Glutamatergic Synaptic R esponses and
L ong-Term Potentiation Are Impaired in
the CA1 Hippocampal Area of Calbindin
ANNE J OUVE NCE AU,1BR IGIT T E POT IE R ,1R E NATA BAT T INI,2ST E FANO F E R R AR I,2
PAT R ICK DUTAR ,1AND J E AN-MAR IE BIL L AR D1*
1LaboratoiredePhysiopharmacologiedu Syste `meNerveux, Paris, France
2Dipartementodi ScienzeBiomediche, Sezionedi chimica Biologica, Universita di Modena, Modena, Italy
KEY WORDS calcium binding protein; long-term potentiation; NMDA receptor;
long-term depression; hippocampus; glutamate
(CaBP) to glutamatergic neurotransmission and synaptic plasticity was investigated in
hippocampal CA1 area of wild-type and antisense transgenic CaBP-deficient mice, with
the use of extracellular recordings in the ex vivo slice preparation. The amplitude of
non-N-methyl-D-aspartate receptor (non-NMDAr)-mediated extracellular field excita-
tory postsynaptic potentials (fEPSPs) recorded in control medium was significantly
greater in CaBP-deficient mice, whereas the afferent fiber volley was not affected. In
contrast, the amplitude of NMDAr-mediated fEPSPs isolated in a magnesium-free
medium after blockade of non-NMDAr and GABAergic receptors was significantly
depressed in these animals. No alteration in the magnitude of paired-pulse facilitation
was found, indicating that the presynaptic calcium mechanisms controlling glutamate
release were not altered in CaBP-deficient mice. The magnitude and time course of the
short-term potentiation (STP) of fEPSPs induced by a 30 Hz conditioning stimulation,
which was blocked by the NMDAr antagonist 2-amino-5-phosphonovalerate acid (2-
APV), was not impaired in the transgenic mice, whereas long-term potentiation (LTP)
induced by a 100 Hz tetanus was not maintained. The long-term depression (LTD)
induced by low-frequency stimulation (1 Hz, 15 min) in the presence of the GABA
antagonist bicuculline was not altered. These results argue for a contribution of CaBP to
the mechanisms responsible for the maintenance of long-term synaptic potentiation, at
least in part by modulating the activation of NMDA receptors. Synapse 33:172–180,
?1999 Wiley-Liss, Inc.
The contribution of the cytosolic calcium binding protein calbindin D28K
INT R ODUCT ION
Theregulation of intracellular calcium concentration
is crucial for neuronal functions and particularly for
synaptic transmission and neuronal plasticity, which
are generally accepted as underlying learning and
memory processes (Baudry and Massicote, 1992; Bliss
and Collingridge, 1993). Neurons use a number of
mechanisms toregulate calcium level, including extru-
sion by ATPases, sequestration into organelles, and
chelation by cytosolic calcium-binding proteins (see
Kostyuk and Verkhratsky, 1995, for review). Among
theseproteins, calmodulin wasfirst suggestedtocontrib-
utetoneuronal plasticity (Malenka et al., 1989; Silva et
al., 1992). Calbindin D28K (CaBP) is another calcium-
binding protein of the ‘‘EF-hand’’ family strongly ex-
pressed in brain areas implicated in learning and
memory such as the cerebral cortex, the thalamus, and
the CA1 field and dentate gyrus of the hippocampus
(Baimbridge and Miller, 1982; Celio, 1990). Although a
neuroprotectiveroleis attributed toCaBP (seeIacopino
et al., 1994), its contribution to influence synaptic
transmission and neuronal plasticity remains to be
determined (Baimbridge et al., 1992; Chard et al.,
1995). We recently reported that mice with reduced
CaBP expression failed to express hippocampal CA1
long-term potentiation (LTP) induced by high-fre-
Contract grant sponsors: Bayer-pharma, France, and Telethon-Italy.
*Correspondence to: J .M. Billard, Laboratoire de Physiopharmacologie du
Syste `meNerveux, INSERM U 161, 2 rued’Ale ´sia, 75014 Paris, France.
