
onds after the last burst. Glutamate was -0
% difference

applied after every tenth spontaneous

Difference

20 ms
1998 Nature America Inc. http://neurosci.nature.com

burst. (b) The experiment shown in (a) -50

1 pA
was repeated in a slice in which CA3 +
control
bursting was induced by tetanic stimula- 0.0 -
-100
tion of the CA3 pyramidal cell layer. The 0 1 2 3 4 5 6 7 8 9 10 0 2 4 6 8
average interburst interval in this prepara- seconds after
Seconds after burst
burst Interevent interval
interval, sec
Interevent (sec)
tion was 13 seconds, which permitted
longer delays between the end of a burst
and glutamate application. As in (a), glutamate application at delays closer to the interburst interval triggered a burst more rapidly, but the ini-
tial response to glutamate was unchanged (inset). Glutamate was applied after every fourth burst in (b) and (c). (c) Summary of initial
response to glutamate application in three experiments in which the average CA3 interburst interval was 10 to 15 seconds. Glutamate
response was calculated as the current averaged over the first 30 ms following glutamate application. Points represent the difference at each
delay value from the mean. There is no significant change in glutamate receptor function during or after CA3 bursts. (d) Cumulative proba-
bility plots of the intervals between spontaneous excitatory postsynaptic currents (EPSCs). EPSCs were recorded for two-minute intervals in
a CA3 pyramidal cell voltage clamped at -50 mV under three conditions: normal extracellular media (control); five minutes after switching to
8.5 mM [K+]o, representing the two-minute interval just prior to the onset of burst activity (-); and five minutes after the onset of bursting in
8.5 mM [K+]o (+). Inset shows averaged EPSCs for each condition. EPSCs were recorded using a potassium electrode solution.

er decrease in frequency versus amplitude (Fig. 3c; n = 4; ref. 29).
This result supports the idea that the mechanism for decreasing
the strength of recurrent excitatory synapses is presynaptic18, but
does not indicate the nature of the mechanism. To evaluate the
possibility that glutamate release is limited by negative feedback via
presynaptic metabotropic glutamate receptors17,19, we blocked
this receptor using the broad-spectrum metabotropic antagonist
MCPG. Although 250 M MCPG produced a 37 11% increase
in the frequency of CA3 bursts induced by 6 mM extracellular
potassium and 100 M picrotoxin, the length of bursts was not
affected (4 2 % increase, n = 8), indicating that metabotropic
receptor feedback was not the critical determinant of the strength
of recurrent synapses during a burst.
To directly assess the supply of releasable glutamate available
at the end of a burst, glutamate release was evoked by applica-
tion of hyperosmotic extracellular fluid (500 mM mannitol) to
the slice30,31. Glutamate release by hyperosmotic media is not
dependent on action potentials or intracellular calcium and is
not affected by known presynaptic inhibitory feedback mecha-
nisms32. Bursts were induced by a tetanic stimulation of the CA3
pyramidal cell layer and recorded during blockade of GABAA and
GABAB postsynaptic conductances and the afterhyperpolariza-
tion. Mannitol substantially increased the frequency of sponta-
neous glutamate release as measured by the frequency of
spontaneous EPSCs (Fig. 5a and b). Despite blocking all inhibito-
ry processes, the EPSC frequency during the two seconds pre-
ceding a burst was 6.5 0.8 times the frequency during the two
seconds immediately following a burst (n = 18 cells in 15 slices
from 13 animals). The time constants describing the rate of
change in EPSC frequency (1.8 to 5 seconds; Fig. 5c) were simi-
lar to the time constants for the change in burst length illustrat-
ed in Fig. 3. The EPSC frequency was not expected to be zero
immediately after a burst because of release of glutamate from
terminals that did not participate in the burst.
REGULATION OF CA3 BURSTING BY PROBABILITY OF RELEASE
These findings support the hypothesis that excitation at CA3
recurrent collateral synapses can be terminated by depletion of
releasable glutamate. When depletion of releasable glutamate
during a burst causes recurrent synaptic strength to fall below
the value at which bursting is supported6, not only is the burst
terminated, but in addition the next burst cannot occur until the
supply of releasable glutamate is replenished. If the probability
of initiating the next burst is proportional to the rate of gluta-
mate release between bursts29, then the next burst requires both
adequate NR so that synapses have the strength to propagate the
burst, and adequate glutamate release (NRPR) to initiate the
burst. Changing the value of PR between bursts would then lead
to a proportionate change in the degree to which NR needed to
recover in order for glutamate release to be sufficient to trigger
another burst; the change in the recovery of NR would be reflect-
ed in the interburst interval.
Thus agents that decrease PR such as adenosine and baclofen33
should increase the interval between bursts34,35, whereas increas-

nature neuroscience volume 1 no 3 july 1998 205

 1998 Nature America Inc. http://neurosci.nature.com

articles

ing PR by increasing external potassium23 should decrease the
interburst interval (Fig. 6ac). The burst length changed in pro-
portion to the burst interval under these conditions, supporting
the idea that the amount of glutamate available for release is
increasing throughout the interburst interval. The effects of these
manipulations are most consistent with a presynaptic locus of
action, as they are the opposite of what would be expected based
on the postsynaptic effects of these manipulations (for instance,
increasing potassium extracellularly would decrease inhibition
due to the decrease the driving force for inhibitory currents36,
and thus would tend to increase burst length as picrotoxin does in
Fig. 2d). By measuring the rate of increase in evoked burst length
with time elapsed from the last burst as in Fig. 3, the
effects of baclofen were shown to be mediated by a aA
decrease in P R rather than a change in the rate of
feedback inhibition should be proportional to the intensity of the
burst, but the opposite relationship is found (Fig. 6e). However, if
the duration of a CA3 population burst is limited by the amount
of releasable glutamate at recurrent synapses, then any factor that
changes the postsynaptic effect of glutamate should affect burst
duration. Thus, although we hypothesize that inhibitory conduc-
tances are neither necessary nor sufficient for burst termination,
they will substantially alter the excitatory effect of released gluta-
mate10. Inhibitory conductances therefore can further shorten a
burst that is limited by the supply of releasable glutamate (Fig. 2cf).
The length of CA3 population bursts described in the litera-
ture varies substantially13,3740. Much of this variation can be

control
Burst

Control
burst
increase of NR (Fig. 6d). Over the wide range of burst
lengths and intervals in the baclofen experiments, burst
length did not predict the time interval until the next
1998 Nature America Inc. http://neurosci.nature.com

burst, but the length of the interval preceding a burst

was correlated with burst length (Fig. 6e). This pro-
vides further evidence for the idea that burst termina-
tion is the result of depletion of an excitatory factor
rather than accumulation of an inhibitory factor.
Mannitol
mannitol

Discussion
During synchronous CA3 population bursts in vitro,
the finite supply of releasable neurotransmitter lim-
its the strength of recurrent synapses. Because pop-
ulation bursting is strongly dependent on the strength
of these positive-feedback synapses, the availability
of releasable glutamate effectively controls the dura-
tion of population bursts. The interval between bursts 500 ms
reflects the replenishment of releasable glutamate. 50 pA
This replenishment increases both the strength of
recurrent synapses and the rate of glutamate release,
bB C
1.0 mannitol 5
which contributes to the initiation of the next burst29,
as schematized in Fig. 7.
(Hz)

4
frequency, Hz

Although we have emphasized burst termination by

EPSCfrequency
Probability

exhaustion of releasable glutamate, inhibitory conduc-

probability

control 3
tances also decrease burst duration (Figs 1b, 2df and
0.5
3c). How do these mechanisms interact to terminate
2
bursts? The experiments in which inhibitory conduc-
EPSC

tances were blocked demonstrate that exhaustion of

1
releasable glutamate is sufficient to terminate bursts (Fig.
2cf). Whether exhaustion of releasable glutamate is also
0.0 0
necessary for burst termination can not be proved as 0 200 400 600 800 1000 0 1 2 3 4 5 6 7 8 9 10
directly, because it is not possible to prevent this exhaus- interevent interval,
Interevent intervalms
(ms) seconds after
Seconds afterend of burst
end of burst
tion. However, three lines of evidence support the
hypothesis that depletion of releasable glutamate is nec-
essary for burst termination, as opposed to providing a Fig. 5. Amount of releasable glutamate available at the end of a burst. (a)
fail-safe mechanism should inhibition fail. First, inhibi- Spontaneous EPSCs recorded in a CA3 pyramidal cell before and after sponta-
neous bursts (bursts are truncated). Bursting was triggered by a tetanic stimula-
tion does not increase at the end of a burst in a manner
tion of the CA3 pyramidal layer; GABAA and GABAB postsynaptic conductances
consistent with a burst termination mechanism10 (Fig.
and IAHP were blocked. Top two records are consecutive bursts that were
1). Second, when extracellular potassium is increased, recorded prior to mannitol application (control); bottom ten records are con-
the driving forces for inhibitory chloride and potassium secutive bursts recorded after application of 500 mM mannitol to the distal api-
currents are decreased36. If inhibition terminated bursts cal dendritic layer of CA3. Following a burst, the EPSC frequency is depressed.
independently of the supply of releasable glutamate, then Average interburst interval was ten seconds. Holding potential was 70 mV,
the decrement in inhibition caused by increased potas- records were low-pass filtered at 1 kHz, and the electrode solution contained
sium should increase burst length, but over a very wide cesium and QX314. (b) Cumulative probability plots of EPSC intervals demon-
range of potassium concentrations, this does not occur strate that EPSCs are increased during application of mannitol. (c) EPSC fre-
(Fig. 6a and b). Finally, if bursts were terminated by feed- quency as a function of time elapsed since the end of the last spontaneous burst.
back inhibition, then the interval following a burst would EPSCs were binned in one-second intervals. Solid line, fit exponential function,
be proportional to burst length, because the amount of time constant, 1.8 seconds.

206 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 6. Presynaptic modulation of CA3 population bursts.

a A
b B (a, b) Effect of [K+]o on CA3 burst frequency and burst
length. Burst interval and length decreased as [K+]o was

(sec)
(sec)

20 3.5 500 25 150

interval

interval, sec
500 ms increased, consistent with an increase in PR and a decrease
interval, sec

ms
length

length

burst length, ms
20

Burst length
interburstinterval
interburst interval

15 2 mV in the degree to which NR had recovered at the start of the

length,
8.5

(ms)

(ms)
15 burst. Points are averaged over all bursts acquired during

Burst
10 30-minute recording periods; bars are standard error.
10

burst

Interburst
Interburst

interval length Experiment in (a) was performed in a transverse brain slice,

5 5
250
and (b) in a coronal brain slice; note the difference in the
0 110 range of burst lengths. (c) Doseresponse relationships for
3 6 9 5 6 7 8 9 10
the decrease in burst frequency and increase in burst length
+
KK+oo (mM)
, mM K+Ko+(mM)
o mM
with baclofen. The length and interval of CA3 population
cC Dd bursts induced by 8.5 mM [K+ ]o increased as baclofen was
(sec)

30 110 120 applied by bath at the indicated concentrations; bursting

interval
stopped in 8 M baclofen. Picrotoxin (100 M) was used in
interval, sec

ms
msms

length
length, ms
interburstinterval

all baclofen and [K+]o experiments to avoid confounding

burst length,
length,

20
effects due to decreased GABA release and altered Cl gra-
burst length,

dients. (d) The decreased burst frequency in baclofen is not

Interburst

burst
burst

10 due to a change in the rate of increase of NR. CA3 popula-

1998 Nature America Inc. http://neurosci.nature.com

tion bursts were induced by 8.5 mM [K+]o, and bursts were

70 50 evoked after every fourth spontaneous burst as described
0 1 2 3 4 0 1 2 3 4 5 6
in Fig. 2b. Under control conditions () and in baclofen
M
M baclofen
baclofen Stimulus
stimulusdelay
delay,(sec)
sec
e 2 M (o), the time constant describing the increase in burst
E f length at increased stimulus delays was essentially the same
1.0 1.0
(solid line). Longer delays could be included in baclofen
interval
burst interval

because the spontaneous burst interval was increased. (e,f)

burstlength
length

The duration of a burst does not predict the time interval

0.5
nextburst

0.5 until the next burst (e) but the interval preceding a burst
Burst

predicts the burst duration (f; compare to d). Using the

Next

data shown in averaged form in (c), normalized time inter-

0.0 0.0
vals and burst lengths are plotted for each concentration of
0.0 0.5 1.0 0.0 0.5 1.0 baclofen (control P , baclofen 1 M p , baclofen 2 M l ,
burst length
Burst length burst
Burst interval
interval baclofen 4 M L ).

explained by differences in PR produced by the experimental pro- vivo is regulated not only by the supply of releasable glutamate
tocols used to induce bursting. The differences in PR will lead to but also by another presynaptic mechanism, synapse-specific vari-
proportional differences in NR at the start of the burst, and burst ation in PR, which may be enhanced by neuromodulators17,33.
duration is limited by NR (Figs 3, 6 and 7). Some of the variation Whether or not the supply of releasable transmitter affects other
in CA3 burst length, for instance the difference noted between modes of network activity, for instance those in which inhibition
septal and temporal slices (compare Fig. 6a and b), is not easily alternates with excitation48,49, will depend on the rate of trans-
explained by differences in PR. This suggests that there may be mitter release, the number of release sites, and the rate of replen-
septo-temporal differences in the number of glutamate release ishment of releasable transmitter during the inhibitory periods.
sites or in the factors that modulate the effect of released gluta- Termination of network activation by exhaustion of releasable
mate, such as glutamate receptor desensitization or re-uptake glutamate rather than postsynaptic inhibition increases the com-
characteristics41. Another potential explanation for the septo-tem- putational flexibility of the network, because network activity is not
poral difference is that the interneuronal connectivity of tempo- constrained by a progressive dampening as inhibitory conductances
ral transverse slices is maximized by the plane of section. This terminate the burst. The causal link between the properties of indi-
connectivity makes possible the prolonged driving of CA3 by other vidual synapses such as PR and NR and the behavior of the network
networks in which epileptiform discharges are induced4244. suggests that the plasticity of synaptic properties should also be
Prolonged CA3 discharges that are not driven by other net- reflected in the behavior of the network. For instance, tetanization of
works35,45,46 consist of a primary burst followed by a series of hippocampal afferents results in an increase in the strength of the
brief bursts at short intervals, which are termed afterdischarges47. activated synapses31 as well as an alteration in the modes of network
Although our data do not address afterdischarges explicitly, the activity43. Regulation of CA3 burst duration by the supply and prob-
relationship between afterdischarge length and the interval since ability of release of neurotransmitter is a novel mechanism for con-
the last discharge seems to follow the relationship shown in Figs trol of a neural network that presents new avenues for understanding
3 and 7, suggesting that afterdischarge duration and interval are the physiology of neurotransmitter release as well as new strategies
also governed by the supply of releasable glutamate. If so, then for the treatment of epilepsy.
series of afterdischarges may represent burst behavior under con-
ditions in which the interburst probability of release remains very Methods
high at the end of the initial burst, for instance as a consequence The duration of reported CA3 bursts varies along the septo-temporal
of prolonged depolarizing conductances47. axis37 and with the method used to induce bursting6,7,1214,3840. In order
In CA3 in vivo, not all pyramidal cells fire during a sharp to ascertain the range of applicability of the present findings, CA3 bursts
wave4,5. This suggests that the activation of the CA3 network in were studied in both septal and ventral CA3, and bursts were induced

nature neuroscience volume 1 no 3 july 1998 207

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 7. Schematic of proposed

presynaptic modulation of burst ter-
a P = 0.25
A
bB cC
P RR0= 0.25 PPR = =.32
0.32
max
max 0 R
0 0
mination and timing. (a, b) The pat-
tern of bursting if the rate of 8 120
glutamate release from the presy-

interval, secsec

ms ms
naptic terminal at recurrent

NRR xPPRR
6 length

interburstinterval,

length,
synapses limits synaptic strength at

Burst length,
these synapses. Glutamate release is

N
burst 4
calculated as the product of PR x

burst
threshold

interburst
NR. The value of NR x PR at which a 2 interval
population burst is likely to be trig- min
min
gered29 and propagated6 is desig- 20
0
nated burst threshold. Bursts stop
max
max
when PR x NR falls below this min PR max

threshold; bursts are possible when

PR x NR exceed threshold. For sim-
NRR

6 100
N

sec sec
plicity we assume bursts occur as

ms ms
interval,
soon as they are possible. In (a) and
min

length,
min

interburst interval,
4
(b), NR x PR, NR, and PR are plotted length
1998 Nature America Inc. http://neurosci.nature.com

burst length,
for 2 different interburst values of 1

interburst
PR (interburst PR = 0.25 in a and 2

burst
0.32 in b). Between bursts, NR interval
R
PR

increases at a rate proportional to

P

0 0
the number of empty release sites
min rate
min rate ofofNNR Rincrease
increase max
max
remaining, so that dNR / dT = krefill * 0
(NRmax - NR), and NR = NRmax x (1- 2 4 6 8 2 4 6 8
max
max burst threshold
burst threshold min
min
e (t / ) ), where = 1 / krefill . NR seconds
seconds seconds
seconds
decreases during a burst at a rate
proportional to the number of releasable vesicles25, so that during a burst dNR / dT = - kdecrease *Nr, and NR = NRmax x e (t / ), where = 1
/ kdecrease. PR between bursts is assumed to have been increased by an experimental manipulation such as increased extracellular [K+]o.
During a burst, Pr is assumed to increase further because of bursts of action potentials in the presynaptic cell. (c) Top panel, the relationships
between interburst PR, burst length, and interburst interval. Burst interval decreases as PR increases because for a given rate of increase in
NR after a burst, NR x PR reaches burst threshold more rapidly for larger values of PR. Burst length decreases as PR increases because NR
does not recover to a very high value before PR x NR exceeds burst threshold, and so there are fewer releasable glutamate vesicles at the
start of the next burst. Bottom panel, the rate at which N R increases between bursts determines when NR x PR reaches burst threshold and
therefore sets the interburst interval. Burst length is not affected, because NR at the start of the burst is not changed. Increasing the burst
threshold increases the burst interval, because in order for NR x PR to cross burst threshold, NR must increase to closer to its maximum
value, where the rate of increase in NR is low (a, middle panel). However, increasing the burst threshold does not change the amount by
which NR decreases during the burst: although N R is higher at the start of the burst, the burst also terminates at a higher value of N R because
NR x PR drops below the increased burst threshold sooner.

using three different manipulations: increased extracellular potassium
([K+]o), GABAA receptor block, and tetanic stimulation of the CA3 pyra-
midal layer43,44.
S LICE PREPARATION . Recordings from area CA3 of the adult rat hip-
pocampus were made from 400-m thick hemibrain slices at 35C. The
brain was cut in the coronal plane except where transverse plane is spec-
ified. For both planes of section, recordings were made from slices per-
pendicular to the long axis of the hippocampus. Therefore, slices from
the septal (rostral) end of the hippocampus were used when the brain
was cut in the coronal plane, and from the temporal (caudal) end when
the brain was sliced transversely. Extracellular solution was saturated
with 95% O 2, 5% CO 2 and contained (in mM) NaCl 126, KCl 2.5,
NaHCO3 26, CaCl2 2, MgCl2 2, NaH2PO4 1.25, glucose 10.
RECORDINGS. Whole-cell recordings were performed using a filling solu-
tion containing (in mM) potassium methylsulfonate 123, MgCl2 2, NaCl 8,
potassium ethylene glycol-bis(b-aminoethyl ether) N,N,N,N-tetraacteic
acid (EGTA) 1, potassium adenosine 5-triphosphate 4, and sodium guano-
sine 5-triphosphate 0.3. The whole-cell solution was buffered with 16 mM
KHCO3 and saturated with 95% O2, 5% CO2. In voltage-clamp experi-
ments, 2 mM QX314 (Astra Pharmaceuticals) was added to the pipette
solution to block voltage-dependent sodium conductances, and where
noted cesium replaced potassium to improve the space clamp. Extracellu-
lar recordings were performed using saline-filled patch pipettes. Record-
ings were accepted when the access resistance remained under 10 M for
voltage-clamp experiments, and under 20 M in experiments where no
current was injected. Recordings using Axoclamp 2B amplifiers were dig-
itized at 2 KHz using routines written in Axobasic (Axon Instruments, Fos-
ter City, California). Burst length rather than burst area was used to estimate
burst intensity to facilitate comparisons between extra and intracellular
recordings (Fig. 3). Burst length was calculated as the time during which the
absolute value of the burst was above a threshold value, generally three
times the baseline noise. Similarly, for stimuli applied after a burst, the end
of a spontaneous burst was defined as 50 ms after the time at which the
burst amplitude was less than a defined value, generally three times the
baseline noise. Intervals between the end of a burst and the stimulus were
randomly generated, then rounded to the nearest 200 ms. Electrical stim-
uli were applied after every fourth burst to the pyramidal cell layer, and
the intensity was set to elicit one population spike in control media.
BURST INDUCTION. Bursting in CA3 was induced by increasing extracellular
potassium to between 6 and 8.5 mM as noted in the text, by blockade of
GABAA inhibition with 100 M picrotoxin or by a single tetanic stimula-
tion43 of the CA3 pyramidal cell layer. The tetanic stimulus consisted of a
100 Hz, one-second train of stimuli that were of sufficient amplitude to
elicit a population spike when delivered at a lower frequency. If a single
tetanus did not induce bursting, it was repeated after a ten-minute inter-
val43,44. When inducing bursting by tetanic stimulation, extracellular solu-
tions were modified43 to Ca2+ 1.3 mM, Mg2+ 0.9 mM, and K+ 3.3 mM.

208 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com
GLUTAMATE AND MANNITOL APPLICATION. Glutamate and mannitol were
applied via a whole-cell pipette using a 40 psi pressure pulse lasting 5 to
50 ms. Intervals between the end of a burst and glutamate application
were randomized to a value between 200 ms and the average interburst
interval. Mannitol was applied every 4 to 30 seconds as described31.
REAGENTS. Drugs were applied by bath. Appropriate precautions were
taken to avoid oxidation of norepinephrine26. Where noted, afterhyper-
polarization was blocked by addition of 40 M norepinephrine, GABAA
conductance by 100 M picrotoxin, GABA B conductance by 1 M
CGP55845A (Ciba Geigy, Basel), and glutamate-receptor desensitization
by 100 M cyclothiazide to the extracellular fluid. Reagents were obtained
from Sigma (St. Louis, Missouri). Ten micromolar acetazolamide was
applied with pentobarbital to diminish GABAA-receptor-mediated den-
dritic depolarization50.
Acknowledgements
We thank Thomas Dunwiddie, Darrell Lewis, Charles Stevens, and Roger Traub
for comments and discussions. This work was supported by the NIH and the
1998 Nature America Inc. http://neurosci.nature.com
Epilepsy Foundation of America.
RECEIVED 16 APRIL: ACCEPTED 7 MAY 1998
1. Li, X.G., Somogyi, P., Ylinen, A. & Buzsaki, G. The hippocampal CA3
network: an in vivo intracellular labeling study. J. Comp. Neurol. 339,
181208 (1994).
2. MacVicar, B.A. & Dudek, F.E. Local circuit interactions in rat hippocampus:
interactions between pyramidal cells. Brain Res. 242, 341344 (1982).
3. Miles, R. & Wong, R.K.S. Single neurones can initiate synchronized
population discharge in the hippocampus. Nature 306, 371373 (1983).
4. Buzsaki, G. Hippocampal sharp waves: their origin and significance. Brain
Res. 398:242252 (1986).
5. Chrobak, J.J. & Buzsaki, G. High-frequency oscillations in the output
networks of the hippocampal-entorhinal axis of the freely behaving rat. J.
Neurosci. 16, 30563066 (1996).
6. Traub, R.D. & Miles R. in Neuronal networks of the hippocampus Ch. 6
(Cambridge Univ. Press, 1991).
7. Traub, R.D., Knowles, W.D., Miles, R. & Wong, R.K.S. Synchronized
afterdischarges in the hippocampus: simulation studies of the cellular
mechanism. Neuroscience 12, 11911200 (1984).
8. Jefferys, J.G.R. in The Epilepsies, 4th edn (eds Laidlaw, J., Richens, A. &
Chadwick, D.W.) 241276 (Edinburgh, Churchill Livingstone, 1993).
9. Wong, R.K.S. & Prince, D.A. Afterpotential generation in hippocampal
pyramidal cells. J. Neurophysiol. 45, 8697 (1981).
10. Bush, P.C. & Sejnowski, T.J. Effects of inhibition and dendritic saturation in
simulated neocortical pyramidal cells. J. Neurophysiol. 71, 21832193 (1994).
11. Miles, R., Wong, R.K.S. & Traub, R.D. Synchronized afterdischarges in the
hippocampus: contribution of local synaptic interactions. Neuroscience 12,
11791189 (1984).
12. Rutecki, P.A., Lebeda, F.J. & Johnston, D. Epileptiform activity in the
hippocampus produced by tetraethylammonium. J. Neurophysiol. 64,
10771088 (1990).
13. Mueller, A.L. & Dunwiddie, T.V. Anticonvulsant and proconvulsant actions
of alpha- and beta-noradrenergic agonists on epileptiform activity in rat
hippocampus in vitro. Epilepsia 24, 5764 (1983).
14. Chamberlin, N.L. & Dingledine, R. Control of epileptiform burst rate by CA3
hippocampal cell afterhyperpolarizations. Brain Res. 492, 337346 (1989).
15. Mayr, O. The origins of feedback control (MIT Press, Cambridge,
Massachusetts, 1970).
16. Martina, M. & Jonas, P. Functional differences in Na+ channel gating
between fast-spiking interneurones and principal neurones of rat
hippocampus. J. Physiol. 505, 593603 (1997).
17. Scanziani, M. et al. Use-dependent increases in glutamate concentration activate
presynaptic metabotropic glutamate receptors. Nature 385, 630634 (1997).
18. Otis, T., Zhang, S. & Trussell, L.O. Direct measurement of AMPA receptor
desensitization induced by glutamatergic synaptic transmission. J. Neurosci.
16, 74967504 (1996).
19. von Gersdorff, H., Schneggenburger, R., Weis, S. & Neher, E. Presynaptic
depression at a calyx synapse: the small contribution of metabotropic
glutamate receptors. J. Neurosci. 17, 81378146 (1997).
20. Dobrunz, L.E. & Stevens, C.F. Heterogeneity of release probability,
facilitation, and depletion at central synapses. Neuron 18, 9951008 (1997).
nature neuroscience volume 1 no 3 july 1998
articles
21. Harris, K.M. & Sultan, P. Variation in the number, location, and size of
synaptic vesicles provides an anatomical basis for the nonuniform
probability of release at hippocampal CA1 synapses. Neuropharmacology 34,
13871395 (1995).
22. Schikorski, T. & Stevens, C.F. Quantitative ultrastructural analysis of
hippocampal excitatory synapses. J. Neurosci. 17, 58585867 (1997).
23. Liu, G. & Tsien, R.W. Properties of synaptic transmission at single
hippocampal synaptic boutons. Nature 375, 404408 (1995).
24. Stevens, C.F. & Tsujimoto, T. Estimates for the pool size of releasable quanta
at a single central synapse and for the time required to fill the pool. Proc. Natl
Acad. Sci. USA 92, 846849 (1995).
25. Murthy, V.N., Sejnowski, T.J. & Stevens, C.F. Heterogenous release properties
of visualized individual hippocampal synapses. Neuron 18, 599612 (1997).
26. Madison, D.V. & Nicoll, R.A. Noradrenaline blocks accommodation of
pyramidal cell discharge in the hippocampus. Nature 299, 636638 (1986).
27. Stevens, C.F. Quantal release of neurotransmitter and long-term
potentiation. Neuron 10(suppl), 5563 (1996).
28. Patneau D.K., Vyklicky L. Jr & Mayer M.L. Hippocampal neurons exhibit
cyclothiazide-sensitive rapidly desensitizing responses to kainate. J Neurosci.
13, 34963509 (1993).
29. Chamberlin, N.L., Traub, R.D. & Dingledine R. Role of EPSPs in initiation of
spontaneous synchronized burst firing in rat hippocampal neurons bathed in
high potassium. J. Neurophysiol. 64, 10001008 (1990).
30. Fatt, P. & Katz, B. Spontaneous subthreshold activity at motor nerve endings.
J. Physiol. 117, 109128 (1952).
31. Manabe T., Renner, P. & Nicoll, R.A. Postsynaptic contribution to long-term
potentiation revealed by the analysis of miniature synaptic currents. Nature
355, 5055 (1992).
32. Rosenmund, C. & Stevens, C.F. Definition of the readily releasable pool of
vesicles at hippocampal synapses. Neuron 16, 11971207 (1996).
33. Capogna, M., Ghwiler, B.H. & Thompson, S.M. Presynaptic inhibition of
calcium-dependent and independent release elicited with ionmycin,
gadolinium, and -latrotoxin in the hippocampus. J. Neurophysiol. 75,
20172028 (1996).
34. Dunwiddie, T.V. Endogenously released adenosine regulates excitability in
the in vitro hippocampus. Epilepsia 21, 541548 (1980).
35. Swartzwelder, H.S., Lewis, D.V., Anderson, W.W. & Wilson, W.A. Seizure-like
events in brain slices: suppression by interictal activity. Brain Res. 410,
362366 (1987).
36. Thompson, S.M. & Ghwiler, B.H. Activity-dependent disinhibition. II.
effects of extracellular potassium, furosemide, and membrane potential on
ECl in hippocampal CA3 neurons. J. Neurophysiol. 61, 512523 (1989).
37. Anderson, C.M. & Jefferys, J.G.R. Ventral CA3 exhibits a lower threshold to
disinhibition than dorsal hippocampus. Epilepsia 38(Supp 8), 16 (1997).
38. Korn, S.J., Giacchino, J.L., Chamberlin, N.L. & Dingledine, R. Epileptiform
burst activity induced by potassium in the hippocampus and its regulation by
GABA-mediated inhibition. J. Neurophysiol. 57, 325340 (1987).
39. Scharfman, H.E. EPSPs of dentate gyrus granule cells during epileptiform
bursts of dentate hilar mossy cells and area CA3 pyramidal cells in
disinhibited rat hippocampal slices. J. Neurosci. 14, 60416057 (1994).
40. Whittington, M.A. & Jefferys, J.G.R. Epileptic activity outlasts disinhibition
after intrahippocampal tetanus toxin in the rat. J. Physiol. 481, 593604
(1994).
41. Tanaka, K. et al. Epilepsy and exacerbation of brain injury in mice lacking the
glutamate transporter GLT-1. Science 276, 16991702 (1997).
42. Barbarosie, M. & Avoli, M. CA3-driven hippocampal-entorhinal loop
controls rather than sustains in vitro limbic seizures. J. Neurosci. 17,
93089314 (1997).
43. Stasheff, S.F., Anderson, W.W., Clark, S. & Wilson, W.A. NMDA antagonists
differentiate epileptogenesis from seizure expression in an in vitro model.
Science 245, 648651 (1989).
44. Rafiq, A., DeLorenzo, R.J. & Coulter, D.A. Generation and propagation of
epileptiform discharges in a combined entorhinal cortex/hippocampal slice.
J. Neurophysiol. 70, 19621974 (1993).
45. Traub, R.D., Borck, C., Colling, S.B. & Jefferys J.G. On the structure of ictal
events in vitro. Epilepsia 37, 879891 (1996).
46. Taylor, G.W., Merlin, L.R. & Wong, R.K. Synchronized oscillations in
hippocampal CA3 neurons induced by metabotropic glutamate receptor
activation. J. Neurosci. 15, 80398052 (1995).
47. Traub R.D., Miles, R. & Jefferys, J.G. Synaptic and intrinsic conductances
shape picrotoxin-induced synchronized after-discharges in the guinea-pig
hippocampal slice. J Physiol. (Lond.) 461, 525547 (1993).
48. Ylinen, A. et al. Intracellular correlates of hippocampal theta rhythm in
identified pyramidal cells, granule cells, and basket cells. Hippocampus 5,
7890 (1995).
49. Kim, U., Sanchez-Vives, M.V. & McCormick, D.A. Functional dynamics of
GABAergic inhibition in the thalamus. Science 278, 130134 (1997).
50. Staley, K.J., Soldo, B. & Proctor W. Ionic mechanisms of neuronal excitation
by inhibitory GABAA receptors. Science 269, 977981 (1995).
209
 1998 Nature America Inc. http://neurosci.nature.com

articles

Input synchrony and the irregular

firing of cortical neurons
Charles F. Stevens and Anthony M. Zador

Howard Hughes Medical Institute and Sloan Center for Theoretical Neurobiology, Salk Institute for Biological Studies, La Jolla, California 92037, USA
Correspondence should be addressed to A.Z. (zador@salk.edu)

Cortical neurons in the waking brain fire highly irregular, seemingly random, spike trains in response
to constant sensory stimulation, whereas in vitro they fire regularly in response to constant current
injection. To test whether, as has been suggested, this high in vivo variability could be due to the
postsynaptic currents generated by independent synaptic inputs, we injected synthetic synaptic cur-
rent into neocortical neurons in brain slices. We report that independent inputs cannot account for
1998 Nature America Inc. http://neurosci.nature.com

this high variability, but this variability can be explained by a simple alternative model of the
synaptic drive in which inputs arrive synchronously. Our results suggest that synchrony may be
important in the neural code by providing a means for encoding signals with high temporal fidelity
over a population of neurons.

