Content uploaded by Bong June Yoon
Author content
All content in this area was uploaded by Bong June Yoon on Oct 16, 2014
Content may be subject to copyright.
ARTICLES
854 VOLUME 6
|
NUMBER 8
|
AUGUST 2003 NATURE NEUROSCIENCE
For decades it has been appreciated that monocular deprivation
during early postnatal life sets in motion a cascade of events in the
primary visual cortex that culminates in severe visual impairment
1
.
At the peak of the sensitive period, even 24 h of MD is sufficient to
cause neurons in visual cortex to lose responsiveness to stimulation
of the deprived eye
2
.This effect of MD is not merely a passive con-
sequence of reduced neural activity. Rather, the residual ‘noise’ in
the visually deprived retina actively stimulates the loss of synaptic
efficacy
3
.The synaptic modifications induced by MD reflect
processes that normally refine cortical receptive fields and visual
capabilities during postnatal development. A major challenge has
been to identify the mechanisms responsible for this type of visual
cortical plasticity.
Inspired originally by a theory of visual cortical plasticity
4
,research
over the past 15 years has also established that in visual cortex and the
CA1 region of hippocampus, long-term synaptic depression can be
induced by activation of postsynaptic NMDA receptors (NMDARs)
5
and a protein phosphatase cascade
6,7
.This synaptic activity leads to a
very specific pattern of altered AMPA receptor (AMPAR) phosphory-
lation, reduced cell-surface receptor expression, and depressed
AMPAR-mediated transmission
8,9
.Precisely how these mechanisms
contribute to experience-dependent synaptic modifications in vivo,
however, has remained uncertain.
One way to connect the effects of MD and the mechanisms of
LTD is to correlate deficits after genetic or pharmacological manip-
ulations. However, correlation does not establish causality, and the
absence of correlation can reflect degeneracy—the ability of ele-
ments that are structurally different to perform the same function
or yield the same output—that is ubiquitous in biological sys-
tems
10
. Showing that the effects of MD survive deletion of one
mechanism for LTD is obviously not grounds for rejecting the
hypothesis that this mechanism normally provides a substrate for
the effects of MD. An alternative approach is to ask the simple
question: Does MD induce LTD in visual cortex? The aim of the
current study was to exploit knowledge of the molecular mecha-
nisms of LTD to address this question, focusing on alterations in
AMPAR phosphorylation and surface expression.
The AMPAR phosphorylation sites that have been monitored dur-
ing LTD are serine residues 831 and 845 on the GluR1 subunit, and
Ser880 on the GluR2 subunit. Although Ser831 is selectively phos-
phorylated by calcium/calmodulin-dependent protein kinase II
(CaMKII) during long-term synaptic potentiation (LTP)
11
, it is unal-
tered during LTD
12,13
.In contrast, Ser845 and Ser880 are unaltered by
LT P, but are selectively and persistently dephosphorylated and phos-
phorylated, respectively, during LTD
12–14
.Dephosphorylation of
Ser845 can lead to synaptic depression, in part, by decreasing AMPAR
channel open probability
15
.However, changes in phosphorylation
also reflect a loss of AMPARs at the synapse. It has been established
that AMPARs are rapidly internalized in response to LTD-inducing
stimulation
16
and that internalized receptors are dephosphorylated at
GluR1 Ser845
17
.Moreover, phosphorylation of GluR2 at Ser880 alters
the relative affinity of the receptor for intracellular anchoring pro-
teins
18
, and current models suggest that this change either stimulates
internalization or causes internalized AMPARs to be retained inside
the neuron. Regardless of their specific roles, however, these unique
changes in AMPAR phosphorylation and the concomitant changes in
cell-surface expression constitute a molecular ‘fingerprint’ for the
occurrence of NMDAR-dependent homosynaptic LTD.
Here we report that 24 h of MD during the sensitive period pre-
cisely mimics LTD with respect to AMPAR phosphorylation, and also
significantly reduces the expression of AMPARs on the surface of
visual cortical neurons. Moreover, the changes induced by MD
occlude the subsequent expression of LTD at synapses ex vivo.
Therefore, we conclude that MD induces LTD in visual cortex. A sub-
set of these results was presented at the 2002 Society for Neuroscience
meeting, and published in abstract form.
1
Howard Hughes Medical Institute, The Picower Center for Learning and Memory, and Department of Brain and Cognitive Sciences, Massachusetts Institute of
Technology, Cambridge, Massachusetts 02139, USA.
2
Howard Hughes Medical Institute and Department of Neuroscience, Johns Hopkins University Medical School,
Baltimore, Maryland 21205, USA.
3
These authors contributed equally to this work. Correspondence should be addressed to M.F.B. (mbear@mit.edu).
Molecular mechanism for loss of visual cortical
responsiveness following brief monocular deprivation
Arnold J Heynen
1,3
,Bong-June Yoon
1,3
,Cheng-Hang Liu
1
,Hee J Chung
2
,Richard L Huganir
2
& Mark F Bear
1
A dramatic form of experience-dependent synaptic plasticity is revealed in visual cortex when one eye is temporarily deprived of
vision during early postnatal life. Monocular deprivation (MD) alters synaptic transmission such that cortical neurons cease to
respond to stimulation of the deprived eye, but how this occurs is poorly understood. Here we show in rat visual cortex that brief
MD sets in motion the same molecular and functional changes as the experimental model of homosynaptic long-term depression
(LTD), and that prior synaptic depression by MD occludes subsequent induction of LTD. The mechanisms of LTD, about which
there is now a detailed understanding, therefore contribute to visual cortical plasticity.
© 2003 Nature Publishing Group http://www.nature.com/natureneuroscience
ARTICLES
NATURE NEUROSCIENCE VOLUME 6
|
NUMBER 8
|
AUGUST 2003 855
RESULTS
LTD in vivo alters AMPAR phosphorylation and visual responses
NMDAR-dependent homosynaptic LTD was first described in hip-
pocampal area CA1, where it can be induced in vitro
5
and in vivo
19
by
prolonged presynaptic stimulation at 1 Hz. A biochemical require-
ment for LTD expression at CA1 synapses is dephosphorylation of
postsynaptic protein kinase A (PKA) substrates
20
.Inhibition of post-
synaptic PKA both mimics and occludes LTD, demonstrating that
dephosphorylation of postsynaptic PKA substrates initially mediates
LTD expression
21
. One PKA substrate is GluR1 Ser845. This site is
highly phosphorylated under basal conditions, and undergoes rapid,
specific and persistent dephosphorylation in response to LTD-induc-
ing stimulation in the hippocampus in vitro
12,13
.Thus, dephosphory-
lation of Ser845 of GluR1 serves as a molecular marker for the
occurrence of NMDAR-dependent LTD by reporting the status of
postsynaptic PKA-dependent phosphorylation.
We felt it was important to confirm that LTD has the same bio-
chemical signature in visual cortex in vivo.To this end, we estab-
lished a preparation in which LTD could be elicited by 1-Hz
stimulation of the dorsal lateral geniculate nucleus (LGN) of anes-
thetized rats at postnatal day (P) 21–25, an age when monocular
deprivation also causes synaptic depression
22
.Concentric bipolar
stimulating electrodes were placed bilaterally in the LGN, and
recording electrodes were inserted into the visual cortex
(Brodmann’s area 17, A17) of each hemisphere at a depth yielding
the maximum negative-going field potential in response to LGN
stimulation, as described previously for adult rats
23
.
