Proc. Natl. Acad. Sci. USA
Vol. 92, pp. 11175-11179, November 1995
Calcium/calmodulin-dependent kinase II and long-term
potentiation enhance synaptic transmission by the
(glutamate receptor/synaptic plasticity/learning)
PIERRE-MARIE LLEDO*, GREGORY 0. HJELMSTADt, SUCHETA MUKHERJIl, THOMAS R. SODERLINGt,
ROBERT C. MALENKA§T, AND ROGER A. NIcOLL*§
Departments of *Cellular and Molecular Pharmacology, §Physiology, and IPsychiatry, and the tNeuroscience Graduate Program, University of California, San
Francisco, San Francisco, CA 94143-0450; and:Vollum Institute, Oregon Health Sciences University, 3181 SW Sam Jackson Park Road, Portland, OR 97201
Contributed by Roger A. Nicoll, August 25, 1995
role in long-term potentiation (LTP). To test the involvement
of Ca2+/calmodulin-dependent kinase II (CaM-K II), a trun-
cated, constitutively active form of this kinase was directly
injected into CAl hippocampal pyramidal cells. Inclusion of
CaM-KI in the recording pipette resulted in a gradual increase
in the size of excitatory postsynaptic currents (EPSCs). No
change in evoked responses occurred when the pipette contained
heat-inactivated kinase. The effects of CaM-K II mimicked
several features ofLTP in that it caused a decreased incidence of
synaptic failures, an increase in the size of spontaneous EPSCs,
and an increase in the amplitude of responses to iontophoreti-
cally applied cv-amino-3-hydroxy-5-methyl-4-isoxazolepropi-
onate. To determine whether the CaM-K 11-induced enhance-
ment and LTP share a common mechanism, occlusion experi-
ments were carried out. The enhancing action ofCaM-K II was
greatly diminished by prior induction of LTP. In addition,
following the increase in synaptic strength by CaM-K II, tetanic
stimulation failed to evoke LTP. These findings indicate that
CaM-K II alone is sufficient to augment synaptic strength and
that this enhancement shares the same underlying mechanism as
the enhancement observed with LTP.
Ca2+-sensitive kinases are thought to play a
Repetitive activation of excitatory glutamatergic synapses re-
sults in a long-lasting enhancement in synaptic strength, re-
ferred to as long-term potentiation (LTP). The most wide-
spread form of LTP requires the activation of postsynaptic
N-methyl-D-aspartate (NMDA) receptors and an increase in
postsynaptic Ca2 . Considerable evidence suggests that the
increase in Ca2+ activates Ca2+-dependent kinases (1-4). In
particular, Ca2+/calmodulin-dependent kinase II (CaM-K II),
which is present at extremely high concentrations in the
postsynaptic density (5, 6), is an attractive candidate for
mediating the effects of Ca2 . Many lines of indirect evidence
support a role for this kinase in mediating the effects of Ca2+
on synaptic strength (2, 3). A direct approach to investigating
a role for CaM-K II in LTP is to determine its effects on
synaptic strength and LTPwhen the concentration ofactivated
kinase is increased in postsynaptic cells. Recently two groups,
one using vaccinia virus in acute hippocampal slices (7) and
one using mouse genetics (8), have expressed constitutively
active forms of this enzyme. In the vaccinia virus experiments
(7), evidence consistent with an enhancement in synaptic
transmission was presented, and LTP induction was impaired.
In contrast, in the transgenic mouse experiments (8), no
change in synaptic transmission or in the ability to generate
LTPwas found. Aside from the apparent discrepancies in these
reports, it is difficult in these experiments to entirely exclude
indirect effects associated with the use of the expression
systems. In the present experiments we have examined the
effects of CaM-K II by directly injecting a constitutively active
form of this enzyme into the postsynaptic cell.
Experiments were performed at room temperature (21-24°C)
on hippocampal slices prepared (9, 10) from 3- to 5-week-old
male Hartley guinea pigs or 2- to 3-week-old Sprague-Dawley
rats. All bathing solutions contained picrotoxin (100 ,tM).