Cortical neurons in the waking brain fire highly irregular spike
trains that have more in common with the ticking of a Geiger
counter than of a clock. What is the source of this irregular fir-
ing? Softky and Koch1 noted a theoretical conundrum posed by
the irregularity of cortical neurons firing at a constant average
rate in vivo. If each cortical neuron were providing independent
input to the other similar cortical neurons it contacts, then the
input to any neuron would just be a shower of statistically inde-
pendent excitatory postsynaptic potentials (EPSPs) with a con-
stant mean rate. Yet such an input, when tested on a theoretical
model of a cortical neuron, gave rise to a firing variability much
less than that observed in vivo. What, then, is the source of the
unexpectedly large variability in spike timing that characterizes
in vivo firing behavior? Is the high variability due to some sub-
tle interplay between the noise resulting from randomly timed
synaptic inputs and the nonlinear spike-generating mechanism,
or instead due to unexpectedly large fluctuations in the synaptic
input itself? Although most earlier workers have argued that the
large variability arises from the interaction between this noisy
synaptic input and the spike-generating mechanism, we will con-
clude that large fluctuations in the synaptic drive, such as might
arise from the synchronous arrival of inputs from many neu-
rons, are necessary to account for the high in vivo variability.
The neuronal spike generator converts input current into
output spike trains with high fidelity2,3. Irregular firing must,
then, reflect fluctuations in the currents that drive the spike gen-
erator, rather than some intrinsic noise in the spike-generating
mechanism itself4. Most previous attempts to identify the source
of irregular firing have focused on the details of the spike-gen-
erating mechanism and on the importance of the inhibitory
synaptic noise57 (Tsodyks et al., Soc. Neurosci. Abstr., 20, 1527,
1994). These workers have concluded that unstructured synap-
tic noise processed by a model spike generator can produce the
unexpected variability in spike timing. However, most of these
proposals have relied on theoretical models of the neuronal spike
generator (but see ref. 8), and thus one may question the extent
to which such models can be relied upon to provide a sufficiently
realistic representation of spike generation.
We therefore set out to determine experimentally whether
the high variability observed in vivo could arise from the
superposition of independent excitatory and inhibitory inputs.
Because reconciliation of the in vivo variability with the synap-
tic drive depends critically on the details of the spike genera-
tion mechanism, we have used a direct approach rather than
rely on theoretical models of the spike generator. We have
injected synthetic synaptic currents through a somatic elec-
trode to drive neurons in neocortical slices to fire. With this
method, we can assess the output variability in response to
any input ensemble. We have complemented this approach by
testing the response to miniature excitatory postsynaptic cur-
rents (EPSCs) whose rate of release was elevated at many
synapses independently through the local application of
hypertonic solution.
We find that when neocortical neurons are driven by a pop-
ulation of independent inputs, the spike variability is consis-
tently lower than that observed in vivo. We thus cannot confirm
the earlier conclusions that unstructured synaptic noise inter-
acting with the spike-generating mechanism can account for
the unexpectedly large variability of neuronal activity in the
neocortex. However, when a population of transiently syn-
chronous inputs is added to the background of independent
inputs, the observed firing variability is within the range
observed in vivo. The high variability observed in vivo is there-
fore inconsistent with the activity of independent excitatory
and inhibitory inputs, but could arise from large rapid fluctu-
ations in the synaptic drive, such as would result from the near-
ly synchronous firing of subpopulations of afferents.
Results
The results are organized as follows. First we show that, in
response to a synthetic synaptic input corresponding to an
ensemble of purely excitatory inputs firing at a constant rate,
the coefficient of variation (CV) of the interspike interval dis-
tribution is well below that observed in vivo. Next we show
that CV increases but remains below the in vivo level when
the input consists of a steady mix of excitatory and inhibitory

210 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

a b 0.5 (mean standard devia-

tion) and a skewed distri-
0.4 bution (Fig. 1b). Under
slice recording conditions,

Probability
Probability
0.3 this spontaneous release
rate was too low to cause
0.2 fluctuations of more than
about a millivolt (Fig. 1c),
0.1 and was never sufficient to
cause spontaneous spiking.
0 We therefore generated
0 5 10 15 20 25
Amplitude(pA)(pA) synthetic synaptic currents
Amplitude
to test whether such EPSCs,
arriving at a sufficiently
c high rate, could account for
Fig. 1. Spontaneous synaptic event recorded in a
the variability observed in
layer 2/3 cortical neuron. (a) A short record
vivo. In order to simulate
showing EPSCs recorded at the soma under volt-
this drive in a neocortical
1998 Nature America Inc. http://neurosci.nature.com

age clamp at a holding potential of -60 mV, in the

presence of tetrodotoxin (TTX). Calibration, 40
ms, 30 pA. (b) The distribution of spontaneous
miniature EPSCs in this neuron. Inset, a simulated
scaled mEPSC is superimposed on a typical
mEPSC from (a). (c) A short record from the
same neuron showing EPSPs. Calibration, 40 ms,
0.5 mV.
inputs. We then repeat the analysis of the response variability
for a complementary measure, the Fano factor. We then con-
firm that the spiking variability in response to miniature
EPSCs evoked by hypertonic solution is still below the in vivo
levels. Finally, we demonstrate that input synchrony can yield
in vivo levels of variability.
PURELY EXCITATORY INPUT
Cortical neurons in vivo are driven to fire by a shower of EPSCs.
An estimate of the size of the unitary EPSCs that comprise this
shower was obtained by recording spontaneous EPSCs from
pyramidal neurons in rat neocortical slices under voltage-clamp
conditions (Fig. 1a). As previously observed911, spontaneous
EPSCs varied in size, with a mean amplitude of 6.4 3.4 pA
slice, synthetic synaptic cur-
rents were generated online
and injected through a
somatic patch electrode.
Chemical synaptic inputs
were pharmacologically
silenced, so that all the drive
was supplied by the elec-
trode. The synthetic cur-
rents were constructed to
mimic the current that
would be generated at the
soma by a population of excitatory inputs, each firing indepen-
dently according to a Poisson process with a constant mean rate.
These currents have the advantage that the synthetic synaptic noise
can be created with known statistical structure. The unitary event
used to construct the synthetic synaptic inputs, 30 pA, was chosen
to be at the extreme high end of the observed range. Because the
sum of independent Poisson processes is itself a Poisson process,
the input statistics were determined solely by the net rate at which
excitatory inputs showered onto the soma. We made no explicit
assumption about the number of synapses or the rate of neuro-
transmitter release from each synapse, so that, for example, 100
synapses each generating 20 EPSCs per second would be indis-
tinguishable from 200 synapses generating 10 EPSCs per second.
Figure 2a shows the results of a typical experiment in which

25
a b
Fig. 2. Fluctuating currents affect the
20
fine structure of the spike train but
(Hz)

not the mean rate. (a) Sample traces

rate (Hz)

15
show typical responses of a layer 2/3
Firingrate

cortical neuron to a constant current

Firing

(left), and a fluctuating current (right) 10

consisting of the sum of independent

Poisson EPSCs firing at a population 5

rate of 2.4 per ms. Calibration, 200

ms, 10 mV, 0.3 nA. (b) The average 0
0 0.05 0.1 0.15 0.2 0.25
number of spikes in a one second Input current
Input (nA)
current (nA)
trial is plotted as a function of the
mean input for the constant (o) and fluctuating (*) currents. Except at very c
low firing rates, the mean spike rate depends only on the mean of the input
current. (c) An example of a spike train recorded extracellularly from area MT
of an alert macaque monkey in response to a constant-velocity visual stimulus
(for details, see ref. 16). Time (ms)

nature neuroscience volume 1 no 3 july 1998 211

 1998 Nature America Inc. http://neurosci.nature.com

articles

1
a b
0.8

distribution
Distribution
0.6

of ISI
0.4

of ISI
C.V.C.V.
0.2

0
0 0.2 0.4 0.6 0.8 1
Ratio
Ratio of of inhibitory to
inhibitory to excitatory
excitatory drive
drive

1
c
0.8
Fig. 3. Variability in response to mixed excitatory
1998 Nature America Inc. http://neurosci.nature.com

and inhibitory input is less than in vivo. (a) Response

Fano factor
0.6
Fano factor

to mixed excitatory and inhibitory input. Typical

responses of a layer 2/3 cortical neuron to a fluctuat-
0.4
ing current (Ri = 0.5) consisting of the sum of a mix of
independent Poisson EPSCs (4.4 per ms) and IPSCs
(2.2 per ms). Calibration, 200 ms, 10 mV, 0.3 nA. (b) 0.2
Dependence of CV, the coefficient of variation of the
interspike interval distribution, on the ratio Ri of inhi- 0
0 0.2 0.4 0.6 0.8 1
bition to excitation. The CV increases with Ri, but
RatioRatio of inhibitoryto
of inhibitory toexcitatory
excitatory drive
drive
remains below the in vivo level even at the largest
value of Ri tested. The dotted line indicates the CV of
a Poisson process. The error bars indicate the standard error. (c) Dependence of the Fano factor F (the
variance divided by the mean of the spike count) on the ratio Ri. The solid line shows the actual Fano factor,
whereas the dashed line shows the Fano factor predicted from F = CV2. Even for high values of Ri, the Fano
factor of the response to synthetic synaptic currents remains far below that observed in vivo. The dotted line
indicates the Fano factor of a Poisson process. The error bars indicate the standard error.

a neuron from layer 2/3 of rat neocortex was driven to fire
with synthetic synaptic currents. In this example, the input
rate of 2.4 EPSCs per millisecond yielded a firing rate of 21
Hz, indicating that about 115 EPSCs were required for each
action potential. For comparison, the response to an injection
of constant current is also shown. The striking difference
between the responses in Fig. 2a is in the fine temporal struc-
ture of the spike trains. In response to a constant input, spikes
arrive at regular intervals (which become longer during the
one-second stimulus as a result of spike adaptation), where-
as in response to the synthetic synaptic current, the interspike
intervals are much more irregular.
Although the fine temporal structure of the spike trains
generated by the synthetic synaptic current was very different
from that generated by the constant stimulus, the mean output
firing rate in response to this ensemble depended only on the
mean input. The fI curves (the firing frequency f, averaged
over the one second stimulus, as a function of the input cur-
rent I) for the two stimulus ensembles are fairly linear, show-
ing some slight saturation at high firing frequencies, and are in
close agreement over the range of currents tested (Fig. 2b).
Is the irregularity of the spike train in Fig. 2a as high as
that seen in vivo? One approach to quantifying spike irregu-
larity is based on the coefficient of variation (CV) of the dis-
tribution of interspike intervals (ISIs). The CV is defined as
the standard deviation isi divided by the mean isi of the ISI
distribution, CV = isi/isi. For a Poisson process, CV is one 12,
but the converse is not true: one cannot conclude that a
process is Poisson simply because CV is one.
The CV of spike trains from cortical neurons recorded in
vivo are generally near or above unity1,1315. The CV depends
on a number of factors, including the firing rate and the degree
of adaptation during a response. In order to provide a specific
benchmark against which to compare the present in vitro
results, we computed the CV of responses of neurons in the
middle temporal (MT) cortex of alert macaque monkeys driven
with constant-motion stimuli 16 . An example of one such
response, emphasizing the markedly irregular activity of neo-
cortical sensory neurons in vivo, is depicted in Fig. 2c. To min-
imize the effect of adaptation within a trial, only responses
after the first 500 ms were considered, and to minimize any
effect of slow trial-to-trial drift, the CV from each trial was
computed separately and then these CVs were averaged. The
mean CV was 1.1 0.16 (range, 0.91 to 1.4; n = 10) for neu-
rons with a mean firing rate of = 17 6.7 Hz (range, 11 to
33 Hz). Based on these data, it is reasonable to consider a CV of
0.8 as a lower limit for the in vivo range at the firing rates con-
sidered here.
The spike trains from the neuron shown in Fig. 2 had a CV
of 0.28. Similar results were obtained in all other neurons for
which similar stimuli were tested (CV = 0.29 0.09, = 21
2.4 spikes per s; n = 9). These values are much lower than the
CV value of 1.1 observed in MT cortical neurons in vivo. In
other experiments, we examined a range of synthetic synaptic

212 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com
1998 Nature America Inc. http://neurosci.nature.com
Fig. 4. Responses elicited by hypertonic solution evoked increases
in the rate of miniature EPSC release. (a) A regular train of action
potentials driven by miniature EPSCs elicited by a one-second appli-
cation of hypertonic solution from a puffer pipette positioned about
40 m from the soma. The variability of these responses (CV = 0.26,
Fano factor F = 0.05) was close to that elicited by comparable syn-
thetic synaptic currents. (b) The synaptic current elicited by hyper-
tonic solution in the same neuron. Calibration, 300 ms, 20 mV
(top), 100 pA (bottom). The bath solution contained standard
Ringers with 100 M picrotoxin, and 30 M cadmium added to pre-
vent recurrent activity. The hypertonic solution consisted of the
bath solution plus 0.5 mM sucrose.
currents and found that, as expected, the CV was smaller for
currents constructed from smaller synthetic EPSCs. For this
reason, we chose to test an EPSC amplitude at the upper limit of
the spontaneous responses recorded in slice (30 pA), because
any smaller EPSC would have yielded a lower CV even more
inconsistent with the in vivo range. We also examined higher
firing rates and found that, as expected, the CV decreased as
the firing rate increased. The CV reported here, then, can be
considered an upper limit on the CV likely to be generated by
the random superposition of independent inputs. Thus a pure-
ly excitatory steady drive from independent inputs does not
account for the observed irregularity of cortical spike trains.
MIXED IPSPS AND EPSPS
The synaptic drive to cortical neurons in vivo contains a sub-
stantial inhibitory component mediated by GABA recep-
tors17,18. It has been suggested that the irregularity observed
in vivo might arise in part from added fluctuations in the drive
caused by inhibitory inputs6,7. Inhibition increases the out-
put variability by increasing the variance in the input drive
associated with a given mean. To test whether the inclusion of
inhibitory inputs could account for the irregularity of spike
trains in vivo, we synthesized synaptic currents consisting of
a mixture of excitatory and inhibitory currents. Figure 3a
shows the response of a neuron to such an input. In this exam-
ple, the total inhibitory current Ri was half the excitatory cur-
rent (Ri = 0.5); the excitatory drive (the average rate of EPSC
occurrence) was increased in order to maintain a high firing
rate. As expected, the addition of this inhibitory component
increased the irregularity, to CV of 0.44. Similar results were
nature neuroscience volume 1 no 3 july 1998
articles
obtained in all other neurons for which similar stimuli were
tested (CV = 0.43 0.09, = 20.7 2.4 spikes per s; n = 9).
Thus although this input does increase spike irregularity, the
resultant values are still substantially below the CV value of
1.1 observed in vivo.
The fraction of the total input to a cortical neuron pro-
vided by inhibition has not been measured in vivo. Could a
higher inhibitory-to-excitatory ratio Ri boost the CV into the
in vivo range? The dependence of the CV on R i is shown in
Fig. 3b. As expected, the CV increased with increasing R i, but
remained well below unity for the highest values tested (CV
= 0.72 0.07, = 19.2 1.2 spikes per s; n = 5 for Ri = 0.9).
Thus at the highest levels of Ri tested, the CV approached, but
remained below, the values observed in vivo. Because ratios
of inhibition to excitation are unlikely to exceed 0.9 (see ref.
17), we conclude that these experiments represent an approx-
imate upper limit on the variability that can arise from synap-
tic noise from uncorrelated sources, and that such noise thus
cannot account for the high CV observed in vivo.
TRIAL-TO-TRIAL VARIABILITY
The CV is one of several standard measures of firing variabil-
ity. A high CV may reflect irregularity in the fine structure of
the spike train, but it may arise from changes arising on a time
scale longer than a typical interspike interval, such as those
induced by spike adaptation. Indeed, the response to the con-
stant stimulus shown in Fig. 2 has a CV of 0.43, which is about
the same as that for the mixed excitatory/inhibitory input with
R i of 0.5. The response to the purely excitatory synthetic
synaptic input also shows strong adaptation.
We therefore also considered a second measure of vari-
ability, the Fano factor. The Fano factor F is defined as the
variance divided by the mean of the spike count N, F = N2/N
where the spike count N is the number of spikes generated on
a (typically one-second) particular trial. The Fano factor is
insensitive to slow changes such as adaptation that occur with-
in a single trial, as long as they occur consistently from one
trial to the next. For any Poisson process (including both
homogeneous and rate-modulated Poisson processes), the
Fano factor is exactly one; spike trains from cortical neurons
in vivo are consistently found to have a Fano factor near or
above one16,19. For example, neurons from the middle tem-
poral (MT) area of alert monkeys were reported to have a
Fano factor of 1.3 (ref. 16).
In certain limiting cases, the Fano factor and the CV are
related by F = CV 2. The main requirement is that the spike
train be a stationary renewal process 12, that is, a process in
which each interspike interval (ISI) is statistically independent
of every other ISI. Although there are many ways in which a
spike train could deviate from a renewal process, in the pre-
sent context deviations are most likely to occur if successive
(or neighboring) ISIs are statistically dependent (for example,
neurons show some degree of bursting or short ISIs tended to
be followed by long ISIs), or if the spike train adapts, for
instance slows down, during the one-second stimulus. Because
the response both to the constant input and to the synthetic
synaptic drive can show considerable adaptation, at least the
possibility of adaptation during the stimulus must be consid-
ered. Thus the Fano factor can provide additional and inde-
pendent information about how well the responses to the
synthetic currents mimic the variability observed in vivo.
The Fano factor of the response to the constant stimulus was
only 0.02, compared with the Fano factor (F = 0.432 = 0.18) pre-
213
 1998 Nature America Inc. http://neurosci.nature.com

articles

a b C.V.
C.V. F.F.
F.F. tic terminals simultane-
ously, it is not a good
model for the asynchro-
1.2 1.2 nous and independent
barrage of synaptic stim-
1 1
ulation under considera-
0.8 0.8
tion here. By contrast,
local perfusion with
0.6 0.6 hypertonic solution ele-
vates the rate of minia-
0.4 0.4 ture EPSC release from
many terminals indepen-
0.2 0.2
dently and thereby gen-
0 0
erates a slowly varying
E E/I sync E E/I sync average current. The
Fig. 5. Input synchrony yields in vivo variability. (a) mechanism of this
Response to an input drive consisting primarily of occa- sucrose-evoked trans-
sional large brief (3050 ms) excitatory events, ran- mitter release is indepen-
1998 Nature America Inc. http://neurosci.nature.com

domly distributed in time, each consisting of the nearly dent of presynaptic

synchronous arrival of on average 100200 EPSCs, action potentials and cal-
yielding transient peak currents as high as 1 nA or cium influx 11,21,22 . The
more. Calibration, 200 ms, 10 mV (top), 0.1 nA (bot- miniature EPSCs re-
tom). (b) Summary of CV and Fano factor. Left, CV for leased under these con-
purely excitatory (E), mixed excitatory/inhibitory (E/I), and synchronous (sync), inputs. Right, same statistics ditions therefore provide
for the Fano factor. For both graphs, the error bars indicate standard errors. The dotted lines indicate the a much closer approxi-
value of unity expected for both CV and Fano factor for a perfect Poisson process and are near the values typ- mation of the drive to a
ically observed from cortical spike trains in vivo. cortical neuron expected
if all inputs were inde-
pendent.
Figure 4 shows the
dicted from the renewal assumption (where F rather than F is used response of a layer 2/3 neuron to a hypertonic solution-evoked
to denote the Fano factor predicted from the CV). The large dis- increase in the miniature EPSC rate. The spike train was very
crepancy indicates that for a constant stimulus the renewal assump- regular, with CV of 0.26 and Fano factor of 0.05. These results
tion is not satisfied: marked adaptation within a single trial leads to are in very close agreement with values obtained for the pure-
a large CV that is not reflected in the trial-to-trial measure F. The ly synthetic current (CV = 0.29 and F = 0.06; see Fig. 2). Thus
Fano factor for the purely excitatory input was only slightly high- the hypertonic solution-evoked responses showed much less
er than for the constant current (F = 0.06 0.03; n = 9), and close variability than observed in vivo.
to that predicted from the renewal assumption (F = 0.292 = 0.08). We were unable to elicit spiking in the absence of blockers
However, for the highest inhibitory/excitatory ratios tested (Ri = of (inhibitory) GABAergic inputs (n = 4), presumably because
0.9), the observed value (F = 0.29 0.12; n = 5) was substantially the hypertonic solution did not activate a sufficiently high
less than the predicted value (F = 0.722 = 0.52) and well below the ratio of excitatory to inhibitory input. This may be due in part
in vivo levels of one or above (Fig. 3c). The finding that the in vitro to the positioning of the puffer pipette near the soma, where
Fano factors, even for responses driven by inputs with a strong the density of GABAergic terminals may be higher. We were
inhibitory component, were consistently lower than those observed therefore unable to use hypertonic solution to validate the
in vivo reinforces the previous conclusion that independent inputs variability measurements for the synthetic inputs that mixed
cannot account for the high variability observed in vivo. excitation and inhibition. Nevertheless, the close agreement
ASYNCHRONOUS EPSC-EVOKED ACTION POTENTIALS between the synthetic and hypertonic solution-evoked vari-
The results presented so far have relied on current injected ability for the purely excitatory input suggests that currents
through a somatic electrode as a model for the synaptic cur- injected at the soma can adequately mimic synaptic inputs.
rents that drive neurons in vivo. Injected current may not,
however, mimic synaptic current in all respects. For example, INPUT SYNCHRONY YIELDS IN VIVO FIRING VARIABILITY
injected current implicitly treats each synapse as a current Because a steady input consisting of independent EPSCs and
source, whereas in fact a synapse is more properly considered IPSCs fails to drive firing that is as irregular as is observed in
as a conductance in series with the driving potential at the vivo, we tested the possibility that some degree of input syn-
subsynaptic membrane. (The dynamic clamp20 can only par- chrony might be required. We considered a simple model in
tially compensate for this error, as most excitatory synapses which the input consisted primarily of occasional large brief
onto cortical neurons are onto dendrites that are electrotoni- (3050 ms) excitatory events, randomly distributed in time
cally remote from the soma and whose subsynaptic voltage (average inter-event interval, 100 ms), each consisting of the
cannot be controlled by a somatic electrode.) We therefore nearly synchronous (3050 ms) arrival of on average about
developed a more direct approach to study the spiking vari- 100 to 200 EPSCs, yielding peak currents of 1 to 2 nA (Meth-
ability in response to synaptic stimulation. ods). Evidence from a variety of sources suggests the impor-
Because conventional extracellular stimulation triggers a tance of correlated firing in the neocortex9,16,2330.
large increase in the rate of transmitter release from all synap- Figure 5a shows a typical experiment in which the syn-

214 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com
chronous input was synthesized according to this simple
model. The irregularity of the resulting spike train was much
higher, with both the CV and the Fano factor in the in vivo
range in all neurons tested (CV = 1.3 0.11; F = 0.94 0.2,
= 19.8 1.0 spikes per s; n = 5). The data (Fig. 5b) demon-
strate that input correlations could account for the high degree
of variability observed in vivo.
Discussion
The observations1 on the paradox posed by the irregularity of
in vivo cortical spike trains have sparked considerable debate
on the underlying mechanism57,3234. Although Softky and
Koch1 raised the possibility that the unexpectedly large vari-
ability might arise from input synchrony, later workers have
argued that statistically stationary synaptic noise, interacting
with the spike-encoding mechanism, can account for the large
variability. All of these previous studies have, however, relied
on simulations of the spike encoder, and the results are sen-
1998 Nature America Inc. http://neurosci.nature.com
sitive to assumptions about the encoder properties.
Our results do not rely on any theoretical model of the
spike encoder. Rather, we have used the actual spike encoder of
neurons in vitro to test directly whether the high variability
observed in vivo could arise from a constant shower of statis-
tically independent synaptic inputs. We conclude that it could
not, and propose that the high in-vivo variability arises from
correlations among the inputs28,29. We illustrate this by match-
ing the in-vivo variability with a simple model of the input
drive, in which occasional large brief synchronous events are
superimposed on a tonic background.
Our conclusions rest on two main assumptions. First, we
have assumed that somatic current injection represents a rea-
sonable model of synaptic drive. Although the close agree-
ment between the somatic current injection and hypertonic
solution-evoked miniature EPSC results (Fig. 4) supports this
notion, it is possible that the agreement would not be as good
when the inhibitory drive was very strong. Second, we have
assumed that the in-vitro biophysical properties of neocortical
neurons can be used to make inferences about their in-vivo
behavior. This supposes, for example, that the spontaneous
miniature EPSCs recorded in slices accurately reflect the in-
vivo synaptic amplitude distribution. In addition, the
inputoutput transformation might be different in vivo,
because of neuromodulatory influences35 for example, or the
activation of large nonlinear conductances 36.
MECHANISM UNDERLYING SPIKING VARIABILITY IN VIVO
Irregular firing in vivo might in principle arise from input
variability or from noise in the spike-generating mechanism.
Because the neuronal spike generator is very reliable, input
variability is likely to be the primary source2,3. In spinal neu-
rons, synaptic noise (fluctuations in membrane potential
arising from a barrage of EPSPs) fully accounts for output
variability 4 . In the neocortex, the in-vivo synaptic drive
inferred from current clamp recordings could, when injected
into neocortical neurons in vitro, reproduce their basic firing
statistics 8 . These latter experiments did not interpret the
injected currents in terms of the underlying synaptic proper-
ties and correlational structure of the inputs, nor did they
establish the properties of the drive necessary or sufficient to
account for the high variability.
In the present study, we therefore began with the assump-
tion that output variability reflects fluctuations in the input
current and asked what form the synaptic noise must take to
nature neuroscience volume 1 no 3 july 1998
articles
give rise to the observed spiking variability. The conundrum is
that the fluctuations in the input drive generated by inde-
pendent excitatory afferents seem too small to account for the
high variability of in-vivo cortical firing. The initial report1
emphasized the difficulty in accounting for the high variabil-
ity at very high firing rates (over 100 Hz). In our experiments,
we found that even at more moderate firing rates (1525 Hz),
independent EPSCs cannot account for the variability
observed in vivo; at higher firing rates, independent EPSCs
induced even lower output variability.
Previously proposed resolutions fall into two main classes.
First, simulations suggested that inhibitory inputs increase the
input fluctuations and thereby the output variability suffi-
ciently to account for the in vivo responses57. Our experi-
mental results, however, demonstrate that the output variability
in response even to a drive with a very strong inhibitory com-
ponent remains well below the in vivo levels, indicating that
some alternative mechanism must be responsible.
The second class of explanation invokes the details of the
spike-generating mechanism. In one proposal7, the reset volt-
age following an action potential was used to match the slope
or gain of the firing-intensity curve for the first interspike
interval in an integrate-and-fire model. In simulations, the
output variability of responses generated by this model was
near that found in vivo. This model failed, however, to pre-
dict the responses of cortical neurons to the synthetic synap-
tic currents we used (C.S. and A.Z., unpublished results). The
initial report1 proposed a more radical modification of the
spike generator. Here it was suggested that the coincident
arrival of just a few EPSPs might activate powerful nonlinear
dendritic conductances that would then trigger a somatic
spike. Although our experiments cannot rule out this mecha-
nism (and indeed, the large events shown in Fig. 5 could in
principle arise from the activation of dendritic nonlineari-
ties), we favor the notion that the dynamics of cortical net-
works, rather than the properties of single neurons, underlie
the requisite large fluctuations in the input drive.
The model of synaptic drive that we have used is by no
means unique in its ability to account for the variability
observed in vivo; countless other input ensembles would cer-
tainly have done as well. The key requirement is that the input
current contain very large fluctuations, much larger than
would be expected if all the synapses were releasing quanta
independently and at a constant rate. These fluctuations could
arise from the brief and coordinated increase in the firing rate
of a large number of excitatory, or perhaps inhibitory 37 ,
synapses. Moreover, if a small number of presynaptic neurons
each made dozens of powerful synaptic contacts onto a sin-
gle postsynaptic target38, then the requisite postsynaptic fluc-
tuations might arise from correlated activity in just this
smaller subpopulation of neurons. Our experiments do not
address the specific network mechanisms that might give rise
to this correlated synaptic activity. Nevertheless, the present
results indicate that, however the input current is generated,
some kind of correlation among the synaptic inputs is likely
to play a critical role in generating the high degree of vari-
ability observed in vivo.
SYNCHRONY AND THE NEURAL CODE
The irregularity of cortical spike trains has led to two rather
different models of how cortical circuitry operates. In the con-
ventional model, this irregularity represents noise around
the signal, a perturbation around a mean firing rate that is
215
 1998 Nature America Inc. http://neurosci.nature.com
articles
obtained by averaging over some period (such as one second)
that is much longer than a typical ISI. In this view, the precise
timing of individual spikes conveys little information, because
it reflects only the noisy activity of other neurons attempting
to signal their own mean rate. This model follows readily from
the idea that the drive to cortical neurons is composed of the
uncorrelated activity of its synaptic inputs; it would be hard
to imagine how the precise timing of spikes could represent
more than noise in this model, because adding or removing
just a few EPSPs might perturb the spike timing appreciably.
In the second model, the irregular spikes reflect modula-
tion on a fine time scale of the neurons output rate. It is now
established that under some conditions, the fine temporal
structure of the spike train (the precise location of each spike)
can carry information16,39,40. For example, when visual stim-
uli have fine temporal structure, the timing of spikes from
neurons in MT cortex can be tightly (25 ms) locked to the
stimulus16,41. Recordings of local field potentials suggest that
1998 Nature America Inc. http://neurosci.nature.com
this high fidelity of spike timing is achieved by rapid comod-
ulation of the rates of the input neurons16; under these con-
ditions, the input correlations encode temporal edges within
the stimulus. However, our results suggest that even when the
sensory stimulus does not have fine temporal structure, spikes
in cortical neurons may nevertheless arise from large events
in the input drive that represent the correlated activity of many
neurons. Such large events are presumably more robust to
noise than the small fluctuations that are posited to drive fir-
ing in the first model. Thus even when the sensory stimulus
is devoid of fine temporal structure, spikes may be encoding
something with high temporal fidelity.
Methods
SLICE PREPARATION. Brain slices were prepared from Long-Evans rats
(postnatal day 1428) that were deeply anesthetized with metofane
and then decapitated. The skull was rapidly opened and the brain
placed in ice-cold Ringer solution. The cooled brain was glued with
cyanoacrylate to the stage of a vibratome and 400-m slices were cut.
Slices were transferred to a holding chamber and incubated for at least
one hour at room temperature in a solution continuously bubbled
with a mixture of 95% O2 and 5% CO2, and then placed in a record-
ing chamber.
PATCH-CLAMP RECORDING. Whole-cell, patch-clamp recordings were
made from neurons in the sensory neocortex visualized through an
upright microscope equipped with infrared light and differential inter-
ference optics42. Recordings were performed at 3335C. Slices were
continuously perfused with a Ringers solution containing (in mM)
NaCl 120, KCl 3.5, CaCl2 2.6, MgCl2 1.3, NaH2PO4 1.25, NaHCO3 26,
glucose 10, pH 7.35. Unless otherwise indicated, recordings were
obtained in the presence of the AMPA receptor antagonist 6-cyano-7-
nitroquinoxaline-2,3-dione (CNQX; 10 M). Recording pipettes were
filled with an internal solution consisting of (in mM) K gluconate,
170; HEPES, 10; NaCl, 10; MgCl 2 , 2; EGTA, 1.33; CaCl 2 , 0.133;
MgATP, 3.5; GTP, 1.0, pH 7.2 and 290300 mOsm. Resistance to bath
was 35 M before seal formation.
In experiments in which the rate of miniature EPSCs was elevated
by locally perfusing with hypertonic solution, the bath Ringers con-
tained 100 M picrotoxin (to block -aminobutyric acid type A recep-
tor responses), 30 M Cd+2 to block synaptic responses that could
cause recurrent activity and no CNQX. To elicit hypertonic-solution-
evoked miniature EPSCs, visual guidance was used to position a 2 m
puffer pipette (containing bath Ringers to which 0.5 mM sucrose had
been added) close to dendrites about 3050 m from the somatic
recording electrode. A picospritzer II was then used to apply 12 sec-
ond pulses (36 psi) of the pipette solution.
216
Recordings were obtained using an Axopatch 200A or 200B (Axon
Instruments). Most recordings were obtained in the current-clamp
mode, without series-resistance compensation. Traces were filtered
at 2 KHz and digitized at 4 KHz. Series resistance, monitored prior
to each trace, was typically in the range of 820 M, and never more
than 30 M. To compensate for any slow drift in membrane potential,
prior to each trace sufficient current was injected to return the mem-
brane to a value determined at the start of the experiment (between -
60 and -75 mV); if the actual rest potential decreased by more than
5 mV from this level, the experiment was terminated. The response
to constant current injection (an FI curve) was also periodically
monitored to assess washout, and if the spike rate at the maximum
intensity decreased by more than 30%, the experiment was termi-
nated (usually 3060 minutes).
Forty-two regular spiking neurons43 from cortical layers 2/3 were
analyzed. In addition, results from two layer 5 neurons were consistent
with the other results and so were included in the averaged data. Cur-
rent stimulation was usually applied for one second every five or ten
seconds; the long interstimulus rest period helped to reduce the influ-
ence of one trial on the next. For the analysis, only neurons whose aver-
age firing rate was between 15 and 25 Hz were used. At higher sustained
firing rates, recording stability tended to degrade more quickly.
Data were acquired using a National Instruments AT-MIO-16-F-
5 A/D card on a 120 MHz Pentium-based computer under the Win-
dow NT (Microsoft) operating system. Software written in Labview
(National Instruments) with Dynamic Data Exchange links to Mat-
lab (Mathworks) allowed convenient online synthesis and injection
of arbitrary synthetic current waveforms.
SYNTHETIC SYNAPTIC CURRENTS. Neurons were driven to fire through
currents injected through a somatic patch electrode. Several models of
correlated synaptic inputs were tested. In the independent EPSC model,
the synaptic drive was given by I e = P(r e ) * I ampa , where P(r e ) is a
sequence of independent Poisson points arriving at a rate re, Iampa =
We et/e is the waveform of the basic EPSC (We = 30 pA, e = 3 ms),
and * indicates convolution. The firing rate under these conditions is
determined by a single free parameter, the rate of synaptic impulses re.
In the mixed model, an inhibitory drive given by Ii = P(ri) * Igaba
was added to the excitatory drive, where as above P(ri) is a sequence
of independent Poisson points arriving at a rate ri, Igaba = Wi et/e is
the waveform of the basic IPSC (Wi = 30 pA, i = 6 ms), and * indi-
cates convolution. The total input under these conditions was given by
Ib = Ii + Ie. In this scenario, the firing rate depended both on ri and re.
The value of ri was chosen to keep the ratio Ri of total inhibitory and
excitatory currents fixed, R i = r i/ r e where = (W e e)/(W m m)
takes into account differences in the synaptic waveforms.
Finally, we examined the effect of synchronous excitatory inputs,
Is = C(t) * Iampa, where C(t) was the stochastic function that describes
the occurrence of synchronous events. C(t) consisted of 3050-ms
periods of elevated input activity; these periods occurred in a ran-
dom (Poisson) fashion at a rate of 10 Hz. During each 3050-ms peri-
od of elevated activity, about 200 EPSCs were distributed in 28
shorter events. The synchronous input Is was added to Ie and Ii.
Acknowledgements
We thank G. Buracas for the MT data and L. Dobrunz, E. Huang, K. Miller,
P. Latham, T. Troyer and K. Zhang for comments. This work was supported
by the Howard Hughes Medical Institute (C.F.S.), National Institutes of
Health grant NS 12961 (C.F.S.) and the Sloan Center for Theoretical
Neurobiology (A.M.Z.).
RECEIVED 9 APRIL: ACCEPTED 25 MAY 1998
1. Softky, W. & Koch, C. The highly irregular firing of cortical cells is
inconsistent with temporal integration of random epsps. J. Neurosci. 13,
334350 (1993).
2. Mainen, Z.F. & Sejnowski, T.J. Reliability of spike timing in neocortical
neurons. Science 268, 15031505 (1995).
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com
3. Zador, A. The impact of synaptic unreliability on the information
transmitted by spiking neurons. J. Neurophysiol. 19, 12301238 (1998).
4. Calvin, W.H. & Stevens, C.F. Synaptic noise and other sources of randomness
in motoneuron interspike intervals. J. Neurophysiol. 31, 574587 (1968).
5. Shadlen, M.N. & Newsome, W.T. Is there a signal in the noise? Curr. Opin.
Neurobiol. 5, 248250 (1995).
6. Shadlen, M.N. & Newsome, W.T. The variable discharge of cortical neurons:
Implications for connectivity, computation, and information coding. J.
Neurosci. 18, 3870-3896 (1998).
7. Troyer, T.W. & Miller, K.D. Physiological gain leads to high ISI variability in a
simple model of a cortical regular spiking cell. Neural Computation 9,
971983 (1997).
8. Nowak, L.G., Sanchez-Vives, M.V. & McCormick, D.A. Influence of low and
high frequency inputs on spike timing in visual cortical neurons. Cereb.
Cortex 7, 487501 (1997).
9. Berretta, B. & Jones, S.G. A comparison of spontaneous EPSCs in layer II and
layer IV-V neurons of the rat entorhinal cortex in vitro. J. Neurophysiol. 76,
10891100 (1996).
10. Burgard, E.C. & Hablitz, J.J. NMDA receptor-mediated components of
miniature excitatory synaptic currents in developing rat neocortex. J.
Neurophysiol. 70, 18411852 (1993).
11. Bekkers, J.M., Richerson, G.B. & Stevens, C.F. Origin of variability in quantal
size in cultured hippocampal neurons and hippocampal slices. Proc. Natl
Acad. Sci. USA 87, 53595362 (1990).
1998 Nature America Inc. http://neurosci.nature.com
12. Feller, W. An Introduction to Probability Theory and its Applications, vol. 2,
2nd edn. (Wiley, New York, 1971).
13. Burns, B.D. & Webb, A.C. The spontaneous activity of neurones in the cats
cerebral cortex. Proc. Royal Soc. Lond. B 194, 211223 (1976).
14. Holt, G.R. Softky, W.R., Koch, C. & Douglas, R.J. Comparison of discharge
variability in vitro and in vivo in cat visual cortex neurons. J. Neurophysiol.
75, 18061814 (1996).
15. Noda, H. & Adey, R. Firing variability in cat association cortex during sleep
and wakefulness. Brain Res. 18, 513526 (1970).
16. Buracas, G., Zador, A., Deweese, M. & Albright, T. Efficient discrimination of
temporal patterns by motion-sensitive neurons in primate visual cortex.
Neuron 20, 959969 (1998).
17. Berman, N.J., Douglas, R.J., Martin, K.A. & Whitteridge, D. Mechanisms of
inhibition in cat visual cortex. J. Physiol. 440, 697722 (1991).
18. Nelson, S., Toth, L., Sheth, B. & Sur, M. Orientation selectivity of cortical neurons
during intracellular blockade of inhibition. Science 265, 774777 (1994).
19. Gershon, E.D., Wiener, M.C., Latham, P.E. & Richmond, B.J. Coding
strategies in monkey V1 and inferior temporal cortices. J. Neurophysiol. 79,
11351144 (1998).
20. Sharp, A.A., ONeil, M.B., Abbott, L.F. & Marder, E. Dynamic clamp:
computer-generated conductances in real neurons. J. Neurophysiol. 3,
992995 (1993).
21. Hubbard, J.I., Jones, S.F. & Landau, E.M. An examination of the effects of
osmotic pressure changes upon transmitter release from mammalian motor
nerve terminals. J. Physiol. 197, 639657 (1968).
22. Stevens, C.F. & Tsujimoto, T. Estimates for the pool size of releasable quanta
at a single central synapse and for the time required to refill the pool. Proc.
Natl Acad. Sci. USA 92, 846849 (1995).
nature neuroscience volume 1 no 3 july 1998
articles
23. Abeles, M. CorticonicsNeural Circuits of the Cerebral Cortex. (Cambridge
Univ. Press, 1991).
24. Alonso, J.M., Usrey, W.M. & Reid, R.C. Precisely correlated firing in cells of
the lateral geniculate nucleus. Nature 383, 815819 (1996).
25. deCharms, R.C. & Merzenich, M.M. Primary cortical representation of
sounds by the coordination of action-potential timing. Nature 381, 610613
(1996).
26. Nelson, J.I., Salin, P.A., Munk, M.H., Arzi, M. & Bullier, J. Spatial and
temporal coherence in cortico-cortical connections: a cross-correlation study
in areas 17 and 18 in the cat. Visual Neurosci. 9, 2137 (1992).
27. Singer, W. & Gray, C.M. Visual feature integration and the temporal
correlation hypothesis. Ann. Rev. Neurosci. 18, 555586 (1995).
28. Usher, M., Stemmler, M., Koch, C. & Olami, Z. Network amplification of
local fluctuations causes high spike rate variability, fractal firing patterns and
oscillatory local field potentials. Neural Computation 6, 795836 (1994).
29. van Vreeswijk, C. & Sompolinsky, H. Chaos in neuronal networks with
balanced excitatory and inhibitory activity. Science 274, 17241726 (1996).
30. Whittington, M.A., Traub, R.D., Faulkner, H.J., Stanford, I.M. & Jefferys, J.G.
Recurrent excitatory postsynaptic potentials induced by synchronized fast
cortical oscillations. Proc. Natl Acad. Sci. USA 94, 1219812203 (1997).
31. Zohary, E., Shadlen, M.N. & Newsome, W.T. Correlated neuronal discharge
rate and its implications for psychophysical performance. Nature 370,
140143 (1994).
32. Shadlen, M.N. & Newsome, W.T. Noise, neural codes and cortical
organization. Curr. Opin. Neurobiol. 4, 569579 (1994).
33. Softky, W.R. Simple codes versus efficient codes. Curr. Opin. Neurobiol. 5,
239247 (1995).
34. Reich, D.S., Victor, J.D., Knight, B.W., Ozaki, T. & Kaplan, E. Response
variability and timing precision of neuronal spike trains in vivo. J.
Neurophysiol. 77, 28362841 (1997).
35. Tang, A.C., Bartels, A.M. & Sejnowski, T.J. Effects of cholinergic modulation
on responses of neocortical neurons to fluctuating inputs. Cereb. Cortex 7,
502509 (1997).
36. Softky, W.R. Sub-millisecond coincidence detection in active dendritic trees.
Neuroscience 58, 1341 (1994).
37. Hausser, M. & Clark, B.A. Tonic synaptic inhibition modulates neuronal
output pattern and spatiotemporal synaptic integration. Neuron 19, 665678
(1997).
38. Markram, H. A network of tufted layer 5 pyramidal neurons. Cereb. Cortex 7,
523533 (1997).
39. Bialek, W., Rieke, F., de Ruyter van Steveninck, R.R. & Warland, D. Reading a
neural code. Science 252, 18541857 (1991).
40. Rieke, F., Warland, D., de Ruyter van Steveninck, R.R. & Bialek, W. Spikes:
Exploring the Neural Code. (MIT Press, 1997).
41. Bair, W. & Koch, C. Temporal precision of spike trains in extrastriate cortex
of the behaving macaque monkey. Neural Computation 8, 11841202 (1996).
42. Stuart, G.J., Dodt, H.U. & Sakmann, B. Patch-clamp recordings from the
soma and dendrites of neurons in brain slices using infrared video
microscopy. Pflug. Archiv. Eur. J. Physiol. 423, 511518 (1993).
43. McCormick, D.A., Connors, B.W., Lighthall, J.W. & Prince, D.A.
Comparative electrophysiology of pyramidal and sparsely spiny stellate
neurons of the neocortex. J. Neurophysiol. 54, 782806 (1985).
217
 1998 Nature America Inc. http://neurosci.nature.com

articles

Nociceptive and thermoreceptive

lamina I neurons are anatomically
distinct
Z.-S. Han1,2, E.-T. Zhang1 and A. D. Craig1