Received 5 May 1998;Accepted 17 November 1998
SYNAPSE 33:172–180 (1999)
?1999 WILEY-LISS, INC.
quency stimulation of afferent fibers (Molinari et al.,
1996). However, the possible mechanisms of this deficit
in LTP expression havenot yet been characterized. It is
now generally accepted that within the hippocampal
CA1 area, LTP and long-term depression (LTD), an
activity-dependent reduction in the strength of synap-
tic transmission, requires the activation of NMDA
glutamate receptors which gate calcium flux into the
postsynaptic cells (Lynch et al., 1983; Malenka et al.,
1988; Dudek and Bear, 1992). Because NMDA receptor
propertiesareregulatedby intracellular calciumconcen-
tration (Ca)i (Legendre et al., 1993; Vyklicky, 1993;
Medina et al., 1994, 1995), possible alterations of these
receptors related to disturbance of (Ca)imight partici-
pate in the deficit of neuronal plasticity in the CaBP-
In the present study, we addressed the status of the
glutamatergic synaptic transmission in the CaBP-
deficient mice and determined to what extent the
synaptic plasticity was impaired in theseanimals.
MAT E R IAL S AND ME T HODS
Experiments were conducted in 3–9-month-old wild-
type (n ? 48) and homozygous (n ? 35) male mice.
Generation of the CaBP-deficient mice using an anti-
sensemethodology has been previously described(Moli-
nari et al., 1996). One control or one transgenic mouse
was studied daily using extracellular recordings in the
ex vivohippocampal slice preparation. The animal was
anesthetized with halothane and decapitated. The hip-
pocampus was quickly removed and placed in a cold
medium. Slices (400 µm thick) from both thedorsal and
ventral part of the hippocampus were cut and placed in
a holding chamber for at least 1 h. A single slice was
then transferred totherecording chamber whereit was
held between two nylon nets. The slice was then
submerged beneath a continuously superfusing me-
dium at room temperature and pre-gassed with 95% O2
and 5% CO2. The composition of the medium was (in
mM): NaCl, 119; K Cl, 3.5; MgSO4, 1.5; CaCl2, 3;
NaHCO3, 26.2; NaH2PO4, 1.0; glucose, 11.
Hippocampal slices were fixed by immersion in 4%
paraformaldehyde in 0.1 M phosphate buffer (pH 7.4)
for 4 h at 4°C and cryoprotected at 4°C in 30% buffered
sucrose. Frontal serial sections were cut at 40 µm on a
freezing microtome, collectedin 0.1 M phosphatebuffer,
and processed for immunohistochemistry. They were
first incubated in normal serum (3% in 0.1 M PBS, 0.3%
Triton X-100) for 30 min and then in the primary
antiserum at 4°C overnight. This antiserum was a
rabbit anti-bovine cerebellar calcium-binding protein
(BaimbridgeandMiller, 1982) andwas usedat 1/40,000.
After washing, the sections were incubated in a second
antisera at room temperature for 1 h. Then the avidin-
biotin-peroxidase technique, with chemicals purchased
from Vector Laboratories (Vectastain; Burlingame, CA)
was used to reveal the antibodies raised against CaBP
was made using a naphthol reaction following the
technical procedure detailed by Menetrey et al. (1992).
Extracellular recordings were obtained from the api-
cal dendritic layer of the CA1 field using glass micropi-
pettes filled with 2 M NaCl and having resistances of
2–6 M?. Field EPSPs (fEPSPs) were evoked every 15
sec by electrical stimulation of the stratum radiatum.
Stimuli (100 µs duration) wereapplied between thetwo
poles of a bipolar electrode, one pole inserted into the
sliceand theother in thefluid just abovetheslice.