Standard extracellular, intracellular, and whole-cell patch
clamp methods were used (10-12). Field electrodes contained
1 M NaCl and intracellular electrodes contained 2 M potas-
sium acetate with either 1,AMactivated CaM-K II or 1 ,tM
heat-inactivated CaM-K II as control. The tips of whole-cell
pipettes were filled with a solution containing 123 mM cesium
gluconate, 15.5 mM CsCl, 10 mM Hepes, 10 mM CsEGTA, 8
mM NaCl, 1 mM CaCl2, 2 mM MgATP, 0.3 mM Na3 GTP, 0.2
mM cAMP, 10mM D-glucose, and 10 ,tM microcystin-LR (pH
7.3 with CsOH, 280-290 mosM). These patch electrodes were
then backfilled with the same solution containing either 200
nM activated CaM-K II or 200 nM heat-inactivated CaM-K II.
Cells were maintained at a membrane potential between -70
and -85 mV. Whole-cell experiments were stopped when the
series resistance was >30 Mfl or when the series resistance
changed >20% during the course of an experiment.
To evoke synaptic responses, stimuli (100-,us duration at a
frequency of 0.05-0.1 Hz) were delivered through fine bipolar
stainless steel electrodes placed in stratum radiatum. For
two-pathway experiments, two stimulating electrodes were
placed on either side of the recording electrodes. To elicit LTP
tetani [four trains of 100 pulses (100 Hz) at 20-s intervals] were
delivered at test stimulus intensity. a-Amino-3-hydroxy-5-
methyl-4-isoxazolepropionate (AMPA) responses were
evoked by iontophoretically applying AMPA (negative 75-150
nA, 1 s) every 120 s with an electrode containing 10 mM
AMPA (pH 8) placed close to the cell body layer. For the first
2-3 min after break-in, AMPA pulses were applied more
frequently and the iontophoretic current was adjusted to
obtain 100- to 200-pA responses. Minimal stimulation record-
ings at a frequency of 1 Hz were obtained under visual control.
Pipette solutions were exchanged using a 2PK+ perfusion kit
(Adams & List, Westbury, NY). Fast green (1%) was included
in the solution to verify diffusion into the cell, which took 8-10
min. Failure rates were estimated by the method of Liao et al.
(13), for successive epochs of 180 stimuli. Unless otherwise
Abbreviations: CaM-K II, Ca2+/calmodulin-dependent protein kinase
II; EPSC, excitatory postsynaptic current; EPSP, excitatory postsyn-
aptic potential; fEPSP, field EPSP; LTP, long-term potentiation;
AMPA, a-amino-3-hydroxy-5-methyl-4-isoxazolepropionate; NMDA,
The publication costs of this article were defrayed in part by page charge
payment. This article must therefore be hereby marked "advertisement" in
accordance with 18 U.S.C. §1734 solely to indicate this fact.
Proc. Natl. Acad. Sci. USA 92 (1995)
stated, values given in the text are means ± SEM and
significance was assessed by either a paired or an unpaired
Student's t test.
Baseline values of excitatory postsynaptic potentials (EP-
SPs), excitatory postsynaptic currents (EPSCs), and ionto-
phoretic responses were obtained from averages of responses
during the first 2-3 min (time 0 on each graph)-or, for the
perfusion experiments, during the 6 min prior to the time of
perfusion-and defined as 100% for subsequent analyses. For
all experiments neurons injected with activated or heat inac-
tivated CaM-K II were interleaved during daily sessions. The
CaM-K II (a subunit) was truncated at residue 316 to give a
monomeric (30- to 40-kDa) enzyme, expressed in baculovirus-
infected Sf9 insect cells, and purified as described (14). The
kinase (2 ,uM) was converted to its constitutively active form
(30-40% of Ca2+-independent activity) by autothiophospho-
rylation (15). Inactivated kinase was heated for 10 min at
100°C prior to addition to the autothiophosphorylation reac-
tion mixture. All kinase samples were diluted 2-fold (intracel-
lular recording) or 10-fold (whole-cell recording) in pipette
solution just before use and were maintained on ice. Aliquots
were renewed within 2 hr after their preparation. After some
experiments, kinase activity ofthe internal solution containing
the diluted activated CaM-K II was measured and found to
retain 80% of its original activity.