1 Divisions of Neurobiology and Neurosurgery, Barrow Neurological Institute, 350 West Thomas Rd., Phoenix, Arizona 85013, USA
2 Present address: RS Dow Neurological Sciences Institute, 1120 NW 20th Ave., Portland, Oregon 97209, USA
Correspondence should be addressed to A.D.C. (bcraig@mha.chw.edu)

Pain and temperature stimuli activate neurons of lamina I within the dorsal horn of the spinal cord,
1998 Nature America Inc. http://neurosci.nature.com

and although these neurons can be classified into three basic morphological types and three major
physiological classes, earlier studies did not establish a structure/function correlation between their
morphology and their physiological responses. We recorded and intracellularly labeled 38 cat
lamina I neurons. All 12 fusiform cells were nociceptive-specific, responsive only to pinch and/or
heat. All 11 pyramidal cells were thermoreceptive-specific, responsive only to innocuous cooling. Of
ten multipolar cells, six were polymodal, responsive to heat, pinch and cold, and four were nocicep-
tive-specific. Five unclassified cells had features consistent with this pattern. These results support
the view that central pain and temperature pathways contain anatomically discrete sets of
modality-selective neurons.

Lamina I of the superficial dorsal horn is an integral component
of the central representation of pain and temperature sensitivi-
ties in mammals13. It receives direct input from A and C pri-
mary afferent fibers, including specific nociceptors and
thermoreceptors, and it is the main source of output from the
superficial dorsal horn. In contrast to the deep dorsal horn with
its modality-ambiguous wide dynamic range (WDR) nocicep-
tive neurons, which have been studied extensively, lamina I con-
tains unique concentrations of modality-selective nociceptive
and thermoreceptive neurons. In primates, the axons of these
neurons project in the lateral spinothalamic tract (which is crit-
ical for pain and temperature sensation) to a dedicated nocicep-
tive- and thermoreceptive-specific relay nucleus in posterolateral
thalamus4 and to other sites3 that together provide input to the
cortical regions that are activated by painful and thermal stim-
uli in humans5,6.
These characteristics support the concept that pain is sub-
served by distinct sets of neurons both peripherally and cen-
trally. This view is not shared by all investigators in the field of
pain research7, in part because numerous intracellular labeling
studies did not uncover distinct types of neurons in the super-
ficial dorsal horn816. This contrasts with the visual and motor
systems, where distinct structure/function correlations have
been identified17,18, and accumulating evidence has suggested
that this issue deserves re-examination. Physiological studies
of lamina I spinothalamic cells have documented three major
classes of modality-selective neurons in cat and monkey: noci-
ceptive-specific (NS) cells, thermoreceptive-specific (COLD)
cells, and polymodal nociceptive (HPC) cells (and also WDR
cells in monkey)1925. These classes have different axonal con-
duction velocities and thalamic terminations, as shown with
antidromic mapping (refs 21, 24, 26; Dostrovsky, J.O. & A.D.C.,
Soc. Neurosci. Abstr., 646.9, 1993), features indicating anatom-
ical differences that could be reflected in the morphology of
their cell bodies in lamina I. In horizontal sections, lamina I
neurons in rat, cat and monkey can be anatomically catego-
rized almost comprehensively into three basic, albeit variegat-
ed, cell types: fusiform, pyramidal, and multipolar 2730. The
fusiform cells appear to have unmyelinated axons, whereas
pyramidal and multipolar cells have myelinated axons; this
matches the physiological difference in conduction velocities
between NS cells and COLD or HPC cells. These physiological
and morphological classifications were not distinguished in
prior intracellular labeling studies of lamina I cells for method-
ological reasons. We directly re-examined the characteristics of
lamina I neurons, using intracellular recording and labeling,
which revealed a clear correspondence between these anatom-
ical and physiological classifications.
Results
PHYSIOLOGICAL CHARACTERISTICS
We recorded intracellularly from a total of 126 cells in the super-
ficial dorsal horn of 41 anesthetized cats. Stable intracellular
recordings were obtained for several minutes to over one hour.
Altogether 76 cells were injected with biocytin, 53 of which were
recovered histologically, and 38 of which were lamina I neurons
that had been characterized physiologically. The remaining recov-
ered cells were uncharacterized, unresponsive and/or were locat-
ed in lamina II or deeper. For some identified cells, there was
apparent injury during penetration, based on the gradual dete-
rioration of the membrane potential and action-potential ampli-
tude, yet the spontaneous activity pattern and the response
properties to natural stimuli remained stable.
All 38 characterized, labeled lamina I neurons were respon-
sive to natural cutaneous stimulation of the ventral hindpaw. The
response characteristics of these neurons were consistent with

218 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

NS cell COLD cell HPC cell

Fig. 1. Intracellular responses of NS, COLD and HPC cells. Bars indicate the application of
stimuli. The intensity of the pinch stimuli was increased during application. The NS cell (mul-
tipolar cell 36-739) had a resting membrane potential of -45 mV and a spike amplitude that
varied with firing rate. It developed a high ongoing discharge rate, which may have occluded a
phasic response to the brief cold stimulus. The COLD cell (pyramidal cell 25-670) had a rest-
1998 Nature America Inc. http://neurosci.nature.com

ing membrane potential of -40 mV, a steady spike amplitude of 45 mV, and an afterhyperpolarization of 10 mV. It had irregular ongoing dis-
charge at room temperature (about 23C) prior to any stimulation. The HPC cell (multipolar cell 44-1000) had a resting membrane potential
of -55 mV and a spike amplitude of 60 mV. The small arrow indicates hyperpolarizing current pulses (0.5 nA, 50 ms, 1Hz) applied to stabilize
the recording.

observations in prior studies2023, and so the cells were classified
as nociceptive-specific (NS) cells responsive only to pinch and/or
heat, thermoreceptive-specific (COLD) cells responsive only to
innocuous cooling and inhibited by warming, and polymodal
nociceptive (HPC) cells responsive to heat and/or pinch and also
to cold. (WDR lamina I neurons are rare in the cat8,9,11,15,21,26.)
Although the need for rapid characterization prevented us from
systematically delimiting the receptive fields of nociceptive cells,
they were generally small and confined, ranging in size from part
of one toe pad to most of the ventral surface of the paw. Many
receptive fields occupied the glabrous pads on one or two digits,
sometimes including a region of hairy skin between the toes or
near the central pad. The receptive fields of COLD cells were gen-
erally larger, including half or more often all of the ventral paw.
Based on their intracellular responses to natural stimuli, 18 of
the 38 identified lamina I neurons were categorized as NS cells,
13 as COLD cells and 7 as HPC cells (examples in Fig. 1). None
of the cells was responsive to low-threshold (weak) stimuli. Of the
18 NS cells, 8 responded only to noxious pinch, and 10 respond-
ed to both pinch and noxious heat (Fig. 1). The NS cells often
showed prolonged afterdischarge following a noxious stimulus
(Fig. 1) and moderate-sized spontaneous excitatory postsynaptic
potentials (Fig. 2d). The 13 COLD cells were readily excited by
cooling but were not excited by pinch or heat; rather, their dis-
charges were usually inhibited by heat or even by the warmth of
the fingers when used for pinching (Fig. 1). One COLD cell (19-
558) showed a paradoxical burst response to heat. Most HPC cells
responded vigorously to noxious heat and in a graded manner to

a c

b d

Fig. 2. Intracellular responses of different classes of lamina I cells to prolonged current injection (0.5 nA, 500600 ms, indicated by the bars).
(a, b) The COLD cell in (a) and the HPC cell in (b) both produced repetitive discharges without significant frequency adaptation. Note the
deep but brief afterhyperpolarization following each COLD cell action potential in (a) and the more shallow but broad and variable afterhy-
perpolarization following each HPC cell action potential in (b). (c, d) In contrast, both of the NS cells responded to current injection with
action potential trains that exhibited pronounced frequency adaptation, and both displayed a prolonged hyperpolarizing potential following
current offset (triangular arrowheads). Both also had spontaneous EPSPs (small downward arrows). The cell in (c) had action potentials rid-
ing on EPSPs and often fired in doublets (upward arrows). The cell in (d) showed variable, brief afterhyperpolarizations.

nature neuroscience volume 1 no 3 july 1998 219

 1998 Nature America Inc. http://neurosci.nature.com

articles

noxious mechanical stimuli; their responses to cold 17-699 pinch + heat, U 22-820 pinch + heat, F
were less brisk than those of COLD cells and were
often delayed (Fig. 1). As in prior work, HPC cells had
varied thresholds to each submodality, and with the 43-451 pinch + heat, F
22-964 pinch + heat, F
rapid characterization methods we employed, three
HPC cells were detected that apparently had high
thresholds to thermal stimulation. Two of these (27-
551 and 44-1000) showed only a transient response
to the brief (3 to 5 second) cold stimuli, which is char- 16-865 pinch, F 28-1007 pinch + heat, F
acteristic of HPC cells that have high noxious cold
thresholds and produce a tonic response to stimuli of
longer duration (30 to 60 seconds) than was practical
for this study21. One (20-910) that responded clearly 35-1169 pinch + heat, F 25-690 pinch + heat, F
to pinch and to cold had only a weak, poorly repro-
duced response to radiant heat, similar to HPC cells
that have high thresholds to maintained heat upon
quantitative examination with a Peltier stimulator21.
We applied antidromic stimulation to the con- 23-795 pinch, F 17-755 pinch + heat, F
1998 Nature America Inc. http://neurosci.nature.com

tralateral thalamus in 22 of the 41 experiments, in an

attempt to identify spinothalamic lamina I neurons26.
Antidromic responses to trains of stimuli were
observed in two of three identified cells in which
extracellular recordings were obtained prior to pen-
etration and staining (25-690 and 28-1007). Howev- 45-794 pinch, M 43-750 pinch, M
er, in intracellular recordings, only one or two
time-locked action potentials were observed with
antidromic stimulation in four identified lamina I
cells (17-755, 19-558, 21-633, and 25-670). This is
consistent with the reported failure of antidromic
somatic invasion in intracellularly recorded lamina I
neurons following stimulation of the midbrain 36-739 pinch, M
parabrachial region9. Thus, whereas only these 6 of
the 24 neurons identified in 22 cats could be consid-
ered putative lamina I spinothalamic cells, many oth- 100 M
ers may also have been spinothalamic.
The intracellular responses to a step-depolarizing Fig. 3. Camera-lucida drawings of the soma and proximal dendrites of each iden-
current pulse (500 milliseconds) corresponded with tified NS cell. For each cell, the response characteristics and the morphological
the classification of the identified neurons in the small classification (F, fusiform; M, multipolar; U, unclassified or transitional) are indi-
subset that was tested. A single action potential was cated. The shading indicates processes that extended dorsally or ventrally out of
the plane of the horizontal section or that were more weakly labeled. Rostral is
normally evoked at threshold intensities (0.2 to 0.4
left, medial is up. Scale bar, 100 m.
nA), but depolarization at suprathreshold intensities
(over 0.5 nA, 500 milliseconds) produced a train of
non-decrementing action potentials with a steady
interspike interval in COLD cells (Fig. 2a, n = 4) and HPC cells (Methods), examined by light microscopy and documented with
(Fig. 2b, n = 3). In contrast, NS cells (n = 4) displayed signifi- photomicrographs and camera lucida drawings. An additional
cant frequency adaptation (Fig. 2c and d). Further, COLD cells eight cells were stained with FITC-avidin (in double-labeling
showed a prominent afterhyperpolarization following each action experiments performed to test various immunohistochemical
potential (Figs 1 and 2a), whereas variable or no afterhyperpo- markers) and thus were not drawn. Most of the identified neu-
larization was generally present in NS or HPC cells (Figs 1 and rons were located in the dorsal cap at the lateral apex of the dor-
2). However, NS cells displayed a prolonged hyperpolarization sal horn, where lamina I spinothalamic cells are concentrated.
following intracellular current injection (Fig. 2c and d). These The somata and dendrites of all labeled cells were well stained
correlations of membrane properties with functional class resem- and generally could be followed longitudinally within lamina I
ble those reported for dorsal horn cells31 and are clearly sugges- for hundreds of microns (see Figs 6 and 7). The shape and ori-
tive of differences between these physiological categories in entation of the somata and proximal dendrites of most cells could
voltage-gated channels. There was also a tendency for the mean be assessed in one section. For a few cells, labeled elements in
half-amplitude spike widths of COLD cells to be greater, although adjacent serial sections needed to be carefully fitted with the use
it was not statistically significant (COLD, 0.91 0.51, n = 8, ver- of tracing paper or digital overlays. Their axons, if observable,
sus NS, 0.55 0.19, n = 6, p = 0.12; COLD versus HPC, 0.59 were very fine and were not included in the drawings.
0.25, n = 7, p = 0.20). We categorized the identified lamina I neurons anatomically
as fusiform, pyramidal, multipolar and unclassified or transition-
MORPHOLOGICAL CHARACTERISTICS al neurons, based on the morphological criteria of somatoden-
Of the 38 physiologically characterized lamina I cells injected dritic shape and orientation in horizontal sections (see below) that
with biocytin, 30 were stained with the ABC/DAB reaction we have used in prior studies of retrogradely labeled lamina I

220 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

23-734 cold, P 31-672 cold, P Of the 18 NS neurons identified, 10 were typ-

ical spindle-shaped, longitudinally oriented,
bipolar fusiform lamina I neurons of various
sizes (Fig. 3). Cell 35-1169 differed slightly in
that its large soma was located at the border
22-903 cold, P 35-1129 cold, P between lamina I and lamina II, but its dendrites
extended dorsally into lamina I. Two other NS
cells (23-795 and 17-755) belonged to the T-
shaped variety of fusiform cells, using the crite-
45-777 cold, P
ria described previously29,30, because they had a
19-558 cold, P
minor dendrite protruding perpendicularly from
the longitudinal, spindle-shaped soma (see Fig.
3). One NS cell (17-699) is an example of a bipo-
lar cell with an eccentric soma, a shape that was
termed unclassified in our prior studies, but
27-176 cold, P 25-670 cold, P which could simply be another subtype of the
fusiform category. The other unclassified NS cell
found was an unusual bipolar cell (FITC-labeled,
1998 Nature America Inc. http://neurosci.nature.com

not shown) that has a very large, round soma

and widely radiating secondar y dendritic
branches. In addition, four NS cells were excep-
tional in that they were multipolar neurons;
43-627 cold, U 27-333 cold, U three of these (45-794, 43-750 and 47-726 [not
shown]) are of the tube-like variety described
earlier29,30, and one (36-739) is of the radiating
subtype. Cell 36-739 was one of only two iden-
tified neurons that had spiny dendrites.
Of the 13 COLD neurons identified, all
showed a general triangular shape, and 11 were
100 M
classified as pyramidal lamina I neurons with
three major dendritic poles and three or four
Fig. 4. Camera-lucida drawings of the soma and proximal dendrites of each identi- dendrites. (Fig. 4 shows eight of these; three were
fied COLD cell. For each cell, the response characteristics and the morphological labeled with FITC.) As in our prior observations
classification (P, pyramidal; U, unclassified or transitional) are indicated. The shading of lamina I spinothalamic cells, the dendrites of
indicates processes that extended dorsally or ventrally out of the plane of the hor- most of these pyramidal COLD neurons issued
izontal section or that were more weakly labeled. Rostral is left, medial is up. Scale few arbors and extended primarily longitudinal-
bar, 100 m. ly, though a few cells had more widespread ram-
ifications (e.g., 22-903 or 27-176). Cell 25-670
differed in that it had only two major dendrites.
In addition, two exceptional COLD cells (43-627
spinothalamic cells in cat and monkey29,30. Even though there was and 27-333) were categorized morphologically as unclassified
variation in the fine details of the morphology of the neurons in or transitional neurons, because even though they had gen-
this sample, which is similar to the variation observed in prior erally triangular shapes, they had multiple dendritic origins.
work2730, overall there was an obvious correspondence between Cell 43-627 also differed because of a protruding bulge at the
the functional category and the somatodendritic shape of these rostral, apical (left in Fig. 4) end of the soma, and because its
cells, which indicates that there are three major structural/func- basilar dendrites on the caudal (right) end were rotated out
tional types of lamina I neurons. Of the 18 NS cells identified, 12 of the horizontal plane.
were fusiform neurons, whereas 4 were multipolar and 2 were Of the seven HPC cells identified, six were multipolar lam-
unclassified. Of the 13 COLD cells, 11 were pyramidal, whereas 2 ina I neurons that had polygonal somata with multiple den-
were unclassified. Of the 7 HPC cells, 6 were multipolar and 1 was drites that arborized both longitudinally and mediolaterally.
unclassified. The converse description reveals this correspondence Of these, three (21-633, 44-1000, 32-740) were of the radiat-
even more clearly. Of the 12 fusiform neurons, all were NS cells. ing variety of multipolar cells and three (24-514, 27-551 and
Of the 11 pyramidal neurons, all were COLD cells. Of the 10 mul- 20-910) were of the tube-like variety29,30. Cells 21-633 and
tipolar cells, 6 were HPC cells and 4 were NS cells. The interme- 44-1000 were clearly similar to the -shaped subtype of mul-
diate characteristics of the unclassified neurons were consistent tipolar lamina I spinothalamic cells described in the mon-
with this pattern. The drawings in Figs 35 show the physiologi- key30, but infrequently observed in the cat29. Cell 32-740 was
cal characteristics, the anatomical classification and the somato- the only cell that resembled somewhat the quadrilateral sub-
dendritic morphology of each cell from the sample of 30 cells type that was relatively commonly observed in both cat and
stained with diaminobenzidine, organized according to the three monkey. In addition, one HPC cell (26-882) was a large,
functional categories identified (NS, COLD, HPC) . The pho- unclassified neuron located on the laminae I-II border that
tomicrographs in Fig. 6 show examples of each morphological cell had two major and several minor radiating, spiny dendrites
type. Figure 7 shows the full dendritic arborization of one repre- extending into lamina I and the dorsal white matter and also
sentative neuron of each major structural/functional category. into lamina II; such a cell has not been observed before.

nature neuroscience volume 1 no 3 july 1998 221

 1998 Nature America Inc. http://neurosci.nature.com

articles

21-633 heat + pinch + cold, M
32-740 heat + pinch + cold, M
27-551 heat + pinch + cold, M
20-910 pinch + cold, M
1998 Nature America Inc. http://neurosci.nature.com
44-1000 heat + pinch + cold, M
24-514 heat + pinch + cold, M
26-882 heat + pinch + cold, U
of a homeomorphic correspondence between these respec-
tive classes. This finding supports the view that there are
three basic types of lamina I neurons with correlated phys-
iological and morphological characteristics: fusiform NS
cells, pyramidal COLD cells, and multipolar HPC cells.
This evidence provides direct support for the concept
that pain and temperature are represented centrally by dis-
tinct sets of modality-selective neurons13. It indicates that
the functional and anatomical characteristics of these neu-
rons are linked and developmentally specified. We infer for
several reasons that these lamina I cell types are a general
mammalian feature. These morphological types have been
observed in rat, cat and monkey. The three physiological
classes have been demonstrated in cat and in monkey. In the
rat, physiological data from a lamina I projection target
(parabrachial nucleus34) indirectly support this idea. Fur-
thermore, a thermoreceptive-(COLD)-specific subnucleus
within trigeminal lamina I of the owl monkey has recently

been identified, and it contains almost entirely pyramidal

trigeminothalamic neurons, whereas fusiform and multi-
polar neurons are present in the adjacent portions of lamina
I, where NS and HPC cells are recorded (Blomqvist, A., E.-
100 M T.Z. & A.D.C. Soc. Neurosci. Abstr. 48.15, 1995).
The correspondence of the three morphological types with
Fig. 5. Camera-lucida drawings of the soma and proximal dendrites of the three major physiological classes validates the biological
each identified HPC cell. For each cell, the response characteristics and relevance of these respective categories. Further, this evidence
the morphological classification (M, multipolar; U, unclassified or transi- for discrete categories of lamina I neurons implies the exis-
tional) are indicated. The shading indicates processes that extended dor-
tence of other pharmacological and neurochemical correlates
sally or ventrally out of the plane of the horizontal section or that were
that, when uncovered, will facilitate the analysis of specific
more weakly labeled. Rostral is left, medial is up. Scale bar, 100 m.
pathways activated by painful and thermal stimuli. For exam-
ple, many NS lamina I cells that respond to substance P in the
rat35 are fusiform cells that bear NK1 receptors36,37, which are
Discussion found on very few pyramidal neurons (Yu, X.H., A.D.C. et al.,
Prior physiological and morphological findings indicate that lam- unpublished observations). Similarly, the COLD-specific subnu-
ina I neurons belong to different classes. There are at least three cleus within trigeminal lamina I of the owl monkey is devoid of
different physiological classes, which differ not only in their fibers immunoreactive for substance P or serotonin, in contrast to
response characteristics but also in the conduction velocities of the remainder of lamina I (Blomqvist, A., E.-T.Z. & A.D.C. Soc. Neu-
their ascending axons, the distribution of their terminals in the rosci. Abstr. 48.15, 1995). In addition, these cell types may contain
thalamus, their susceptibility to descending controls, and their different peptides in the rat38. Conversely, physiological inferences
responses to pharmacological modulation20,21,23,32,33. Similarly, can now be made based on anatomical knowledge of the types of
there are three basic morphological types of neurons that differ in lamina I cells that terminate in a region. For example, retrograde
their somatodendritic shape, the myelination of their axons and labeling work indicates that all three types of lamina I cells project to
their longitudinal distribution in the spinal cord2730. In the pre- the autonomic cell column in the ventrolateral medulla in cat; this
sent study, we re-examined the possibility of a correlation suggests that, in addition to nociceptive sensitivity39, neurons in this
between cell morphology and response characteristics with intra- region may respond to innocuous cooling, which was recently con-
cellular labeling in the cat, and we have obtained strong evidence firmed (Krout, K. & A.D.C., unpublished observations).

Fig. 6. Digital photomicrographs of intracellularly stained, identified lamina I neurons. F1, fusiform NS cell 28-1007, with a small arrowhead
indicating a second faintly labeled cell; F2, fusiform NS cell 22-820; P1, pyramidal COLD cell 35-1129; P2, pyramidal COLD cell 23-734; M1,
multipolar HPC cell 21-633; M2, multipolar NS cell 45-794; T1, T-shaped fusiform NS cell 17-755; U1, unclassified or transitional COLD cell
27-333. Scale bar, 100 m.