Input/output (I/O) curves were first constructed in
control medium to assess the responsiveness of the
non-NMDA glutamate receptor subtype to electrical
stimulation in both groups of mice. The slope of three
averaged fEPSPs were measured for different intensi-
ties of stimulation (from25–125 µA) andplottedagainst
theslopeof thepresynaptic fiber volley (PFV). Sensitiv-
ity of these fEPSPs to the non-NMDA receptor antago-
nist 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX, 10
µM) was assessed. In another set of experiments, I/O
curves were constructed in a magnesium-free medium
alsocontaining CNQX (10 µM) and theGABAAreceptor
antagonist bicuculline (50 µM). In this case, a knife cut
separating CA3 and CA1 was introduced toprevent the
propagation of epileptiform discharges. In theseexperi-
ments, the slope of three averaged fEPSPs was deter-
mined for stimulus intensities from 25–125 µA and
plotted against the PFV slope in both groups of ani-
mals. These responses were characterized as mediated
by NMDA receptor activation by their sensitivity to
bath application of the selective antagonist 2-amino-5-
phosphonovalerateacid 2-APV (30–50 µM).
The paired-pulse facilitation (PPF) of synaptic trans-
mission was induced by paired pulses (40–75 µA; 100
µs) and monitored across a range of interstimulus
intervals (ISI) from 30 to500 ms. The PPF was quanti-
fied by normalizing the amplitude of the second re-
sponse by the amplitude of the first one and plotted
In order toinvestigate short- and long-term synaptic
plasticity, test stimuli were adjusted to get a baseline
fEPSP of 0.3 mV.s-1slope. Three successive fEPSPs
were then averaged, the slope determined, and plotted
against timeon-lineusing theAcquis 1software(CNRS,
France). Control traces were recorded for 15–20 min
Short-term potentiation (STP) of synaptic transmis-
sion was induced by a high-frequency train (30 Hz for
0.5 sec at twice test intensity) and testing with single
pulses was resumed for at least 30 min. For LTP, the
conditioning stimulation consisted of two high-fre-
quency trains (100 Hz for 1 sec at twice or 125% test
SYNAPTIC PLASTICITY AND CALBINDIN DEFICIENCY
intensity) separated by 20-sec intervals; testing with
single pulses was resumed for 90 min to determine the
level of stable LTP. Potentiation was defined as an
increase in the fEPSP slope of at least 25%. Topharma-
cologically characterize the mechanisms of LTP, the
conditioning stimulation was induced in wild-type and
CaBP-deficient mice in the presence of the NMDA
receptor antagonist 2-APV (40–60 µM).
For LTD, the GABAAreceptor antagonist bicuculline
(50 µM) was addedtothesuperfusing medium through-
out the recordings to facilitate development of the
depression (Wagner and Alger, 1995). Because bicucul-
the intensity of stimulation was monitored to yield a
response similar to predrug control. The baseline was
recorded for 15 min and LTD was induced by a low-
frequency stimulation (LFS, 900 pulses, 1 Hz at test
intensity). Testing with singlepulses was then resumed
for 40 to45 min todeterminethelevel of stableLTD.
All results are expressed as the mean ? SEM. The
degree of statistical significance was calculated using
multivariate analysis of variance (ANOVA) in order to
take into account the correlation inherent in repeated
measures data. Differences were considered significant
when P ? 0.05.
R E SULT S
Hippocampal immunoreactivity of CaBP in
wild-type and homozygous mice
As shown in Figure 1A, the hippocampal CaBP
immunoreactivity displayed a characteristic laminar
pattern in wild-type mice. While the immunoreactivity
was particularly strong in the dentate gyrus, the high-
est density was found in the CA1/CA2 pyramidal layer
when considering the hippocampus proper. At this
level, the soma and apical dendrites of the pyramidal
cells were intensively immunoreactive by contrast to
theCA3 area whereCaBP immunoreactivity was found
in mossy fibers underlying unstained pyramidal cell
bodies. In addition, CaBP was also observed in a
moderate number of interneurons localized in stratum
radiatum and lacunosum-moleculare of all hippo-
campal subfields. In the homozygous mice, CaBP ex-
pression was clearly reduced (Fig. 1B). This was
particularly pronounced in CA1 area where the immu-
noreactivity of the pyramidal cells as well as the
interneurons was nearly absent. Interestingly, the im-
munoreactivity appeared poorly affected within the
CA2 area while, although significantly decreased, a
slight CaBP immunoreactivity persisted in the dentate
F ield properties
As illustrated in Figure 2A, the stimulation of the
stratum radiatum induced a PFV followed by a fEPSP,
both increasing with stimulus intensity. I/O curves of
PFV and fEPSP slopes were constructed in control
medium in wild-type (27 slices in 12 animals) and
transgenic mice (25 slices in 13 animals). The fEPSPs
were totally suppressed by CNQX (10 µM, Fig. 2A),
thus indicating the activation of the non-NMDA sub-
and CaBP-deficient mice. Bright-field microphotograph of frontal
sections through thehippocampal formation in a slicefrom a wild-type
(A) anda CaBP-deficient mouse(B) showing thedecreaseof CaBP-like
immunoreactivity in the transgenic animal (scale bar ? 500 µM).