We first tested the effects of activated CaM-K II by adding it
to the whole-cell pipette solution (200 nM) and measuring the
amplitude ofEPSCs over time. While a clear growth in the size
of the EPSCwas seen in some cells, the effect was variable and
could not be systematically studied. In an attempt to increase
Inact. CaM-K 11
I & 40 min
1 & 40 min
the reliability of this effect, we included the phosphatase
inhibitor microcystin-LR (10 ,uM) in the pipette solution.
Under these conditions a very consistent growth in the size of
the EPSC was observed (Fig. 1). The size began to increase
significantly within 6-8 min after establishment of the whole-
cell recording, and the maximum effect occurred at 15-30 min
(Fig. 1B, *). The average size of the EPSC reached 164 ± 16%
(n = 8) of the initially recorded EPSC. CaM-K II had no effect
on the cells' passive membrane properties. To ensure that the
enhancement resulted from the CaM-K II and not from
nonspecific effects or from the microcystin-LR, interleaved
recordings were made with the identical internal recording
solution, except that heat-inactivated kinase (200 nM) was
used instead of the active kinase. Under these recording
conditions, no enhancement was observed (105 ± 6%, n = 15)
(Fig. 1B, 0).
To gain insight into the mechanism(s) involved in the
CaM-K 1I-induced enhancement, we carried out experiments
using minimal stimulation so that we could record failures of
evoked responses (16-18). In these experiments, the pipette
was internally perfused under visual control to deliver the
CaM-K II in a temporally controlled manner (seeMethods). In
the example shown in Fig. 2A the stimulus intensity was
initially adjusted so that it produced mostly failures. After a
stable baseline was established, perfusion of the pipette with
CaM-K II was initiated, and about 10 min later the failure rate
began to decrease, accompanied by an increase in the mean
size ofthe response (Fig. 2A2).A summary of five experiments
with the active kinase (Fig. 2B1, 0) demonstrates that the
failure rate on average decreased by 27 ± 10% (P < 0.02),
whereas the heat-inactivated kinase had no effect on either the
failure rate (Fig. 2B1, 0) or the mean size of the EPSC (Fig.
o InacL CaM-K I
Current amplitude (-pA)
recorded in hippocampal CAl neurons. (A) Average of six consecutive
EPSCs obtained during whole-cell recordings at the indicated times
following break-in. Recordings were made with pipette solutions
containing activated (upper traces) or heat-inactivated (lower traces)
CaM-K IT. EPSCs are superimposed to the right. (B) Summary data
illustrating the time course of EPSC amplitudes (see Methods for
normalization procedure) in the presence of activated (-) or heat-
inactivated (0) CaM-K IT.
Activated form of CaM-K II potentiates evoked EPSCs
change in the number of synaptic failures and the size of spontaneous
EPSCs. (Al) An individual example of the decrease in failure rate
following exchange of the pipette solution with one containing acti-
vated CaM-K II. Time 0 is defined as when the solution reaches the
tip of the electrode. (A2) Average of 100 (upper) or composite of 6
consecutive (lower) traces at times indicated. (B) Summary of failures
(Bil) and average amplitude (B2) for experiments using activated (-)
or heat-inactivated (0) CaM-K II. (Cl) Cumulative amplitude distri-
butions comparing events collected 2-10 min (0) and 40-48 min (-)
after break-in with a CaM-K II-containing pipette. (C2) Averages of
events shown in Cl are superimposed (upper traces) and scaled (lower
CaM-K II-induced potentiation is associated with both a
Neurobiology:Lledo et aL
Proc. Natl. Acad. Sci. USA 92 (1995)
2B2) (n = 5). Consistent with the decrease in failure rate, we
also found that, associated with the increase in the evoked
response (Fig. 1), the frequency of spontaneous EPSCs in-
creased 32 ± 6% (n = 4) [compared with 6 ± 4% (n = 3)] for
the heat-inactivated CaM-K TI (P < 0.02) (data not shown).