After collecting stable baseline responses, the LGN of one hemi-
sphere received 1-Hz stimulation; the other hemisphere served as a
within-animal control. We found that reliable LTD in this prepara-
tion required two 30-min episodes of 1-Hz stimulation (Fig. 1a).
One hour later, the animals were killed with an overdose of anes-
thetic, and the brains were removed. The visual cortex of each hemi-
sphere was then probed for changes in GluR1 phosphorylation, using
quantitative immunoblot analysis as previously described
12,13
.
Important features of this and all other biochemical assays in this
study were (i) that the experimenters performing the analyses were
always blind to the stimulation history of the tissue, and (ii) that
there was always a yoked, within-animal control sample run at the
same time under identical conditions, allowing for pair-wise statisti-
cal comparisons. We found that 1-Hz stimulation of the LGN
Figure 1 LTD in vivo is accompanied by a
dephosphorylation of GluR1 at Ser845 and a
decrease in visually evoked potentials.
(a) Application of LFS to the LGN elicited LTD of
field potentials in A17 in vivo. After a 30-min
baseline recording period, application of two
episodes of LFS (1,800 pulses, 1 Hz; arrows
LFS1 and LFS2) to the LGN of one hemisphere
produced LTD of the negative-going component
of LGN-evoked field potentials (FP) recorded in
A17 (n = 10, 77 ± 8% of baseline; P < 0.01,
paired t-test). No change in FP amplitude was
observed in contralateral control A17 that
received only baseline stimulation (101 ± 10% of
baseline; P > 0.7). Traces here and in c are taken
from one representative animal in each group
and are averages of ten consecutive FPs obtained
at times indicated by numerals. Scale bars (a,c):
0.5 mV, 10 ms. (b) Immunoblots were made of
membrane fractions from A17 ipsilateral and
contralateral (Con) to the LGN receiving LFS
using antibodies against phosphorylated Ser845
(S845) or Ser831 (S831) on GluR1. Blots were
stripped and reprobed with an antibody
recognizing the C-terminal end of GluR1.
Extraneous bands on the S845 immunoblot
are not related to GluR1
50
. The bar graph
summarizes the immunoblot analysis: LFS-
induced LTD was accompanied by a significant
decrease in Ser845 phosphorylation in A17.
Ser-845 phosphorylation in A17 undergoing LTD
was 87 ± 4% of contralateral control A17
(n = 10; P < 0.05 versus 100 ± 9%), whereas no
change in Ser831 phosphorylation was observed
(P > 0.6, 97 ± 11% versus 100 ± 15%).
(c,d) LTD induction in vivo depresses VEP
amplitude. (c) LGN-evoked FPs are reduced in
A17 ipsilateral to application of LFS to the LGN
(n = 5, closed circles, 61 ± 18% of baseline;
P < 0.01), whereas no change in FP amplitude is
observed in contralateral control A17 (open circles; 95 ± 16% of baseline; P > 0.6). Shaded areas indicate periods when VEPs in response to grating
stimuli were collected prior to (VEP 1) and following (VEP 2) induction of LTD. (d) Summary of VEP amplitude data for animals undergoing LTD (same
animals as in c). In A17 in which LTD was induced, binocular VEPs were significantly depressed relative to contralateral control A17. Representative VEPs
were obtained from one animal before (VEP1) and following (VEP2) induction of LTD.
© 2003 Nature Publishing Group http://www.nature.com/natureneuroscience
ARTICLES
856 VOLUME 6
|
NUMBER 8
|
AUGUST 2003 NATURE NEUROSCIENCE
produces a modest but significant dephosphorylation of Ser845
(normalized to total GluR1) in the ipsilateral A17 compared to con-
tralateral control A17 (Fig. 1b). In contrast, there was no significant
change in phosphorylation of Ser831. Neither LTD nor dephospho-
rylation of Ser845 occurred after 1-Hz stimulation in animals treated
with an NMDAR antagonist (data not shown).
The finding of modest but selective dephosphorylation of Ser845
in A17 in vivo mimics what has previously been reported for hip-
pocampal slices following induction of LTD. To understand the
meaning of a change of this magnitude, it is helpful to consider two
points. First, bath application of NMDA to hippocampal slices,
which induces LTD in the entire population (100%) of modifiable
synapses, produces an approximately 50% dephosphorylation of
GluR1 Ser845. Second, LTD induced in the same slice preparation by
electrical stimulation of the Schaffer collaterals with a microelec-
trode produces a phosphorylation change of about 10%
12,13
.
Therefore, it can be estimated that the electrical stimulation induces
LTD in ∼20% of the modifiable synapses in the hippocampal slice.
Applying the same logic to A17, our data suggest that approximately
25% of the modifiable synapses may have undergone LTD as a conse-
quence of 1-Hz stimulation of the LGN.
To gain some insight into the functional consequences of inducing
LTD in the geniculocortical pathway, we performed an additional
series of experiments in which field potentials evoked by high-contrast
visual grating stimuli were monitored before and after conditioning
stimulation of the LGN. This study revealed that the amplitude of the
maximum negative-going visually evoked potential (VEP), which
reflects a synaptic current sink in deep layer 3 of visual cortex
23,24
,is
also significantly depressed by induction of LTD (Fig. 1c,d).
Brief MD depresses VEPs contralateral to the deprived eye
Because of the significant decussation of the retinofugal projection
in rodents, even the binocular regions of rat visual cortex are domi-
nated by input from the contralateral eye
25
(Fig. 2a). Thus, in prin-
ciple, MD should depress transmission in a substantially larger
fraction of synapses in A17 contralateral to the deprived eye as com-
pared with the ipsilateral cortex, which would enable within-animal
comparisons of AMPAR changes associated with naturally occur-
ring synaptic modifications.
To investigate the possibility of an interhemispheric asymmetry in
the effects of MD in rats, we recorded VEPs from the surface of pri-
mary visual cortex in precisely matched locations in the two hemi-
spheres. Recording positions were chosen to maximize the response
from the ipsilateral eye, thus ensuring that we were sampling from the
binocular region of A17. In control urethane-anesthetized rats
(P21–25), we confirmed that the responses evoked by full-field pat-
tern-reversing sinusoidal gratings are symmetrical in the two hemi-
spheres when both eyes are open (Fig. 2b). When each eye was tested
individually, the VEP through the contralateral eye was dominant, as
expected (contralateral/ipsilateral (C/I) VEP ratio, 2.4 ± 0.3; n = 5).
Because of our interest in studying the early molecular events
responsible for rapid ocular dominance (OD) plasticity, we chose to
investigate the minimal period of MD that produces reliable
changes in A17. We found that 24 h of MD in P21 rats was sufficient
to produce a substantial OD shift (C/I VEP ratio contralateral to the
deprived eye = 0.63 ± 0.04, n =6;P < 0.01 compared to non-
deprived controls, Mann-Whitney U-test). More importantly, we
found that this period of MD was sufficient to depress responses to
binocular stimulation in the hemisphere contralateral to the
deprived eye as compared with the ipsilateral A17 (Fig. 2c).
Although the binocular VEP ipsilateral to the deprived eye was com-
parable in magnitude to the VEP evoked in control animals, the
response contralateral to the deprived eye was significantly reduced.
Thus, 24 h of MD is sufficient to produce a net depression of synap-
tic transmission in A17 contralateral to the deprived eye. These data
are also consistent with the conclusion that depression of deprived-
eye responses occurs very rapidly following MD and precedes
increases in open-eye responses
2,26
.