222 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com
Fusiform NS
28-1007
Pyramidal COLD
23-734
Multipolar HPC
24-514
1998 Nature America Inc. http://neurosci.nature.com
100 M
Fig. 7. Camera lucida reconstructions of the full dendritic extent of one exam- founded by unintentional thermal stimulation. Both
ple of each of the identified functional/morphological categories of lamina I neu- COLD cells and HPC cells with high cold sensitivity
rons. Nearly all of the dendrites were contained in two or three serial can be misconstrued as WDR cells if the temperatures
horizontal sections for each cell. Rostral is left, medial is up.
It must be noted that, although the morphological charac-
teristics of each of the three basic categories are generally con-
sistent, there is nonetheless variation within each. This may
reflect a degree of ambiguity, or incomplete selectivity, in devel-
opmental specification. This may also imply that subdivisions
of these basic categories of lamina I neurons remain to be iden-
tified. We consider this latter possibility rather likely, in part
because lamina I receives inputs from all tissues of the body
(including muscles, joints, viscera, cornea and teeth26,40; for
other references, see ref. 3), whereas we stimulated only skin,
and in part because subtypes of each category are consistently
recognizable morphologically2730. For example, in the present
sample we obtained almost no quadrilateral multipolar cells,
although these are common in the cat, which indicates that such
neurons may not receive cutaneous input.
Accordingly, the four exceptional NS cells that are multipo-
lar neurons could represent cells that received additional input
from untested tissues or that were sensitive to untested sub-
modalities that evince polymodal C fiber activity, such as chem-
ical or metabolic stimuli 4143 . It is also possible that these
multipolar cells were polymodal nociceptive HPC cells with high
thermal thresholds that were incompletely characterized, which
would conform with the remaining observations. The HPC lam-
ina I spinothalamic cells can have different sensitivities to the
submodalities of heat, pinch and cold, and some respond to cold
only below 15 C or to heat only above 50 C when tested with a
Peltier stimulator21,22 (unpublished observations). Such noxious
thermal stimuli can require a longer stimulus application than
nature neuroscience volume 1 no 3 july 1998
articles
was deemed practical during intracellular characteri-
zation in the present study.
The correspondence we have identified was not dis-
covered in numerous previous attempts to correlate
function and morphology in lamina I with intracellu-
lar HRP labeling816, probably due to methodological
reasons. First, sagittal or transverse sections were cut in
all of the previous studies. This precluded recognition
of the morphological types differentiated in the pre-
sent study, which can be consistently observed only in
the horizontal (or tangential) plane in which lamina
I neurons arborize. Second, innocuous and noxious
cold stimuli were not applied in most of the previous
studies. This precluded identification of the COLD
and HPC functional classes. Notably, cold stimuli were
used in the earliest study8, and the single lamina I cell
identified as specifically sensitive to innocuous cool-
ing was described as a pyramidal cell, although in
transverse sections. The one cell that was identified as
responsive to heat, pinch and cold (explicitly ascribed
to polymodal C nociceptor input, albeit perhaps sen-
sitized) had several dendrites extending horizontally
into lamina I, but its shape cannot be discerned from
the sagittal reconstruction.
In almost all of the other previous studies, cold
stimuli were not used, and lamina I neurons were char-
acterized only as NS or WDR cells. Significantly, the
physiological classification of lamina I cells can be con-
of the probes used for characterization are uncon-
trolled, because even a room-temperature probe can
evoke a graded low-threshold response in such cells,
whereas a neutral probe (warmed to skin temperature)
does not. The resemblance between the thermal sensitivity of
some WDR lamina I spinothalamic cells24 and that of HPC lam-
ina I spinothalamic cells22,23 supports this possibility. In addi-
tion, NS lamina I cells can also be misinterpreted as WDR cells,
because they can become sensitized to weak stimulation follow-
ing repeated noxious or C-fiber stimuli, such as commonly used
for searching26,44. These considerations underscore the recogni-
tion of three major physiological/morphological categories of
lamina I cells. Nevertheless, the possibility remains that WDR
lamina I cells, which have been observed more often in monkey
than in cat, might yet be distinguishable, and this should be
directly re-examined in light of the present observations.
Finally, the observation of distinct types of nociceptive and
thermoreceptive lamina I neurons is consistent with the concept
that these neurons supply substrates for discrete sensory chan-
nels for pain and temperature sensibilities. Prior comparative
psychophysical and physiological results obtained using the ther-
mal grill illusion (in which a sensation of burning, ice-like pain is
evoked by interlaced cool and warm stimuli) indicated that the
activity of COLD lamina I spinothalamic neurons can be asso-
ciated directly with cold sensation, whereas the activity of HPC
lamina I spinothalamic neurons can be associated with the burn-
ing pain caused by noxious cold or unmasked by the thermal
grill6,22. Furthermore, other work indicates that these neurons
can provide the input for modality-selective somato-autonomic
(homeostatic) reflexes to painful and thermal stimuli45,46. These
findings have led to the proposal that the fundamental role of
lamina I neurons is to distribute modality-selective sensory infor-
223
 1998 Nature America Inc. http://neurosci.nature.com
articles
mation on the physiological status of the tissues of the entire
mammalian body to both sensory and homeostatic substrates. The
present observations provide firm evidence supporting this con-
cept and argue that modality-selective lamina I projection neu-
rons may justifiably be considered to be anatomically specified.
Methods
GENERAL PREPARATION. These procedures were approved by the local
institutional animal care and use committee. Acute recording experi-
ments were performed in 41 adult cats of either sex (2.5-4.5 kg) pre-
pared as described2022. Briefly, the animals were anesthetized with
pentobarbital (40 mg per kg i.p.). Cannulae were inserted in the right
cephalic vein (for infusions) and left carotid artery (for the blood pres-
sure monitor). Dexamethasone (10 mg i.v.) was administered. Sup-
plemental pentobarbital (approximately 4 mg per kg per h) was given
as needed to maintain areflexia and pupillary constriction and blood
pressure near 120 mmHg. A tracheostomy was performed, the animal
was artificially ventilated with 25% O 2 and 75% air, and a bilateral
1998 Nature America Inc. http://neurosci.nature.com
pneumothorax was induced. End-tidal CO 2 was maintained at
3.54.5%. Body temperature was maintained at 37.5C with a heating
pad and a feedback-controlled infrared lamp. Pancuronium was given
every 6090 min for paralysis; withdrawal reflexes were tested between
doses. A dorsal laminectomy exposed the lumbosacral enlargement,
and a craniectomy was performed to provide access to the right thal-
amus. The animal was mounted in a stereotaxic holder, and the spinal
cord was stabilized with vertebral clamps. The dura was reflected, and
the cord was covered with a pool of mineral oil or Tyrodes solution
that was maintained at 3738C with a DC heater element.
ANTIDROMIC STIMULATION. Extracellular recordings were made with
tungsten-in-glass microelectrodes to identify the right somatosenso-
ry thalamus in 22 of these cats. Based on these coordinates, an array of
eight concentric bipolar electrodes was implanted in order to
antidromically activate lamina I spinothalamic neurons, as in prior
studies. Stimulus trains consisting of four pulses (0.12 mA, 2 ms) at
200 or 250 Hz were delivered to individual electrodes or between pairs
of electrodes. Spinal cells that responded to such trains in a one-for-
one, time-locked manner at a fixed latency with a distinct threshold
were recognized as spinothalamic neurons; we considered this suffi-
cient, because lamina I cells that fulfill this criterion have in all our
prior studies demonstrated a strict antidromic collision interval when
tested. Extracellular recordings were made from the lumbosacral
spinal cord at sites in L7S1 exposed by small openings in the pia, in
order to ensure that lamina I spinothalamic cells were antidromical-
ly activated by the thalamic array and to localize regions containing
lamina I cells with receptive fields on the ventral hindpaw for subse-
quent intracellular recording.
INTRACELLULAR RECORDINGS. Recordings were made with glass micro-
electrodes fabricated from 1.0 mm standard or 1.2 mm borosilicate glass
pipettes and an Axoclamp 2-A (Axon Instruments) amplifier in bridge
mode. The electrodes were back-filled with 2% biocytin (Sigma) in 2 M
potassium methylsulphate; their resistance ranged from 40 to 80 M.
The electrodes were lowered into the superficial dorsal horn near the
dorsal root entry zone in 2.0 or 2.5 m steps. Cells were sought at a depth
of 200-800 m below the surface. Extracellular recordings were obtained
infrequently, using antidromic stimulation as a search method. Standard
methods of current application through the microelectrode were used
to facilitate cell impalement and to clear debris from the electrode tip.
Intracellular recording was signaled by a sudden large action potential,
a resting membrane potential of at least -40 mV and action potentials of
45 to 65 mV. Hyperpolarizing current pulses (0.5 nA, 50 ms) were often
applied in an effort to stabilize the recording.
RESPONSE CHARACTERIZATION. Following cell impalement, antidromic
activation was tested and natural stimuli were quickly used for unit char-
acterization2023. Innocuous stimuli (light touch, tapping, gentle pres-
sure, limb movement) were generally tested first, followed by thermal
stimuli (cooling with a wet ice cube or cold beaker, warming with the
hand or a radiant heat lamp) and then noxious stimuli (pressure and
224
pinch applied with fingers or a smooth forceps, noxious heat). The cuta-
neous stimuli were applied at various sites on the ventral hindpaw for
35 s; the intensity of the radiant heat lamp was adjusted such that, held
at a fixed distance, it became painful to the investigators after about 5 s
(about 45C) and intolerable after about 8 s (about 50C). Responses
were monitored through a speaker and an oscilloscope and were record-
ed on magnetic tape.
Compromise between the competing needs for proper response char-
acterization and for quick initiation of current injection for intracellular
staining meant that stimuli were repeated only three or four times to
demonstrate reproducibility, that receptive fields were regionally but not
precisely delimited, and that thermal stimuli were not always maintained
sufficiently long (cold, 3060 s; heat, 510 s) to generate the noxious
stimulus required by some HPC cells2123. Cells that responded to low-
threshold cutaneous stimuli, cells with high ongoing discharge, and cells
that were silent or unresponsive were also encountered, but these were
not studied in detail in these experiments, primarily because prior stud-
ies indicated that lamina I spinothalamic cells in cat do not include such
cells. These cells were occasionally stained, however, and they were sub-
sequently found to lie in lamina II or deeper.
INTRACELLULAR STAINING. Following physiological characterization of
a cell, depolarizing current pulses (0.5 to 5.0 nA, 500-ms duration at
1 Hz for 1 to 7 min) were applied to inject biocytin. In most experi-
ments, several intracellular injections were attempted at sites separat-
ed longitudinally by 0.5 to 5 mm that were carefully mapped for later
reconstruction. A period of one to eight hours was allowed following
the injections in order to permit diffusion of the biocytin into cell
processes. The animals were then given an overdose of barbiturate and
perfused transcardially with a rinse of 0.1 M PBS (pH 7.4, room tem-
perature), a fixative consisting of two liters of 4% paraformaldehyde
and 0.1% glutaraldehyde in 0.1 M PB, and then one liter of 10%
sucrose in 0.1 M PB. The lumbosacral spinal cord was removed and
placed in 30% sucrose in PB overnight. Serial 50-m frozen sections
cut in the horizontal plane were rinsed, incubated in avidin-biotiny-
lated horseradish peroxidase complex (ABC, 1:100, Vector) in 0.05 M
Tris-buffered saline (TBS) at 4 C overnight, and reacted with 0.05%
3,3-diaminobenzidine (DAB) containing 0.01% hydrogen peroxide
in TBS for 510 min. Labeled cells were identified in the light micro-
scope as lamina I neurons if they were located in the most superficial
aspect of the gray matter, dorsal to the substantia gelatinosa, which is
clearly recognizable with light- and dark-field illumination. Pho-
tographs were made of all cells with Kodak Technical Pan film and
Ektachrome film using 20x and 40x apochromatic objectives. Digital
microscopic images were made directly with a Leaf Microlumina scan-
ner (at 3380 x 2253 pixels) and a 20x objective, and Adobe Photoshop
was used to enhance contrast and apply labels. Camera lucida draw-
ings were made using 40x and 60x (oil) objectives.
In most cases, only one biocytin-labeled cell was found at the
appropriate location, but for 9 of the 30 identified cells, one to four
neighboring cells were also labeled. These were usually small, light-
ly labeled fusiform neurons (Fig. 6, F1, arrowhead), in contrast to the
darkly labeled cell that we identified as the characterized neuron.
(Interestingly, this occurred most often, in five of these nine instances,
when the identified cell was also a fusiform neuron.) In two instances,
we found two adjacent, well labeled cells that had similar shapes (once
fusiform neurons, once multipolar neurons), but one was more dark-
ly stained and had signs of penetration (a darkly stained, disrupted
section of membrane). In one instance, four cells with differing shapes
were labeled with similar intensity and without obvious signs of pen-
etration in any particular one, and thus the physiologically charac-
terized cell could not be identified and the data were excluded.
Acknowledgements
We thank Elizabeth OCampo, Maribeth Tatum, and Jan Carey for assistance.
This work was supported by NIH grants NS 25616 and DA 07402 and the
James R. Atkinson Pain Research Fund administered by the Barrow
Neurological Foundation.
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com
1. Perl, E.R. in Handbook of Physiology, Section 1, The Nervous System, Volume
III, Sensory Processes (ed. Darian-Smith, I.) 915975 (American Physiological
Society, Bethesda, 1984).
2. Willis, W.D. The Pain System. (Karger, Basel, 1985).
3. Craig, A.D. in Somesthesis and the Neurobiology of the Somatosensory Cortex
(eds Franzen, O., Johansson, R. & Terenius, L.) 2739 ( Birkhuser, Basel,
1996).
4. Craig, A.D., Bushnell, M.C., Zhang, E.-T. & Blomqvist, A. A thalamic nucleus
specific for pain and temperature sensation. Nature 372, 770773 (1994).
5. Coghill, R.C. et al. Distributed processing of pain and vibration by the human
brain. J.Neurosci. 14, 40954108 (1994).
6. Craig, A.D., Reiman, E.M., Evans, A. & Bushnell, M.C. Functional imaging of
an illusion of pain. Nature 384, 258260 (1996).
7. Wall, P.D. Pain in the brain and lower parts of the anatomy. Pain 62, 389391
(1995).
8. Light, A.R., Trevino, D.L. & Perl, E.R. Morphological features of functionally
defined neurons in the marginal zone and substantia gelatinosa of the spinal
dorsal horn. J. Comp. Neurol. 186, 151172 (1979).
9. Light, A.R., Sedivec, M.J., Casale, E.J. & Jones, S.L. Physiological and
morphological characteristics of spinal neurons projecting to the
parabrachial region of the cat. Somatosens. Mot. Res. 10, 309325 (1993).
10. Bennett, G.J., Abdelmoumene, M., Hayashi, H., Hoffert, M.J. & Dubner, R.
Spinal cord layer I neurons with axon collaterals that generate local arbors.
Brain Res. 209, 421426 (1981).
1998 Nature America Inc. http://neurosci.nature.com
11. Molony, V., Steedman, W.M., Cervero, F. & Iggo, A. Intracellular marking of
identified neurones in the superficial dorsal horn of the cat spinal cord. Q. J.
Exp. Physiol. 66, 211-223 (1981).
12. Hoffert, M.J., Miletic, V., Ruda, M.A. & Dubner, R. Immunocytochemical
identification of serotonin axonal contacts on characterized neurons in
laminae I and II of the cat dorsal horn. Brain Res. 267, 361364 (1983).
13. Woolf, C.J. & Fitzgerald, M. The properties of neurones recorded in the
superficial dorsal horn of the rat spinal cord. J. Comp. Neurol. 221, 313328
(1983).
14. Miletic, V., Hoffert, M.J., Ruda, M.A., Dubner, R. & Shigenaga, Y.
Serotoninergic axonal contacts on identified cat spinal dorsal horn neurons
and their correlation with nucleus raphe magnus stimulation. J. Comp.
Neurol. 228, 129141 (1984).
15. Steedman, W.M., Molony, V. & Iggo, A. Nociceptive neurones in the
superficial dorsal horn of cat lumbar spinal cord and their primary afferent
inputs. Exp. Brain Res. 58, 171182 (1985).
16. Hylden, J.L.K., Hayashi, H., Dubner, R. & Bennett, G.J. Physiology and
morphology of the lamina I spinomesencephalic projection. J. Comp. Neurol.
247, 505515 (1986).
17. Moschovakis, A.K., Burke, R. & Fyffe, R. The size and dendritic structure of
HRP-labeled gamma motoneurons in the cat spinal cord. J. Comp. Neurol.
311, 531545 (1991).
18. Tamamaki, N., Uhrich, D.J. & Sherman, S.M. Morphology of physiologically
identified retinal X and Y axons in the cats thalamus and midbrain as revealed
by intraxonal injection of biocytin. J. Comp. Neurol. 354, 583607 (1995).
19. Christensen, B.N. & Perl, E.R. Spinal neurons specifically excited by noxious
or thermal stimuli: marginal zone of the dorsal horn. J. Neurophysiol. 33,
293307 (1970).
20. Craig, A.D. & Hunsley, S.J. Morphine enhances the activity of
thermoreceptive cold-specific lamina I spinothalamic neurons in the cat.
Brain Res. 558, 9397 (1991).
21. Craig, A.D. & Serrano, L.P. Effects of systemic morphine on lamina I
spinothalamic tract neurons in the cat. Brain Res. 636, 233244 (1994).
22. Craig, A.D. & Bushnell, M.C. The thermal grill illusion: unmasking the burn
of cold pain. Science 265, 252255 (1994).
23. Dostrovsky, J.O. & Craig, A.D. Cooling-specific spinothalamic neurons in the
monkey. J. Neurophysiol. 76, 36563665 (1996).
24. Ferrington, D.G., Sorkin, L.S. & Willis, W.D. Responses of spinothalamic
tract cells in the superficial dorsal horn of the primate lumbar spinal cord. J.
Physiol. (Lond.) 388, 681703 (1987).
nature neuroscience volume 1 no 3 july 1998
articles
25. Price, D.D., Hayes, R.L., Ruda, M. & Dubner, R. Spatial and temporal
transformations of input to spinothalamic tract neurons and their relation to
somatic sensations. J. Neurophysiol. 41, 933947 (1978).
26. Craig, A.D. Jr. & Kniffki, K.-D. Spinothalamic lumbosacral lamina I cells
responsive to skin and muscle stimulation in the cat. J. Physiol. (Lond.) 365,
197221 (1985).
27. Gobel, S. Golgi studies of the neurons in layer I of the dorsal horn of the
medulla (trigeminal nucleus caudalis). J. Comp. Neurol. 180, 375394 (1978).
28. Lima, D. & Coimbra, A. A Golgi study of the neuronal population of the
marginal zone (lamina I) of the rat spinal cord. J. Comp. Neurol. 244, 5371
(1986).
29. Zhang, E.T., Han, Z.S. & Craig, A.D. Morphological classes of spinothalamic
lamina I neurons in the cat. J. Comp. Neurol. 367, 537549 (1996).
30. Zhang, E.T. & Craig, A.D. Morphology and distribution of spinothalamic
lamina I neurons in the monkey. J. Neurosci. 17, 32743284 (1997).
31. Lopez-Garcia, J.A. & King, A.E. Membrane properties of physiologically
classified rat dorsal horn neurons in vitro: Correlation with cutaneous
sensory afferent input. Eur. J. Neurosci. 6, 9981007 (1994).
32. Dostrovsky, J.O., Shah, Y. & Gray, B.G. Descending inhibitory influences
from periaqueductal gray, nucleus raphe magnus, and adjacent reticular
formation. II. Effects on medullary dorsal horn nociceptive and
nonnociceptive neurons. J. Neurophysiol. 49, 948960 (1983).
33. Mokha, S.S., Goldsmith, G.E., Hellon, R.F. & Puri, R. Hypothalamic control
of nocireceptive and other neurones in the marginal layer of the dorsal horn
of the medulla (trigeminal nucleus caudalis) in the rat. Exp. Brain Res. 65,
427436 (1987).
34. Menendez, L., Bester, H., Besson, J.M. & Bernard, J.F. Parabrachial area:
Electrophysiological evidence for an involvement in cold nociception. J.
Neurophysiol. 75, 20992116 (1996).
35. De Koninck, Y. & Henry, J.L. Substance P-mediated slow excitatory
postsynaptic potential elicited in dorsal horn neurons in vivo by noxious
stimulation. Proc. Natl Acad. Sci. USA 88, 1134411348 (1991).
36. Brown, J.L. et al. Morphological characterization of substance P receptor-
immunoreactive neurons in the rat spinal cord and trigeminal nucleus
caudalis. J. Comp. Neurol. 356, 327344 (1995).
37. Marshall, G.E., Shehab, S.A.S., Spike, R.C. & Todd, A.J. Neurokinin-1 receptors
on lumbar spinothalamic neurons in the rat. Neuroscience 72, 255263 (1996).
38. Lima, D., Avelino, A. & Coimbra, A. Morphological characterization of
marginal (lamina I) neurons immunoreactive for substance P, enkephalin,
dynorphin and gamma-aminobutyric acid in the rat spinal cord. J. Chem.
Neuroanat. 6, 4352 (1993).
39. Sun, M.-K. & Spyer, K.M. Nociceptive inputs into rostral ventrolateral
medulla-spinal vasomotor neurones in rats. J. Physiol. (Lond.) 436, 685700
(1991).
40. Cervero, F. & Tattersall, J.E.H. Somatic and visceral inputs to the thoracic
spinal cord of the cat: marginal zone (lamina I) of the dorsal horn. J. Physiol.
(Lond.) 383, 383395 (1987).
41. MacIver, M.B. & Tanelian, D.L. Activation of C fibers by metabolic
perturbations associated with tourniquet ischemia. Anesthesiology 76,
617623 (1992).
42. Pickar, J.G., Hill, J.M. & Kaufman, M.P. Dynamic exercise stimulates group
III muscle afferents. J. Neurophysiol. 71, 753760 (1994).
43. Schmelz, M., Schmidt, R., Bickel, A., Handwerker, H.O. & Torebjrk, H.E.
Specific C-receptors for itch in human skin. J. Neurosci. 17, 80038008
(1997).
44. Woolf, C.J., Shortland, P. & Sivilotti, L.G. Sensitization of high
mechanothreshold superficial dorsal horn and flexor motor neurones
following chemosensitive primary afferent activation. Pain 58, 141155
(1994).
45. Craig, A.D. Propriospinal input to thoracolumbar sympathetic nuclei from
cervical and lumbar lamina I neurons in the cat and the monkey. J. Comp.
Neurol. 331, 517530 (1993).
46. Craig, A.D. Distribution of brainstem projections from spinal lamina I
neurons in the cat and the monkey. J. Comp. Neurol. 361, 225248 (1995).
225
 1998 Nature America Inc. http://neurosci.nature.com

articles

Cortically induced thalamic plasticity in

the primate somatosensory system
E.R. Ergenzinger1,2, M.M. Glasier1, J.O. Hahm3 and T.P. Pons1

1 Department of Neurosurgery and 2Program in Neuroscience, Wake Forest University School of Medicine, Winston-Salem, North Carolina 27127, USA
3 Department of Neurosurgery, Georgetown University, Washington, D.C. 20057, USA
Correspondence should be addressed to T.P.P. (tpons@wfubmc.edu)

The influence of cortical feedback on receptive field organization in the thalamus was assessed in
the primate somatosensory system. Chronic and acute suppression of neuronal activity in primary
somatosensory cortex resulted in a striking enlargement of receptive fields in the ventroposterior
1998 Nature America Inc. http://neurosci.nature.com

thalamus. This finding demonstrates a dramatic top-down influence of cortex on receptive field
size in the somatosensory thalamus. In addition, this result has important implications for studies of
adult neuronal plasticity because it indicates that changes in higher-order areas of the brain can
trigger extensive changes in the receptive field characteristics of neurons located earlier in the
processing pathway.

Although the descending projection from the primary consistent with those reported in previous studies by us and oth-
somatosensory cortex to the ventroposterior nucleus (VP) of the ers in normal animals1012. In accordance with previous stud-
thalamus is known to be seven to ten times greater than the ies9, our recordings throughout the cortical hand representation
ascending projection from VP to cortex1, the functional role of one to five months after D-APV administration revealed that
this massive corticofugal projection has remained elusive2. Pre- only 10% of cortical recording sites were responsive to somatic
vious studies on the contribution of neural feedback connections stimulation (Table 1), compared to 100% in control animals.
to the processing of sensory information have reported relative- Analysis of the responsive 10% of recording sites in the cortex of
ly minor and inconsistent effects on receptive field size or the experimental animals showed that these neurons exhibited
response properties of thalamic relay neurons following cortical RFs comparable in size to those found in the cortex of the control
perturbations37. Although the connectivity of the system sug- animals. There was no evidence of degeneration in the thalamus
gests a substantial influence from cortex over the processing of or cortex in any of the animals used in the present study.
somatosensory information at the level of the thalamus, there These results on RF size obtained at the cortical level in the
has been no definitive demonstration of a major cortical influ- chronic D-APV treated animals contrasted sharply with those
ence on receptive field size, response properties or somatotopic found at the thalamic level of these same animals, where RFs
organization in VP. were greatly enlarged compared to those seen in controls, fre-
A majority of corticofugal projection cells in primary quently encompassing more than one digit and often half or
somatosensory cortex (area 3b) express NMDA receptors8, and more of the entire glabrous hand (Fig. 1c). In some instances
chronic systemic blockade of NMDA receptors produces large
unresponsive areas in somatosensory cortex9. In this study, we
determined that chronic administration of an NMDA receptor Table 1. Responsivity of recording sites in cortex and
antagonist directly into the area 3b cortical hand representa- thalamus following chronic cortical administration of D-
tion results in a suppression of activity in area 3b and an enor- APV or saline1
mous enlargement of receptive fields (RFs) in the VP hand Control D-APV
representation. In addition, acute cortical suppression also Cortex (Area 3b)
resulted in an enlargement of RFs in VP. These findings force a % Responsive Sites
re-evaluation of traditional bottom-up models of sensory pro- Infusion Zone (Hand) 100%(197) 10%(754)
cessing, which view the thalamus as a simple relay nucleus to Other 100%(47) 99%(82)
the cortex, and they also have important implications for stud-
Fringe Zone (Hand & Other) N.A. 78%(262)
ies of adult neuronal plasticity.
% Large RF Sites 0%(197) 6%(78)
Results Thalamus (VP)
Over a period of one to five months, either saline (two control % Responsive Sites
monkeys, Macaca mulatta) or the NMDA receptor antagonist D- Hand 100%(102) 99%(388)
2-amino-5-phosphonovaleric acid (D-APV; five monkeys, Maca- Other 99%(116) 100%(243)
ca mulatta) was infused directly into the area 3b hand
% Large RF Sites 2%(102) 58%(383)
representation. The percentage of responsive sites and RF size for
the cortex and VP in the saline control group was completely 1see Methods for details on data analysis.

226 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

a b c

Fig. 1. Representative RFs in VP following chronic administration of D-APV to somatosensory cortex. (a) Representative RFs, indicated by
thinner black lines, recorded from sites localized within the hand region of VPL from saline controls. (b) Representative electrode tracks
through a coronal section of the thalamus in a macaque that had received D-APV to the area 3b hand representation (CM, central median; LP,
lateral posterior; VPM, ventroposterior medial; VPI, ventroposterior inferior; VPL, ventroposterior lateral). Tick marks on tracks indicate
recording sites. Red, recording sites with RFs on the face. Blue, recording sites with RFs on or including the hand. Green, recording sites with
1998 Nature America Inc. http://neurosci.nature.com

RFs on the forearm. Colors are matched with RFs in (c). (c) Representative RFs recorded from sites localized within VP from a monkey with
D-APV administered to area 3b. All RFs were contralateral to VP. Arrow denotes rotation of the hand and forearm to allow visualization of
their ventral surfaces.

these RFs included the entire distal forearm and hand (Fig. 1c), a
type of RF that is never seen in normal animals. Such large RFs
were recorded from 58% of the recording sites within the VP
hand representation of our experimental animals. The RF
enlargement was specific in that these RFs were observed only
for recording sites within the VP hand representation and not in
body-part representations outside of the D-APV infusion zone
in cortex (Fig. 1c).
These findings indicate that chronic D-APV administration
in the area 3b hand representation blocks much of the stimulus-
driven activity at the cortical level and tremendously expands the
size of RFs in the portion of the VP thalamus that normally rep-
resents the hand (within the ventroposterior lateral nucleus or
VPL). This expansion of RF size occurred for the vast majority
of hand-responsive recording sites throughout VPL (Fig. 2). Such
70 Control
D-APV
60
% recording sites
an enormous expansion of RF size after delivery of pharmaco-
logical agents to either the cortex or thalamus has not been
reported previously37.
We next determined if acute administration of D-APV would
produce results similar to those from chronic treatment. We first
recorded from the cortex and thalamus of two additional mon-
keys before administering D-APV. We then delivered injections
of D-APV directly to the cortex of these same animals via a
Hamilton syringe and immediately assessed the results. Acute
cortical recordings were similar to those found in the chronical-
ly administered animals, with 93% of 83 sites unresponsive to
somatic stimulation. Recordings in the thalamus of the acutely
administered animals showed an increase in RF size, though not
as great as that observed in the chronically administered animals.
Although a comparable number of sites (45% of 55 sites) showed
RFs that encompassed one or more digits, compared to only 2%
of 51 sites before acute D-APV administration, the magnitude
of the RF enlargement was not as great as that observed in chron-
ically administered animals, in that no RFs larger than half of the
hand were observed (Fig. 3).

50
Discussion
40 The present findings demonstrate that, at least under our exper-
imental conditions, top-down projections from cortex can induce
30
large-scale reorganization of RFs at a relatively early processing
20 station in the somatosensory system. This effect seems most like-
ly to be mediated directly by corticothalamic projections or indi-
10
rectly via corticocuneate projections. Additionally, the increase
0 in RF size for VP neurons could occur through a loss of excita-
1 Pad < 1 Digit > 1 Pad 1 Digit tory corticothalamic input on inhibitory interneurons within VP
or on the thalamic reticular nucleus. Of course, unknown changes
in additional somatosensory cortical areas such as SII, the insu-
Fig. 2. Distribution histogram of RF sizes for recording sites local- la and/or posterior parietal cortex could conceivably mediate the
ized within the VPL hand representation between control and D- RF enlargement through complex corticothalamic processing,
APV groups. RFs were categorized as to whether they were 1 pad, though this latter possibility seems less likely. That these results
< 1 digit, > 1 pad, or 1 digit. RFs that were classified as > 1 pad might be explained by retrograde degeneration seems highly
were restricted to the pads and did not encompass any digits of the unlikely, given that lesions of the somatosensory cortex, which
hand. RFs that were classified as 1 digit were those RFs that incor- are known to produce massive retrograde degeneration, do not
porated at least one digit, and included RFs as large as the entire result in expanded thalamic RFs13. Furthermore, the expansion of
hand and forearm (see Fig. 1). RFs immediately after delivery of D-APV to the cortex would

nature neuroscience volume 1 no 3 july 1998 227

 1998 Nature America Inc. http://neurosci.nature.com
articles
a 80 Pre
Post
b hydrochloride (15 mg per kg) fol-
70
% recording sites
60
50
40
30
20
10
0 maintained within normal limits.
1 Pad < 1 Digit > 1 Pad 1 Digit
Fig. 3. RF size in VPL before and after acute cortical D-APV administration. (a) Distribution histogram of
RF sizes for recording sites localized within the VPL hand representation before (pre) and after (post)
acute cortical D-APV administration. RFs were categorized as in Fig. 2. (b) Representative RFs recorded
from sites localized within VPL before (pre) and after (post) acute cortical D-APV administration.
1998 Nature America Inc. http://neurosci.nature.com
seem to further eliminate retrograde degeneration as a possible
mechanism explaining our results.
Although some earlier electrophysiological studies of the nor-
mal thalamus in monkeys have reported extra-lemniscal neu-
rons with large RFs 11,12 , such RFs were localized almost
exclusively in the VPLo (oralis) nucleus11, far anterior to the VPLc
(caudalis)14 where our recordings were made. In addition, record-
ings lateral and medial to the affected thalamic zone indicated
neurons had normal RF size, further corroborating that we were
recording within VPLc, and not in the extralemniscal pathway.
Finally, histological analysis from our cases confirmed that our
recordings were from the hand region in VPLc. Importantly, there
was no evidence of any degeneration of neurons in the thalamus
or cortex. Regardless of the precise mechanism(s) responsible for
the expansion, however, the magnitude of the RF expansion is
completely unexpected and highlights the contribution of feed-
back processing loops on RF properties.
Many previous studies of adult neural plasticity after periph-
eral perturbations have focused attention on the earliest point in
the ascending pathway where plastic changes could occur (spinal
cord15,16, brainstem17,18, or thalamus1921), with the interpreta-
tion that plastic changes at early stations are simply relayed to
cortex. The timing, nature and magnitude of our present find-
ings challenge this view of the system as a simple hierarchical
pathway22,23, providing a definitive demonstration of a dramat-
ic and substantial role for the cortex on neuronal processing early
in the somatosensory pathway. This study demonstrates that
some RF characteristics within somatosensory pathways result
from a series of interconnected dynamic loops, with changes at
any given level capable of triggering extensive changes in the RFs
of neurons at both earlier and later stations in the processing
chain. Such a view of feedback connections is entirely consistent
with a recent hypothesis24 that recognizes a substantial role for
corticothalamic feedback loops on the modification of receptive
properties but not for the major driving of thalamic neurons by
the cortex. The extent to which such feedback connections mod-
ulate activity and the precise mechanisms responsible for such
modulation should be ripe areas for future research.
Methods
SURGICAL PROCEDURES. All procedures in the present study were approved
by the Wake Forest University Animal Care and Use Committee. All mon-
keys used in this study were Macaca mulatta. Monkeys were anesthetized
228
with an initial dose of ketamine
lowed by intubation and admin-
istration of isoflurane (0.5% to
3%) to effect. Throughout
surgery, which was performed
using aseptic precautions, the ani-
mals received an intravenous drip
of a solution of 5% dextrose and
0.45% sodium chloride, and their
heart rate, respiratory rate, and
temperature were monitored and
Pre Post
For the chronic administration
of D-APV, we implanted a tran-
scranial catheter directly into the
hand representation in area 3b and
attached the catheter to an osmot-
ic pump containing 50 M D-APV
or physiological saline delivered at
a rate of 2.5 l per h. D-APV or
saline was administered from
between one and five months, and the osmotic pump was replaced every
28 days. Ketamine was not used during osmotic pump replacements because
of potential interactions with glutamatergic neurotransmission. Instead,
valium was used to make the animal receptive to isoflurane anesthesia.
For the acute administration of D-APV, a Hamilton syringe was used
to inject D-APV directly into the hand representation in area 3b. In order
to assure that the entire hand representation was affected by the D-APV
as in the chronic study, multiple injections of 23 l of D-APV were made
approximately 1 mm apart across the mediolateral extent of the hand
representation in cortex (56 injections total). The injections consisted of
50 M D-APV.
ELECTROPHYSIOLOGICAL RECORDING PROCEDURES. A recording chamber and
head fixation device were attached to the skull. The chamber was made
from dental acrylic and positioned to provide maximum access for
recording from VP and area 3b. The bone within the chamber was
removed to expose the dura, the dura then reflected and a high resolu-
tion picture of brain vasculature was then obtained with a CCD camera
using appropriate filters. The brain picture was used to help in the place-
ment of electrode penetrations during recording, with the RF and respon-
sivity characteristics defined for each recording point.
The mapping and recording procedures were similar to those used
previously25. In the chronic preparation, electrode penetrations were
placed at 0.30.5 mm intervals across the mediolateral extent of VP and
area 3b and at 0.30.5 mm intervals across its rostrocaudal extent.
Approximately 2040 electrode penetrations were made across VP in
each hemisphere studied, and approximately 5060 electrode penetra-
tions were made through area 3b in each hemisphere studied. In the acute
preparation, 45 penetrations were made across VP, and 56 penetra-
tions were made across area 3b before and after cortical D-APV injec-
tions. Penetrations across VP were placed at 0.30.5 mm intervals
mediolaterally and rostrocaudally. Penetrations across area 3b were placed
approximately 1 mm apart mediolaterally. The location of pre- and post-
injection penetrations were the same. Microelectrodes were hydraulical-
ly advanced through cortex and thalamus until single- or multiple-unit
responses to mechanical stimulation of the body could be isolated. Small
marker lesions (10 A for 10 s) were placed in the thalamus and cortex at
strategic points to assist in locating the recordings sites post mortem.
H ISTOLOGICAL ANALYSIS . On completion of all electrophysiological
recording experiments, animals were given a lethal dose of pentobar-
bital and then perfused intracardially with 0.9% saline followed by 4%
paraformaldehyde. Brains were cut in the coronal plane on a freezing
microtome into 50 m sections, and every fifth section was mounted
and stained for Nissl substance to assess the placement of recording
tracks, marker lesions, and placement of the cannulae in cortex.
Recording sites were identified by reconstructing electrode tracks and
marker lesions as described25.
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com
DATA ANALYSIS. Percent responsive sites was determined as the number of
responsive sites per total number of sites. Percent large RF sites are given
as number of sites with RFs encompassing one or more digits (includ-
ing RFs encompassing the hand and forearm) per number of hand-
responsive sites. In the chronic D-APV animals the infusion zone was
defined as the region of cortex surrounding the cannula site in which
responsiveness was largely suppressed (90% of recording sites) and cor-
responded to a 5 mm mediolateral expanse of cortex within the hand
representation. The fringe zone was defined as the region in which
responsiveness was only partially suppressed (25% of sites) and extend-
ed approximately 1 mm medial and lateral to the infusion zone. Corti-
cal recordings indicated that the face representation was not affected by
the D-APV administration.
Acknowledgements
This research was supported by NIH grants MH11950-01, MH53369-02 and
NS35246-01.
1998 Nature America Inc. http://neurosci.nature.com
RECEIVED 20 MAY: ACCEPTED 26 MAY 1998
1. Liu, X.B., Honda, C.N. & Jones, E.G. Distribution of four types of synapse on
physiologically identified relay neurons in the ventral posterior thalamic
nucleus of the cat. J. Comp. Neurol. 352, 6991 (1995).
2. Jones, E.G. The Thalamus. (Plenum, New York, 1985).
3. Yuan, B., Morrow, T.J. & Casey, K.L. Responsiveness of ventrobasal thalamic
neurons after suppression of S1 cortex in the anesthetized rat. J. Neurosci. 5,
29712978 (1985).
4. Ghosh, S., Murray, G.M., Turman, A.B. & Rowe, M.J. Corticothalamic
influences on transmission of tactile information in the
ventroposterolateral thalamus of the cat: effect of reversible inactivation of
somatosensory cortical areas I and II. Exp. Brain. Res. 100, 276286 (1994).
5. Burchfiel, J.L. & Duffy, F.H. Corticofugal influences upon cat thalamic
ventrobasal complex. Brain Res. 70, 395411 (1974).
6. Shin, H.C. & Chapin, J.K. Mapping the effects of SI cortex stimulation on
somatosensory relay neurons in the rat thalamus: direct responses and
afferent modulation. Somatosens. Motor Res. 7, 421434 (1990).
7. Tsumoto, T. & Nakamura, S. Inhibitory organization of the thalamic
ventrobasal neurons with different peripheral representations. Exp. Brain
Res. 21, 195210 (1974).
nature neuroscience volume 1 no 3 july 1998
articles
8. Conti, F. & Minelli, A. in Excitatory Amino Acids and the Cerebral Cortex. (eds
Conti, F. & Hicks, T.P.) 8198 (MIT Press, Cambridge, Massachusetts, 1996).
9. Garraghty, P.E. & Muja, N. NMDA receptors and plasticity in adult
primate somatosensory cortex. J. Comp. Neurol. 367, 319326 (1996).
10. Pons, T.P., Wall, J.T., Garraghty, P.E., Cusick, C.G. & Kaas, J.H.
Consistent features of the representation of the hand in area 3b of
macaque monkeys. Somatosensory Res. 4, 309331 (1987).
11. Poggio, G.F. & Mountcastle, V.B. The functional properties of
ventrobasal thalamic neurons studied in unanesthetized monkeys. J.
Neurophysiol. 26, 775806 (1963).
12. Loe, P.R., Whitsel, B.L., Dreyer, D.A. & Metz, C.B. Body representation in
ventrobasal thalamus of macaque: a single-unit analysis. J. Neurophysiol.
40, 13391355 (1977).
13. Bava, A., Fadiga, E. & Manzoni, T. Extralemniscal reactivity and
commisural linkages in the VPL nucleus of cats with chronic cortical
lesions. Arch. Ital. Biol. 106, 204226 (1968).
14. Kaas, J.H. & Pons, T.P. in Comparative Primate Biology, Neurosciences.
(eds Steklis, H.P. & Erwin, J.) 421468 (Alan R. Liss, New York, 1988).
15. Dostrovsky, J.O., Millar, J. & Wall, P.D. The immediate shift of afferent
drive of dorsal column nucleus cells following deafferentation: A
comparison of acute and chronic deafferentation in gracile nucleus and
spinal cord. Exp. Neurol. 52, 480495 (1976).
16. McMahan, S. B. & Wall, P.D. Plasticity in the nucleus gracilis of the rat.
Exp. Neurol. 80, 195207 (1983).
17. Pollin, B. & Albe-Fessard, P. Organization of somatic thalamus in
monkeys with and without section of dorsal spinal track. Brain Res. 173,
431449 (1979).
18. Garraghty, P.E. & Kaas, J.H.. Functional reorganization in adult monkey
thalamus after peripheral nerve injury. Neuroreport 2, 747750 (1991).
19. Merzenich, M.M. et al. Topographic reorganization of somatosensory
cortical areas 3b and 1 in adult monkeys following restricted
deafferentation. Neuroscience 8, 3355 (1983).
20. Merzenich, M.M. et al. Somatosensory cortical map changes following
digit amputation in adult monkeys. J. Comp. Neurol. 224, 591605 (1984).
21. Pons, T.P. et al. Massive cortical reorganization after sensory
deafferentation in adult macaques. Science 252, 18571860 (1991).
22. Pons, T.P., Garraghty, P.E., Friedman, D.P. & Mishkin, M. Physiological
evidence for serial processing in somatosensory cortex. Science 237,
417420 (1987).
23. Pons, T.P. in Somesthesis and the Neurobiology of the somatosensory cortex,
(eds Franzn, O., Johansson, R. & Terenius, L.)187195 (Birkhuser,
Basel, Switzerland, 1996).
24. Crick, F. & Koch, C. Constraints on cortical and thalamic projections:
the no-strong-loops hypothesis. Nature 391, 245250 (1998).
25. Pons, T.P., Garraghty, P.E., Cusick, C.G. & Kaas, J.H.. The somatotopic
organization of area 2 in macaque monkeys. J. Comp. Neurol. 241,
445466 (1985).
229
 1998 Nature America Inc. http://neurosci.nature.com

articles

Strengthening of horizontal cortical

connections following skill learning
Mengia-S. Rioult-Pedotti1, Daniel Friedman1, Grzegorz Hess1,2 and John P. Donoghue1

1 Department of Neuroscience, Brown University, Providence, Rhode Island 02912, USA

2 Permanent address: Jagiellonian University, Institute of Zoology, 31-102 Krakow, Poland
Correspondence should be addressed to M.-S. R.-P. (mengia_rioult@brown.edu)

Learning a new motor skill requires an alteration in the spatiotemporal pattern of muscle activation.
Motor areas of cerebral neocortex are thought to be involved in this type of learning, possibly by
functional reorganization of cortical connections. Here we show that skill learning is accompanied
1998 Nature America Inc. http://neurosci.nature.com

by changes in the strength of connections within adult rat primary motor cortex (M1). Rats were
trained for three or five days in a skilled reaching task with one forelimb, after which slices of motor
cortex were examined to determine the effect of training on the strength of horizontal intracortical
connections in layer II/III. The amplitude of field potentials in the forelimb region contralateral to
the trained limb was significantly increased relative to the opposite untrained hemisphere. No
differences were seen in the hindlimb region. Moreover, the amount of long-term potentiation (LTP)
that could be induced in trained M1 was less than in controls, suggesting that the effect of training
was at least partly due to LTP-like mechanisms. These data represent the first direct evidence that
plasticity of intracortical connections is associated with learning a new motor skill.