The results presented thus far can be explained by an
increase in the probability of transmitter release (18) and/or
an all-or-none upregulation of clusters of AMPA receptors
(13, 19). Indeed, we also found that CaM-K II increased the
size of the spontaneous EPSCs (Fig. 2C) [active, 36 ± 3%, n
= 4; inactive, 6 ± 4%, n = 3 (P < 0.002)]. This observation
suggested that CaM-K II might be enhancing the EPSCs, at
least in part, by increasing the sensitivity ofAMPA receptors
to synaptically released glutamate. This possibility was tested
directly by monitoring the responses of CAl cells to inoto-
phoretically applied AMPA. The response to AMPA slowly
increased when the recording pipette contained CaM-K TI
(151 ± 10% measured at 40-50 min, n = 8) (Fig. 3Al and B,
*), but not when it contained the inactive form (101 ± 14%,
n = 8) (Fig. 3A2 and B, 0) (P < 0.02). These results indicate
that CaM-K II applied directly into the cell can increase the
sensitivity of AMPA receptors, an effect that most likely
contributes to the enhancement of the EPSC.
Are the effects ofCaM-K II related to LTP? To address this
question, we designed occlusion experiments in which the
effects of CaM-K II on control synapses were compared with
its effects on synapses expressing LTP. In these experiments,
two independent inputs onto the same population ofpyramidal
cells were monitored with field electrode recordings (Fig. 4A).
A saturating level of LTP was induced in one of the pathways
(Si), while the other pathway (S2) served as a control (Fig.
4A). Once LTP was observed to be stable (typically within 60
CaM-K 11(1 min)
Inact. CaM-K 11(1 min)
CaM-K 11(45 min)
Inact. CaM-K 11(45 min)
Inact. CaM-K 11
(A) Chart records of membrane current from voltage-clamped CAl
pyramidal cells held at -75 mV show responses to brief iontophoretic
pulses of AMPA. The presence of activated (Al), but not heat-
inactivated (A2), CaM-K II potentiatedAMPA responses 45 min after
recordings were initiated. (B) Plots ofsummarized data illustrating the
time courses of the mean amplitude of AMPA-induced current
presence of either activated (-) or heat-inactivated (0) CaM-K II.
Postsynaptic sensitivity toAMPA is increased by CaM-K II.
as percentage of the mean control amplitude, in the
min), a whole-cell recording was made from a cell within the
population sampled by the field electrode (Fig. 4B). The effect
of CaM-K II on the LTP-expressing pathway was then com-
pared with the effect on the control pathway (Fig. 4 Bl and
B2), while we continued to monitor fEPSPs from both path-
ways (Fig. 4B3). CaM-K II had its normal enhancing effect on
the control pathway but was without effect on the pathway
expressing LTP (Fig. 4Bl andB2).A summary graph of all the
experiments (Fig. 5) demonstrates that the enhancing action of
CaM-K II was markedly attenuated on synapses expressing
LTP (an increase of 17 ± 9%, n = 6 vs. 67 ± 16%, n = 6,
measured at 40-55 min from tetanized and untetanized path-
ways, respectively, P < 0.0001) (Fig. 5B1). The average am-
plitude of the CaM-K II-induced potentiation was less than
that for saturated LTP (compare Fig. SAl and BJ). This most
likely was due to the variable loading of the cells with the
kinase, as suggested by the fact that the potentiation observed
in some ofthe loaded cellswas equivalent to that observedwith
saturated LTP (data not shown). To verify that this selective
effect was, in fact, due to CaM-K II, a series of control
experiments with heat-inactivated enzyme were interleaved
with the above experiments (Fig. 5B2) (n = 6) and no change
in the EPSCs in either pathway occurred. The simultaneously
recorded extracellular responses to both the control and
LTP-expressing pathways remained constant in all of these
experiments (Fig. 5A2).