Brief MD alters AMPAR phosphorylation in visual cortex
We took advantage of this interhemispheric asymmetry in the effect
of MD to test the hypothesis that deprivation-induced synaptic
depression, like LTD, is associated with specific changes in AMPAR
phosphorylation. After 24 h of MD, the visual cortex of both hemi-
Figure 2 Brief MD leads to depression of synaptic transmission in A17
contralateral to the deprived eye (DE). (a) Schematic diagram of rat visual
pathways illustrating the predominance of contralateral retinal input to
A17. (b) Representative visually evoked potentials (VEPs) recorded
simultaneously from recording electrodes on the dural surfaces of left and
right A17 from control and MD animals (P21–25). Traces are averages of
100 consecutive stimulus presentations (scale bar: 50 µV, 100 ms).
(c) Summary of VEP amplitude data for control and MD animals (n =7,
respectively). In control animals, there was no difference in VEP amplitude
between the hemispheres (left A17, 165 ± 34 µV; right A17, 164 ± 37 µV;
P > 0.60, Wilcoxon signed-ranks test). In rats that had been monocularly
deprived for 24 h before recording, binocular VEPs contralateral to the
deprived eye were significantly depressed (106 ± 17 µV, ∗) relative to
control (198 ± 36 µV; P < 0.02, Wilcoxon signed-ranks test).
© 2003 Nature Publishing Group http://www.nature.com/natureneuroscience
ARTICLES
NATURE NEUROSCIENCE VOLUME 6
|
NUMBER 8
|
AUGUST 2003 857
spheres was removed and prepared for quantitative immunoblot
analysis. Using phosphorylation site-specific antibodies, the phos-
phorylation state of GluR1 and GluR2 in the cortex contralateral to
the deprived eye was compared with the ipsilateral cortex.
We found that MD reliably induces dephosphorylation of Ser845
(normalized to total GluR1) in the visual cortex contralateral to the
deprived eye (Fig. 3a,b). The magnitude of this change (10–12%) is
comparable to that observed after induction of LTD in A17 and hip-
pocampus with 1-Hz synaptic stimulation, suggesting that approxi-
mately 20% of the modifiable synapses may undergo LTD as a
consequence of 24 h of MD. The observed decrease in VEP amplitude
could be fully accounted for by this change, especially if the affected
synapses are concentrated in the superficial layers (1–4), as is sug-
gested by additional experiments presented below.
If the change in AMPAR phosphorylation contributes to the
synaptic depression following MD, we expect this response to be
absent under conditions where MD is without effect. Therefore, we
investigated whether similar changes occur when MD is initiated in
mature animals aged beyond the critical
period for OD plasticity. Unlike the effect
observed in young animals, MD did not
induce any changes in Ser845 phosphoryla-
tion in P90 rats (Fig. 3a,b). Another condi-
tion in which MD does not trigger OD
plasticity is when cortical NMDARs are
blocked
27
.Therefore, we treated P21 ani-
mals with the NMDAR antagonist CPP
(5.0 mg/kg) or with saline at 5-h intervals
during the period of MD. Although saline-
injected animals still showed significant
dephosphorylation of Ser845 contralateral
to the deprived eye after 24 h of MD, CPP-
injected animals did not (Fig. 3c,d). Thus,
the deprivation-induced dephosphoryla-
tion of GluR1, like the deprivation-induced
synaptic depression, is restricted to a criti-
cal period of postnatal development and is
dependent on activation of NMDARs.
As mentioned above, not all phosphoryla-
tion sites on AMPARs respond the same way
during LTD. We therefore used antibodies
against GluR1 phospho-Ser831 and GluR2
phospho-Ser880 to examine the selectivity of
the change in AMPAR phosphorylation in
visual cortex after 1 d of MD. This analysis
showed a variable but statistically significant
increase in Ser880 phosphorylation, and no
change in Ser831 phosphorylation, following
MD (Fig. 3e,f), similar to what occurs follow-
ing induction of LTD.
Because immunoblot analyses of all three
phosphorylation sites were typically per-
formed on tissue from the same animals, the
entire A17 of each hemisphere was used to
obtain enough protein. However, this proce-
dure left us vulnerable to the criticism that
the true modifications in AMPARs might
actually occur ipsilateral to the deprived eye,
where they would be the opposite of what
occurs during LTD. Although we considered
this possibility to be exceedingly unlikely
based on the interhemispheric asymmetry in VEPs after MD, we
were nevertheless compelled to rule it out. The ideal within-animal
control tissue is the monocular segment of visual cortex ipsilateral
to the deprived eye, because it is still area 17, but MD has no effect
on activity in this region of cortex (Fig. 2a). Therefore, we repeated
the experiment, performing an analysis of GluR1 Ser845 phospho-
rylation in monocular and binocular segments, ipsilateral and con-
tralateral to the deprived eye.
This analysis confirmed that there is no significant change in
Ser845 phosphorylation in the binocular segment of A17 ipsilateral
to the deprived eye and that Ser845 is significantly dephosphory-
lated in the binocular segment contralateral to the deprived eye
(Fig. 4a,b). However, we also observed dephosphorylation of Ser845
in the monocular segment contralateral to the deprived eye. This
finding suggests that brief deprivation induces synaptic depression
in both monocular and binocular segments contralateral to the
deprived eye. To address this possibility, we repeated the analysis of
VEPs and discovered that significant deprived-eye response depres-
Figure 3 Brief MD alters AMPAR phosphorylation in visual cortex. (a) Representative immunoblots of
membrane (Memb) and synaptoneurosome (Synapto) fractions from A17 contralateral (C) and
ipsilateral (I) to the deprived eye using an antibody against phosphorylated Ser845 (S845; PKA site) on
GluR1. Immunoblots were stripped and reprobed with an antibody recognizing the C-terminal end of
GluR1. Extraneous bands on the S845 immunoblot are other PKA substrates not related to GluR1
50
.
(b) Summary of immunoblot analysis showing that 24 h of MD produced a significant decrease in
Ser845 phosphorylation in A17 contralateral to the deprived eye in juvenile (P21) but not adult (P90)
animals. Asterisk in this and subsequent figures indicates significant (P < 0.05) decrease/increase as
compared to A17 ipsilateral to the deprived eye (paired t-tests). Ser845 phosphorylation contralateral
to the deprived eye was 88 ± 3% (n = 18; P < 0.02 versus 100 ± 15) and 90 ± 5% (n = 18; P < 0.03
versus 100 ± 7) of the ipsilateral A17 for Memb and Synapto preparations, respectively.
(c) Representative immunoblots of membrane fractions from A17 of saline- and CPP-treated animals.
(d) Summary of immunoblot analysis showing absence of an MD-associated decrease in Ser845
phosphorylation (107 ± 6% of ipsilateral (100 ± 7%) A17; n = 8) in animals treated during the time of
deprivation with the NMDA receptor antagonist CPP. Saline-injected animals again showed a significant
decrease (91 ± 3% of ipsilateral (100 ± 5%) A17; n = 8; P < 0.05). (e) Representative immunoblots of
membrane fractions using antibodies against phosphorylated Ser880 on GluR2 and phosphorylated
Ser831 on GluR1. Immunoblots were stripped and reprobed with an antibody recognizing the
N terminus of GluR2 and C terminus of GluR1. Extraneous bands on the S831 immunoblot are not
related to GluR1
50
and do not show consistent differences following MD. (f) Summary of immunoblot
analysis: 24 h of MD produced a significant increase in Ser880 phosphorylation (132 ± 11% of
ipsilateral (100 ± 7%) A17; n = 30; P < 0.02), whereas no change in Ser831 phosphorylation was
observed (106 ± 4% of ipsilateral (100 ± 9%) A17; n = 19).