DG ? dentategyrus.
CaBP immunoreactivity in the hippocampus of wild-type
A. J OUVENCEAU ET AL.
type of glutamate receptors. Comparison of the I/O
curves revealed that in CaBP deficient mice, PFVs were
not significantly altered (Fig. 2B). However, plotting
thefEPSP slopeagainst thePFV slope(Fig. 2C) showed
a significant increase of non-NMDA receptor-mediated
responses in the absence of CaBP [F(1,28) ? 6.46, P ?
In another set of experiments, I/O curves were con-
structed in slices perfused with a magnesium-free
medium and where CNQX and bicuculline was added
20 min before the recordings. These experiments were
performed in 14 slices from six wild-type and 13 slices
from six transgenic mice. In these conditions where
CA3 area was severed from CA1 (see Materials and
Methods), the electrical stimulation of the glutamater-
gic afferents induced a fEPSP which differed from that
recorded in control medium by its slow onset, its
prolonged duration, and its sensitivity to2-APV (30–50
µM, Fig. 3A). No significant differences in the PFV
slope were found between groups [F(1,20) ? 1.48, P ?
0.24]. On thecontrary, plotting thefEPSP slopeagainst
PFV slope (Fig. 3C) revealed that NMDA receptor-
mediated fEPSPs were significantly weaker in CaBP
deficient mice[U’ ? 90, P ? 0.05, Mann Whitney].
In view of the results described above and consider-
ing the key role of glutamate receptors in neuronal
plasticity (see Malenka et al., 1992), one would predict
some alterations in synaptic plasticity of CaBP-
We first investigated the paired-pulse facilitation
(PPF) of synaptic transmission in wild-type (16 slices,
eight animals) and CaBP-deficient mice (20 slices, nine
animals). PPF is generally considered as a model of
neuronal plasticity with a presynaptic origin (Creager
et al., 1980; Hess et al., 1987). Moreover, paired-pulse
stimulation paradigm is also used to assess possible
alteration ofthepresynapticcalcium-dependent mecha-
nisms which control glutamate release. As illustrated
in Figure4A, thesmaller theISI, thegreater thesecond
response pulse. For instance, 20.6 ? 3.5% increase of
thesecond responseamplitudewas obtained for 200 ms
type and CaBP-deficient mice. A: Examples of PFVs and extracellular
fEPSPs induced in control medium by stimulation of the Schaffer
collaterals with increasing current intensities in wild-type and CaBP-
Non-NMDA receptor-mediated synaptic responses in wild-
deficient micebefore(left) andafter bath application of thenon-NMDA
receptor antagonist CNQX (10 µM, right). B: Plot of the mean PFV
slope against the stimulus intensity. C: Plot of the mean fEPSP slope
against thePFV slopein wild-typeand CaBP-deficient mice.
SYNAPTIC PLASTICITY AND CALBINDIN DEFICIENCY
ISI vs. 74.2 ? 7.8% for 30 ms ISI. Comparison of the
magnitude of the PPF between wild-type and CaBP-
deficient mice did not reveal significant differences at
any ISI (Fig. 4B).