In a final set of experiments we performed the reverse
occlusion experiment, in which we first potentiated the syn-
apses with CaM-K II and then asked whether this potentiation
affected the ability to induce LTP. Whole-cell recording
techniques could not be used for this experiment, because the
ability to induce LTP washes out over the 30-40 min required
for the CaM-K II effect to stabilize. We therefore loaded sharp
microelectrodes with the enzyme (1 ,uM). Because this method
ofloading neurons with CaM-K II was less successful than that
used with the whole-cell approach, we examined only the
ability to generate LTP in those cells inwhich CaM-K II caused
an enhancement. Fig. 6A1 shows a summary graph of these
cells (n=5). We also simultaneously monitored field poten-
tials, so that the magnitude of LTP in the field could be
compared with that generated in the cell. While tetanization
caused substantial LTP in the field (Fig. 6A2), it produced no
LTP in the cells loaded with CaM-K II (Fig. 6A1) (208 ± 40%
vs. 79 ± 33% of baseline, P < 0.0001, for field potential and
intracellular recordings, respectively). In a series of control
experiments with the heat-inactivated enzyme, the magnitude
of the LTP in the cells (Fig. 6B1) was similar to that recorded
in the field (Fig. 6B2) (n=5). These findings provide further
evidence that LTP and CaM-K II enhance EPSCs by the same
The hypothesis that CaM-K II mediates the effects of the
NMDA receptor-dependent increase in postsynaptic Ca2+ that
triggers LTP is attractive and is based on a variety of exper-
imental approaches. (i) Biochemical studies have shown that
the kinase can act as a molecular switch, conferring properties
that are advantageous for long-lasting storage of changes
initiated by brief Ca2+ signals (2, 20, 21). (ii) Manipulations
that interfere with CaM-K II function interfere with LTP.
These experiments include the use of inhibitors (12, 22-24)
and knockout of the a subunit of CaM-K II in mice (25). (iii)
LTP is associated with an increase in the activity of the
Ca2+-independent form of CaM-K II (26). (iv) One of the
targets for CaM-K II is the AMPA receptor itself. Thus,
activated CaM-K II can phosphorylateAMPA receptors in the
postsynaptic density and can enhance responses to AMPA
receptor agonists in cultured hippocampal neurons (15) and
acutely isolated dorsal root ganglion neurons (27). Responses
Neurobiology:Lledo et al.
Proc. Natl. Acad. Sci. USA 92 (1995)
0 o0 00000000
4r¶. * w
S2) onto the same population ofCAl neurons were alternatively stimulated. LTP was induced by tetanic stimulation of S1; S2 served as the control
input. (Al) Field potential recordings (average of 15 consecutive traces) from a representative experiment taken at the indicated times. (A2)
Normalized field EPSP (fEPSP) slopes evoked in the tetanized pathway (S1; 0) and in the untetanized control pathway (S2; 0). Two tetanizations
were given in this experiment (arrows). (B) Simultaneous recordings of whole-cell EPSCs obtained with a pipette containing CaM-K II (Bl and
B2) and field potentials (B3) 80 min after having induced LTP (shown inA2). EPSCs and fEPSPs were normalized to the initial responses at the
beginning of the whole-cell experiment (dotted baseline). Bl shows representative EPSCs recorded at the indicated times from the two independent
pathways. A voltage step was given 85 ms before Si to monitor series and input resistance.
CaM-K II-induced potentiation of evoked EPSCs is occluded at synapses expressing saturated LTP. Two independent inputs (Si and
evoked by activating GluRl receptors expressed in oocytes are
also enhanced by CaM-K II (28).