The primary motor area (M1), a cortical region necessary for
skilled voluntary movements, seems also to participate in learn-
ing motor skills. This conclusion is based largely upon findings
that adult M1 representations are modifiable14, that dendritic
morphology of M1 pyramidal neurons is altered by experience5
and that connections among M1 neurons are capable of activi-
ty-dependent, long-term changes in efficacy611. Despite these
suggestive findings, there is no direct evidence that learning is
accompanied by functional modifications of M1 circuits. Modi-
fications outside of the cortex, which have been repeatedly doc-
umented, might account for cortical motor or sensory map
changes that can be produced by experience or nerve lesions12,13.
Further, there is no evidence that morphological changes, such
as an increase in the number of dendritic branches, actually alter
functional interactions in the cortex. Finally, there has been no
definitive evidence that LTP-like mechanisms are engaged with-
in cortex during any form of learning. Demonstration of synap-
tic modification in conjunction with learning is an essential step
towards understanding how cortical circuits support motor skills
or other forms of learning.
The intrinsic horizontal pathways are a potential substrate
for experience-dependent reorganization of relationships among
M1 neurons. Layer II/III pyramidal cells form a broad, intrin-
sic horizontal projection system in M1, and their intracortical
pattern correlates with sites that reorganize after nerve
lesions14,15. Pharmacological adjustments of the excitatory-
inhibitory balance within M1 restructures motor representa-
tions, apparently by uncovering latent horizontal pathways16. In
addition, horizontal connections are capable of LTP, providing a
potential activity-dependent mechanism for synaptic modifica-
tion3,7,9,10,17. These findings raise the possibility that changes in
horizontal connection strength may accompany motor learn-
ing. Here we show that field potentials evoked by stimulation of
rat M1 horizontal connections increase after learning and prac-
ticing a skilled reaching task. The amount of LTP that could be
induced by electrical stimulation was also reduced after learn-
ing, implying that the observed strengthening of horizontal con-
nections may involve an LTP-like mechanism. Plasticity of M1
connections may therefore create cortical circuits needed to
acquire or perform new motor behaviors.
Results
Rats were trained to reach through a hole in a food box with a
single forepaw in order to retrieve small food pellets using a
grasping motion. Training and subsequent practice lasted three
(n = 1) or five (n = 13) successive days with one training session
(approximately one hour) per day. Successful performance of
this skill occurred in the first one or two sessions, and the remain-
ing sessions consisted of repeated practice and refinement of the
skill. By the final two days of training, all rats achieved a perfor-
mance of about 1.5 pellet retrievals per minute, with few errors in
the reach, grasp or retrieval actions. A group of comparably han-
dled, age- and sex-matched cage mates (termed paired controls,
n = 14) and another group of naive rats (unpaired controls,
n = 12) served as controls.
The strength of intrinsic horizontal synaptic connections
within layers II/III was evaluated ex vivo using coronal brain
slices containing both hemispheres (Fig. 1a); experimenters were
blind to whether the animal was trained, and if so, which fore-
limb was trained. Slices were taken 2045 hours after the last
training session to rule out effects that might persist immedi-
ately after practice of the task. Field-potential responses evoked
in the horizontal pathway by electrical stimuli were examined
simultaneously through stimulating and recording electrodes
that were mirror symmetrically positioned in layer II/III of the
left and right M1 (Fig. 1a). Thus, one side in each trained rat

230 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 1. Conse-
quences of motor
a b c Trained
rec Trained n=7
skill learning on rec
field-potential stim stim
responses evoked in
layer II/III horizontal
connections of M1. Control
Control
(a) Mirror-symmet- n=20
ric placement of
stimulating (stim)
and recording (rec)
microelectrodes
bilaterally in layers
II/III of M1 in a coronal slice containing both hemispheres. wm, white matter. (b) Single-case examples of field potentials (averages of five
sweeps), evoked at 60% maximum stimulation intensity from a single trained (top) and a single paired-control (bottom) animal. Dark lines
represent the trained M1 or left M1, hatched lines, the untrained M1 or right M1. (c) Group average responses for trained (top, n = 7) and
control (bottom, n = 20, paired and naive) rats at 60% maximal stimulation intensity, illustrating enhanced field potential in the horizontal
pathway of M1 contralateral to the limb used in the reaching task. Same format as (b).
1998 Nature America Inc. http://neurosci.nature.com

provided a within-animal control because the majority of
engaged M1 neurons are located in the hemisphere contralater-
al to the limb they influence18. For trained animals, we term M1
contralateral to the trained limb the trained M1, and its coun-
terpart on the other side the untrained M1. In all control ani-
mals, the terms left M1 and right M1 are used. stimulation intensity. The average log ratio obtained for trained
Stimulation evoked an initially negative-going field potential
of similar shape in all rats as previously described7. However,
for each rat that had learned the skilled-reaching task, field
potentials evoked in the trained M1 were consistently larger in
amplitude than in the untrained M1 (Fig. 1b and c). Ampli-
tudes in the trained M1 were also larger than those observed
for the control animals. Amplitude differences between trained
and untrained M1 were not a result of stimulus intensity
because absolute current intensities used on the two sides were
not significantly different (p = 0.23). Indeed, in 71% of the
cases, the stimulation intensity was slightly larger
(27.24 3.38%, n = 10) on the untrained side.
The amplitude differences between trained M1 and
untrained M1 were specific to the region of the M1 forelimb
representation. Layer II/III field-potential measurements from
the hindlimb region of the trained M1 and untrained M1 in an
additional group of trained rats showed no significant side-to-
side amplitude differences at any stimulation intensity
(p = 0.20.8, n = 9), whereas slices taken from the forelimb
region of the same animals showed a larger response in the
trained than untrained M1. Field potentials in the hindlimb
and forelimb areas were similar in shape. Peak amplitudes for
the hindlimb at 2.5 and 5 times threshold intensity were
0.92 0.12 mV and 1.47 0.19 mV in the trained M1 and
0.97 0.13 mV and 1.45 0.19 mV in the untrained M1.
The relationship between stimulus intensity and response
amplitude was evaluated systematically using two different
approaches (Fig. 2ac) to rule out the possibility that these effects
were a consequence of the particular intensities used. One series of
7 trained and 20 control (8 paired, 12 unpaired) rats was tested
with stimuli that were a constant fraction of the stimulus inten-
sity evoking a maximum response (absolute intensity less than or
equal to 220 A). A second series of seven trained and six paired
control rats was tested with stimuli that were a constant multiple
of the stimulus intensity evoking a minimal response of about 0.1
mV (absolute threshold intensity 1230 A). In both series, the
average responses for the trained M1 were larger than the
untrained M1 at every stimulation intensity, whereas there were no
differences between left and right sides in the control groups
(Fig. 2a and b). To compare the relative change between trained
and untrained M1, the common logarithm of the response-ampli-
tude ratio between sides was calculated for each animal at each
animals was significantly different from zero (p < 0.05) over a
broad range of stimulus intensities, reflecting larger peak ampli-
tudes in the trained M1. This ratio was not different in control
animals (p > 0.05; Fig. 2c). Figure 2d plots the distribution of the
entire data set; the trained animals show a marked skew toward
larger responses on the trained side, whereas untrained controls
show no difference between the two hemispheres.
Amplitude differences observed in the trained M1 might arise
by several mechanisms. Larger responses in the horizontal path-
ways could result from newly formed synapses, from postsynap-
tic excitability increase or from engaging an LTP-like mechanism
that increased the strength of existing synapses. A larger initial
slope for field potentials evoked at 60% maximal stimulation
intensity in the trained M1 (trained M1, 0.735 0.079 mV per
ms; untrained M1, 0.439 0.079 mV per ms; p < 0.025) suggest-
ed that modifications occurred by a change in synaptic efficacy,
rather than an excitability change. Because layer II/III horizontal
connections are capable of LTP, we compared the ability to poten-
tiate this pathway in trained M1 and untrained M1. We postulat-
ed that if learning recently engaged an LTP-like process in the
trained pathway, further electrically induced potentiation of this
pathway might be occluded. LTP induction was attempted by
theta-burst stimulation simultaneously delivered during a tran-
sient reduction of GABAA-receptor-dependent synaptic inhibi-
tion by focal application of small amounts of bicuculline at each
of the two recording sites6,7, an established method to obtain LTP
reliably in neocortex. In accord with previous results7, LTP was
readily induced in the untrained M1 (24.1 11.7% increase, n = 6)
and in both hemispheres of control rats (left M1, 39.7 10.5%;
right M1, 36.1 12.8%; p = 0.8, n = 7). The difference between
untrained M1 and left and right M1 of controls was not signifi-
cant (p = 0.35). In contrast, the average increase in field potential
amplitude in the trained M1 was only 6.5 4.2% (n = 6; Fig. 2e),
which was significantly less than the untrained M1 (p<0.035,
n = 6). Thus, although identical stimulation and recording pro-
cedures were applied to both hemispheres, the amount of LTP
that could be induced in the trained hemisphere was reduced.

nature neuroscience volume 1 no 3 july 1998 231

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 2. Response differ-

ences in trained and
a Trained
b Trained d Trained

control rats. (a) Field-

potential amplitudes for
trained (top, n = 7) and
paired- and naive-con-

FP amplitude (mV)
log (trained M1/untrained M1)
trol (bottom, n = 20)
animals. Stimulation Control
intensities are plotted Control Control
as a percentage of the
stimulus evoking the
maximal response.
Filled symbols repre-
sent the trained M1 or
log (left M1/right M1)
left M1, open symbols
the untrained M1 or (% max) (x threshold)
e
right M1. Note the Trained
Stimulation intensity
larger response magni-

FP amplitude (% baseline)
tude in the trained M1 c
1998 Nature America Inc. http://neurosci.nature.com

across intensities.
(b) Field-potential
amplitudes for trained
Log ratio

(top, n = 7) and paired- Control

control (bottom, n = 6)
animals. Stimulation
intensities were varied
as multiples of thresh-
old intensity inducing a (% max) (x threshold)
Time
minimal response Stimulation intensity
(about 0.1 mV).
Symbols as in (a).
(c) Comparison of average ( standard error) log-response ratios for trained/untrained M1 (filled symbols) and left/right M1 (open symbols)
for series in (a) (left) or series in (b) (right). Asterisks indicate statistical differences between control and experimental ratios. (d) Log ratio of
field-potential amplitudes for trained (top) and control (bottom) groups. The histograms show the distributions of all log ratios for all rats at all
stimulation intensities (entire data set). The mean value for trained animals (0.17) is significantly different from zero (p<0.001), demonstrating
a shift towards the trained M1. (e) Effect of training on electrically induced LTP. Each point indicates the relative field-potential amplitude
before and after LTP induction in both hemispheres (arrow) in trained (top, n = 6) and paired control (bottom, n = 7) animals. Averages after
LTP induction are compiled from the last 50 min of recordings because of a variable duration of the transient bicuculline effect (interruption in
the x-axis). Open symbols (+ standard error) for right M1 or untrained M1, filled symbols (- standard error) for left M1 or trained M1.

Discussion
These results demonstrate that learning and practicing a motor
skill is accompanied by an increased efficacy of horizontal con-
nections in motor cortex. Although there is extensive data show-
ing that motor skill learning modifies cortical
representations1,2,19,20 and alters dendritic morphology5 and that
cortical connections are capable of activity-dependent strength
changes, our results provide the first direct evidence for a func-
tional change of a cortical connection associated with motor skill
learning. Our results compare with changes recently observed in
the amygdala following fear conditioning, a markedly different
form of learning21,22 and might provide a basis for correlation-
strength changes that have been observed during auditory con-
ditioning23. Plasticity of horizontal connections could contribute
to the reorganization of motor cortical representations that
accompanies motor skill learning, because information from one
region of M1 would be spread more effectively to other regions.
This hypothesis is consistent with recent findings demonstrating
that only the parts of M1 receiving strong horizontal inputs reor-
ganize immediately after nerve lesions in the rat14 and that M1
representations in monkeys and humans enlarge or rearrange
during motor skill learning4,2426. Our data cannot differentiate
between changes that result from the early modifications in
behavior that occur when the task is first achieved and the slow-
er improvements in skill that occur with subsequent practice.
Both can be considered as forms of learning. Increased efficacy of
horizontal connections would not seem to be a consequence of
movement alone. Although the number of movements might be
different for the practiced limb, the actual number is small when
considered as a fraction of the total number of movements made
over the one to two days after training ended. The movements
made were at a low rate (about 1.6 per minute), required little
force and were mixed with many overlearned bilateral move-
ments in consummatory actions, yet enhanced horizontal-con-
nection strength persisted in M1 in the region of the forelimb.
Changes in horizontal connections are also not widespread in
M1 because learning did not modify the layer II/III horizontal
pathway in the hindlimb area.
The marked effect of learning upon field-potential amplitude
within the M1 forelimb region suggests that a large number of
connections within this area have been modified in conjunction
with skill learning. It is difficult to conceptualize how such a gen-
eralized effect can provide a substrate for implementing the
detailed pattern of movements acquired. Studies of human
motor-skill learning using transcranial magnetic stimulation sug-
gest that motor-cortical maps shrink again after explicit knowl-

232 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com
edge of the task is gained25 . Thus, the seemingly generalized
changes we observe after five days of training may eventually lead
to more specific circuits suitable for producing the skill. By con-
trast, fMRI studies during human motor-skill learning indicate
that representations continue to enlarge with repeated prac-
tice24,26. We will need to examine additional time points to deter-
mine what happens to field-potential increases and decrements in
LTP before we can make more definitive statements of the role
of these modification in acquiring new motor skills.
Modifications in horizontal connections seem to result from
changes in synaptic efficacy, perhaps through LTP-like mecha-
nisms, rather than other means. Learning seems to occlude LTP
induction, suggesting they share a similar mechanism. The
increase in the initial slope of field potentials from the trained
M1 is consistent with the hypothesis that learning occurred
through synaptic modification. Although compelling evidence
is lacking, LTP is a strong candidate mechanism for many forms
of learning because it can lead to long-lasting modification of
1998 Nature America Inc. http://neurosci.nature.com
activated synapses27 at a number of sites, including horizontal
cortical connections7,17. Other mechanisms are potentially plau-
sible but are not fully consistent with our results. For instance,
a different form of learning, involving classical conditioning,
leads to increases in the membrane excitability of M1 neurons28,
rather than synaptic modification. If increases in horizontal
field potentials were due to excitability changes in postsynap-
tic neurons, no initial slope changes would be expected and
tetanization would likely produce greater postsynaptic depo-
larization and hence, a larger amount of LTP29. Growth of new
synaptic connections, which has been suggested to occur in
adult cortex30,31, could also lead to larger responses after learn-
ing. However, these new synapses would have to be unable to
undergo LTP to be consistent with the finding of less LTP in the
trained M1. We therefore think it unlikely that these mecha-
nisms underlie the increased field-potential amplitudes seen
after skill learning, although additional studies are required to
identify the exact mechanism involved.
We only examined the horizontal pathway after three or five
days of practice following skill acquisition. Although synaptic
modification seems to occur during this period, other mecha-
nisms may operate during initial skill acquisition or during later
skill improvement. Intracellular studies and measurements of the
time course of change may help to clarify what leads to such dra-
matic increases in this pathways efficacy. It is likely that this effect
is not peculiar to M1. Plasticity of synapses formed by horizon-
tally oriented axon collaterals may operate throughout many areas
of the cerebral cortex to restructure various representation pat-
terns. For example, within visual cortex, filling-in phenomena
and reorganization of visual receptive fields after lesions or other
perturbations appear to be mediated via horizontal connec-
tions32. The common occurrence of these connections in all cor-
tical areas suggests that plasticity of synapses formed by horizontal
pathways may be an important contributor to learning-related
processes throughout the cerebral cortex.
Methods
Animals were cared for in accordance with National Institute of Health
guidelines for laboratory animal welfare. All experiments were approved
by the Brown University Institutional Animal Care and Use Committee.
Forty adult female Sprague-Dawley rats (150225 g) housed on a nor-
mal 12:12 light-dark cycle were used for this study. Twenty-eight animals
were housed in pairs and were food restricted to maintain their body
weight at roughly 85% of their free-feeding weight. Water was provided
ad libitum. One rat from each pair was placed in an operant test cage
(22.8 cm cube), which contained a Plexiglas food box (3.2 x 4.5 x 5 cm)
nature neuroscience volume 1 no 3 july 1998
articles
with a 1.3 cm diameter hole through which the food was retrieved. Small
food pellets (45 mg; Noyes Precision Food Pellets) were placed on the
floor of this food box within reaching distance. Rats learned to reach into
the food box with their preferred paw to retrieve food pellets using a
grasping action. It was impossible for the rats to reach in the food box
with both paws, although initial attempts sometimes involved trying to
reach with both paws or attempting to place the snout in the food box
and reach with the tongue. All but one animal, which was excluded from
the analysis, selected the right forelimb to perform the task. Animals
received one training session per day lasting for one hour. The training
and practice period lasted five days for thirteen rats and three days for
one rat. Because there was no readily apparent difference in the reach-
ing behavior nor in the electrophysiological results after the three- and
five-day training, data were grouped together. The second animal from a
pair served as a paired control, received a comparable amount of han-
dling and was given similar numbers of food pellets. The remaining
twelve animals were used as naive controls.
The experimenter was unaware of the rats training condition until
data analysis of the pair was completed. Coronal brain slices containing
the region of the M1 forelimb representation33, 12 mm anterior to the
bregma, were prepared as described7 and superfused with artificial cere-
brospinal fluid (ACSF) of the following composition (in mM): 126 NaCl,
3 KCl, 1.25 NaH2PO4, 26 NaHCO3, 1 MgSO4, 2 CaCl2, and 10 glucose,
bubbled with a 95% O2, 5% CO2 mixture at 35 0.5C. The humidified
atmosphere over the slices was saturated with 95% O2, 5% CO2. Coro-
nal slices from the hindlimb region were cut at the level of the anterior
part of the hippocampal formation.
Field potentials were recorded using glass micropipettes placed in layer
II/III, 200350 m below the pial surface in the region of the M1 fore-
limb representation (22.2 mm lateral to the midline). Concentric bipo-
lar stimulating electrodes were displaced horizontally by 500 m from
each recording electrode (Fig. 1a). Consistent mirror-symmetrical place-
ment of the electrodes at identical locations in both hemispheres was
achieved with a reticle. Slices were not attached by the corpus callosum,
but remained attached to each other during tissue slicing. For stimula-
tion, constant current pulses (0.2 ms) were delivered at 0.033 Hz. We
used the amplitude of the field potential evoked in the layer II/III hori-
zontal pathway to measure of the population excitatory synaptic response
because it reflects a monosynaptic current sink9 and correlates well with
intracellular excitatory postsynaptic potentials evoked in this pathway7.
Inputoutput curves for a range of stimulation intensities were con-
structed for each stimulationrecording pair. In the first set of experi-
ments, we averaged three to five sweeps evoked at 20, 40, 60, 80, 100 and
120% of the intensity inducing a maximum response. In the second set,
we averaged three responses to stimuli of 2, 2.5, 3, 3.5, 4, 4.5 and 5 times
the intensity that evoked a 0.1 mV (threshold) response. Stimulation did
not influence the contralateral hemisphere because the slices used did
not contain the corpus callosum. Field-potential peak amplitudes were
calculated from averages of three to five waveforms, and the common
logarithm of the left /right ratio was calculated for each animal. Because
the log of this ratio fit a Gaussian distribution, parametric testing was
used (paired t-test). Similar input-output curves were constructed for
field potentials recorded from the hindlimb area of trained animals.
After establishing a 20 min period of stable response amplitudes using
a stimulation intensity 50-60% of maximum, LTP induction was attempt-
ed with an established and reliable protocol for MI7. Prior to tetanic stim-
ulation, the GABAA receptor antagonist bicuculline (3.5 mM) was applied
within 100 m from the recording electrodes using a glass pipette; the
time of application on each side was separated by less than two minutes,
and the same pipette was used for each side of a slice. The bicuculline
pipette was retracted as soon as field-potential responses to test stimu-
lation increased to about 150200% of baseline (typically within 1030 s).
Immediately following bicuculline application, LTP was attempted by
delivering theta-burst stimulation (5 sequences of 10 bursts 10 seconds
apart; 1 burst is 5 pulses at 100 Hz) at double test intensity simultane-
ously through both stimulating electrodes7. The LTP effect was recorded
for at least 30 min after it reached a stable plateau.
233
 1998 Nature America Inc. http://neurosci.nature.com
articles
Acknowledgements
We thank Drs. Barry Connors, Mark Bear, and Marc G. Rioult for critical
comments. This work was supported by NIH grant NS22517. G. H. is an
international scholar of the Howard Hughes Medical Institute.
RECEIVED 21 JANUARY: ACCEPTED 26 MAY 1998
1. Cohen, L.G., Brasil, N.J., Pascual-Leone, L.A. & Hallett, M. Plasticity of
cortical motor output organization following deafferentation, cerebral
lesions, and skill acquisition. Adv. Neurol. 63, 187200 (1993).
2. Donoghue, J.P. Plasticity of sensorimotor representations. Curr. Opin.
Neurobiol. 5, 749754 (1995).
3. Donoghue, J.P., Hess, G. & Sanes, J.N. in Acquisition of Motor Behavior (eds.
Bloedel, J., Ebner, T. & Wise, S.P.) 363386 (MIT Press, Cambridge, 1996).
4. Nudo, R.J., Milliken, G.W., Jenkins, W.M. & Merzenich, M.M. Use-
dependent alterations of movement representations in primary motor cortex
of adult squirrel monkeys. J. Neurosci. 16, 785807 (1996).
5. Greenough, W.T., Larson, J.R. & Withers, G.S. Effects of unilateral and
bilateral training in a reaching task on dendritic branching of neurons in the
rat motor-sensory forelimb cortex. Behav. Neural Biol. 44, 301314 (1985).
6. Hess, G. & Donoghue, J.P. Long-term potentiation of horizontal connections
1998 Nature America Inc. http://neurosci.nature.com
provides a mechanism to reorganize cortical motor maps. J. Neurophysiol. 71,
25432547 (1994).
7. Hess, G., Aizenman, C.D. & Donoghue, J.P. Conditions for the induction of
long-term potentiation in layer II/III horizontal connecitons of the rat motor
cortex. J. Neurophysiol. 75, 17651778 (1996).
8. Hess, G. & Donoghue, J.P. Long-term depression of horizontal connections
in rat motor cortex. Eur. J. Neurosci. 8, 658665 (1996).
9. Aroniadou, V.A. & Keller, A. Mechanisms of LTP induction in rat motor
cortex in vitro. Cereb. Cortex 5, 353362 (1995).
10. Castro-Alamancos, M.A., Donoghue, J.P. & Connors, B.W. Different forms of
synaptic plasticity in somatosensory and motor areas of the neocortex. J.
Neurosci. 15, 53245333 (1995).
11. Asanuma, H. & Pavlides, C. Neurological basis of motor learning in
mammals. Neuroreport 8, ivi (1997).
12. Kaas, J.H. Plasticity of sensory and motor maps in adult mammals. Annu.
Rev. Neurosci. 14, 137167 (1991).
13. Merzenich, M.M., Recanzone, G., Jenkins, W.M., Allard, T.T. & Nudo, R.J. in
Neurobiology of Neocortex (eds Rakic, P. & Singer, W.) 4167 ( Wiley, New
York, 1988).
14. Huntley, G.W. Correlation between patterns of horizontal connectivity and
the extent of short term representational plasticity in rat motor cortex. Cereb.
Cortex 7, 143156 (1997).
234
15. Donoghue, J.P. Limits of reorganization in cortical circuits. Cereb. Cortex 7, 9799
(1997).
16. Jacobs, K. & Donoghue, J. Reshaping the cortical map by unmasking latent
intracortical connections. Science 251, 944947 (1991).
17. Hirsch, J. & Gilbert, C. Long-term changes in synaptic strength along specific
intrinsic pathways in the cats visual cortex. J. Physiol. (Lond) 461, 247262
(1993).
18. Price, A.W. & Fowler, S.C. Deficits in contralateral and ipsilateral forepaw
motor control following unilateral motor cortical ablation in rats. Brain Res.
205, 8190 (1981).
19. Garraghty, P.E. & Kaas, J.H. Dynamic features of sensory and motor maps.
Curr. Opin. Neurobiol. 2, 522527 (1992).
20. Sanes, J.N. & Donoghue, J.P. Motor areas of the cerebral cortex. J. Clin.
Neurophysiol. 11, 382396, (1994).
21. Rogan, M.T., Staeubli, U.V. & LeDoux, J.E. Fear conditioning induces associative
long-term potentiation in the amygdala. Nature 390, 604607 (1997).
22. McKernan, M.G. & Shinnick-Gallagher, P. Fear conditioning induces a lasting
potentiation of synaptic currents in vitro. Nature 390, 607611 (1997).
23. Ahissar, E. et al. Dependence of cortical plasticity on correlated activity of
single neurons and on behavioral context. Science 257, 14121415 (1992).
24. Grafton, S.T. et al. Functional anatomy of human procedural learning
determined with regional cerebral blood flow and PET. J. Neurosci. 12,
25422548 (1992).
25. Pascual-Leone, A., Grafman, J. & Hallett, M. Modulation of cortical motor
output maps during development of implicit and explicit knowledge. Science
263, 12871289 (1994).
26. Karni, A. et al. Functional MRI evidence for adult motor cortex plasticity
during motor skill learning. Nature 377, 155158 (1995).
27. Morris, R.G.M., Davis, S. & Butcher, P. in Long Term Potentiation: A Debate of
Current Issues (eds Baudry, M. & Davis, J.L.) 267300 (MIT Press,
Cambridge, 1991).
28. Woody, C.D., Gruen, E. & Birt, D. Changes in membrane currents during
Pavlovian conditioning of single cortical neurons. Brain Res. 539, 7684
(1991).
29. Yoshimura, Y. & Tsumoto, T. Dependence of LTP induction on postsynaptic
depolarization: a perforated patch-clamp study in visual cortical slices of
young rats. J. Neurophysiol. 71, 16381645 (1994).
30. Darian, S.C. & Gilbert, C.D. Axonal sprouting accompanies functional
reorganization in adult cat striate cortex. Nature 368, 737740 (1994).
31. Kleim, J.A., Lussnig, E., Schwarz, E.R., Comery, T.A. & Greenough, W.T.
Synaptogenesis and Fos expression in the motor cortex of the adult rat after
motor skill learning. J. Neurosci. 16, 45294535 (1996).
32. Gilbert, C.D. Rapid dynamic changes in adult cerebral cortex. Curr. Opin.
Neurobiol. 3, 100103 (1993).
33. Donoghue, J.P. & Wise, S.P. The motor cortex of the rat: cytoarchitecture and
microstimulation mapping. J. Comp. Neurol. 212, 7688 (1982).
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com

articles

Retinotopy and color sensitivity in

human visual cortical area V8
Nouchine Hadjikhani, Arthur K. Liu, Anders M. Dale, Patrick Cavanagh
and Roger B. H. Tootell

Nuclear Magnetic Resonance Center, Massachusetts General Hospital, 149 13th Street, Charlestown, Massachusetts 02129, USA
Correspondence should be addressed to N.H. (nouchine@nmr.mgh.harvard.edu)

Prior studies suggest the presence of a color-selective area in the inferior occipital-temporal region
of human visual cortex. It has been proposed that this human area is homologous to macaque area
V4, which is arguably color selective, but this has never been tested directly. To test this model, we
1998 Nature America Inc. http://neurosci.nature.com

compared the location of the human color-selective region to the retinotopic area boundaries in the
same subjects, using functional magnetic resonance imaging (fMRI), cortical flattening and
retinotopic mapping techniques. The human color-selective region did not match the location of
area V4 (neither its dorsal nor ventral subdivisions), as extrapolated from macaque maps. Instead
this region coincides with a new retinotopic area that we call V8, which includes a distinct
representation of the fovea and both upper and lower visual fields. We also tested the response to
stimuli that produce color afterimages and found that these stimuli, like real colors, caused preferen-
tial activation of V8 but not V4.

In Old World primates such as macaque monkeys and humans,
visual information about color is processed in anatomically
segregated columns, layers, channels or areas. It is important
to know to what extent color is processed in separate versus
convergent visual information pathways, because the added
dimension of color is so rich in visual information. For exam-
ple, we can discriminate about fifteen hundred different levels
of luminance1, whereas we can make several million discrimi-
nations if we also consider variations in color2. It is likely that
this glut of color information is incorporated into the labeled
lines of the neural architecture in some organized way.
In macaque monkeys, an anatomical segregation between chro-
matic-opponent versus achromatic-opponent cells has been report-
ed as early as the lateral geniculate nucleus. Color-specific anatomical
segregation has also been described in primary (V1) and secondary
(V2) visual cortex. In V1, prominent populations of color-selective
cells have been reported in specific layers35 and in the cytochrome-
oxidase blobs46, though the latter claim has been disputed7,8. Sim-
ilar (and equally controversial) claims have been made about the
prominence of color-opponent cells in the thin stripes in area V2,
to which the V1 blobs project (ref. 9,10, but see 11).
However, the most prominent controversy about the
anatomical segregation of color-selective neurons occurs at a
higher level, in cortical area V4. According to different reports,
a high percentage of color-selective cells is either present1215 or
absent16 in the largest and best-studied portion of that area,
dorsal V4 (V4d). A high percentage of color-selective cells has
not been reported in the smaller, ventral subdivision of V4
(V4v). More recent evidence suggests that brain mechanisms
critical for color selectivity are located not in macaque V4, but
rather in areas anterior to it (ref. 1719, Vanduffel et al. Soc.
Neurosci. Abstr. 23, 845, 1997).
This controversy about color selectivity in V4 has now been
extended to human visual cortex. Based on human neuroimag-
ing studies, a small patch of color-selective activity near the mid-
dle of the collateral sulcus has been named V4 (ref. 2022).
This choice of name presupposes that (1) an area homologous to
macaque V4 exists in humans, (2) V4 is color-selective, and (3)
this region in or near the collateral sulcus is the macaque V4
homolog. However, in humans, the location of this color-selec-
tive region has not yet been compared with the map of retino-
topic areas, to see whether color selectivity is really is in a
retinotopically defined human area V4. Furthermore, the degree
of color selectivity in macaque V4 is itself controversial1719.
This issue is not just of academic interest. In an intriguing
clinical syndrome (achromatopsia), human patients report
that the visual world becomes colorless following damage to a
cortical region that apparently includes this color-selective
area in the collateral sulcus2325. This suggests that the con-
scious percept of color involves that area, although it is
known that physical wavelength-dependent differences are
coded throughout prior levels of the visual system as well. If
we can define better which area this is in humans, we can learn
something about where conscious perceptions of color arise.
Accurate localization in humans should also make it possible
to study the homologous area in macaques using more inci-
sive (but invasive) classical neurobiological techniques.
We have attempted to clarify these issues in humans using
functional magnetic resonance imaging (fMRI). Technical details
were similar to those described elsewhere26, except that here we
manipulated the color content of the visual stimuli. We also used
a high-field MRI scanner and other improvements to substan-
tially increase the sensitivity of the retinotopic maps (Methods).
Results
COLOR- VERSUS LUMINANCE-VARYING STIMULI
First, we compared the activity produced by color variations to
that produced by variations in luminance, in the same sub-

nature neuroscience volume 1 no 3 july 1998 235

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 1. Topography of
color-selective activity in
a b
human visual cortex.
(a, b) The inferior,
inflated cortex, with
posterior to the left and
anterior to the right, in
two subjects. (c, d) The
posterior portion of c d
cortex in fully flattened
format, another view of
the same data shown in
(a) and (b), respectively.
In all panels, gyri from
the original brain are
shown as light gray and
sulci as dark gray. The
fundus of the collateral
sulcus (cs) is indicated
1998 Nature America Inc. http://neurosci.nature.com

by the dashed black line.

The borders of previ-
ously described retinotopic areas (V1, V2, V3, VP, V3A, and V4v) are indicated in white (horizontal meridians, solid lines, upper vertical
meridians, dotted lines, lower vertical meridians, dashed lines). Typically, color-varying stimuli produced relatively higher activation in the
foveal representation of V1 and often V2 and V3/VP and a distinctive patch of color-selective activation approximately midway in the collat-
eral sulcus. When present, the latter patch was always located just anterior to the horizontal meridian representation marking the anterior
border of area V4v, rather than within V4v.

jects (Fig. 1). Our stimuli consisted of slowly moving sinu-
soidal radial gratings (pinwheels) of low spatial frequency,
defined by either color or luminance contrast (Methods). As
shown earlier in V1 and V2 (ref. 27), we found that both
color- and luminance-varying stimuli produced robust acti-
vation in many areas of visual cortex, when compared with a
uniform gray field (data not shown).
Here we focused on those locations where the color stimuli
produced more activation than the luminance stimuli. In the
classically retinotopic visual areas (V1, V2, V3/VP, V3A and
V4v), we found prominent color-selective activation in the rep-
resentations of the fovea (center of gaze) but not in peripheral
representations (Fig. 1). A foveal color bias has not been report-
ed in previous imaging studies, perhaps because it is more obvi-
ous in our flattened maps. However, such a foveal color bias is
consistent with the well known predominance of cone pho-
toreceptors, and the corresponding absence of rods, in the fovea
of the retina. A similar foveal color bias is found in routine clin-
ical perimetry and in numerous psychophysical studies.
In 25 of 26 hemispheres (13 subjects) tested, we found an
additional region that responded preferentially to color, locat-
ed midway along the length of the collateral sulcus. Based on
the anatomical location and the nature of the functional com-
parison used here, this collateral color-selective patch appears
equivalent to the previously reported area involved in achro-
matopsia20,22, which has been proposed as the human homo-
logue of macaque area V4 (refs 2022, 2830).
However, when we compared the location of that collateral
color-selective patch to the retinotopic borders in the same sub-
jects, we found that the color-selective patch was consistently
located just beyond the most anterior retinotopic area defined
previously, area V4v. Earlier reports 26,3133 suggested that
human V4v is a quarter-field representation of the contralater-
al upper visual field. The more sensitive retinotopic mapping

Fig. 2. Retinotopic features of area V8 by fMRI a b

mapping. (ac) Retinotopy of polar angle in the infe-
rior row of cortical areas, from three flattened
hemispheres. From left to right, each panel shows
the representations of the contralateral upper quar-
ter field (red through blue or vice versa; see
pseudocolor logo) in inferior V1, then inferior V2,
then VP, then V4v. To the right of (anterior to) V4v
is the distinctive half-field representation comprising
V8 (green through blue through red, from upper to
lower in this figure). (d) Retinotopic representation c d
of eccentricity (the other dimension in polar space),
from the same hemisphere shown in (c). The repre-
sentations of central-through-more-peripheral
eccentricities are coded in red-through-blue-
through-green, respectively (see pseudocolor logo,
bottom right). The representations of the center of gaze are indicated with an asterisk. Area V8 has its own representation of the fovea, quite
distinct (and 3.5 cm) from the foveal representation in adjacent area V4v.