© 2003 Nature Publishing Group http://www.nature.com/natureneuroscience
ARTICLES
858 VOLUME 6
|
NUMBER 8
|
AUGUST 2003 NATURE NEUROSCIENCE
Homosynaptic LTD, by definition, requires presynaptic activity
(i.e., the release of glutamate) for induction. Presynaptic activity is
also known to be important for deprivation-induced synaptic depres-
sion in A17, as the consequences of brief monocular inactivation with
tetrodotoxin (TTX) are much less severe than those of MD by lid clo-
sure
3
.We were therefore interested to compare the consequences of
MD and monocular inactivation on AMPAR surface expression in rat
visual cortex. Under anesthesia, animals received unilateral vitreal
injections of saline (2 µl; n = 7) or TTX (0.4 mM, 2 µl; n = 7) and the
lid over this eye was sutured closed. Twenty-four hours later, slices of
A17 were prepared as described above. These experiments replicated
the finding that MD induces a significant ∼20% reduction in AMPAR
surface expression (saline; Fig. 5d). However, in striking contrast,
depriving the contralateral cortex of all retinally driven activity had
no effect (TTX; Fig. 5e). Therefore, the decrease in AMPAR surface
expression caused by MD is not merely a passive consequence of
reduced input activity; rather it is caused by the uncorrelated noise
that drives deprivation-induced synaptic depression.
Changes induced by brief MD occlude synaptic expression of LTD
Altered AMPAR phosphorylation and/or surface expression could
contribute substantially to the reduced responsiveness of visual corti-
cal neurons to deprived eye stimulation following MD, if these
changes actually occur at synapses. The fact that phosphorylation
changes are observed in synaptoneurosomes (Fig. 3a,b) provides
some support for the conclusion that synaptic AMPARs are modified.
Another way to address this question is to exploit the fact that synap-
tic expression of NMDAR-dependent LTD uses the same, saturable
modifications in AMPARs. Thus, if MD partially exhausts the same
expression mechanism as LTD, then we predict that LTD magnitude
at saturation would be reduced in visual cortex that has already
undergone deprivation-induced synaptic depression.
sion also occurs in the monocular segment (Fig. 4c,d). This finding
replicates with brief MD what has been observed previously in rats
after longer periods of deprivation
25
.
Brief MD alters AMPAR surface expression in visual cortex
There is considerable evidence that an expression mechanism for
NMDAR-dependent LTD is the internalization of synaptic AMPARs,
and internalized receptors show changes in phosphorylation that are
similar to those observed following MD. We therefore hypothesized
that MD triggers the loss of surface-expressed AMPARs in visual cor-
tex. To test this idea, young rats were monocularly deprived for 1 d,
and surface receptors were labeled with biotin in slices of visual cortex
prepared from the hemispheres ipsilateral and contralateral to the
deprived eye. To enrich for modified synapses in the superficial layers,
the slices were bisected and the deep layers (5 and 6) were discarded.
Biotinylated receptors were then precipitated and the ratio of surface
to total AMPAR protein was determined by quantitative western blot-
ting in each hemisphere. Control experiments showed that only sur-
face proteins are biotinylated in the slice preparation (Fig. 5a).
This biochemical analysis (again performed ‘blind’) showed,
remarkably, that the number of AMPARs on the surface of visual corti-
cal neurons is significantly reduced, by approximately 20%, in A17
contralateral to the deprived eye in P21 rats (Fig. 5b). The magnitude
of this change is comparable to that observed following induction of
LTD in hippocampus in vivo
28
.Both the GluR1 and GluR2 subunits
showed similar reductions (80.7 ± 6.4% and 81.5 ±4.1% of ipsilateral
controls; n =7 rats), but the same analysis of GABA
A
receptor surface-
expression revealed no significant difference. MD in adult (P90) rats
failed to alter AMPAR surface expression (97.7 ± 5.8% and
93.1 ± 6.2% of ipsilateral controls for GluR1 and GluR2, respectively;
n =7 rats), showing that this response to MD, like synaptic depression,
is restricted to a sensitive period of postnatal development (Fig. 5c).
Figure 4 Brief MD produces synaptic depression
and dephosphorylation of GluR1 at Ser845 in
monocular and binocular segments of A17
contralateral to the deprived eye. (a) Dorsal
view of a rat brain showing the monocular and
binocular segments of A17 dissected for
biochemical analysis. (b) Summary of
immunoblot analysis showing that 24 h of
MD produced a significant decrease in Ser845
phosphorylation in monocular and binocular
regions of A17 contralateral to the deprived eye
(Contra). Data are expressed as a percentage of
monocular segment ipsilateral to the deprived
eye (Ipsi), in which visual experience is
unaffected by MD. Asterisks indicate significant
decrease as compared to corresponding region of
A17 ipsilateral to the deprived eye (paired
t-test). Ser845 phosphorylation in A17
contralateral to the deprived eye was 89 ± 3%
(n = 28; P < 0.001 versus 100 ± 11%) and
92 ± 4% (n = 28; P < 0.002 versus 100 ± 11%)
of the ipsilateral A17 for monocular and
binocular segments, respectively. Note that
Ser845 phosphorylation in the monocular region
ipsilateral to the deprived eye of MD animals did not differ from monocular segment of nondeprived control animals (105 ± 9%; P > 0.6, unpaired t-test).
(c) Representative binocular VEPs recorded from electrodes on the dural surfaces of the monocular and binocular segments of A17 in the hemispheres
ipsilateral and contralateral to the deprived eye of one animal. Traces are averages of 100 consecutive stimulus presentations (scale bars: monocular A17,
25 µV, 100 ms; binocular A17, 50 µV, 100 ms). (d) Summary of binocular VEP amplitude data for monocular and binocular segments of A17 (expressed
as a percentage of values ipsilateral to the deprived eye, n = 8). In rats that had been monocularly deprived for 24 h before recording, VEPs contralateral to
the deprived eye were significantly depressed in both monocular and binocular segments relative to the corresponding regions of A17 ipsilateral to the
deprived eye (monocular A17: 29 ± 7 versus 46 ± 11 µV, P < 0.02; binocular A17: 67 ± 12 versus 191 ± 54 µV, P < 0.02; Wilcoxon signed-ranks tests).
© 2003 Nature Publishing Group http://www.nature.com/natureneuroscience
ARTICLES
NATURE NEUROSCIENCE VOLUME 6
|
NUMBER 8
|
AUGUST 2003 859
We tested this prediction by preparing slices of A17 after 1 d of MD
and comparing LTD ipsilateral and contralateral to the deprived eye.
Synaptic field potentials were recorded from layer 3 in response to
layer-4 stimulation, as described previously
7
.Three episodes of LFS
(1 Hz, 900 pulses) spaced at 25-min intervals were used to saturate
LT D. There was a striking difference in LTD magnitude at saturation
in the two hemispheres (Fig. 6). Consistent with our prediction, LTD
was diminished in the cortex contralateral to the deprived eye (ipsi-
lateral A17, 66.4 ± 5.0% of baseline; contralateral A17, 81.4 ± 4.0% of
baseline). Repeating the experiment in non-deprived animals con-
firmed that there is normally no interhemispheric asymmetry in
LT D, and further, that the magnitude of LTD ipsilateral to the
deprived eye is not different from controls (left A17, 68.8 ± 10.3% of
baseline; right A17, 67.3 ± 7.4% of baseline). Together, the data
strongly support the conclusion that prior MD, by triggering inter-
nalization of AMPARs, partially occluded LTD in A17 contralateral
to the deprived eye. We note that this effect of MD on LTD is
observed in layer 3, supporting the notion that the superficial layers
are a site for early OD plasticity
29,30
.