We then focused on glutamate receptor-dependent
synaptic plasticity. STP of excitatory transmission was
first investigated in wild-type(10 slices, seven animals)
and CaBP-deficient mice(12 slices, eight animals). Test
stimuli wereadjustedtoget a baselineslopeof fEPSP of
0.3 mV.s-1(0.31 ? 0.01 V.s-1in wild-type vs. 0.29 ? 0.02
mV.s-1in transgenic mice) to obtain a level where the
intensity of stimulation was not significantly different
in the twogroups of animals considering the I/O curve.
As shown in Figure 5A, a single 30 Hz conditioning
stimulation (see Materials and Methods) resulted in a
rapidincreasein fEPSP slope, which then progressively
decayed to baseline over 30 min in wild-type mice.
When a second conditioning stimulation was applied in
the presence of 2-APV (40–60 µM), no potentiation of
the synaptic transmission was observed (Fig. 5A),
indicating that STP was induced through activation of
NMDA receptors. In CaBP-deficient mice, a similar
fEPSP potentiation was recorded in response to the
conditioning stimulation (Fig. 5A). Nosignificant differ-
encein themagnitude[F(1,20) ? 0.239, P ? 0.63] or the
timecourse[F(60,20) ? 0.805, P ? 0.85] of theSTP was
found between thetwogroups of animals.
In order to investigate possible effects of CaBP-
deficiency on other forms of synaptic plasticity, a LTP
paradigm was first induced using two successive 100
Hz conditioning stimulation (see Materials and Meth-
ods). Here again, test stimuli were adjusted to get
baselineslopeof fEPSP of 0.3 mV.s-1(0.29 ? 0.02 mV.s-1
in wild-type vs. 0.28 ? 0.02 mV.s-1in transgenic mice).
LTP of the synaptic transmission was observed in 17
slices among 21 recorded in wild-type mice, whereas it
could be induced in only five slices among the 20 in
CaBP-deficient mice. As illustrated in Figure 5B, a
robust posttetanic potentiation (PTP) developed imme-
diately after the conditioning stimulation, which then
decayed in both groups of animals during the first 20
min.After this decay, thepotentiation then stabilizedat
and CaBP-deficient mice. A: Examples of PFVs and extracellular
fEPSPs inducedby stimulation of theSchaffer collaterals with increas-
ing current intensities in a magnesium-freemedium in thepresenceof
CNQX (10 µM) andbicuculline(50 µM). Left andright: traces recorded
NMDA receptor-mediated synaptic responses in wild-type
in wild-typeandCaBP-deficient micebeforeandafter bath application
of the NMDA receptor antagonist 2-APV (30 µM), respectively. B: Plot
of the mean PFV slope against the stimulus intensity. C: Plot of the
mean fEPSP slope against the PFV slope in wild-type and CaBP-
A. J OUVENCEAU ET AL.
around 180% of thebaselineslopein wild-typemice. By
contrast, it still decreased in CaBP-deficient mice so
that 65 min posttetanus, the fEPSP slope was not
significantly different from that of the baseline (Fig.
5B). Consequently, 90 min after theconditioning stimu-
lation a robust LTP was still observed in wild-typemice
(mean fEPSP slope: 180.9 ? 13.8% of the average slope
before tetanic stimulation) but not in CaBP-deficient
mice (mean fEPSP slope: 100.9 ? 4.72%). Statistical
analysis of the total responses recorded in the two
groups of animals (including slices with and without
persistent LTP) indicated that both the magnitude
[F(1,35) ? 20.98, P ? 0.001] and the time course
[F(109,35) ? 1.48, P ? 0.001] were significantly differ-
In addition, the conditioning stimulation was in-
duced in slices from wild-type and CaBP-deficient mice
pretreated for 20 min with 2-APV (40–60 µM). In all
cases, a robust PTP occurred after the stimulation
which only persistedfor about 1 min in both strains (not
Finally, LTD was studied in eight slices from three
wild-type and in 11 slices from 4 CaBP-deficient mice.