236 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

a e In conventional Talairach coordi-

nates, foveal V4v (as defined retino-
topically in this study) was centered at
32, -87, -16, whereas foveal V8 was
centered at 33, -65, -14. When we
b f averaged the Talairach coordinates of
the color-selective area V4 described
in previous studies ( 26, -67, -9), we
found that it was about twice as close
to the location of our retinotopically
c g defined V8, compared to our retino-
topically defined V4v. This supports
all the other evidence that the color-
selective activity is located in area V8,
rather than in V4.
d h In each hemisphere, V8 comprises
a continuous representation of the
entire contralateral half of the visual
1998 Nature America Inc. http://neurosci.nature.com

field. Although the authors did not

attempt polar-coordinate retinotopy or
flattened mapping, a prior study also
Fig. 3. Detailed retinotopy of the polar angle representation, from the same hemisphere shown
in Fig. 2a. This figure shows the peak fMRI response (noisy white lines) corresponding to polar
concluded that this human color-selec-
angle gradients of approximately 20o, superimposed on the standard pseudocolor rendering of tive region included a representation
areas V1, V2, VP and V4v (drawn in a). To the right of each panel is a logo indicating the specific of upper and lower visual fields 22 .
polar angle (white line) stimulated. The complete contralateral visual field is represented in V8, Together with V8 in the opposing
from the lower visual field (a and b), across the horizontal meridian (ce) to the upper visual hemisphere, the entire visual field is
field (fh). Note in (eh) that the upper visual field representation in V8 can clearly be distin- thus represented. Such a complete rep-
guished from, and is mirror symmetric to, that in adjacent V4v. resentation of the visual field would be
appropriate in an area processing high-
er-order, large-field color information.

methods used here confirmed that V4v represents just that HUMAN VERSUS MACAQUE MAPS
quarter field, with its foveal representation located superiorly Because so much of the historical controversy about corti-
alongside that of adjacent areas V3/VP (Figs 2 and 3). cal color processing arose in studies of macaque monkey, it is
The improved retinotopic methods also revealed additional natural to wonder which area in macaque corresponds to
retinotopic features anterior to V4v. Taken together, these fea- area V8 in humans. To clarify the topographic relationship
tures indicate the presence of an additional retinotopic map, of the human and macaque maps, we first averaged together
comprising a previously undifferentiated cor-
tical area that we call V8. This continues the
naming scheme begun by Zeki and colleagues, a b
who identified areas V1 through V61215,21,22,34.
(We also identified an area V7, which is a rep-
resentation of the contralateral lower visual
field anterior to human V3A.)
Area V8 has a unique polar angle retinotopy
and a distinctive foveal representation. This
contrasts significantly with the three extrastri-
ate representations posterior to V8 (V4v, VP,
and the inferior wing of V2), all of which are
quarter-field representations of the contralat-
eral upper visual field. Although the polar-
angle retinotopy in V8 includes an additional
representation of this quarter field, it also
extends further to include a lower-field repre-
sentation as well (Figs 2 and 3). These three
extrastriate visual areas (V4v, VP and inferior
V2) also share a contiguous representation of Fig. 4. Comparison of the polar angle retinotopy in human visual cortex, relative to
the fovea, at the top (superior) end of this row that reported in macaque monkeys. In both species, visual cortex is shown in flat-
of areas (Fig. 2d). However, the foveal repre- tened format, with visual area boundaries and polar angle continua as in Fig. 2ac.
sentation in V8 is not part of this contiguous Area MT is shown in gray. In macaque, dorsal area V4 is also indicated (V4d). The
foveal band; instead it is located about 3.5 cen- retinotopy of V8 is similar to that reported in area TEO, in that both areas are located
timeters away along the cortical surface, at the immediately adjacent to area V4v. However, the two areas differ in overall shape, and
anterior border of V8 (Fig. 2d). the retinotopy of V8 is rotated approximately 90o relative to that reported in TEO.

nature neuroscience volume 1 no 3 july 1998 237

 1998 Nature America Inc. http://neurosci.nature.com

articles

a Alternating Constant b
colors colors

Fixation Fixation Fixation

% MIRI signal
% MIRI isgnal

Color Fixation

Time (s) Time (s)

1998 Nature America Inc. http://neurosci.nature.com

Fig. 5. The time course of V8 activity is related to the perception of color afterimages. (a) Time course from all voxels in retinotopically
defined V8 that responded (p<0.00001) to the colored stimuli, relative to the initial presentation of the uniform gray stimulus, averaged
across 16 MR scans, showing the response to both the actual and the illusory color stimuli. Epochs in which subjects viewed a uniform grey
field, or an illusory afterimage on a gray background, are indicated with a white background; epochs when the subjects viewed colored stim-
uli (alternating- or constant-colored) are indicated with a gray background. After the color stimulus, subjects viewed a uniform gray field, or
an illusory afterimage on a gray background. During a color afterimage, the fMRI response was quite prolonged, consistent with the time
course of the percept of the illusory colors. (b) The MR afterimage is shown more directly by subtracting signal during constant color and
subsequent gray period from alternating color and subsequent grey period.

six of the most robust human retinotopic maps, using digital
morphing techniques (Fig. 4b) as described26. These aver-
aged maps could then be compared to the map of macaque
retinotopy, as estimated from single-unit mapping 35
(Fig. 4a). This comparison suggests that human area V8
shows some retinotopic and topographic similarity to
macaque area TEO. Furthermore, TEO and/or more anteri-
or areas have appeared strongly color selective in recent stud-
ies of macaque visual cortex (refs 1719, 28, Vanduffel et al.
Soc Neurosci. Abstr. 23, 845, 1997).
However, the retinotopic similarity between V8 and TEO is
far from exact (Fig. 4). Furthermore, other investigators have
proposed different area boundaries in this region of macaque
cortex 3638 . Unfortunately, those alternative models of the
macaque maps are even less similar to the empirical human
maps in this region of cortex. Thus it is not clear which macaque
area is homologous to human area V8. However, the flat maps
in Fig. 4 do make it clear that this human collateral color area is
not topographically similar to macaque V4 (neither dorsal nor
ventral subdivisions). Human V8 is also topographically incon-
sistent with the location of subdivisions proposed in macaque
dorsal V4, such as V4t (ref. 39) and V4A (e.g. ref. 14).
COLOR AFTERIMAGES
Another way to assess functional selectivity is by measuring the
fMRI responses during visual aftereffects, rather than during
the fMRI effects produced by the visual stimuli themselves. In
other dimensions such as motion40 and orientation41, such indi-
rect aftereffects have ironically proven to be functionally more
selective than the effects themselves. Here we tested whether
illusory color would also activate V8, as did real color stimuli.
If one stares for a time at a saturated color, then looks away
at a uniform gray field, one sees an illusory percept of the com-
plementary color. Unlike motion or orientation aftereffects,
these negative color afterimages are thought to arise primarily
in the retina42,43. However, they also presumably trigger activ-
ity at higher levels, as would a real stimulus that was similarly
stabilized on the retina29,44. To test for the presence of fMRI
responses to these illusory colors, we produced such color after-
images in the MR scanner, along with control stimuli that were
very similar but did not produce color afterimages.
Figure 5 shows the time course of fMRI activity produced
by these stimuli in area V8. As expected, the colored stimuli
produced robust fMRI activity in V8, whether alternating or
constant. However, only the constant-colored stimuli pro-
duced a perceptual color afterimage and a corresponding fMRI
aftereffect in V8 during subsequent viewing of the uniform
gray field (Fig. 5a). The alternating-colored stimuli produced
neither a perceptual color afterimage nor a prolongation of
the normal fMRI return to baseline during the subsequent
viewing period (Fig. 5a). The duration of the isolated fMRI
color aftereffect was prolonged, consistent with the prolonged
duration of the illusory color percept (Fig. 5b). Overall, this
strongly suggests that these fMRI responses were related to
the processing of the illusory colors.
One unexpected finding was that the stimuli with alter-
nating colors produced slightly more activity than the stim-
uli in which color remained constant (Fig. 5a). This may
reflect the fact that the hues in the constant-colored stimuli
become progressively less saturated (less densely colored) with
time because of chromatic adaptation45. Essentially one begins
to see the mixture of the color afterimage and the actual color,
Because these two colors are complementary, they produce a
less saturated, more washed-out color. This decreased fMRI
response to the decreasingly saturated colors supports the
other evidence that V8 is involved in color perception.
Findings similar to our fMRI afterimages in V8 were
reported from scattered voxels in single-slice imaging through
nearby posterior fusiform gyrus29, but the location of those
voxels was not localized to any specific cortical area. However,

238 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com
1998 Nature America Inc. http://neurosci.nature.com
Fig. 6. The perception of color afterimages produces relatively
higher activation in cortical area V8, compared with other cortical
areas. The activation shown here represents all regions that
responded significantly more (p0.00001) during viewing of the uni-
form gray stimulus following the constant color stimulus, compared
with viewing of the same gray stimulus following the alternating
color stimulus.
none of these data address the possibility that similar fMRI
afterimages occur nonspecifically throughout much wider
areas of visual cortex. This could arise from retinal color pro-
cessing that is transmitted passively to cortex, or from glob-
ally increased attention when viewing the color afterimages.
To test this, we reanalyzed our data to find the areas in an
activity map that respond differentially during the presence
of the visual afterimage (Fig. 6). At lower levels of significance
than shown here, a number of additional visual areas (e.g. V1,
V2) do respond more to color afterimages. However, at the
more strict significance threshold used in Fig. 6, the activity
in V8 was more prominent than in any other area. In partic-
ular, the wider areas of foveal color selective activities includ-
ing areas such as V1, V2, etc., were relatively less prominent
in the afterimage test, compared to the direct comparisons of
color versus luminance (see Fig. 1c and Fig. 6, showing the
same hemisphere).
Discussion
The retinotopic maps make it clear that an additional area
(V8) exists beyond those areas described previously in human
visual cortex. Area V8 is retinotopically distinct from the pre-
viously described area V4v, based on at least four different cri-
teria. First, V4v and V8 have separate foveal representations,
approximately 3.5 cm apart along the cortical surface. Second,
V4v and V8 each include separate representations of the upper
visual field, separated from each other by a representation of
the horizontal meridian. Third, V8 differs from V4v in its
global functional properties, including but not limited to color
sensitivity. Fourth, the nature of the retinotopy in V8 is dif-
ferent from that in V4v.
The direct comparisons between color- and luminance-vary-
ing stimuli (Fig. 1) indicate that area V8 responds slightly better
than neighboring cortical regions to colored stimuli. However,
we also found that the color-varying stimuli produce preferen-
nature neuroscience volume 1 no 3 july 1998
articles
tial activation in the foveal representation of all retinotopic areas.
Thus, V8 may seem especially biased for color stimuli merely
because its foveal representation sets it topographically apart
from the conjoined foveal color responses of its neighbors (Figs
2 and 3). This is a relatively trivial explanation for the color selec-
tivity reported earlier, but we cannot rule it out completely. This
idea is further supported by the fact that area V8 responds at
reasonable levels to a wider variety of visual stimuli.
However, other evidence argues that V8 is involved in
wavelength-dependent processing and perhaps in the con-
scious perception of color itself. The robust and selective
response to illusory colors (Figs 5 and 6) strongly supports
this idea. Also, the anatomical colocalization of V8 compared
with the previous clinical data makes it likely that area V8 is
damaged in achromatopsic patients2325.
What do these human data tell us about macaque visual
cortex? This question is constrained by several factors. The
human data are based on clinical and neuroimaging data,
whereas the macaque data are derived from different tech-
niques (e.g., single units, lesions and DG imaging), which
could conceivably produce different results. Also, there may
be significant biological differences between the cortical orga-
nization of color sensitivity in humans compared with
macaques. As we learn more about human and macaque visu-
al cortex, the number of differences between these species are
increasing correspondingly26,4648.
Despite these caveats, the data suggest that the area of
macaque cortex that is homologous to the human achro-
matopsia area should be located in or anterior to TEO, rather
than in V4. This is supported by data from macaque 1719 as
well as the present data from human V8.
Methods
GENERAL PROCEDURES. Except for modifications described below, the
methods in this study are similar to those described26. Informed writ-
ten consent was obtained for each subject prior to the scanning ses-
sion, and all procedures were approved by Massachusetts General
Hospital Human Studies Protocol numbers 907227 and 967464.
Normal human subjects, with (or corrected to) emmetropic vision,
were scanned in General Electric magnetic resonance (MR) scanners
retrofitted with ANMR echo-planar imaging. Most scans were
acquired in a high-field (3 T) scanner, but some early scans were
acquired in a scanner of conventional (1.5 T) field strength. Based on
signal-to-noise ratios obtained during otherwise comparable condi-
tions, four functional scans at 1.5 T were found to be approximately
equal to one functional scan at 3 T, so this was the ratio used to equate
data acquired from the two scanners. Head motion was minimized
by using bite bars with deep, individually molded dental impressions.
The subjects task in all experiments was to fixate the center of each
type of visual stimulus throughout the period of scan acquisition.
MR images were acquired using a custom-built quadrature surface
coil, shaped to fit the posterior portion of the head. MR slices were
3-4 mm thick, with an in-plane resolution of 3.1 x 3.1 mm, oriented
approximately perpendicular to the calcarine fissure. Each scan took
either 4 min 16 s (color-versus-luminance and color afterimage scans)
or 8 min 32 s (retinotopy), using a TR of either two or four seconds,
respectively. Each scan included 2,048 images, comprised of 128
images per slice in 16 contiguous slices.
Improved retinotopic maps were obtained from 32 subjects (79
scans polar angle, 79 scans eccentricity, 323,584 images total). Among
them, 13 subjects were also tested for color-versus-luminance (112
scans, 229,376 images total). Of these, five subjects were tested exten-
sively for color afterimages (100 scans; 204,800 images total). In most
subjects, additional scans were done to clarify the location of area
MT and other visual areas.
239
 1998 Nature America Inc. http://neurosci.nature.com
articles
VISUAL STIMULI. The goal of the first color-related experiment (Fig. 1)
was to map the MR activity produced by color- versus luminance-
varying stimuli throughout visual cortex, using conventional psy-
chophysical stimuli. Prior to scanning, the equiluminance values for
different color combinations (red-cyan, CIE x and y coordinates 0.645,
0.345 and 0.185, 0.248 respectively, or green-purple, CIE x and y coor-
dinates 0.277, 0.684 and 0.351, 0.220, respectively) were measured for
each subject, outside the scanner. Equiluminance was measured using
a motion-null test, with the same stimulus projector (NEC model MT
800), lens and color software used subsequently in the MR experi-
ments. In the first experiment, both color- and luminance-varying
stimuli were produced using slowly moving (0.5 Hz) sinusoidal radi-
al gratings (pinwheels) of low spatial frequency (3 cycles per revo-
lution). The gratings varied either in achromatic luminance
(maximum, 95% luminance contrast) or in equiluminant color (at
maximum available saturations of the display device within con-
straints of approximately 140 cd per m2 mean luminance and approx-
imately white mean chromaticity), in direct alternation, in 16-second
epochs, using 16 epochs per scan.
1998 Nature America Inc. http://neurosci.nature.com
In a second experiment (Figs 5 and 6), color afterimages were pro-
duced by showing subjects colored adaptation patterns. Subjects
adapted to two general types of stimuli; one produced a pronounced
color afterimage when subjects subsequently viewed a uniform gray
field, but a very similar control stimulus did not produce such a color
afterimage. There were five epochs in each scan, presented in the fol-
lowing order: (1) uniform gray, (2) alternating-color (control adap-
tation), (3) uniform gray, (4) constant-color (experimental
adaptation) and (5) uniform gray. All epochs were 48-s long, except
that the final fixation period was prolonged 16 s to reveal the final
traces of the MR afterimage. In the analysis for Fig. 5b, the last 16 s
in that final epoch was truncated to match the duration of all other
epochs. Both the constant- and the alternating-colored patterns were
comprised of complementary colors (red-cyan or green-purple), spa-
tially arranged in opposed quarter-fields (i.e. two-cycle-per-revolu-
tion polar square waves), akin to interleaved bow ties. All hues were
presented at equal luminance, based on motion-null tests in each sub-
ject. Following adaptation to the constant colors, subjects initially
experienced a prominent color afterimage against the uniform gray
background, which faded over tens of seconds. The afterimage was
retinotopically similar to the adaptation stimulus, but of comple-
mentary color. As controls, we presented stimuli equivalent to the
constant colors, except that the colors alternated between color and
complementary color, reversing every 2 s. The latter condition pro-
duced no perceptual color afterimages during the subsequent epoch of
uniform gray stimulation.
Stimuli for retinotopic mapping were slowly moving, phase-encod-
ed thin rays or rings comprised of counterphasing black and white
checks, scaled according to polar coordinates, similar to those
described26,31,32,50. However, to produce the most informative retino-
topic maps possible, several stimulus modifications and new proce-
dures were implemented. First, all retinotopic measurements were
made in the 3 T scanner. This increased the MR amplitudes by a factor
near four, and the physiological signal-to-noise ratio by a factor near
two. Second, we signal-averaged the information from 412 scans
(8,19224,576 MR images) of polar angle or eccentricity. Data were
also combined from different slice prescriptions on the same cortical
surface, to reduce intervoxel aliasing. Third, the retinotopic stimuli
were increased in extent both foveally and peripherally, to extend from
0.2 through 1830. This activated correspondingly more surface in
each cortical area. Fourth, the visual stimuli were presented using a
new LCD projector of higher spatial resolution (800 x 600), using bet-
ter optics than previously (aperture lens, bypassing shielding screen,
etc.). Fifth, the retinotopic stimuli varied in color as well as luminance,
to better activate any color-selective cells in the region. The sum of all
these manipulations produced very robust retinotopic maps.
DATA ANALYSIS. Data from two-condition experiments (e.g., color-versus-
luminance comparisons) and phase-encoded retinotopic experiments were
240
initially analyzed by doing a fast Fourier transform on the MR time course
from each voxel. Statistical significance was calculated by converting the
Fourier magnitude of the response to an f-statistic. The phase of the sig-
nal at the stimulus frequency was used to track stimulus location in the
case of retinotopic stimuli, and to distinguish between positive- or negative-
going MR fluctuations in the case of two-condition stimulus comparisons.
Scans comparing more than two stimuli (e.g., the color afterimage
data) were analyzed by selective averaging of two conditions. This was
followed by statistical comparison using a t-test of the difference of the
first seconds following onset of the next epoch (here stimulus offset).
For topographic clarity, all data were analyzed and displayed in cor-
tical surface format, as described 36,43,44 . This made it possible to
extract the MR time courses from voxels in specific cortical areas,
which were defined in the same subjects. The specific areas sampled
were V1, V2, V3/VP, V3A, V4v, MT+, and V8. Area MT+ was defined
on the basis of additional scans comparing moving and stationary
stimuli26,40. All other areas were based on retinotopic criteria.
For ease of comparison, all hemispheres are shown in right hemi-
sphere format. Above a minimum threshold, the statistical significance
of the displayed pseudocolor range has been normalized according to
the overall sensitivity of each subject, as described elsewhere.
Acknowledgments
This work was supported by grants from the Human Frontiers Science
Foundation and NEI EY07980 to R.B.H.T., NEI EY09258 to P.C. and Swiss
Fonds National de la Recherche Scientifique to N.H. We thank Terry
Campbell and Mary Foley for scanning and participation in these
experiments, Robert Savoy, Ken Kwong, Bruce Fischl and Kevin Hall for
advice, Tommy Vaughan for coil design and manufacture and Martin Sereno
for modifying pilot stimuli. Wim Vanduffel, Ekkehardt Kustermann and
Irene Tracy also helped in preliminary versions of this experiment.
RECEIVED 16 APRIL: ACCEPTED 21 MAY 1998
1. Cornsweet, T.N. Visual Perception (Academic Press, New York, 1970).
2. Judd, D.B. & Wyszecki, G. Color in Business, Science, & Industry 3rd edn
388 (Wiley, New York, 1975).
3. Dow, B.M. Functional classes of cells and their laminar distribution in
monkey visual cortex. J. Neurophysiol. 37, 927946 (1974).
4. Livingstone, M.S. & Hubel, D.H. Anatomy and physiology of a color
system in the primate visual cortex. J. Neurosci. 4, 309356 (1984).
5. Tootell, R.B.H., Silverman, M.S., Hamilton, S.L., De Valois, R.L. &
Switkes, E. Functional anatomy of macaque striate cortex: III. Color J.
Neurosci. 8, 15691593 (1988).
6. Tso, D.Y., Frostig, R.D., Lieke, E.E. & Grinvald, A. Functional
organization of primate visual cortex revealed by high resolution optical
imaging. Science 249, 417420 (1990).
7. Lennie, P., Krauskopf, J. & Sclar, G. Chromatic machanisms in striate
cortex of macaque. J. Neurosci. 10, 649669 (1990).
8. Leventhal, A.G., Thompson, K.G., Liu, D., Zhou, Y. & Ault, S.J.
Concommitant sensitivity to orientation, direction and color of cells in
layers 2, 3 and 4 of monkey striate cortex. J. Neurosci. 15, 18081818 (1995).
9. Hubel, D.H. & Livingstone, M.S. Segregation of form, color and stereopsis
in primate area 18. J. Neurosci. 7, 33783415 (1987).
10. Tootell, R.B.H. & Hamilton, S.L. Functional anatomy of the second
cortical visual area (V2) in the macaque. J. Neurosci. 9, 26202644 (1989).
11. Gegenfurtner, K.R., Kiper, D.C. & Fenstemaker, S.B. Processing of color,
form and motion in macaque area V2. Vis. Neurosci. 13, 161172 (1996).
12. Zeki, S.M. Colour coding in rhesus monkey prestriate cortex. Brain Res.
27, 422427 (1973).
13. Zeki, S.M. Colour coding in the superior temporal sulcus of rhesus
monkey visual cortex. Proc. R. Soc. Lond. B 197, 195223 (1977).
14. Zeki, S. Uniformity and diversity of structure and function in rhesus
monkey prestriate visual cortex. J. Physiol. (Lond.) 277, 273290 (1978).
15. Zeki, S. The distribution of wavelength and orientation selective cells in
different areas of monkey visual cortex. Proc. R. Soc. Lond. B 217, 449470
(1983).
16. Schein, S.J., Marrocco, R.T. & de Monasterio, F.M. Is there a high
concentration of color-selective cells in area V4 of monkey visual cortex?
J. Neurophysiol. 47, 193213 (1982).
17. Heywood, C.A., Gadotti, A. & Cowey, A. Cortical area V4 and its role in
the perception of color. J. Neurosci. 12, 40564065 (1992).
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com
18. Heywood, C.A., Gaffan, D. & Cowey, A. Cerebral achromatopsia in
monkeys. Eur. J. Neurosci. 7, 10641073 (1995).
19. Cowey, A. & Heywood, C.A. Theres more to colour than meets the eye.
Behav. Brain Res. 71, 89100 (1995).
20. Lueck, C.J. et al. The colour centre in the cerebral cortex of man. Nature
340, 386389 (1989).
21. Zeki, S. et al. A direct demonstration of functional specialization in
human visual cortex. J. Neurosci. 11, 641649 (1991).
22. McKeefry, D.J. & Zeki, S. The position and topography of the human colour
centre as revealed by functional magnetic resonance imaging. Brain 120,
22292242 (1997).
23. Pearlman, A.L., Birch, J. & Meadows, J.C. Cerebral color blindness: An
acquired defect in hue discrimination. Ann. Neurol. 5, 253261 (1979).
24. Damasio, A., Yamada, T., Damasio, H., Corbett, J. & McKee, J. Central
achromatopsia: Behavioral, anatomic, and physiologic aspects. Neurology 30,
10641071 (1980).
25. Zeki, S. A century of cerebral achromatopsia. Brain 113, 17211777
(1990).
26. Tootell, R.B.H. et al. Functional analsis of V3A and related areas in human
visual cortex. J. Neurosci. 17, 70767078 (1997).
27. Engel, S., Zhang, X. & Wandell, B. Colour tuning in human visual cortex
measured with functional magnetic resonance imaging. Nature 388, 6871
(1997).
28. Kennard, C., Lawden, M., Morland, A.B. & Ruddock, K.H. Colour
1998 Nature America Inc. http://neurosci.nature.com
identification and colour constancy are impaired in a patient with
incomplete achromatopsia associated with prestriate cortical lesions.
Proc. R. Soc. Lond. B 260, 169175 (1995).
29. Sakai, K. et al. Functional mapping of the human colour centre with echo-
planar magnetic resonance imaging. Proc. R. Soc. Lond. B 261, 8998 (1995).
30. Kleinschmidt, A., Lee, B.B., Requardt, M. & Frahm, J. Functional mapping
of color processing by magnetic resonance imaging of responses to selective
P- and M-pathway stimulation. Exp. Brain Res. 110, 279288 (1996).
31. DeYoe, E.A. et al. Mapping striate and extrastriate visual areas in human
cerebral cortex. Proc. Natl. Acad. Sci. USA 93, 23822386 (1996).
32. Sereno, M.I. et al. Borders of multiple visual areas in humans revealed by
functional magnetic resonance imaging. Science 268, 889893 (1995).
33. Tootell, R.B.H., Dale, A.M., Sereno, M.I. & Malach, R. New images from
human visual cortex. Trends Neurosci. 19, 481489 (1996).
34. Galletti, C. Fattori, P., Battaglini, P.P., Shipp, S. & Zeki, S. Functional
demarcation of a border between areas V6 and V6A in the superior
parietal gyrus of the macaque monkey. Eur. J. Neurosci. 8, 3052 (1996).
nature neuroscience volume 1 no 3 july 1998
articles
35. Boussaoud, D., Desimone, R. & Ungerleider, L.G. Visual topography of
area TEO in the macaque. J. Comp. Neurol. 306, 554575 (1991).
36. Felleman, D.J. & Van Essen, D.C. Distributed hierarchical processing in
the primate cerebral cortex. Cereb. Cortex 1, 147 (1991).
37. Zeki, S. Are areas TEO and PIT of monkey visual cortex wholly distinct
from the fourth visual complex (V4 complex)? Proc. R. Soc. Lond. B 263,
15391544 (1996).
38. Maguire, W.M. & Baizer, J.S. Visuotopic organization of the prelunate
gyrus in rhesus monkey. J. Neurosci. 7, 16901704 (1984).
39. Van Essen, D.C., Maunsell, J.H. & Bixby J.L. The middle temporal visual
area in the macaque: Myeloarchitecture, connections, functional
properties and topographic organization. J. Comp. Neurol. 199, 293326
(1981).
40. Tootell, R.B.H. et al. Visual motion aftereffect in human cortical area MT
revealed by functional magnetic resonance imaging. Nature 375, 139141
(1995).
41. Tootell, R.B.H. et al. Functional analysis of primary visual cortex (V1) in
humans. Proc. Natl. Acad. Sci. USA 95, 811817 (1998).
42. Craik, K.J.W. Origin of visual afterimages. Nature 145, 512 (1940).
43. Brindley, G.S. Two new properties of foveal afterimages and a
photochemical hypothesis to explain them. J. Physiol. 164, 168179
(1962).
44. Schiller, P.H. & Dolan, R.P. Visual aftereffects and the consequences of
visual system lesions on their perception in the rhesus monkey. Vis.
Neurosci. 11, 643665 (1994).
45. Jameson, D., Hurvich, L.M. & Varner, F.D. Receptoral and postreceptoral
processes in recovery from chromatic adaptation. Proc. Natl. Acad. Sci.
USA 76, 30343038 (1979).
46. Tootell, R.B.H. & Taylor, J.B. Anatomical evidence for MT and additional
cortical visual areas in humans. Cereb. Cortex 5, 3955 (1995).
47. Tootell, R.B.H. et al. Functional analysis of human MT and related visual
cortical areas using magnetic resonance imaging. J. Neurosci. 15,
32153230 (1995).
48. Jacobs, G.H. & Deegan, J.F. Spectral sensitivity of macaque monkeys
measured with ERG flicker photometry Vis. Neurosci. 14, 921928 (1997).
49. DeYoe, E.A., Felleman, D.J., Van Essen, D.C. & McClendon, E. Multiple
processing streams in occipitotemporal visual cortex. Nature 371,
151154 (1994).
50. Engel, S.A., Glover, G.H. & Wandell, B.A. Retinotopic organization in
human visual cortex and the spatial precision of functional MRI. Cereb.
Cortex 7, 181192 (1997).
241
 1998 Nature America Inc. http://neurosci.nature.com

articles

Complete sparing of high-contrast color

input to motion perception in cortical
color blindness
Patrick Cavanagh1, Marie-Anne Hnaff2, Franois Michel2, Theodor Landis3,
Tom Troscianko4 and James Intriligator5

1 Department of Psychology, Harvard University, 33 Kirkland Street, Cambridge, Massachusetts 02138, USA
2 INSERM U-280, 151 cours A.Thomas, Lyon, 69003, France
3 Department of Neurology, University Hospital Geneva, rue Micheli-du Crest, CH - 1211 Genve 14, Switzerland
4 Department of Experimental Psychology, University of Bristol, 8 Woodland Road, Bristol, BS8 1TN, UK
1998 Nature America Inc. http://neurosci.nature.com

5 Department of Neurology, Beth Israel Deaconess Medical Center, 330 Brookline Avenue, Boston, Massachusetts, 02215, USA
Correspondence should be addressed to P.C. (patrick@wjh.harvard.edu)

It is widely held that color and motion are processed by separate parallel pathways in the visual
system, but this view is difficult to reconcile with the fact that motion can be detected in
equiluminant stimuli that are defined by color alone. To examine the relationship between color and
motion, we tested three patients who had lost their color vision following cortical damage (central
achromatopsia). Despite their profound loss in the subjective experience of color and their inability
to detect the motion of faint colors, all three subjects showed surprisingly strong responses to high-
contrast, moving color stimuli equal in all respects to the performance of subjects with normal
color vision. The pathway from opponent-color detectors in the retina to the motion analysis areas
must therefore be independent of the damaged color centers in the occipitotemporal area. It is
probably also independent of the motion analysis area MT/V5, because the contribution of color to
motion detection in these patients is much stronger than the color response of monkey area MT.

In central achromatopsia, patients lose the sensation of color
as a result of damage to the cortex and often report that the
world appears solely in shades of gray, as in a black-and-white
movie 14. Despite this almost total loss of color sensation,
there are several reports that central achromats paradoxical-
ly preserve other contributions of color to visual performance.
Some of these preserved functions may indicate that the dam-
age to the color analysis areas was less than total. For exam-
ple, tests of sensitivity to color increments and color contrast
in several central achromats59 and in one patient with par-
tial achromatopsia10 have, up to now, revealed performance
quite similar to that of subjects with normal color vision. In
contrast, the achromats in our experiments all show profound
losses on these tests, indicating more extensive damage to color
processing areas. On the other hand, it seems that all central
achromats, including those in our experiments, can see color-
defined shapes due to the preserved visibility of borders
between colors they no longer distinguish8,1114. A likely can-
didate for this border response is the magnocellular, or non-
opponent pathway, which responds to color transitions15,16.
Both the perception of motion for equiluminous color
stimuli11 and a color-specific motion aftereffect 5 have been
demonstrated in one central achromat. However, neither study
established the relative strength of the preserved motion
response. If motion responses to color stimuli were only weak-
ly preserved, we could attribute this to the partial sparing of
some color-analysis areas adjacent to the cortical damage. Any
particular spared response could mediate other responses to
color stimuli. For example, to an observer with normal color
vision, a saturated color appears brighter than its luminance
would predict17. If an achromat lost the experience of color
but retained this supplementary brightness response8,11, equi-
luminant color bars would have a residual brightness differ-
ence. In this case, the visibility of the bars and their motion
could be reduced but never eliminated. Conversely, if motion
responses to color are totally preserved in central achro-
matopsia, it would argue against a residual, spared mecha-
nism and suggest that colors contribution to motion takes a
route that is anatomically remote from the damaged areas.
We now resolve this question with a population of central
achromats who show little or no preserved opponent-color
processing in spectral sensitivity, detection or motion tasks at
threshold. At suprathreshold levels, however, these same
achromats surprisingly demonstrate normal levels of motion
responses to moving colored gratings.
To evaluate this performance, we included two control
groups: normal subjects and congenitally red/green-deficient
observers (where the site of loss is in the retina). These indi-
viduals are commonly described as color blind; however, they
suffer only a partial loss of color experience limited to the
red/green dimension, a loss much less extensive than that
experienced by the central achromats. Because these observers
have little chromatic response to red/green stimuli, however,
we can use their performance to evaluate the level of residual