Time course of MD-induced changes in AMPAR phosphorylation
In a final series of experiments, we tracked the time course of
changes in AMPAR phosphorylation following MD. The trends of
decreased GluR1 phosphorylation at Ser845 and increased GluR2
phosphorylation at Ser880 were already apparent after 6 h of MD,
but disappeared after 2 d of MD (Fig. 7a,b). This transient change in
receptor phosphorylation following MD was expected, as receptors
dephosphorylated at GluR1 Ser845 are degraded
17
.It was also
expected that the markers for net LTD in A17 contralateral to the
deprived eye would disappear with time as the ipsilateral inputs grow
in strength. Nonetheless, at a minimum the data suggest that the
mechanisms of LTD contribute to the rapid loss of visual responsive-
ness that occurs during the first 2 d of MD.
With long periods of MD, there are well-documented changes in
axonal arbors in visual cortex
1
,but these changes may lack the speed
necessary to account for rapid OD plasticity
31
.Therefore, the mecha-
nisms of LTD can bridge the early functional consequences of MD
with the later anatomical modifications. However, we further suggest
that the mechanisms of LTD—particularly the loss of postsynaptic
glutamate receptors—actually serve to trigger the anatomical loss of
visually deprived connections
32
.This hypothesis is supported by
work performed on synapse elimination at the neuromuscular junc-
tion, where it has been shown that the loss of receptors postsynap-
tically precedes and triggers the withdrawal of the presynaptic axon
33
.
DISCUSSION
There are well established criteria that must be met to conclude that
two manipulations induce plasticity by the same mechanism
21
.The
first is mimicry. Here we show that the effects of MD in the contralat-
eral visual cortex (relative to the ipsilateral cortex) mimic LTD with
respect to (i) depressed synaptic transmission, (ii) decreased phospho-
rylation of GluR1 Ser845, (iii) increased phosphorylation of GluR2
Figure 5 Brief MD is accompanied by a loss of
surface-expressed AMPARs in juvenile animals.
(a) Representative immunoblots of the
biotinylated surface protein of A17 from P21 MD
animals. Blots were probed with an antibody,
stripped and reprobed with additional antibodies.
Immunoblot with anti β-tubulin antibody shows
that biotin specifically labeled cell surface
proteins (Ipsi, ipsilateral hemisphere to the
deprived eye; Contra, contralateral hemisphere to
the deprived eye; T, total cell homogenate;
S, biotinylated surface protein). (b) Summary of
surface protein biotinylation assays of P21 MD
animals (n = 7). Surface levels of both GluR1
(80.7 ± 6.4%, P < 0.05, t-test) and GluR2
(81.5 ± 4.1%, P < 0.01, t-test) were
significantly lower in the visual cortex
contralateral to the deprived eye. However, there
was no significant difference in the surface level
of GABA
A
receptor α1 subunit. (c) Summary of
surface protein biotinylation assays of P90 MD
animals (n = 7). In contrast to juvenile animals,
no significant change in surface levels of both
GluR1 (97.7 ± 5.8%, P > 0.90, t-test) and
GluR2 (93.1 ± 6.2%, P > 0.30, t-test) were
observed in the visual cortex contralateral to the
deprived eye of adult animals. Similarly, no
significant difference in the surface level of
GABA
A
receptor α1 subunit was observed.
(d) Summary of surface protein biotinylation
assays of animals in which the deprived eye was
injected with saline prior to lid closure (n = 7).
Both GluR1 (86.6 ± 4.1%, P < 0.035, t-test)
and GluR2 (81.8 ± 5.3%, P < 0.019, t-test)
showed the similar loss of surface level as shown
in c. (e) Summary of surface protein biotinylation
assays of animals in which the deprived eye was injected with TTX prior to lid closure (n = 7). TTX injection prevented the loss of surface protein of both
GluR1 (98.4 ± 4.8%) and GluR2 (99.8 ± 8.4%) normally caused by MD from lid closure.
© 2003 Nature Publishing Group http://www.nature.com/natureneuroscience
synaptic depression induced by MD in vivo reduces the level of synap-
tic depression that can be attained by the mechanism of LTD in vitro.
We therefore conclude that MD induces LTD in visual cortex. We has-
ten to add that we are not suggesting that LTD is the only mechanism
for initiating OD plasticity. For example, there is evidence for
NMDAR-independent mechanisms of response depression
27
, and
responses driven by the non-deprived eye eventually potentiate
2,26
.
However, a possible contribution of other mechanisms does not
diminish the significance of the present findings. There is now a very
detailed understanding of the molecular basis for NMDAR-dependent
LT D. We can therefore reconstruct, for the first time, a molecular chain
of events that is set in motion by MD in visual cortex.
The data support a model in which the activity in the deprived
retina, relayed to the visual cortex by the lateral geniculate nucleus,
weakly activates postsynaptic NMDARs. The activation of NMDARs
is ‘weak’ because it rarely correlates with responses evoked by visual
stimulation of the open eye.Activated NMDARs admit Ca
2+
ions into
the postsynaptic neuron that, in turn, regulate a network of protein
phosphatases and kinases. Among the consequences of the modest
rise in intracellular calcium is dephosphorylation of postsynaptic
PKA substrates, including Ser845 of the AMPAR GluR1 subunit, and
the phosphorylation of Ser880 on the GluR2 subunit. These changes
in phosphorylation alter the binding of intracellular scaffolding pro-
teins, leading to net endocytosis, retention, and eventual degradation
of synaptic glutamate receptors. Consequently, the deprived eye no
longer effectively drives synaptic excitation in the visual cortex.
Ser880, (iv) unaltered phosphorylation of GluR1 Ser831, (v) decreased
surface expression of GluR1, (vi) decreased surface expression of
GluR2, (vii) unaltered surface expression of GABA
A
receptor protein,
(viii) sensitivity to NMDAR blockade and (ix) dependence on presy-
naptic activity. The second criterion is occlusion. Here we show that
ARTICLES
860 VOLUME 6
|
NUMBER 8
|
AUGUST 2003 NATURE NEUROSCIENCE
Figure 7 Time course of changes in AMPAR phosphorylation following
monocular deprivation. Changes in phosphorylation at Ser845 of GluR1 (a)
and Ser880 (b) of GluR2 were apparent after 6 h of MD, but disappeared
after two days of MD. No significant change in phosphorylation of Ser831
on GluR1 was observed at any time point (c).
Figure 6 Prior monocular deprivation occludes LFS-induced decreases in synaptic strength. (a) Slices of A17 were prepared following 24 h of MD in vivo.