The present results demonstrate that in hippocampal pyra-
midal cells in situ, CaM-K II enhances EPSCs and that this
enhancement is due, at least in part, to an enhancement of
AMPA receptor sensitivity, thus extending previous results in
cultured neurons (15) and spinal cord neurons (27). CaM-K II
also caused an increase in the frequency of spontaneous
EPSCs and a reduction in the number of failures of evoked
responses. These results can be explained by a presynaptic en-
hancement in transmitter release (18) and/or an all-or-none
upregulation ofclusters ofAMPA receptors (13,19). The finding
U) ,& 150
occlusion of the CaM-K II-induced
potentiation by LTP. The time
course of changes in extracellular
fEPSPs(A)and evoked whole-cell
EPSCs (B) is shown. Al illustrates
that LTP was induced in S1, but not
S2, prior to whole-cell recordings;
andA2 shows the stability offEPSPs
during the whole-cell recordings
(normalized to the initial recordings
at the beginning ofwhole-cell exper-
iments). Arrow indicates the time of
the first tetanus given to S1. (B)
Graph illustrating potentiation of
EPSC amplitudes by CaM-K II only
in the untetanized control pathway
(S2; 0) (Bl). EPSCs did not change
when the pipette solution contained
heat-inactivated CaM-K II (B2). Be-
cause there was no difference in the
-g w wmagnitudeof LTP induced in the
two sets ofexperiments (activeki-
nase, 117 ± 11%, n = 6; inactive
kinase, 102 ± 10% increase, n =6),
the data from these experiments
were combined in A.
Summary data showing
Neurobiology:Lledo et al.
; I[..I.Lil Ai..l
Proc. Natl. Acad. Sci. USA 92 (1995)
_ _ _ }_1300
Inact. CaM-K 11
clusion experiments examining
LTP at synapses already potenti-
ated by activated CaM-K II. (A)
Time course of EPSP changes re-
corded with intracellular elec-
trodes containing CaM-K II (Al)
with simultaneous recordings of
field potentials (A2). (B) Similar
experiments using heat-inactivated
CaM-K II as control. Arrow indi-
cates the tetanus given at time 0.
Slopes of EPSPs in LTP were then
renormalized 5 min before the tet-
anus (dotted baseline). Insets show
averaged EPSPs taken at indicated
Summary data from oc-
that the enhancement of responses to exogenously applied
AMPA was of similar magnitude to that observed with the
evoked EPSC favors the latter hypothesis. Most importantly,
occlusion experiments establish that the synaptic enhancement
evoked by CaM-K II shares the same underlying mechanisms as
LTP. Two recent reports examined the effects of overexpression
of constitutively active CaM-K II on LTP. One study found that
overexpression with vaccinia virus in the hippocampal slice
impaired the ability to induce LTP (7), whereas the other study,
using transgenic mice, found no impairment in LTP (8), even
though significant effects onlong-term depression were observed.
The results ofour acute experiments are more consistentwith the
Inhibitor studies have led to the proposal that protein kinase
C activity (23, 29-31) and protein-tyrosine kinase activity (32)
are also required for LTP. Since our results indicate that
CaM-K II alone mimics LTP, either CaM-K II can, in some
unknown manner, activate these enzymes, or certain levels of
basal protein kinase and tyrosine kinase activity are required
for CaM-K II to cause a synaptic enhancement.
In summary, the present results indicate that the enhance-
ment of synaptic strength caused by postsynaptic injection of
activated CaM-K II shares the same underlying mechanism(s)
as LTP. It appears that elucidating the substrates targeted by
CaM-K II should shed considerable light on the molecular
mechanisms underlying LTP. A likely target is the AMPA
receptor itself, which is colocalized in the postsynaptic density
with CaM-K II. Thus, defining the mechanism by which
CaM-K II increases AMPA receptor function should be par-
We thank Drs. D. Copenhagen and D. Dixon for comments on the
manuscript. P.-M.L. was supported by the Centre National de la
Recherche Scientifique and by a North Atlantic Treaty Organization
fellowship. G.O.H. was supported by a National Science Foundation
predoctoral fellowship. T.R.S. was supported by National Institutes of
Health Grant NS27037. R.C.M. and R.A.N. were supported by grants
from the National Institutes ofHealth. R.A.N. is amember ofthe Keck
Center for Integrative Neuroscience and the Silvio Conte Center for
Neuroscience Research. R.C.M. is a member of the Center for
Neurobiology and Psychiatry.