242 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com
JPC light
Log relative sensitivity
Log relative sensitivity
Wavelength (nm)
WM light
Log relative sensitivity
Log relative sensitivity
1998 Nature America Inc. http://neurosci.nature.com
Wavelength (nm)
Fig. 1. Sensitivity for increment on light and dark fields as a function
of wavelength for two central achromats, JPC and WM, and one
normal, TT. The form of the function on a dark field was similar for
both achromats and for the normal, and only one plot is shown
(JPC, top right). Only the normal observer shows the typical three-
peaked function on the light field. Both achromats show a single-
peaked function on light and dark fields, indicating that there is little
or no contribution of opponent-color mechanisms to these detec-
tion thresholds.
luminance signals in our red/green color stimuli.
Our tests reveal that motion responses to high-contrast
color gratings are at normal levels in the three patients but
absent in the congenitally red/green-deficient observers. These
results suggest that the cortical damage that produces achro-
matopsia spares a particular cortical route for high-contrast
color information, allowing it to contribute to motion per-
ception without contributing to color sensations.
Results
CASE HISTORIES
The three patients with central achromatopsia were similar in
that they all had bilateral lesions, made several errors in read-
ing the colored digits of the Ishihara text and, when asked to
order several colored disks according to their hue (Farnsworth-
Munsell test), they made virtually random orderings. All three
also had visual agnosia and prosopagnosia (deficits in object
and face recognition). Two had upper-field loss (WM, JPC)
and one (JPN) had only one full quadrant of vision (lower
right). Visual acuity and contrast sensitivity for achromatic
tests in the preserved visual fields showed that at least coarse
spatial resolution was retained for all three patients2,12.
At the time of testing, JPC was 43, a former florist, and his
lesions were the result of an assault that caused bilateral hem-
orrhage in the occipital-temporal regions. Magnetic resonance
imaging (MRI) revealed lesions principally in both fusiform
gyri but sparing the lingual gyri and the left posterior fusiform
gyrus. WM was 74, a former electrician who suffered a severe
nature neuroscience volume 1 no 3 july 1998
articles
JPC dark reduction of vision following a brief loss of consciousness.
(WM is described in more detail in ref. 18.) MRI and com-
puter tomography subsequently revealed extensive occipi-
totemporal lesions bilaterally in the territory of both posterior
cerebral arteries. JPN was 41 and had suffered bilateral strokes
of the occipital region, producing a large lesion of the right
occipital pole and a lesion of the left lingual and fusiform gyri.
JPN reported no conscious sensations of color and showed
chance-level color naming in a forced-choice task. JPC noticed
only reds (and some yellows) as being weakly tinted but he
Wavelength (nm) did not name them as red because the tint differed from the
red he remembered. WM reported no conscious sensations of
TT light normal color but when forced to guess, he demonstrated better-than-
chance naming of red and yellow.
SPECTRAL SENSITIVITY
The increment threshold spectral sensitivity task is a measure
of the degree of function of the opponent-color pathways19,20.
A chromatic stimulus of varying intensity is presented briefly
on a white background, and the observer reports whether any-
thing has been presented. In the normal observer, the sensi-
tivity peaks at three different wavelengths, which are
Wavelength (nm)
characteristic of opponent-cone interactions20. If the tests are
presented on a dark background, the luminance mechanism
is more sensitive than any of the opponent mechanisms, and
the sensitivity curve has a single peak. Both of these tests were
performed on two of the patients, JPC and WM, and a nor-
mal observer, TT, one of the authors. The data for WM have
been reported elsewhere13.
Both patients showed single-peaked functions for thresh-
olds on the light and dark backgrounds. The normal observer
showed the expected three-peaked function with light back-
grounds and single-peaked function with dark background.
These results (Fig. 1) indicate that the two patients have sig-
nificant losses in the opponent-color pathways, which mediate
broad ranges of the typical three-peaked threshold function
seen in normal subjects.
This experiment was very similar to previous experi-
ments 7,8 on two other central achromats. Both of those
patients showed the normal, three-peaked function. One
explanation of these earlier results8 was that some color-oppo-
nent response was spared and contributed to a brightness per-
cept for the stimuli but not a color percept. We conclude that
the previously tested patients had some sparing of high-level
color centers, whereas WM and JPC have a more profound
loss with no evidence of preserved color-opponent contribu-
tion to increment thresholds.
CONTRAST THRESHOLDS
These tests evaluated the color contrast that allow observers
to detect a red/green sinusoidal grating or to determine its
direction of motion. Both smoothly moving gratings and grat-
ings moving in 90 steps were used. The 90 steps11,21,22 effec-
tively control for motion cues from the borders, which central
achromats are known to detect well. With each step, the new
positions of the red/green transitions fall exactly halfway
between the previous positions, rendering the border cue
ambiguous.
All three central achromats showed very high discrimina-
tion (Fig. 2) and detection thresholds. The thresholds for WM
were exceptionally high. These thresholds suggest that WM,
JPC and JPN have severe losses in their response to color. We
did not find any evidence of a preserved brightness response
243
 1998 Nature America Inc. http://neurosci.nature.com
articles
Threshold color contrast (%)
Motion jump ( of phase)
Fig. 2. Threshold color contrast for the three central achromats
for direction discrimination of the moving color gratings. The range
of values measured for two congenitally red/green-deficient
1998 Nature America Inc. http://neurosci.nature.com
observers is shown by a yellow band and that for two normals by a
green band. Detection thresholds are not shown but are similar to
the discrimination thresholds for all subjects, except for the nor-
mals, in which detection thresholds are lower than discrimination
thresholds by a factor of two to three.
to color, which was proposed as the source of the preserved
chromatic sensitivity in MS and other previously tested achro-
mats11.
There were no marked differences between the thresholds
for smoothly moving or jumping stimuli, indicating that the
response to the chromatic borders in these patients was not a
significant factor in producing these threshold responses.
Despite the elevated thresholds, the red/green equiluminance
values for the achromats were in the normal range, indicating
normal function of at least the red- and green-sensitive cone
classes 21 . Of the two congenitally red/green-deficient
observers, the deutan (with losses in the green-sensitive cone
class) showed an equiluminance setting in the normal range,
whereas the protan (with losses in the red-sensitive cone class)
required significantly more luminance for red.
MOTION NULLING
Like a subject (MS) in a previous study11, the three patients
tested in the previous experiment could report the direction
of a color grating (once it was above threshold). However,
unlike MS, these three patients have severe losses in chromat-
ic sensitivity and require very high contrasts for accurate
judgements. A luminance mechanism, which is nominally but
never perfectly insensitive to color, may start to respond to
these very strong color stimuli. Therefore when the three
patients tested here do respond to color at
high contrasts, it may be due to a small
response of the luminance pathway to color Superimposed
or it may be due to the response of a color- gratings
specific mechanism.
To resolve this question, we used a stim-
ulus (Fig. 3) that dissociates the two com-
ponents, color and luminance, at
suprathreshold levels21,2325. This gives us
a measure of the strength of the color con-
tribution to motion for the patients and for
Color
the normal observers. This test evaluates grating
the low-level motion response to color as
244
opposed to a high-level, tracking response21,25. To make the
measurements, color and luminance gratings were superim-
posed and set in motion in opposite directions. At a particu-
lar balance of the contrast of the two gratings, their motions
cancelled and ambiguous motion or flicker was seen. The con-
trast of the luminance grating that just cancelled the motion of
the color grating was taken as the equivalent luminance con-
trast of the color grating21. Moreover, by again using con-
genitally red/green-deficient subjects as a control group, we
were able to estimate the baseline response of the luminance
mechanisms to residual luminance components in the color
stimulus.
Despite catastrophic loss in color sensitivity, when tested
with these suprathreshold color gratings, the three achro-
matopsic patients have equivalent luminance contrasts that
lie near or within the range of values found in normal sub-
jects (Fig. 4). The responses for the 90 jump conditions can
only be mediated by color-specific motion analysis because
the red/green borders are producing ambiguous motion sig-
nals in this stimulus.
The congenitally red/green-deficient observers showed
motion strengths in the 2% to 4% range, indicating that the
sum of their weak color response and luminance artifacts from
monitor and optical sources was less than 4%. The difference
between these values and the strengths measured for the cen-
tral achromats is therefore a lower bound on the actual
strength of the color-specific input to motion for these
patients. The high levels of equivalent luminance contrasts
found for the three patients, as well as their normal equilu-
minance settings, again demonstrate that they have normal
red- and green-sensitive cone function21.
It is nevertheless possible that some residual cues might
differentiate the red and the green of the gratings for the cen-
tral achromats and allow them to track individual grating bars.
It is unlikely, however, that these strategies (which the patients
did not report) could produce equivalent luminance contrasts
in the normal range seen here. On the other hand, JPC does
report some residual color sensations for red stimuli, and so all
the tests were repeated with purple/green gratings (differen-
tially stimulating principally the blue-sensitive cones) for both
WM and JPC. The patients again showed the same motion
strength as normal subjects.
How can these patients have normal levels of equivalent
luminance contrast but red/green-deficient levels of color-
contrast threshold? How can the response to color be absent
up to a significantly elevated threshold and then jump to
robust, normal levels once above threshold? We tested one
patient (JPC) for blindsight 26, to test the possibility that he
might be able to report the direction of motion for stimuli he
claims not to see. After instructions to guess, his reports of
Fig. 3. A sinusoidal, equiluminous,
color grating (red/green) and a sinu-
Luminance
soidal luminance grating (light and
grating
dark yellow) were superimposed,
moving in opposite directions. (The
stimuli are depicted here as square-
wave gratings for convenience.) At a
particular contrast of the luminance
grating, the two motions null each
other, establishing the equivalent
luminance contrast of the color
grating.
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com
Equivalent luminance contrast (%)
Motion jump ( of phase)
Fig. 4. Equivalent luminance contrasts for color gratings moving in
smooth (10.8) and 90 jumps. Data are shown individually for the
three central achromats. The range of values measured for six con-
1998 Nature America Inc. http://neurosci.nature.com
genitally red/green-deficient observers is shown as a yellow band
and that for normals (five observers) as a green band.
motion direction for red/green gratings were random for stim-
uli in the 8% to 15% range of color contrast just below his
detection threshold (but still about ten times higher than nor-
mal thresholds). It seems that the loss of performance at low
color contrast was real and is independent of the performance
at high color contrast.
We also tested his responses in the motion-nulling test
(described above) at low contrasts. The color grating had no
effect once it was below threshold; the motion of the opposing
luminance grating was the only direction the patient reported.
Above his color threshold, however, the contrast of the lumi-
nance grating required to null the motion of the color grat-
ing rose rapidly. Indeed, a normal subject (PC) showed the
same, rapid rise in equivalent luminance contrast above 20%
color contrast. However, the normal observer also had a weak
residual equivalent luminance contrast (1% to 2%) for lower
chromatic contrasts, whereas the achromat showed none. The
patient seems to share with the normal subject a motion
response to color that rapidly gains strength at high contrast.
The patient has lost a more sensitive process that operates at
low contrasts in normal subjects and apparently depends on
the functioning of higher color centers.
Discussion
At high color contrasts, the strength of the color contribution
to motion for the three central achromats was equivalent to
that for normal subjects, demonstrating that there is a direct
path for color information from the retina to cortical motion
detectors that is spared in central achromatopsia. Motion
responses mediated by this pathway were fully preserved, sug-
gesting that the pathway was completely independent of the
lesioned color centers of the ventral surface of the occipi-
totemporal region.
Previously tested central achromats have shown residual
opponent-color processing as indicated by a three-peaked
spectral sensitivity curve and normal detection thresh-
olds79,11. In contrast, in our study, both patients who were
tested for spectral sensitivity had only a single-peaked func-
tion, and all three had significantly elevated color-detection
thresholds. The damage to the color centers in these three
nature neuroscience volume 1 no 3 july 1998
articles
patients appears to be the most profound yet tested. The pre-
served motion responses to color were found despite this
extreme loss of color processing.
To gauge the magnitude of colors contribution to motion
in our patients, consider that the maximum red/green color
contrast we used produced 14% and 35% of cone contrast for
the red- and green-sensitive cones, respectively. This is the
same order of magnitude as the luminance contrast required
to null it, 12% to 23% in these patients (and the normal sub-
jects). In other words, the contribution of a color stimulus to
motion, where the cone responses are out of phase, seems to
be quite similar to the contribution of a luminance stimulus,
where the same cone contrasts are in phase.
This result seems at first at odds with the many reports of
degraded motion perception for chromatic stimuli. Howev-
er, recent studies have shown that thresholds for motion of
gratings defined by color are actually lower than those for
luminance-defined gratings 27,28. The explanation is that the
threshold for seeing a color-defined stimulus is lower still, the
lowest of all stimuli29, so that color gratings can be seen before
their motion is noticed. The result is the phenomenon of
stopped motion30,31, which led to the erroneous belief that
color was a poor source of motion information. More recent
studies demonstrate a strong21 and independent32 contribu-
tion of color to motion, and our results here show a contri-
bution at a level that could not be mediated by any secondary,
residual effect. The secondary, residual responses probably
account for the eventual detection of the color stimuli at the
very high thresholds found in the patients and the congeni-
tally red/green-deficient observers. This suggest that the sec-
ondary cues have about one-tenth the power of the color
signal.
One possible site for mediating the response to color is MT,
an area specialized in motion analysis, which is located far
from the areas damaged in achromatopsia. Physiological
recordings in monkeys 3335 and fMRI studies in humans 36
have shown responses to equiluminous chromatic gratings in
area MT. Experiments with a 90 stimulus like that used here
have shown that this response can be based on color infor-
mation, not just on the color borders 33 . However, this
response is very weak, with an equivalent luminance contrast
of only 2.5%, within the range that congenitally red/green-
deficient observers show and therefore attributable to distor-
tions in the monitor, the eye or the non-opponent pathway.
This level of response is far too low to explain the performance
of normal subjects and patients whose equivalent luminance
contrasts were in the range of 12 to 23%.
If MT is not the site of the robust, low-level motion
response to chromatic stimuli that we report here, where could
it be? The site of cortical damage in humans with achro-
matopsia invariably includes the ventral surface of occipi-
totemporal region, and although this deficit is often claimed to
involve the human homologue of monkey V4, recent lesion
studies in monkeys suggest that this may not be the case37.
In humans, the most recent fMRI studies38 suggest a more
anterior site, V8, for color analysis in a location consistent
with the damage in the achromatopsic patients. The human
homologue to V4 on the ventral surface (V4v) is probably also
damaged in these patients but it includes only a representa-
tion of the upper visual fields39. The matching area that rep-
resents the lower visual field has not been identified with
certainty. If there is a dorsal equivalent to V4 with a map of
the lower visual fields, we speculate that it might offer a pos-
245
 1998 Nature America Inc. http://neurosci.nature.com
articles
sible site for the strong contribution of color to low-level
motion in these patients and in normals. In monkeys, V4 does
have a fair percentage of cells that are directionally selective40.
Moreover, V4 in monkeys is directly involved in maintaining
a representation of motions in a delayed match-to-sample
task 41 . Although the site of the robust motion response to
high-contrast color remains to be determined, our results
demonstrate that the path from the retina to this area does
not pass through the damaged ventral area of these patients.
This suggests that color information does not flow as a single,
hierarchically organized stream through the visual system but
rather projects in parallel to several different centers, only one
of which leads to the conscious experience of color.
Methods
SPECTRAL SENSITIVITY. The observers were two central achromats, WM
and JPC, and one normal subject, TT, one of the authors. Observers in
this and following experiments gave informed written consent before
1998 Nature America Inc. http://neurosci.nature.com
the experiments, which were approved by the Ethisches Kommitee
der neurologischen Klinik des Universittsspital Zrich, by the Comit
consultatif de protection des personnes dans la recherche biomdi-
cale du centre Lon Brard, Lyon and by the F.A.S. Human Subjects
Committee, Harvard University.
The uniform light surround was 20 cd per m 2 and the dark sur-
round was 0.3 cd per m 2. The test stimuli were presented for 0.5 s
with a gap of 2.5 s between stimuli. The target was a slightly blurred
rectangle subtending about 2.2 by 1.6. Fixation was not monitored
during the experiment. The chromatic tests were provided by a mono-
chromator with a bandwidth of 10 nm.
After adapting to the light or dark background for several minutes,
testing began at the shortest wavelength. Increment thresholds were
determined using a method of limits, and the procedure was repeat-
ed three times, reversing the order of the wavelengths each time. The
mean of the three values was used to calculate the sensitivity for each
wavelength.
CONTRAST THRESHOLDS . The observers were the three central achro-
mats, two congenitally red/green-deficient (one deutan and one
protan) and two normal subjects. The red/green-deficient subjects
each failed at least 18 of the 24 Ishihara plates, whereas neither normal
subject made any errors.
The stimulus was a rotating wheel of eight cycles of a color grat-
ing, varying sinusoidally between red and green (Fig. 2). The wheel
was set in rotation either smoothly (because of the 66.7 Hz raster
rate, the actual step was 10.8 on each frame but it appeared to move
smoothly) or in 90 jumps. The rate of rotation was fixed at 2 Hz in all
cases (0.25 revolutions per second). The outer diameter of the wheel
was eight degrees of visual angle, and its inner radius was three
degrees. Its mean luminance was approximately 45 cd per m2, and the
surround was dark. The control of luminance was linearized through
gamma correction look-up tables.
The chromatic contrast of a grating was defined in terms of the
percentage of the maximum chromatic modulation obtainable with
the phosphors involved. 100% color contrast of the monitor produces,
at equiluminance, 14% cone contrast for the red-sensitive cones and
35% cone contrast for the green-sensitive cones.
The colors were set to equiluminance42 individually for each observ-
er. Thresholds were measured at that value and often two adjacent
values bracketing the equiluminance setting as well. In every instance
tested, the highest thresholds occurred at the predetermined equilu-
minance setting. Observers were instructed to fixate the central bulls-
eye, and the tests were presented on the display until a response was
given. Observers reported whether a stimulus was absent or present in
the detection conditions or moving clockwise versus counterclock-
wise in the direction-discrimination condition. Detection thresholds
were taken as the chromatic contrast for which the subject reported
present on 50% of the trials. Discrimination thresholds were taken as
246
the chromatic contrast for which the subject reported the direction
correctly on 75% of the trials.
MOTION NULLING. The observers were the three central achromats, six
congenitally red/green-deficient observers and five normal observers.
The red/green-deficient subjects (two deutans, four protans) each
failed at least 18 of the 24 Ishihara plates, whereas the normal sub-
jects made no more than three errors.
The stimulus was a rotating wheel of eight cycles of a red/green
sinewave as before but now (Fig. 4) superimposed additively on a sim-
ilar grating with eight cycles of a light/dark yellow sinewave rotating
in the opposite direction. The color contrast was set at the maximum
available between the red and green phosphors with allowance for the
superimposed luminance grating. The resulting color contrast was
between about two (WM) and five (JPC, JPN) times threshold con-
trast. The stimulus was otherwise identical in all respects to that used
in the previous experiment.
In a forced-choice procedure, while fixating the bulls-eye, the
observers reported the direction of the global motion seen in the com-
bined stimulus. The luminance contrast that produced a motion null
(equal frequency of reports favoring luminance and color directions)
was taken as the equivalent luminance contrast of the color grat-
ing21. We also tested red/green luminance ratios bracketing the equi-
luminance setting and, in every instance, the minimum equivalent
luminance contrast occurred at the predetermined setting.
Acknowledgments
We thank Manfred Fahle, Anne Kurtenbach and Lukas Rttiger for the loan
of the spectral-sensitivity measuring equipment and its calibration and Seth
Hamlin for technical assistance. Supported by NEI grant EY09258 to PC,
Swiss National Science Foundation Grant NR: 2100-045699.95 to TL, and
UK Defense Research Agency grant D/ER1/9/4/2034/102/RARDE to TT.
RECEIVED 9 MARCH: ACCEPTED 27 MAY 1998
1. Damasio, A., Yamada, T., Damasio, H., Corbett, J. & McKee, J. Central
achromatopsia: Behavioral, anatomic, and physiologic aspects. Neurology
30, 10641071 (1980).
2. Grsser, O.-J. & Landis, T. in Vision and Visual Dysfunction, Vol 12: Visual
Agnosias. (ed. Cronly-Dillon, J.) 394400 (MacMillan, London, 1991).
3. Pearlman, A.L., Birch, J. & Meadows, J.C. Cerebral color blindness: An
acquired defect in hue discrimination. Ann. Neurol. 5, 253261 (1979).
4. Zeki, S.M. A century of cerebral achromatopsia. Brain 113, 17211777
(1990).
5. Mollon, J.D., Newcombe, F., Polden, P.G. & Ratcliff, G. in Colour Vision
Deficiencies (ed. Verriest, G.) 130135 (Hilger, Bristol, 1980).
6. Wooten, B.R. Partial cerebral achromatopsia with selective hue loss. Doc.
Ophthalmol. Proc. Series 33, 139144 (1982).
7. Jaeger, W., Krastel, H. & Braun, S. Cerebral achromatopsia (symptoms,
course, differential diagnosis and examination strategy). II. [Article in
German] Klin. Monatsbl. Augenheilkd. 194, 3236 (1989).
8. Heywood, C.A., Cowey, A. & Newcombe, F. Chromatic discrimination in
a cortically colour blind observer. Eur. J. Neurosci. 3, 802812 (1991).
9. Heywood, C.A., Nicholas, J.J. & Cowey, A. Behavioural and
electrophysiological chromatic and achromatic contrast sensitivity in an
achromatopsic patient. J. Neurol. Neurosurg. Psychiatry 60, 638643
(1996).
10. Victor, J.D., Maiese, K., Shapley, R., Sidtis, J. & Gazzaniga, M.S. Acquired
central dyschromatopsia: Analysis of a case with preservation of color
discrimination. Clin. Vis. Sci. 4, 183196 (1989).
11. Heywood, C.A., Cowey, A. & Newcombe, F. On the role of parvocellular
(P) and magnocellular (M) pathways in cerebral achromatopsia. Brain
117, 245254 (1994).
12. Henaff, M.A. & Michel, F. A case of central achromatopsia: Possible
implicit treatment of hue. Invest. Ophthalmol. Vis. Sci. 34, 745 (1993).
13. Troscianko, T. et al. Human colour discrimination based on a non-
parvocellular pathway. Curr. Biol. 6, 200210 (1996).
14. Barbur, J.L., Harlow, A.J. & Plant, G.T. Insights into the different exploits
of colour in the visual cortex. Proc. R. Soc. Lond. B 258, 327334 (1994).
15. Lee, B.B., Martin, P.R. & Valberg, A. The physiological basis of
heterochromatic flicker photometry demonstrated in the ganglion cells of the
macaque retina. J. Physiol. (Lond.) 404, 323347 (1988).
16. Schiller, P.H. & Colby, C.L. The responses of single cells in the lateral
geniculate nucleus of the rhesus monkey to color and luminance contrast.
Vision Res. 23, 16311641 (1983).
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com
17. Wagner, G. & Boynton, R.M. A comparison of four methods of heterochromatic
photometry. J. Opt. Soc. Am. 62, 15081515 (1972).
18. Scheidler, W., Landis, T., Rentschler, I., Regard, M. & Baumgartner, G. A pattern
recognition approach to visual agnosia. Clin. Vis. Sci. 7, 175193 (1992).
19. Harwerth, R.S. & Sperling, H.G. Effects of intense visible radiation on the
increment-threshold spectral sensitivity of the rhesus monkey eye. Vision Res.
15, 11931204 (1975).
20. King-Smith, P.E. & Carden, D. Luminance and opponent-color contributions to
visual detection and adaptation and to temporal and spatial integration. J. Opt.
Soc. Am. 66, 709717 (1976).
21. Cavanagh P. & Anstis S.M. The contribution of color to motion in normal and
color-deficient observers. Vision Res. 31, 21092148 (1991).
22. Dobkins, K.R. & Albright, T.D. What happens if it changes color when it moves?:
psychophysical experiments on the nature of chromatic input to motion
detectors. Vision Res. 33, 10191036 (1993).
23. Chichilnisky, E.J., Heeger, D. & Wandell, B.A. Functional segregation of color
and motion perception examined in motion nulling. Vision Res. 33, 21132125
(1993).
24. Teller, D.Y. & Palmer, J. Infant color vision: motion nulls for red/green vs
luminance-modulated stimuli in infants and adults. Vision Res. 36, 955974
(1996).
25. Cavanagh, P. Attention-based motion perception. Science 257, 15631565
(1992).
26. Weiskrantz, L. Blindsight revisited. Curr. Opin. Neurobiol. 6, 215220 (1996).
1998 Nature America Inc. http://neurosci.nature.com
27. Metha, A.B., Vingrys, A.J. & Badcock, D.R. Detection and discrimination of
moving stimuli: the effects of color, luminance, and eccentricity. J. Opt. Soc. Am.
11, 16971709 (1994).
28. Stromeyer, C.F. III, Kronauer, R.E., Ryu, A., Chaparro, A. & Eskew, R.T. Jr
Contributions of human long-wave and middle-wave cones to motion
detection. J. Physiol. (Lond.) 485, 221243 (1995).
29. Chaparro, A., Stromeyer, C.F. III, Huang, E.P., Kronauer, R.E. & Eskew, R.T. Jr
Colour is what the eye sees best. Nature 361, 348350 (1993).
nature neuroscience volume 1 no 3 july 1998
articles
30. Ramachandran, V.S. & Gregory, R. Does colour provide an input to
human motion perception? Nature 275, 5556 (1978).
31. Cavanagh, P., Tyler, C.W. & Favreau, O.E. Perceived velocity of moving
chromatic gratings. J. Opt. Soc. Am. 1, 893899 (1984).
32. Cropper, S.J. & Derrington, A.M. Rapid colour-specific detection of
motion in human vision. Nature 379, 7274 (1996).
33. Dobkins, K.R. & Albright, T.D. What happens if it changes color when it
moves?: the nature of chromatic input to macaque visual area MT. J.
Neurosci. 14, 48544870 (1994).
34. Saito, H., Tanaka, K., Isono, H. Yasuda, M. & Mikami, A. Directionally
selective response of cells in the middle temporal area (MT) of the
macaque monkey to the movement of equiluminous opponent color
stimuli. Exp. Brain Res. 75, 114 (1989).
35. Gegenfurtner, K.R. et al. Chromatic properties of neurons in macaque MT.
Vis. Neurosci. 11, 455466 (1994).
36. Ffytche, D.H., Skidmore, B.D. & Zeki, S. Motion-from-hue activates area
V5 of human visual cortex. Proc. R. Soc. Lond. B 260, 353358 (1995).
37. Heywood, C.A., Gaffan, G. & Cowey, A. Cerebral achromatopsia in
monkeys. Eur. J. Neurosci. 7, 10641073 (1995).
38. Hadjikhani, N.K., Liu, A.K., Cavanagh, P., Dale, A.M. & Tootell, R.B.H.
Retinotopy and color sensitivity in human visual cortical area V8. Nature
Neurosci., 1, 235241 (1998).
39. Tootell, R.B., Dale, A.M., Sereno, M.I. & Malach, R. New images from
human visual cortex. Trends Neurosci. 19, 481489 (1996).
40. Zeki, S.M. Uniformity and diversity of structure and function in rhesus
monkey prestriate visual cortex. J. Physiol. (Lond.) 277, 273290 (1978).
41. Ferrera, V.P., Rudolph, K.K. & Maunsell, J.H. Responses of neurons in the
parietal and temporal visual pathways during a motion task. J. Neurosci.
14, 61716186 (1994).
42. Cavanagh, P., Anstis, S.M. & MacLeod, D.I.A. Equiluminance: Spatial and
temporal factors and the contribution of blue-sensitive cones. J. Opt. Soc.
Am. 4, 14281438 (1987).
247
 1998 Nature America Inc. http://neurosci.nature.com

articles

The effects of frontal eye field and

dorsomedial frontal cortex lesions
on visually guided eye movements
Peter H. Schiller and I-han Chou

Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA
Correspondence should be addressed to P.H.S (schiller@wccf.mit.edu)

In the frontal lobe of primates, two areas play a role in visually guided eye movements: the frontal
eye fields (FEF) and the medial eye fields (MEF) in dorsomedial frontal cortex. Previously, FEF lesions
1998 Nature America Inc. http://neurosci.nature.com

have revealed only mild deficits in saccadic eye movements that recovered rapidly. Deficits in eye
movements after MEF ablation have not been shown. We report the effects of ablating these areas
singly or in combination, using tests in which animals were trained to make saccadic eye movements
to paired or multiple targets presented at various temporal asynchronies. FEF lesions produced large
and long-lasting deficits on both tasks. Sequences of eye movements made to successively presented
targets were also impaired. Much smaller deficits were observed after MEF lesions. Our findings indi-
cate a major, long-lasting loss in temporal ordering and processing speed for visually guided
saccadic eye movement generation after FEF lesions and a significant but smaller and shorter-lasting
loss after MEF lesions.

In primates with a small foveal region of densely packed pho-
toreceptors in their retinae that provide high acuity vision, each
eye is moved about by six extraocular muscles, innervated by
neurons in the brainstem oculomotor centers. Rapid saccadic eye
movements are executed two to four times per second to bring
visual objects into central view, and slower, smooth-pursuit eye
movements are made to keep objects in view when either the
observer or the object is in motion. In the frontal lobe, two areas
have been shown to be significant in visually guided eye move-
ments: the frontal eye fields (FEF) and the medial eye fields
(MEF), both of which make direct projections to brainstem ocu-
lomotor centers111. The MEF, located in the dorsomedial frontal
cortex, are also known as the supplementary eye fields8. Single-
cell recording and microstimulation studies have shown that these
two areas perform different coding operations for the generation
of saccadic eye movements3,7,11. Electrical stimulation of the FEF
elicits saccades that have specific directions and amplitudes; pro-
longed electrical stimulation yields a staircase of identical sac-
cades with intervening fixations. By contrast, electrical
stimulation of the MEF produces saccades that take the eyes to
a particular orbital position; prolonged stimulation keeps the
eyes at that position. These findings indicate that the FEF carry a
vector code and the MEF a place code10,11. To shed further light
on the functions of these two areas, we examined the effects of
ablating them, either singly or in combination, and studied the
effects on visually guided eye movements. Most studies have
shown only mild, temporary deficits after FEF lesions1216.
Deficits in visually guided saccadic eye movements have not pre-
viously been demonstrated after MEF lesions in the monkey.
Here we present results obtained on four tasks. The first task
was saccadic eye movements to single targets. Following fixation
of a small central spot, a single target appeared in one of several
locations on the monitor. Execution of an accurate saccadic eye
movement to the target was rewarded with a drop of apple juice.
The second task was saccadic eye movements to paired targets.
Paired targets were presented with various temporal onset asyn-
chronies. Monkeys were rewarded for making a saccadic eye
movement to either target. The onset asynchrony between paired
targets was varied to determine the temporal delay required to
produce equal probability of eye movements to each target. The
third task was saccadic eye movements to sequential targets. On
each trial, two targets were presented in rapid sequence with var-
ious temporal durations and with various delays between them.
The task was to repeat the order of the target presentations by
making successive saccadic eye movements to the target loca-
tions. The fourth task was saccadic eye movements to targets in
arrays. Following fixation, eight stimuli were presented equidis-
tant from the fixation spot; one of the stimuli, the target,
appeared at various times prior to the other seven stimuli. The
monkey was rewarded only for eye movements directed to the
target stimulus. In all four tasks tested, we observed prominent
deficits following FEF lesions. Considerably smaller deficits that
recovered more rapidly were observed after MEF lesions. The
effects of combined FEF and MEF lesions were no greater than
FEF lesions alone.
Results
Figure 1 shows the performance of two monkeys on the single-
target task. Each panel shows the distribution of rightward and
leftward saccadic movement latencies made to individual visual
targets presented randomly in one of four locations. Panels (a)
and (b) compare the distributions before and after a left FEF
lesion. The insets display eye-movement records that show some-
what less accurate saccades to the right after the left FEF lesion; a
significant change in saccadic peak velocities accompanies this
for right saccades from a preoperative 534 to a postoperative 459

248 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 1. Distribution of sac-

cadic latencies to single tar-
a a
Pre-op
Pre-op
b b
3 wks post LFEF c
3 wks post LFEF
gets before and after lesions.
Data are shown from two 100
140
monkeys with (a) and (d) rep- 80
resenting pre-operative data. 100
60
(b) shows the distribution of
left and right saccadic eye 60 40

Numberofofsaccades
saccades
movements three weeks after 20
left FEF lesion; (e) shows simi- 20
lar data two weeks after a 60 100 140 180 220 60 100 140 180 220
right MEF lesion in the second
animal. (f) shows data col- d Pre-opd e 2
e
wks post RMEF f 2 wks post MEF and LFEF
f
Number

Pre-op 2 wks post RMEF 2 wks post MEF and LFEF

lected after a left FEF lesion
120
had been made in the same 140 250
animal subsequent to bilateral 100
200
MEF lesions. Eye-movement 100 80 left saccades
records before and after the 150
60 right saccades
left FEF lesion appear in the
1998 Nature America Inc. http://neurosci.nature.com

60 100 40
inset of (a) and (b). Mean left
and right saccadic latencies 50 20
20
were as follows: (a) L, 125; R,
121, (b) L, 112; R, 166, (d) L, 60 100 140 180 220 60 100 140 180 220 60 100 140 180 220
130; R, 130, (e) L, 141; R, 130, Latency in milliseconds
Latency milliseconds
(f) L, 121; R, 175. The 11-ms
difference in mean left and right saccadic latencies two weeks after the right MEF lesion (e) is significant at the .05 level. Differences after FEF
lesions (b and f) are significant beyond the .001 level. (c) Reconstruction of an MEF and FEF lesion from a third monkey. The major sulci are
labeled as are the locations of the FEF and MEF. Ant: anterior portion of the brain. Number of trials per histogram: (a) n = 1200, (b) n = 720,
(d) n = 1120, (e) n = 2160, (f) n = 1092.

degrees per second (p < .001, t-test). In (b) an increase of 45 ms
is evident in the mean latency of saccadic eye movements made to
the right as compared with pre-operative performance (p < .001,
t-test); the lesion also produced a broadening in the latency dis-
tribution for rightward saccades. Saccadic eye movements made
to the left after the left FEF ablation yielded significantly shorter
latencies than pre-operatively by 13 ms (p < .05). Panels (d), (e),
and (f) show pre- and post-operative latency distributions from
another monkey. The right MEF lesion produced a small, but
significant increase in latencies for leftward saccadic eye move-
ments (11 ms, p < .05, t-test). Saccadic accuracy and velocity were
unaffected by the MEF lesions. The left FEF lesion made in this
animal subsequent to the bilateral MEF lesion (Fig. 1f) produced
deficits similar to those obtained with a single left FEF lesion,
yielding a difference of 54 ms in the mean latencies between left
and right saccades. Four months after the left FEF lesion in mon-
key 1, the latency difference between saccades made to the left
and right dropped to 25 ms; four months after the paired FEF
and MEF lesions in monkey 2, the difference dropped to 24 ms
(both with p < .001, t-test). Saccadic velocity differences for left-
ward and rightward saccadic eye movements were also dimin-
ished but remained significant.
The paired-target task we devised permitted us to quantify the
degree to which target selection was biased by the cortical lesions.
We presented paired targets with different onset asynchronies to
determine the interval required to produce equal probability of
target selection appearing in the right and left hemifields. Results
obtained on this task using targets with an angular separation of
90 degrees relative to the fixation spot appear in Fig. 2. The paired
targets appeared either above or below the fixation point, with
one stimulus falling in the left and the other in the right visual
hemifield. Paired targets were interspersed with single targets that
appeared twice as often; conditions were configured to have the
monkeys execute equal numbers of saccadic eye movements to
the left and to the right during each session, thereby forestalling
the emergence of position biases.
The inset in Fig. 2 shows preoperatively collected eye-move-
ment records for paired targets that appeared either simultane-
ously or with asynchronies favoring either the left or the right
stimulus in the pair. With simultaneous presentation, the prob-
ability of making saccadic eye movements to the left and right
targets was equal, whereas when one target preceded the other
by 33 ms, most saccades were made to the stimulus that had
appeared first. The percent of saccades made to the left target at
different asynchronies are plotted in the center of the graph
labeled pre-op. Also plotted is the percent of saccadic eye move-
ments made into the left visual field at various times after a left
FEF and a right MEF lesions in two different animals. Two weeks
after the left FEF lesion, the curve shifts to the far left. To achieve
an equal probability of left and right saccades (50% crossover
point), the target in the affected right hemifield had to be pre-
sented 116 ms prior to the target in the left hemifield. In the
weeks following the lesion, the size of the asynchrony required
for equivalent target choice decreased gradually, as shown by the
successive curves on the left side of the graph in Fig. 2.
The shift in target choice after a right MEF lesion, as shown
by the curves on the right in Fig. 2, was much smaller in magni-
tude. Two weeks after the lesion, an asynchrony of 31 ms was
required for the monkey to make saccades with equal probabili-
ty to either target. Recovery after the MEF lesion was complete
by the 16th week. At a comparable time after the left FEF lesion,
there was still a strong bias in favor of the ipsilateral target, requir-
ing a 54 ms target-onset asynchrony for equal probability per-
formance. At comparable post-operative times, the asynchronies
required to produce equal probability choice were more than
twice as long as the latency differences between left and right-

nature neuroscience volume 1 no 3 july 1998 249

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 2. Paired

targets
targets pre- Pre-op eye movements made to paired targets
100 LFEF lesion pre-op
pre-op Pre-op eye movements made to paired targets

targets
sented with var- RMEF lesion L,
L, 33ms
33 ms 00ms
ms R, 33
R, 33ms
ms
ied asynchronies. 80

leftleft
wk 3
wk3
wk
wk22
The percent of wk16
wk 16
to to
60
saccades made saccades
to the left target 40 wk16
wk 16 wk 2
saccades
wk2
as a function of
20
temporal offset
Percent

between the 0
Percent

Right target first Left target first

paired targets is
plotted. Data are 220 200 180 160 140 120 100 80 60 40 20 0 20 40 60 80 100 120 140 160 180 220 220
shown at various Onset
Onset asynchrony in milliseconds
milliseconds
times after
recovery from FEF and MEF lesions as well as pre-operatively (pre-op). Squares and solid lines depict data from the animal with the FEF
lesion; circles and dotted lines depict data from the animal with the MEF lesion. Records of saccadic eye movements made to paired targets
with various temporal offsets in the intact animal appear as an inset. For each of the post-operative weeks plotted (wk 2wk 16), data were
collected for 5 or 6 successive days.
1998 Nature America Inc. http://neurosci.nature.com

ward saccades made to single targets (see Fig. 1). There was no
further improvement on the paired target task when the mon-
key was tested five months after FEF lesions.
When two targets have an angular separation of 90 degrees,
most of the time monkeys generate accurate saccadic eye move-
ments to one target or the other, as shown in the inset in Fig. 2.
However, when the target separation is decreased, animals
begin to make averaging saccades that land the center of gaze at
intermediate positions between the two targets17. Such vector-
averaged saccades are presumably the product of simultane-
ously arriving commands from the cortex to the superior
colliculus or the brainstem to move the eyes to each target.
Studies have shown that concurrent electrical stimulation of
two sites in the superior colliculus of the FEF produces sac-
cadic eye movements that are a vector average of saccades elicit-
ed by stimulating each site alone 18,19 . Figure 3 shows data
collected with targets that had an angular separation of 40
degrees; under such conditions vector- averaged saccades
become common. In the unlesioned monkey, vector-averaged
saccades occur most frequently when the targets appear simul-
taneously; the frequency of such saccades falls off rapidly as
target-onset asynchrony is increased. The consequence of an
FEF lesion is a dramatic shift in the asynchrony between the
targets at which vector-averaged saccades occur most com-
monly; as a result of the lesion there are now no vector-aver-
aged saccades when the targets are simultaneous; instead, they
become most frequent when the target in the hemisphere con-
tralateral to the lesion is presented 67100 ms prior to the tar-
get in the left hemisphere. These findings suggest that FEF
lesions retard the rate at which information can be processed
for the generation of visually guided eye movements.
The dorsomedial frontal area within which the MEF resides
has been implicated in the temporal processing of events, par-
ticularly in the execution of sequential motor acts2022. We there-
fore wanted to determine how MEF and FEF lesions alter the
production of a sequence of eye movements to successively pre-
sented targets. In Fig. 4, data from the sequential task are pre-
sented. Four different sequences with several different durations
were presented in randomized order. Pre-operative and post-
operative data collected at various times after the lesions are
shown. The results demonstrate a significant but mild deficit
after MEF lesions, which is consistent with findings in humans20.
However, a much larger deficit was evident after the FEF lesion.
The inset shows eye-movement records collected two months
after a left FEF lesion using four sets of sequentially presented
targets with a sequence duration of 117 ms. Performance on

a bb
averaged

50 Intact monkey LFEF lesion

vectoraveraged

40
Percent

30
Percent

Intact monkey
saccadesvector

20
L, 67 ms 0 ms R, 67 ms R, 100 ms R, 133 ms 10
saccades

0
LFEF lesion
60 40 20 0 20 40 60 80 100 120 140
Left target first Right target first
Onset
Onsetasynchrony
asynchronybetween
between paired targets
targets

Fig. 3. Data collected with paired targets having a 40% angular separation. (a) Eye-movement records obtained at various target asyn-
chronies; text shows which target appeared first (L, left; R, right) and by how many milliseconds (ms). (b) Plot of the frequency of vector-
averaged saccades as a function of target asynchrony. The data were obtained from an intact animal and from an animal with a left FEF lesion.
We counted as vector averaged those saccades that fell within plus or minus nine angular degrees of the midpoint between the two targets.