Layer-3 field potentials were recorded in response to layer-4 stimulation, and LTD was induced with multiple episodes of LFS. (b) Summary of the effects
of MD on the saturation level of LTD in A17 contralateral (n =7 slices from 6 rats) and ipsilateral (n =7 slices from 5 rats) to the deprived eye. For
statistical analysis, the magnitudes of the field potentials obtained during the last 5 min of the baseline recording period and the final 5 min following
each LFS period were compared. A two-way repeated measures ANOVA showed a significant difference between hemispheres (F
1,36
= 3.29; P < 0.03) of
MD animals. Subsequent Student-Newman-Keuls post hoc tests showed that the magnitude of LTD contralateral to the deprived eye was saturated after
the first LFS episode, as no significant increase in LTD magnitude was observed after the second and third LFS periods. In contrast, in the ipsilateral A17,
a significant increase in LTD magnitude was still observed following the third LFS episode (P < 0.05). (c) Representative layer 2/3 field potentials obtained
from A17 contralateral (Contra A17) and ipsilateral (Ipsi A17) to the eye of deprivation. Traces are averages of ten consecutive field potentials obtained at
times indicated by numerals in a. (d) Summary of LTD saturation levels in A17 of the right (n =4 slices from 4 rats) and left (n =4 slices from 4 rats)
hemispheres of control animals. No significant difference in the magnitude of LTD was observed between left and right A17 (ANOVA; P > 0.9).
(e) Representative layer 2/3 field potentials obtained from A17 of the right and left hemispheres of control animals.
© 2003 Nature Publishing Group http://www.nature.com/natureneuroscience
ARTICLES
NATURE NEUROSCIENCE VOLUME 6
|
NUMBER 8
|
AUGUST 2003 861
The biochemical and functional changes caused by brief MD are
homosynaptic—they are driven by activity generated in the
deprived retina—and are expressed initially in both the monocular
and binocular segments of visual cortex. It should be noted that
these findings are entirely compatible with the notion of ‘binocular
competition’
34
if it is assumed that the activity of the non-deprived
eye influences the potential for subsequent recovery of deprived-eye
responses via regulation of the LTD/LTP modification thresh-
old
35–37
,for which there is now ample evidence
38–40
.Regardless,
even under conditions where cortical neurons are not under the
influence of a competing input (e.g.,binocular lid closure), it is well
established that deprivation induces a rapid and profound visual
impairment in both cats and rodents
2,41,42
.
The homosynaptic model for deprivation-induced synaptic
depression provides a new framework to address several fundamen-
tal unanswered questions about experience-dependent cortical
development; for example, how OD plasticity is regulated by age. We
show here that MD is no longer sufficient to dephosphorylate or
internalize AMPARs in adults. One possibility is that the expression
of key signaling molecules is altered during development, rendering
synapses incapable of supporting LTD
43
.A second possibility is that
the retinal activity required to drive the NMDAR-dependent process
of LTD fails to reach the modifiable synapses in A17 as a consequence
of the late maturation of cortical inhibition
44
.We note in this context
that the developmental decline in OD plasticity is accelerated in
genetically modified mice displaying a precocious maturation of
inhibition
45
.The model also accounts for the seemingly contradic-
tory finding that global reductions in cortical inhibition can reduce
OD plasticity
46,47
.Under conditions of heightened cortical excitabil-
ity, activity from the deprived eye is more likely to correlate with
evoked postsynaptic responses, and thus fails to satisfy the condi-
tions necessary to trigger LTD (i.e.,‘weak’ NMDAR activation). The
model also suggests novel explanations for how OD plasticity is
modulated by behavioral state (LTD is powerfully modulated by
acetylcholine and norepinephrine
48
).
Studies of LTP and LTD have revealed much about the mechanisms
of bidirectional synaptic plasticity; yet the questions of how, when,
and where these mechanisms contribute to brain function remain
largely unanswered.We have now shown that brief MD sets in motion
the same molecular and electrophysiological changes as LTD, and that
prior synaptic depression by MD occludes subsequent induction of
LTD in visual cortex. Thus, the mechanisms of LTD contribute to the
rapid loss of visual responsiveness after brief MD. LTD also appears to
participate in the loss of responses in somatosensory cortex following
whisker deprivation
49
.It therefore seems likely that these same mech-
anisms also contribute to the processes that refine cortical connec-
tions during early postnatal development.
METHODS
Animals. Juvenile (P21–25) and adult (P90 or older) Long Evans black-
hooded rats (Charles River Labs) were used. Animals were group-housed, with
food and water available ad libitum and maintained on a 12:12 h light:dark
cycle. All protocols were approved by the Brown University Institutional
Animal Care and Use Committee.
LTD of geniculocortical transmission in vivo. Electrophysiological recordings
of LGN-evoked field potentials were performed in urethane-anesthetized ani-
mals. Briefly, monopolar recording electrodes were positioned bilaterally in
the supragranular cell layers of A17, and concentric bipolar stimulating elec-
trodes were positioned bilaterally in the LGN. Final positions of the stimulat-
ing and recording electrodes were adjusted to maximize the amplitude of the
field potential recorded in A17 in response to pulsed electrical stimulation of
the LGN. Field potentials were elicited using square-wave pulses of 0.2 ms
duration and 0.2–0.5 mA intensity. Evoked responses were amplified, filtered
(0.1 kHz and 3.0 kHz), digitized (160 kHz), then stored and analyzed using
Experimenter’s Workbench. Before each experiment, a full input-output series
was performed and a stimulation intensity yielding a field potential amplitude
50–60% of maximum was used for the remainder of the experiment. Baseline
responses were collected every 30 s for 30 min before application of low-fre-
quency stimulation (LFS; 1,800 pulses at 1 Hz). Electrophysiological data are
expressed as a percentage of the mean response magnitude recorded during
the baseline recording period.
Monocular deprivation and VEP recordings. MD was performed by eyelid
suture between the hours of 9:00 and 11:00 in the morning. For VEP record-
ings, rats were anesthetized with urethane (2 g/kg, i.p.) and positioned in a
modified stereotaxic frame which allowed for unobstructed viewing of visual
stimuli. Monopolar recordings from the dural surface of A17 were made rela-
tive to a ground screw inserted into the bone overlying the frontal cortex.
Maximum amplitude VEPs were obtained with the recording electrodes posi-
tioned >7.0 mm posterior to Bregma and ≥4.0 mm lateral to the midline, cor-
responding to the binocular region of A17 in the rat. Visual stimuli consisted
of full-field, horizontally oriented sinusoidal gratings (0.05 c.p.d.) presented
on a computer monitor placed 16 cm from the eyes, in a darkened room. Mean
luminance of the monitor was 28 cd/m
2
. VEPs were amplified, filtered
(0.1 kHz and 1.0 kHz), digitized at 20 kHz, and averaged (>500 events per
block) in synchrony with stimulus contrast reversal using a PC. Transient VEPs
in response to abrupt contrast reversing stimuli (1 Hz) were analyzed by meas-
uring the amplitude of the positive peak of the major response component.
Immunoblot analyses. Membrane fractions (or, in some cases, synaptoneuro-
somes) of A17 were prepared and immunoblotted as previously described
12,13,28
.
The signals were quantified using NIH Image 1.62 software or ImageQuant soft-
ware (Molecular Dynamics). Relative amounts of AMPAR phosphorylation were
ascertained by determining the ratio (P/G) of the signal detected using the phos-
phorylation-specific antibody (P) and the phosphorylation-independent anti-
body (G). All analyses were performed blind. The P/G ratios from the
experimental hemispheres (ipsilateral to LFS of the LGN, or contralateral to the
deprived eye) were compared with the values in the yoked control hemispheres
using paired two-tailed t-tests. For display purposes, changes in AMPAR phos-
phorylation levels for experimental A17 are expressed as a percentage of control
A17, according to previously established conventions
13
.We confirmed that
absolute GluR1 levels do not differ across hemispheres after 24 h MD
(GluR1 normalized to total protein contralateral to the deprived eye, 103 ±7% of
ipsilateral value; n = 7).