Bliss, T. V. P. & Collingridge, G. L. (1993) Nature (London) 361,
Lisman, J. (1994) Trends Neurosci. 17, 406-412.
Nicoll, R. A. & Malenka, R. C. (1995) Nature (London) 377,115-118.
Larkman, A. U. & Jack, J. J. B. (1995) Curr. Opin. Neurobiol. 5,
Kennedy, M. B., Bennett, M. K. & Erondu, N. E. (1983) Proc. Natl.
Acad. Sci. USA 80, 7357-7361.
Kelly, P. T., McGuinness, T. L. & Greengard, P. (1984) Proc. Natl.
Acad. Sci. USA 81, 945-949.
Pettit, D. L., Perlman, S. & Malinow, R. (1994) Science 266, 1881-
Mayford, M., Wang, J., Kandel, E. R. & O'Dell, T. J. (1995) Cell 81,
Nicoll, R. A. & Alger, B. E. (1981) Science 212, 957-959.
Castillo, P. E., Weisskopf, M. G. & Nicoll, R. A. (1994) Neuron 12,
Kullmann, D. M., Perkel, D. J., Manabe, T. & Nicoll, R. A. (1992)
Neuron 9, 1175-1183.
Wyllie, D. J. A. & Nicoll, R. A. (1994) Neuron 13, 635-643.
Liao, D., Hessler, N. A. & Malinow, R. (1995) Nature (London) 375,
Mikherji, S., Brickey, D. A. & Soderling, T. R. (1994) J. Biol. Chem.
McGlade-McCulloh, E., Yamamoto, H., Tan, S. E., Brickey, D. A. &
Soderling, T. R. (1993) Nature (London) 362, 640-642.
Kullmann, D. M. & Nicoll, R. A. (1992) Nature (London) 357, 240-
Liao, D., Jones, A. & Malinow, R. (1992) Neuron 9, 1089-1097.
Stevens, C. F. & Wong, Y. (1994) Nature (London) 371, 704-707.
Isaac, J. T. R., Nicoll, R. A. & Malenka, R. C. (1995) Neuron 15,
Miller, S. G. & Kennedy, M. B. (1986) Cell 44, 861-870.
Lisman, J. E. (1985) Proc. Natl. Acad. Sci. USA 82, 3055-3057.
Malenka, R. C., Kauer, J. A., Perkel, D. J., Kelly, P. T., Nicoll, R. A.
& Waxham, M. N. (1989) Nature (London) 340, 554-557.
Malinow, R., Schulman, H. & Tsien, R. W. (1989) Science 245,
Ito, I., Hidaka, H. & Sugiyama, H. (1991)Neurosci. Lett. 121,119-121.
Silva, A. J., Stevens, C. F., Tonegawa, S. & Wang, Y. (1992) Science
Fukunaga, K, Stoppini, L., Miyamoto, E. & Muller, D. (1993) J. Biol.
Chem. 268, 7863-7867.
Kolaj, M., Cerne, R., Cheng, G., Brickey, D. A. & Randic, M. (1994)
J. Neurophysiol. 72, 2525-2531.
Yakel, J. L., Vissavajjhala, P., Derkach, V. A., Brickey, D. A. &
Soderling, T. R. (1995) Proc. Natl. Acad. Sci. USA 92, 1376-1380.
Wang, J.-H. & Feng, D.-P. (1992) Proc. Natl. Acad. Sci. USA 89,
Linden, D. J. & Routtenberg, A. (1989) Brain Res. Rev. 14, 279-296.
Hvalby, 0., Hemmings, H. J., Paulsen, O., Czernik, A., Nairn, A.,
Godfraind, J., Jensen, V., Raastad, M., Storm, J., Andersen, P. &
Greengard, P. (1994) Proc. Natl. Acad. Sci. USA 91, 4761-4765.
O'Dell, T. J., Kandel, E. R. & Grant, S. G. N. (1991) Nature (London)
Neurobiology:Lledo et al.