250 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 4. Pre and post-operative data

for percent correct performance are
a b Post RMEF
RMEF lesion
lesion
Preop Post Eye movements, post LFEF
shown for left and right saccadic eye wk
wk55 Eye movements, post LFEF
100 100
movements made to sequential tar- L
gets before and after a right MEF 80 80
lesion (a, b) and before and after a R
60 60
left FEF lesion (c, d). Two stimuli wk 1
wk1
appeared in succession using four dif- 40 40
ferent sets of locations and several
L
different overall durations. The inset 20 20

shows eye movements made to one

Percent correct
Percent Correct
set of target sequences after a left FEF 125 150 175 200 225 250 125 150 175 200 225 250 275
lesion, when presented for a
sequence duration of 117 ms, which cc Preop d Post LFEF lesion
is completed well before the initiation 100
L 100
L
of the first eye movements. The four
sets of sequential targets, as indicated 80 80
wk11
wk 11
R
in the inset, appeared at positions 60 60 wk
wk33
wk 6
wk6
A2B1, E2D1, A4B5 and E4D5.
1998 Nature America Inc. http://neurosci.nature.com

40 40
Each trial began with a central fixation
R
spot followed by the appearance of 20 20
one sequence. The eye movements wk
wk22
0 0
shown in the inset demonstrate cor-
rect sequences made to the intact left 125 150 175 200 225 250 125 150 175 200 225 250 275 300 325 350 375 400 425 450 475 500
side and mostly incorrect eye move- Sequenceduration
Sequence Duration in
in milliseconds
milliseconds
ments made for sequences presented
to the right.

sequences presented to the right was severely impaired. The most
common error was the execution of a single saccade, instead of
a sequence of saccades, that landed the center of gaze near either
of the target locations or between them.
Examination of saccadic latencies made to the first target in
the sequence suggests that target reaction times make only a small
contribution to the deficit. For example, three weeks after the left
FEF lesion, the mean latencies to the first target in the sequences
we used were 130 ms to the left and 198 ms to the right (a laten-
cy difference of 68 ms). Yet the comparable performance to the
left and right occurs with a sequence duration difference of more
than 200 ms (Fig. 4).
Full recovery on the sequential task is evident five weeks
after MEF lesions. Following FEF lesions, considerable recov-
ery occurs by week 11. Combined MEF and FEF lesions pro-
duced deficits and recovery times similar to those obtained
from the FEF lesion alone. The results from the sequential task
a a lesion
LFEF lesion b b
MEFand
MEF
100 100
correct
suggest that for the generation of sequences of eye movements,
the FEF are more important than the MEF.
The paired-target and sequential tasks do not directly test the
ability of animals to make a temporal discrimination. To assess
deficits in ascribing temporal order to successively appearing
events, we devised a task that required the animals to discrimi-
nate stimuli on the basis of the order in which they appeared.
Eight identical stimuli were presented equidistant from fixation,
one of which preceded the others by various times. Only a direct
saccadic eye movement made to the stimulus that appeared first
was rewarded. This task explicitly requires that the animal dis-
criminate the relative onset of the stimuli. Data collected four
months after a left FEF and paired MEF and left FEF lesions
appear in Fig. 5a and b. A major impairment is evident for eye
movements made into the contralateral visual field. This deficit is
not due to loss of perception of high-frequency temporal infor-
mation per se, as the monkeys were unimpaired on flicker sen-
sitivity. This suggests that the
deficit lies not in perceiving tem-
& LFEF
LFEFlesions
lesions poral discontinuity but in a dis-
ability to assess the temporal order
of presentations for the generation
of saccadic eye movements.

80 80
Percent correct

left hemifield In addition to the four tests

60 60 described here, the performance of
right hemifield
Percent

40 40 monkeys was also assessed for eye

movements made to targets that
20 20
moved at various velocities, on
0 N = 3220 0 N = 2720 sensitivity to stimuli of varied con-
trasts, on their ability to discrimi-
50 100 150 200 250 300 50 100 150 200 250 300
Target onset asynchrony in milliseconds
nate targets that were of a different
Target onset asynchrony in milliseconds
size, shape or color from other,
Fig. 5. Performance on the discrimination of time differences. Percent correct performance is shown simultaneously appearing stimuli,
separately for targets appearing in the left and right hemispheres. (a) Data from a monkey with a left and their ability to discriminate
FEF lesion. (b) Data from a monkey after paired MEF and FEF lesions. (a) n = 3220, (b) n = 2720. flickering stimuli from steadily illu-

nature neuroscience volume 1 no 3 july 1998 251

 1998 Nature America Inc. http://neurosci.nature.com
articles
minated ones. Except for an increase in saccadic latencies, per-
formance on these tasks was largely unaffected by the lesions.
Discussion
In foveate animals, with each shift in gaze not only are stimuli in
central vision analyzed but a decision has to be made as to where
to look next. The findings we report here suggest that the FEF
are central to this process of target selection. Particularly, the FEF
are involved in the timing of saccadic eye-movement generation
and in optimizing the speed with which eye movements can be
initiated to objects in the visual scene. In intact animals, there is
cooperative interaction between the two FEF23. Following a uni-
lateral FEF lesion, target selection and processing speed are biased
in favor of the intact hemisphere. Even five months after the
lesion, sizeable deficits remained on our tasks that tapped this
competition. In contrast, the MEF lesions yielded much smaller
deficits suggesting that this area is not as centrally involved in
target selection and in the execution of sequences of eye move-
1998 Nature America Inc. http://neurosci.nature.com
ments. This is consistent with single-cell recording studies show-
ing that neuronal activity in the FEF is more predictive of saccade
generation and timing than is the MEF (ref. 24, 25, Patterson,
W.F. II & Schall, J.D. Soc. Neurosci. Abstr., 185.8, 1997).
Our paired-stimulus task is similar to a task used to study
patients with brain lesions in which paired visual or tactile stimuli
are presented in the right and left hemifields or parts of the body.
The tendency to ignore the stimulus that appears in space con-
tralateral to the lesion has been termed the extinction phenome-
non.26 Adding precisely defined temporal asynchronies to the
presentation of paired targets, as we did in this study, makes it pos-
sible to obtain exact measurements of the temporal factors involved.
As a result, it is possible to ascribe an exact value in time for the
magnitude of the deficit incurred and its recovery. Our results raise
the possibility that the extinction phenomenon seen after frontal
and parietal lesions is due not to a loss of attention per se, but to a
loss in the speed of processing in the impaired hemisphere.
What brain structures might be responsible for bringing about
the gradual improvement in performance found on the tasks we
had used? It has been proposed that two major cortical streams
contribute to the control of visually guided saccadic eye move-
ments, the anterior and the posterior 27. Because even after
removal of both major areas of the anterior stream, the FEF and
the MEF, there is notable improvement in performance over time,
it is probable that the posterior stream is involved in the recovery.
This stream originates in the occipital and parietal lobes and
reaches the brainstem oculomotor areas predominantly through
the superior colliculus2830.
Methods
Four monkeys were trained on a variety of visually guided eye-movement
tasks that allowed us to assess the relative effects of frontal eye field, dor-
somedial frontal cortex and combined lesions on the execution of sac-
cadic eye movements to single and paired target stimuli. Following
training, in the first animal, a unilateral lesion was made of the left FEF. In
the second animal, successive lesions of the left dorsomedial frontal cor-
tex and of the right FEF were made several months apart. In the third ani-
mal, three lesions were made, also several months apart; initially the left
dorsomedial frontal cortical area containing the left MEF was ablated,
then the right MEF area, and finally the left FEF. The monkeys were test-
ed extensively after each lesion for several months before the second and
third lesions were made. The fourth monkey served as a control animal.
All but one of the lesions were made by aspiration under aseptic con-
ditions in anesthetized animals with the aid of a surgical microscope.
The FEF was removed by aspirating the anterior bank of the arcuate sul-
252
cus 6 mm medial and 6 mm lateral from the posterior tip of the princi-
pal sulcus as well as the gray matter between the arcuate and the posterior
tip of the principalis. The fundus and the posterior bank of the arcuate
involved in smooth pursuit eye movements were spared3133. Anteriorly
these lesions encroached upon area 46. In the second animal, the left
dorsomedial cortex area was destroyed after we recorded and stimulated
the area extensively to establish the exact location of the MEF; this par-
ticular lesion was produced by repeated, closely spaced lidocaine injec-
tions until neural activity was permanently shut down, resulting in a
lesion verified histologically and shown in Fig.1c. Subsequently in this
animal, the right FEF was removed by aspiration and was also verified
histologically. In the third animal, the same region of dorsomedial frontal
cortex was removed as in the second animal. The lesions extended 8 mm
laterally from the midline and 6 mm anterior and posterior from a point
that was in the same coronal plane as the posterior tip of the principal
sulcus. Detailed photographs were taken during surgery before and after
each aspiration. Electrophysiological mapping was carried out only in
the second animal. All protocols were approved by the MIT Animal Care
Committee and followed NIH guidelines.
The animals were tested using a color monitor placed at a distance of
57 cm. The head was restrained during testing, and eye movements were
measured using the scleral search coil method. Monkeys readily per-
formed 8003000 trials per day. Background luminance for the single,
paired and multiple target tasks was 2.26 cd per m2. The targets were
small, 0.34 degree of visual angle squares with a luminance of 90.27 cd per
m2. Saccadic latencies were computed from the time of target onset to
the time of departure of the eye-movement trace from an electronic win-
dow set around the fixation spot. The targets most commonly appeared
at an eccentricity of 12 degrees from fixation.
The sequential task was made distinct from the other tasks in four
ways. First, twenty-five equally spaced small outline squares of low con-
trast were present on the screen throughout. Second, the targets appeared
inside the selected squares and were bright red in color. Third, the tar-
gets flashed on briefly in rapid succession, with a gap between them.
Fourth, the animal was rewarded only when two correct successive sac-
cadic eye movements were made to the successively appearing target
positions. The target durations and the delays introduced between the
first and second target were arrived at experimentally and were chosen
to optimize the ability of monkeys to perform the sequential task suc-
cessfully. Several different sets or paired target locations were studied,
which were run in blocks always using four pairs arranged in a mirror-
image fashion to allow us to compare performance in the left and right
hemispheres. The first target duration ranged between 33 and 300 ms,
the interval between 50 and 117 ms and the second target between 33
and 83 ms using 16.7 ms steps (the frame rate for the monitor). Within
each block the sequence durations as well as the mirror-imaged pair loca-
tions appeared in randomized order. (See inset in Fig. 4 showing eye
movements made to one set of four pairs used in a block.)
For the temporal-discrimination task, eight stimuli appeared, one of
which preceded the others by various times ranging from 16 ms to 300
ms. The eight stimuli were identical and the same size and luminance as
the targets presented in the single- and paired-target case. All the stimuli
were equally spaced around a circle with a radius of 12 degrees centered
on the fixation spot.
Acknowledgements
We thank J. Colby and W. Slocum for their technical assistance.
RECEIVED 3 MARCH: ACCEPTED 18 MAY 1998
1. Bizzi, E. Discharge of frontal eye field neurons during eye movements in
unanesthetized monkeys. Science 157, 15881590 (1967).
2. Bruce, C.J. & Golberg, M.E. Primate frontal eye fields: I. Single neurons
discharging before saccades. J. Neurophysiol. 53, 603635 (1985).
3. Chen, L.L. & Wise, S.P. Supplementary eye field contrasted with frontal eye
field during acquisition of conditional oculomotor associations. J.
Neurophysiol. 73, 11221134 (1995).
4. Mann, S.E., Thau, R. and Schiller, P.H. Conditional task-related responses in
monkey dorsomedial frontal cortex. Exp. Brain Res. 69, 460468 (1988).
nature neuroscience volume 1 no 3 july 1998
 1998 Nature America Inc. http://neurosci.nature.com
5. Robinson, D.A. & Fuchs, A.F. Eye movements evoked by stimulation of
frontal eye fields. J. Neurophysiol. 32, 637648 (1969).
6. Russo, G.S. & Bruce, C.J. Neurons in the supplementary eye field of the
rhesus monkeys code visual targets and saccadic eye movements in an
oculocentric coordinate system. J. Neurophysiol. 76, 825848 (1996).
7. Schall, J.D. Neuronal activity related to visually-guided saccades in the frontal
eye fields of rhesus monkeys: comparison with supplementary eye fields. J.
Neurophysiol. 66, 530579 (1991).
8. Schlag, J. & Schlag-Rey, M. Evidence for a supplementary eye field. J.
Neurophysiol. 57, 179200 (1987).
9. Shook, B.L., Schlag-Rey, M. & Schlag J. Primate supplementary eye field: I.
Comparative aspects of mesencephalic and pontine connections. J. Comp.
Neurol. 301, 618642 (1990).
10. Tehovnik, E.J. The dorsomedial frontal cortex: eye and forelimb fields. Behav.
Brain Res. 67, 147163 (1995).
11. Tehovnik, E.J. & Lee, K.M. The dorsomedial frontal cortex of the rhesus
monkey. Topographic representation of saccades evokes by electrical
stimulation. Exp. Brain Res. 96, 430442 (1993).
12. Deng, S-Y., Goldberg, M.E., Segraves, M.A., Ungerleider, L.G. & Mishkin, M.
in Adaptive Processes in the Visual & Oculomotor Systems (eds Keller, E. & Zee,
D. S.) 201208 (Pergamon, Oxford, 1986).
13 Dias, E.C., Kiesau, M. & Segraves, M.A. Acute activation and inactivation of
macaque frontal eye field with GABA-related drugs. J. Neurophysiol. 74,
27442748 (1995).
1998 Nature America Inc. http://neurosci.nature.com
14. Latto, R.A. & Cowey, A. Visual field defects after frontal eye field lesions in
monkeys. Brain Res. 30, 124 (1971).
15. Schiller, P.H., Sandell, J.H. & Maunsell, J.H.R. The effect of frontal eye field
and superior colliculus lesions on saccadic latencies in the rhesus monkey. J.
Neurophysiol. 57, 10331049 (1987).
16. Sommer, M.A. & Tehovnik, E.J. Reversible inactivation of macaque frontal
eye field. Exp. Brain Res. 116, 229249 (1997).
17. Ottes, F.P., Van Gisbergen, J.A.M. & Eggermont, J.J. Metrics of saccade responses
to visual double stimuli, two different modes. Vision Res. 24, 11691197 (1984).
18. Robinson, D.A. Eye movements evoked by collicular stimulation in the alert
monkey. Vision Res. 12, 17951808 (1972).
19. Schiller, P.H., True, S.D. & Conway, J.L. Paired stimulation of the frontal eye
fields and the superior colliculus of the rhesus monkey. Brain Res. 179,
162164 (1979).
nature neuroscience volume 1 no 3 july 1998
articles
20. Gaymard, B., Pierrot-Deseilligny, C. & Rivaud, S. Impairment of sequences
of memory-guided saccades after supplementary motor area lesions. Annals
Neurol. 28, 622626 (1990).
21. Mushiake, H., Inase, M. & Tanji, J. Neuronal activity in the primate premotor,
supplementary and precentral motor cortex during visually guided and
internally determined sequential movements. J. Neurophysiol. 66, 705718
(1991).
22. Tanji, J. & Shima K. Role for supplementary motor area cells in planning
several movements ahead. Nature 371, 413416 (1994).
23. Schlag, J., Dassonville, P. & Schlag-Rey, M. Interaction of the two frontal eye
fields before saccade onset. J. Neurophysiol. 79, 6472 (1998).
24. Hanes, D.P. & Schall, J.D. Neural control of voluntary movement initiation.
Science 274, 427430 (1996).
25. Hanes, D.P., Patterson, W. F. & Schall, J.D. The role of frontal eye fields in
countermanding saccades: visual, movement, and fixation activity. J.
Neurophysiol. 79, 817834 (1998).
26. Bisiach, E. & Vallar, G. in Handbook of Neuropsychology Vol. 1 (eds Boller, E. &
Grafman, J.) 195222 (Elsevier, Amsterdam, 1988).
27. Schiller, P.H. in Cognitive Neuroscience of Attention. (ed Richards, J.) 350
(Lawrence Erlbaum, London, 1998).
28. Keating, E.G., Gooley, S.G., Pratt, S. & Kelsey, J. Removing the superior
colliculus silences eye movements normally evoked from stimulation of the
parietal and occipital eye fields. Brain Res. 269, 145148 (1983).
29. Schiller, P.H. The effect of superior colliculus ablation on saccades elicited by
cortical stimulation. Brain Res. 122, 154156 (1977).
30. Tehovnik, E.J., Lee, K-M. & Schiller, P.H. Stimulation-evoked saccades
from the dorsomedial frontal cortex of the rhesus monkey following
lesions of the frontal eye fields and superior colliculus. Exp. Brain Res. 98,
179190 (1994).
31. Gottlieb, J.P., Bruce, C.J. & MacAvoy, M.G. Smooth eye movements elicited
by microstimulation in the primate frontal eye field. J. Neurophysiol. 69,
786799 (1993).
32. Keating, E.G. Lesions of the frontal eye field impair pursuit eye movements
but preserve the predictions driving them. Behav. Brain Res. 53, 91104
(1993).
33. Lynch, J.C. Saccade initiation and latency deficits after combined lesions of
frontal and posterior eye fields in monkeys. J. Neurophysiol. 68, 19131916
(1992).
253
 1998 Nature America Inc. http://neurosci.nature.com

articles

Top-down influences on stereoscopic

depth-perception
Isabelle Blthoff1, Heinrich Blthoff1 and Pawan Sinha2

1 Max-Planck-Institut fr biologische Kybernetik, 72076 Tbingen, Germany

2 Department of Psychology, University of Wisconsin, Madison, Wisconsin 53706, USA
Correspondence should be addressed to I.B. (isabelle.buelthoff@tuebingen.mpg.de)

The interaction between depth perception and object recognition has important implications for the
nature of mental object representations and models of hierarchical organization of visual
processing. It is often believed that the computation of depth influences subsequent high-level
1998 Nature America Inc. http://neurosci.nature.com

object recognition processes, and that depth processing is an early vision task that is largely immune
to top-down object-specific influences, such as object recognition. Here we present experimental
evidence that challenges both these assumptions in the specific context of stereoscopic depth-
perception. We have found that observers recognition of familiar dynamic three-dimensional (3D)
objects is unaffected even when the objects depth structure is scrambled, as long as their two-
dimensional (2D) projections are unchanged. Furthermore, the observers seem perceptually
unaware of the depth anomalies introduced by scrambling. We attribute the latter result to a top-
down recognition-based influence whereby expectations about a familiar objects 3D structure over-
ride the true stereoscopic information.

Our visual system can often recognize objects not only on the basis
of their static appearance but also by observing how they move.
In some cases, impoverished image sequences can be recognized
from their pattern of motion despite the fact that no single frame
has enough figural information to support recognition (Fig. 1).
Image sequences such as those devised by Johansson1, which con-
vey vivid impressions of humans engaged in various dynamic activ-
ities, are elegant and powerful demonstrations of this fact.
Since the first reports of this work, the issue of how motion
sequences are interpreted by the primate brain has been exten-
sively studied. Previous work2 has shown that observers can
identify human individuals from their gait alone. Coding the-
ory principles have also been applied to model gait percep-
tion3. A model for the interpretation of these sequences has
been developed4 by adapting structure-from-motion ideas5.
Similar approaches have also been developed by other
researchers 6,7. Neurons that respond selectively to biological
motion sequences have been described in the visually respon-
sive areas of the primate temporal cortex 8 , a region that is
known to be involved in object recognition.
Most of the studies that have attempted to explain human
perception of biological motion sequences have done so largely in
terms of bottom-up mechanisms, in which the object geometry
is extracted from low-level features without recourse to higher-
level internal representations of objects. In this report, however,
we examine the possibility that internal object representations
may also play a top-down role in this process. Our stimuli were
stereo views of walking human figures that were defined by a
small number of dots; we have used depth-distorted versions of
these figures to study the interactions between depth cues and
recognition. Our results provide two pieces of evidence in this
regard. First, they suggest that the anomalous stereo-depth cues
do not significantly influence the recognizability of the stimuli.
Second, they show that top-down recognition-based influences
can strongly alter depth perception, such that expectations about
a familiar objects 3D structure override the true stereoscopic
information. Consistent with this hypothesis, we have found that
reducing the recognizability of objects reduces the magnitude of
the top-down influence on depth perception.
Results
As stimuli, we used variants of previously described biologi-
cal motion sequences1 (see Fig. 1). They were 3D stereo ani-
mations showing twelve points on a male human (three points
positioned at the joints of each limb) as he walked on a tread-
mill at a normal pace. We compared the normal version of this
stimulus with a depth-scrambled version in which the depth
positions of the joints were randomly altered in the z-axis. This
rendered their 3D trajectories arbitrary within the volume
defined by the original structure, while leaving their 2D pro-
jections on the retina unchanged (discounting the minor posi-
tional shifts induced by the need to incorporate depth-disparity
information). The degree of arbitrary depth scrambling of the
scrambled walker was a continuously variable parameter. We
defined the extent of added depth noise as a function of the
depth-extent (say, D) of the original undistorted walker. Thus,
a noise level of 0 % corresponded to an undistorted sequence,
whereas a noise level of 50 % implied a sequence wherein the
individual body points could assume any depth value (with
uniform probability) within a bound of D/2 about their orig-
inal depth position. Additionally, we generated random ver-
sions of both the normal and the depth-scrambled walkers by
scrambling the x and y positions of their constituent points.
Unlike the purely depth-scrambled version, this created dis-
tortions of the 2D retinal projection. In all sequences, the
added offsets were kept constant from frame to frame.

254 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com

articles

a would be expected to increase in going from a viewing direction

Fig. 1. Example of stimulus. (a) Stereogram
of a single frame of a motion sequence of a
b
human walker represented by dots only. The
structure of a human walker is only recog-
nizable if the dots begin to move accordingly
1998 Nature America Inc. http://neurosci.nature.com
perpendicular to the distorted depth axis (90) to one parallel
to it (0), as the monocular view comes to resemble more close-
ly a 2D human figure.
The key result (Fig. 3) is that irrespective of the depth struc-
ture of the sequences, viewpoints preserving the normal 2D pro-
jections yielded biological motion percepts (high ratings). Other
viewpoints for the scrambled sequence yielded percepts of ran-
domly moving dots (low ratings). The data strongly indicate that
the recognition process used by the subjects in this task is heav-
ily biased towards 2D traces. Stereo-depth information does not
seem to contribute significantly to the recognition processes.
The most surprising result of this experiment is that depth-
scrambled motion sequences that had normal 2D traces were
rated as highly as unscrambled sequences. There are at least two
possible explanations for this. Either subjects might be percep-
tually aware of the depth scrambling but decide nevertheless to
base their ratings on the similarity of the 2D projection, or they

to their biological motion pattern for as

short as 100 to 200 ms1. (b) We illustrate
the structure of the human figure by adding
connecting lines to the dots.
In our first experiment (recognizability experiment), we
asked whether randomizing the depth structure of the moving
figure while preserving its 2D traces would adversely affect its
recognizability as a human. Subjects viewed in stereo the dis-
torted walker sequences interspersed with random and human
(unscrambled) sequences. The viewing position was varied
between 0 (the figure is seen walking in place facing right; depth
distortion does not affect the positions of the points in the image
plane) and 90 (the walker is walking toward the observer; the
same distortion now results in displaced points in the new image
plane, see Fig. 2). The subjects rated all sequences for their struc-
tural goodness as a human on a scale from 1 (completely ran-
dom) to 5 (completely human). We expected that if stereo-depth
information were critical for the recognition processes, subjects
would perceive our depth-scrambled sequences as random
objects from all viewpoints because the 3D structure of the
depth-distorted walker would be completely different from a
human figure. The ratings assigned to these sequences would,
therefore, be uniformly low for all viewing positions. If, howev-
er, recognition required merely 2D congruence, then the rating
might be perceptually unaware of the depth scrambling, possi-
bly due to a top-down object-specific influence that actively
imposes the expected structure on the input and thus suppress-
es the perception of 3D anomalies.
To distinguish between these possibilities, we designed a sec-
ond experiment (depth-plane experiment) to test for the exis-
tence of any recognition-dependent influences that might serve to
suppress information about depth-anomalies being provided by
low-level stereo processes. To assess observers ability to perceive
the true depth structure of these sequences, we designed a simple
task that required them to report whether three indicated points
in the structure were in the same fronto-parallel depth plane.
Stereo viewing was used throughout the experiment. Each exper-
imental trial commenced with a presentation of either a depth-
scrambled walker or a random pattern for the duration of one
walk-cycle, this being sufficient time to allow the moving figure
to be recognized. As the presentation continued into the next
cycle, three of the points were highlighted by thin red outlines.
After two-thirds of the duration of a walk-cycle, the screen then
turned blank. Subjects had been instructed to report whether the
three red dots lay in the same fronto-parallel depth plane. In 50 %
of the sequences, the three dots were in the same plane and in
50% they were not. In the depth-distorted figures, we could vary
the depth of the highlighted dots independent of their positions
on the limbs (same versus different limb); this allowed us to ask
whether the perceived depth was influenced by the expectation
that points on the same limb would be at the same depth.

a (a)
Y
b Y

(b)

X X
Z Z
90 90
0 0

Fig. 2. In the recognizability experiment, subjects viewed depth-scrambled and normal sequences in stereo from different viewing positions
indicated by the arrows along the equator at waist level in (a) and (b). (a) Undistorted walker. (b) Depth-scrambling a biological motion
sequence involves adding depth noise to the positions of the joints while leaving their 2D positions (in the xy-plane) largely unchanged. From
other viewing positions (e.g., in the yz projection shown on the right), the original 2D pattern of a human figure is severely distorted. In all fig-
ures, connecting lines serve to enhance recognizability of the human figure and are not shown in the experiments.

nature neuroscience volume 1 no 3 july 1998 255

 1998 Nature America Inc. http://neurosci.nature.com

articles

Fig. 3. Results of the recognizability experiment

a human
In Fig. 4a and b, the 5
averaged across 22 subjects. The following stimulus

a human
false-alarm rate (a response
sequences were presented: human walker (no dis-
of in same plane when the
4 z-distortion tortion), depth-distorted walker (a constant z-dis-

as as
z-dist
dots are in fact not in the xz-distortion
xz-dist
tortion was applied from frame to frame, the amount

goodness
no-dist
no distortion
same plane) is shown as a

Structuralgoodness
n=22 of depth noise added was 100%) and random pattern
function of depth-noise 3
(constant xz-distortion). The subjects rated them on
level and depth-disparity a scale from 1 (very random) to 5 (very human).
respectively. The curves 2 Abscissa: viewing position in degrees, at 0 the
Structural
show data for three stimu- walker is seen walking to the right with its depth axis
lus types: (1) human 1 parallel to the viewing axis; at 90 the walker is seen
sequences with dots on the 0 9 18 27 36 45 54 63 72 81 90 walking towards the observer with its depth axis
same limb, (2) human Viewingposition
Viewing position(deg)
(deg) perpendicular to the viewing axis. P z-distortion, R
sequences with dots on dif- xz-distortion, H no distortion. n = 22.
ferent limbs, and (3) ran-
dom sequences. If the
perceived depth is affected
by prior expectations based on object recognition, the dots on experiment suggests that the recognition process is based large-
the same limb would be most likely to be perceived as copla- ly on matching the stimulus to an internal representation of the
1998 Nature America Inc. http://neurosci.nature.com

nar, and those on different limbs would be least likely. The false
alarm rate is highest for the same limb condition. In Fig. 4c,
the hit-rate (correct responses of in same plane) is shown as
a function of depth noise. It is uniformly high for the same
limb condition (average value, 94%, SE 1%), lower for the ran-
dom sequences (average value, 72%, SE 1%) and lowest for the
different limbs condition. Thus, the perceived depth is influ-
enced by prior expectations about the depth structure of the
image, which are in turn determined by its recognizability.
Discussion
Our results have important implication for the nature of the men-
tal representations for dynamic objects. The limited influence of
depth information on object recognition observed in the first
b believe, however, that this is
a b
100%
rate
objects 2D trace-structure rather than its 3D geometry. Consis-
tent with this idea, the monkey infero-temporal cortex, which is
known to be involved in object recognition, has recently been
reported to contain view-tuned neurons, which respond to 3D
objects only when they are seen from a certain viewpoint9. An
alternative possibility, which we must consider, is that recogni-
tion might involve structure-from- motion processes. This idea
is based on the fact that a vivid perception of 3D structure can
arise from the 2D projection of a rotating object, in the absence
of stereoscopic depth cues. Structure-from-motion perception
can occur independent of object recognition, because even unfa-
miliar rotating objects give rise to a 3D percept. Our subjects
might therefore have derived a 3D structure from the moving
2D projection of the walking figure and matched this to an inter-
nal 3D representation. We
unlikely, because it has been10
100% demonstrated that recogni-
tion based on 2D cues pro-
rate

80% 80%
ceeds unhindered even when
alarmrate

alarmrate

60% 60% the 3D structure suggested by

False alarm

Falsealarm

n =n=11
11
structure-from-motion
40% 40%
processes is inconsistent with
False

False

nn=11
= 11
20% 20%
the object identity.
We interpret the results of
0% 0% the depth-plane experiment
0% 50% 100% 150% 200% 5 10 15 20 25 30 as pointing to the existence of
Noise level
Noise level Disparity (pixels)
Disparity (pixels)
a top-down influence capable
c of modulating the informa-
Fig. 4. Results of the depth-plane experiment aver- c tion provided by the early
aged across 11 subjects. The false alarm rate (a depth-perception processes
100%
response of in same plane when the dots are physi- based on binocular disparities.
cally not in the same plane) is plotted against the 80% There are other related exam-
maximum random depth-distortion allowed in the
ples, e.g. the hollow mask
rate

sequence in (a) and against the maximum disparity in

Hitrate

60%
nn=11
= 11 effect 11 and the cyclopean
pixels (each pixel subtends 0.015 degrees of visual 12 (in which an
Necker-cube
Hit

40%
angle) between the three highlighted dots in (b). For
comparison the hit rate (a correct response when ambiguous 2D image gives
the three dots are in the same plane) is shown in (c).
20%
rise to two alternating per-
Three conditions are plotted in each graph. 0%
cepts with different depth
__ Human figure with the 3 marked dots on the 0% 50% 100% 150% 200% structures in conflict with the
same limb; ._._._. Human figure with the 3 marked Noise
Noise level
Level disparity given by the stere-
dots on different limbs; .. Random figure. The ogram), which argue in favor
viewing position for all trials was 0 (the walker is seen walking to the right); this viewing position insured good of the influence of high-level
recognizability of the moving figure. The specific noise levels we used for the distorted walker in this experiment cues on depth perception.
were 0, 25, 50, 100, 150, and 200 %, which is around the noise level used in the first experiment. This influence can, in turn, be

256 nature neuroscience volume 1 no 3 july 1998

 1998 Nature America Inc. http://neurosci.nature.com
modulated by factors that change the recognizability of a stimu-
lus. This hypothesis would explain why the human sequences give
rise to false depth perception so much more frequently than do
the random sequences. Other important factors that need to be
considered in interpreting these results include grouping induced
by common motion or proximity. That is, dots may be more like-
ly to be perceived as coplanar if they move in synchrony or if they
are close together. We believe, however, that in our experiments
the effect of such factors would be relatively limited, because the
sequences in the different conditions had very similar mid-level
attributes such as motion and density distributions. Specifically
the differences in performance between the walker and the ran-
dom conditions suggest the importance of object-specific, recog-
nition-based influences over general configurational ones. The
motion trajectories of the individual dots were the same in both
conditions, and only their 2D offsets were randomized. Thus, two
dots of the walker that moved in phase continued to do so in the
random stimulus, thereby largely maintaining the mid-level group-
1998 Nature America Inc. http://neurosci.nature.com
ing cues and the articulation geometry in the two conditions. Yet,
the suppression of binocular disparity perception occurs only when
the walker is recognizable.
The idea that top-down influences can affect perception is
certainly not a new one. Several well known visual illusions,
such as the Dalmatian dog picture 13 or the mother-in-
law/daughter-in-law figure14 demonstrate the significance of
top-down expectations in interpreting ambiguous stimuli.
Recent computational models of the neocortex have argued that
feedback cortico-cortical projections might allow top-down
influences to propagate from the higher cortical areas that are
involved in object recognition back to the earlier areas that sup-
port lower-level processes15,16. Our study now provides evi-
dence that even the very low-level process of stereo-depth
perception, which was previously considered to be a purely bot-
tom-up process17,18, is in fact susceptible to top-down influ-
ences. Additionally, our experimental results provide indirect
evidence that dynamic three-dimensional objects might be rec-
ognized by the visual system based on their 2D traces rather
than on their 3D structural descriptions.
Methods
The biological motion sequences used in our experiments were based on
data collected at the Gait Analysis Laboratory of the Spaulding Rehabil-
itation Hospital in Boston, Massachusetts. The data comprised the 3D
positions of twelve points on a male human as he walked in place. The
point positions were updated 39 times over the course of one complete
nature neuroscience volume 1 no 3 july 1998
articles
walk cycle. The stimuli were generated by displaying each point as a bright
dot (0.03 degree of visual angle) on a gray background (mean luminance:
20 cd per m2). The experiments were conducted on a Silicon Graphics
Indigo 2 workstation. All sequences were presented in stereo using a pair
of StereoGraphics Crystal-Eyes (TM) LCD shutter glasses synchronized
with the display. Two views were generated for each frame to allow stereo-
scopic vision. All subjects were tested to ensure that they had functioning
stereoscopic ability, two subjects were rejected and all subjects were naive
as to the purpose of the experiments.
Acknowledgments
We wish to thank N. Logothetis, B. Tjan and D. Kersten for insightful comments
on earlier versions of the manuscript and P. Lipson for providing the biological
motion data set.
RECEIVED 29 JANUARY: ACCEPTED 22 MAY 1998
1. Johansson, G. Visual perception of biological motion and a model of its
analysis. Percept. Psychophys. 14, 201211 (1973).
2. Cutting, J. E. & Kozlowski, L. T. Recognition of friends by their walk. Bull.
Psychonom. Soc. 9, 353356 (1977).
3. Cutting, J. E. Coding theory adapted to gait perception. J. Exp. Psychol. Hum.
Percept. Perform. 7, 7187 (1981).
4. Hoffman D. D. & Flinchbaugh B. E. The interpretation of biological motion.
Biol. Cybern. 42, 195204 (1982).
5. Ullman S. The Interpretation of Visual Motion (MIT Press, Cambridge, 1979).
6. Shibata T., Sugihara, K., & Sugie, N. Recovering three-dimensional structure
and motion of jointed objects from orthographically projected optical flow.
Trans. IECE 68-D, 16891696 (1985).
7. Webb J. & Aggarwal, J. Structure from motion of rigid and jointed objects.
Artif. Intell. 19, 107131 (1982).
8. Oram, M. W. & Perrett, D. I. Responses of anterior superior temporal
polysensory (STPa) neurons to biological motion stimuli. J. Cog. Neurosci. 6,
99116 (1994).
9. Logothetis, N. K., Pauls, J. & Poggio, T. Shape recognition in the inferior
temporal cortex of monkeys. Curr. Biol. 5, 552563 (1995).
10. Sinha, P. & Poggio, T. Role of learning in three-dimensional form perception.
Nature 384, 460463 (1996).
11. Gregory, R. L. in Illusion in Nature and Art (eds Gregory, R. L. & Gombrich, E.
H.) 4996 (Duckworth, London, 1973).
12. Julesz, B. Foundations of Cyclopean Perception (Univ. of Chicago, Chicago,
1971).
13. Goldstein, B. Sensation and Perception (Brooks/Cole, Pacific Grove, 1996).
14. Boring, E. G. A new ambiguous figure. Am. J. Psychol. 42, 444445 (1930).
15. Ullman, S. Sequence seeking and counter streams: a computational model for
bidirectional information flow in the visual cortex. Cereb. Cortex 5, 111
(1995).
16. Mumford, D. On the computational architecture of the neocortex. II. The
role of cortico-cortical loops. Biol. Cybern. 66, 241252 (1992).
17. Julesz, B. Early vision and focal attention. Rev. Mod. Phys. 63, 735772
(1991).
18. Nakayama, K., Shimojo, S. & Silverman, G. H. Stereoscopic depth: Its relation
to image segmentation, grouping, and the recognition of occluded objects.
Perception 18, 5568 (1989).
257