Cortical slice biotinylation. With the experimenter blind to the deprivation
history of the rat, A17 from both hemispheres were dissected into ice-cold dis-
section buffer containing 2 µM jasplakinolide (Molecular Probes) and sec-
tioned simultaneously into 400-µm-thick slices using a vibratome. Slices were
incubated in ice-cold, oxygenated artificial cerebrospinal fluid (ACSF) con-
taining 2 µM jasplakinolide for 30–60 min, and then incubated in 1 mg/ml
biotin (EZ-link, sulfo-NHS-SS-Biotin, Pierce) solution in ACSF on ice for
20 min. After three washes with ACSF, the superficial layers were microdis-
sected, homogenized in RIPA buffer containing 0.5% SDS, and the protein
concentration of the RIPA soluble fraction determined. Samples were mixed
with NeutrAvidin (Pierce) and incubated at 4 °C overnight. Biotinylated pro-
tein-avidin complex was then spun down and washed 3× with RIPA before
resuspension in SDS sample buffer. Samples of total protein and the pulled-
down fraction were loaded side by side on 7.5% SDS gels. The average values
from triplicate runs of total protein and surface protein were used to deter-
mine the relative level of surface protein. The surface/total ratio of the hemi-
sphere contralateral to the deprived eye was then normalized to the ratio of the
ipsilateral hemisphere to obtain a percentage of control value. It should be
noted that it is not necessary for this analysis to assume that biotin fully pene-
trates the slice. Broken membranes do not contribute significantly to the signal
because the intracellular protein β-tubulin was not biotinylated. We estab-
lished the validity of this approach by showing that induction of LTD by brief
bath application of NMDA
13
in visual cortical slices decreases surface expres-
© 2003 Nature Publishing Group http://www.nature.com/natureneuroscience
ARTICLES
862 VOLUME 6
|
NUMBER 8
|
AUGUST 2003 NATURE NEUROSCIENCE
sion of GluR1 to approximately 56.5 ± 14.3% of control (n = 5), consistent
with the well-established finding that AMPARs are internalized during LTD.
Cortical slice recordings. Animals that had been monocularly deprived for
24 h were anesthetized with Beuthanasia-D (Schering-Plough), and their
brains were removed and dissected in ice-cold dissection buffer containing
212.7 mM sucrose, 2.6 mM KCl, 1.23 mM NaH
2
PO
4
, 26 mM NaHCO
3
,3 mM
MgCl
2
,1 mM CaCl
2
and 10 mM dextrose. Slices of 400 µm thickness contain-
ing the primary visual cortices were isolated and allowed to recover for 1–2 h at
31 °C in ACSF containing 124 mM NaCl, 5 mM KCl, 1.23 mM NaH
2
PO
4
,
26 mM NaHCO
3
,1 mM MgCl
2
,2 mM CaCl
2
and 10 mM dextrose. ACSF and
dissection buffer were saturated with 95% O
2
/5% CO
2
.Recordings were per-
formed in a submersion chamber with the experimenter blind to the hemi-
sphere from which slices were obtained. Extracellular recording electrodes
were placed in layer 2/3 of the binocular region of A17 to monitor field poten-
tials evoked by stimulating electrodes placed at the border of layer 4 and upper
layer 5, as described previously
7
.In every experiment, a full input-output
curve was generated, and baseline responses were obtained every 30 s with a
stimulation intensity that yielded a half-maximal response. To induce LTD,
three episodes of low-frequency stimulation (LFS) consisting of 900 pulses at
1 Hz were applied, each separated by a 25-min interval. Baseline recordings
continued for ≥45 min after the conclusion of the final LFS period.
ACKNOWLEDGMENTS
We thank C. Orsini from the laboratory of L. Maffei (Pisa, Italy) for providing
VEP software, and E. Sklar, K. Clayton, K. Miller and S. Meagher for assistance.
The work was funded by the Howard Hughes Medical Institute and the National
Eye Institute.
COMPETING INTERESTS STATEMENT
The authors declare that they have no competing financial interests.
Received 13 May; accepted 17 June 2003
Published online 28 July 2003; doi:10.1038/nn1100
1. Wiesel, T.N. Postnatal development of the visual cortex and the influence of the
environment. Nature 299, 583–592 (1982).
2. Mioche, L. & Singer, W. Chronic recordings from single sites of kitten striate cortex
during experience-dependent modifications of receptive-field properties.
J. Neurophysiol. 62, 185–197 (1989).
3. Rittenhouse, C.D., Shouval, H.Z., Paradiso, M.A. & Bear, M.F. Monocular depriva-
tion induces homosynaptic long-term depression in visual cortex. Nature 397,
347–350 (1999).
4. Bear, M.F., Cooper, L.N. & Ebner, F.F. A physiological basis for a theory of synaptic
modification. Science 237, 42–48 (1987).
5. Dudek, S.M. & Bear, M.F. Homosynaptic long-term depression in area CA1 of hip-
pocampus and effects of N-methyl-D-aspartate receptor blockade. Proc. Natl. Acad.
Sci. USA 89, 4363–4367 (1992).
6. Mulkey, R.M., Endo, S., Shenolikar, S. & Malenka, R.C. Calcineurin and inhibitor-1
are components of a protein-phosphatase cascade mediating hippocampal LTD.
Nature 369, 486–488 (1994).
7. Kirkwood, A. & Bear, M.F. Homosynaptic long-term depression in the visual cortex.
J. Neurosci. 14, 3404–3412 (1994).
8. Kandler, K., Katz, L.C. & Kauer, J.A. Focal photolysis of caged glutamate produces
long-term depression of hippocampal glutamate receptors. Nat Neurosci 1,
119–123 (1998).
9. Dodt, H., Eder, M., Frick, A. & Zieglgansberger, W. Precisely localized LTD in the
neocortex revealed by infrared-guided laser stimulation. Science 286, 110–113
(1999).
10. Edelman, G.M. & Gally, J.A. Degeneracy and complexity in biological systems. Proc.
Natl. Acad. Sci. USA 98, 13763–13768 (2001).
11. Barria, A., Muller, D., Derkach, V., Griffith, L.C. & Soderling, T.R. Regulatory phos-
phorylation of AMPA-type glutamate receptors by CaM-KII during long-term potenti-
ation. Science 276, 2042–2045 (1997).
12. Lee, H.K., Kameyama, K., Huganir, R.L. & Bear, M.F. NMDA induces long-term
synaptic depression and dephosphorylation of the GluR1 subunit of AMPA receptors
in hippocampus. Neuron 21, 1151–1162 (1998).
13. Lee, H.K., Barbarosie, M., Kameyama, K., Bear, M.F. & Huganir, R.L. Regulation of
distinct AMPA receptor phosphorylation sites during bidirectional synaptic plastic-
ity. Nature 405, 955–959 (2000).
14. Kim, C.H., Chung, H.J., Lee, H.K. & Huganir, R.L. Interaction of the AMPA receptor
subunit GluR2/3 with PDZ domains regulates hippocampal long-term depression.
Proc. Natl. Acad. Sci. USA 98, 11725–11730 (2001).
15. Banke, T.G. et al. Control of GluR1 AMPA receptor function by cAMP-dependent
protein kinase. J. Neurosci. 20, 89–102 (2000).
16. Malinow, R. & Malenka, R.C. AMPA receptor trafficking and synaptic plasticity.
Annu. Rev. Neurosci. 25, 103–126 (2002).
17. Ehlers, M.D. Reinsertion or degradation of AMPA receptors determined by activity-
dependent endocytic sorting. Neuron 28, 511–525 (2000).
18. Chung, H.J., Xia, J., Scannevin, R.H., Zhang, X. & Huganir, R.L. Phosphorylation of
the AMPA receptor subunit GluR2 differentially regulates its interaction with PDZ
domain-containing proteins. J. Neurosci. 20, 7258–7267 (2000).
19. Heynen, A.J., Abraham, W.C. & Bear, M.F. Bidirectional modification of CA1
synapses in the adult hippocampus in vivo. Nature 381, 163–166 (1996).
20. Kameyama, K., Lee, H.K., Bear, M.F. & Huganir, R.L. Involvement of a postsynaptic
protein kinase A substrate in the expression of homosynaptic long-term depression.
Neuron 21, 1163–1175 (1998).
21. Malenka, R.C. & Nicoll, R.A. Long-term potentiation—a decade of progress?
Science 285, 1870–1874 (1999).
22. Fagiolini, M., Pizzorusso, T., Berardi, N., Domenici, L. & Maffei, L. Functional post-
natal development of the rat primary visual cortex and the role of visual experience:
dark rearing and monocular deprivation. Vision Res. 34, 709–720 (1994).
23. Heynen, A.J. & Bear, M.F. Long-term potentiation of thalamocortical transmission in
adult visual cortex in vivo. J. Neurosci. 21, 9801–9813 (2001).
24. Kenan-Vaknin, G. & Teyler, T.J. Laminar pattern of synaptic activity in rat primary
visual cortex: Comparison of in vivo and in vitro studies employing the current source
density analysis. Brain Res. 635, 37–48 (1994).
25. Yinon, U. & Auerbach, E. Deprivation of pattern vision studied by visual evoked
potentials in the rat cortex. Exp. Neurol. 38, 231–251 (1973).
26. Sawtell, N.B. et al. NMDA receptor-dependent ocular dominance plasticity in adult
visual cortex. Neuron 38, 977–985 (2003).
27. Bear, M.F., Kleinschmidt, A., Gu, Q. & Singer, W. Disruption of experience-depend-
ent synaptic modifications in striate cortex by infusion of an NMDA receptor antago-
nist. J. Neurosci. 10, 909–925 (1990).
28. Heynen, A.J., Quinlan, E.M., Bae, D.C. & Bear, M.F. Bidirectional, activity-depend-
ent regulation of glutamate receptors in the adult hippocampus in vivo. Neuron 28,
527–536 (2000).
29. Kossut, M. & Singer, W. The effect of short periods of monocular deprivation on exci-
tatory transmission in the striate cortex of kittens: a current source density analysis.
Exp. Brain Res. 85, 519–527 (1991).
30. Trachtenberg, J.T., Trepel, C. & Stryker, M.P. Rapid extragranular plasticity in the
absence of thalamocortical plasticity in the developing primary visual cortex.
Science 287, 2029–2032 (2000).
31. Antonini, A., Fagiolini, M. & Stryker, M.P. Anatomical correlates of functional plas-
ticity in mouse visual cortex. J. Neurosci. 19, 4388–4406 (1999).
32. Bear, M.F. & Rittenhouse, C.D. Molecular basis for induction of ocular dominance
plasticity. J. Neurobiol. 41, 83–91 (1999).
33. Colman, H., Nabekura, J. & Lichtman, J.W. Alterations in synaptic strength preced-
ing axon withdrawal. Science 275, 356–361 (1997).
34. Guillery, R. Binocular competition in the control of geniculate cell growth. J. Comp.
Neurol. 144, 117–130 (1972).
35. Bienenstock, E.L., Cooper, L.N. & Munro, P.W. Theory for the development of neuron
selectivity: Orientation specificity and binocular interaction in visual cortex.
J. Neurosci. 2, 32–48 (1982).
36. Blais, B.S., Shouval, H.Z. & Cooper, L.N. The role of presynaptic activity in monoc-
ular deprivation: comparison of homosynaptic and heterosynaptic mechanisms.
Proc. Natl. Acad. Sci. USA 96, 1083–1087 (1999).
37. Mitchell, D.E., Gingras, G. & Kind, P.C. Initial recovery of vision after early monocu-
lar deprivation in kittens is faster when both eyes are open. Proc. Natl. Acad. Sci.
USA 98, 11662–11667 (2001).
38. Kirkwood, A., Rioult, M.G. & Bear, M.F. Experience-dependent modification of
synaptic plasticity in visual cortex. Nature 381, 526–528 (1996).
39. Philpot, B.D., Sekhar, A.K., Shouval, H.Z. & Bear, M.F. Visual experience and depri-
vation bidirectionally modify the composition and function of NMDA receptors in
visual cortex. Neuron 29, 157–169 (2001).
40. Philpot, B.D., Espinosa, J.S. & Bear, M.F. Evidence for altered NMDA receptor func-
tion as a basis for metaplasticity in visual cortex. J. Neurosci. 23, 5583–5588 (2003).
41. Freeman, R.D., Mallach, R. & Hartley, S. Responsivity of normal kitten striate cortex
deteriorates after brief binocular deprivation. J. Neurophysiol. 45, 1074–1084 (1981).
42. Prusky, G.T. & Douglas, R.M. Developmental plasticity of mouse visual acuity. Eur. J.
Neurosci. 17, 167–73 (2003).
43. Dudek, S.M. & Friedlander, M.J. Developmental down-regulation of LTD in cortical
layer IV and its independence of modulation by inhibition. Neuron 16, 1–20 (1996).
44. Rozas, C. et al. Developmental inhibitory gate controls the relay of activity to the
superficial layers of the visual cortex. J. Neurosci. 21, 6791–6801 (2001).
45. Huang, Z.J. et al. BDNF regulates the maturation of inhibition and the critical period
of plasticity in mouse visual cortex. Cell 98, 739–755 (1999).
46. Ramoa, A.S., Paradiso, M.A. & Freeman, R.D. Blockade of intracortical inhibition in
kitten striate cortex: effects on receptive field properties and associated loss of ocu-
lar dominance plasticity. Exp. Brain Res. 73, 285–296 (1988).
47. Hensch, T.K. et al. Local GABA circuit control of experience-dependent plasticity in
developing visual cortex. Science 282, 1504–1508 (1998).
48. Kirkwood, A., Rozas, C., Kirkwood, J., Perez, F. & Bear, M.F. Modulation of long-term
synaptic depression in visual cortex by acetylcholine and norepinephrine. J.
Neurosci. 19, 1599–1609 (1999).
49. Allen, C.B., Celikel, T. & Feldman, D.E. Long-term depression induced by sensory
deprivation during cortical map plasticity. Nat. Neurosci. 6, 291–299 (2003).
50. Mammen, A.L., Kameyama, K., Roche, K.W. & Huganir, R.L. Phosphorylation of the
alpha-amino-3-hydroxy-5-methylisoxazole4- propionic acid receptor GluR1 subunit
by calcium/calmodulin-dependent kinase II. J. Biol. Chem. 272, 32528–32533
(1997).
© 2003 Nature Publishing Group http://www.nature.com/natureneuroscience