In these experiments, test stimuli were adjusted to
get a fEPSP baseline of 0.35 mV.s-1(0.34 ? 0.002
mV.s-1in wild-typevs. 0.36 ? 0.004 mV.s-1in transgenic
mice). A robust LTD was achieved in five out of eight
slices in wild-type and in 6 out of 11 in CaBP-deficient
mice. Forty-five minutes after LTD induction, the aver-
age fEPSP slope was 78 ? 7% of the average slope
before LFS in wild-type and 85 ? 3.5% in CaBP-
deficient mice (Fig. 5C). Statistical analysis indicated
that both the magnitude [F(1,8) ? 1.3, P ? 0.29] and
the time course [F(179,8) ? 0.79, P ? 0.98] of LTD were
not significantly different in wild-type and transgenic
PPF of fEPSP induced by two successive stimuli (arrows) with two
different interstimulus intervals (250 ms, a, and 70 ms, b) in a
wild-type mouse. B: Comparison of the averaged PPF of the fEPSP
amplitudein wild-type(n ? 16) and CaBP-deficient mice(n ? 20).
PPF in wild-type and CaBP-deficient mice. A: Examples of
mice. A: STP expressed as the percent change in fEPSP slope versus
time in slices obtained from wild-type (open squares) and CaBP-
deficient mice (filled circles) before and after 30 Hz conditioning
stimulation of the stratum radiatum (arrow). Note the absence of
significant differences in STP magnitude between the groups at any
time after stimulation. STP was prevented by application of the
NMDA receptor antagonist 2-APV (40 µM). B: LTP expressed as the
percent change in fEPSP slope vs. time in slices obtained from
wild-type(open squares) andCaBP-deficient mice(filledcircles) before
and after 100 Hz conditioning stimulation. Note that the transgenic
mice failed to maintain LTP of synaptic transmission normally
occurring in wild-type mice. C: LTD expressed as the percent change
in fEPSP slope vs. time in slices obtained from wild-type (open
squares) and CaBP-deficient mice (filled circles) before and after
low-frequency conditioning stimulation (900 pulses at 1 Hz) stimula-
tion. Notetheabsenceof significant differences in LTD magnitudeand
Synaptic potentiation in wild-type and CaBP-deficient
SYNAPTIC PLASTICITY AND CALBINDIN DEFICIENCY
The results of the present study have been obtained
from CaBP-deficient mice generated by the insertion of
theantisenseCaBP geneplacedunder thecontrol of the
L neurofilament gene promotor (Molinari et al., 1996).
In this way, theactivity of thetransgenecould interfere
with the normal gene, thus reducing the mRNA and
protein expressions. This antisense methodology has
been successfully used to reduce the expression of the
type II glucocorticoid receptor in mice brain (Pepin et
al., 1992; see Weiss et al., 1997). In the present experi-
ments, CaBP immunoreactivity is localized within the
pyramidal cell layer, in a population of interneurons of
the CA1 and CA3 hippocampal areas, as well as in the
granule cells of the dentate gyrus. These observations
further increase the knowledge obtained in species
including humans (Celio, 1990; Seress et al., 1992). The
immunoreactivity of the protein was dramatically re-
duced in the hippocampus of the antisense transgenic
mice when compared to labeling in wild-type animals.
Thedecreasereached particularly low levels in theCA1
area where the electrophysiological recordings were
performed. This reduction in protein immunoreactivity
fits with the absence of mRNA for CaBP previously
observed with in situ hybridization (Molinari et al.,
1996). Interestingly, experiments in progress suggest
that this decrease in CaBP immunoreactivity is not
compensated by upregulation of other calcium binding
proteins of related structure (P.C. Emson, personal
communication). Complete loss of CaBP expression has
recently been obtained in cerebellar Purkinje cells of
CaBP null mutant mice (Airaksinen et al., 1997). In
this model, too, theabsenceof theprotein content is not
compensated by an increase in parvalbumin content,
another calcium-binding protein coexpressed in these
Our electrophysiological investigations in the hippo-
campal CA1 have shown that the glutamatergic neuro-
transmission is altered in the absence of CaBP. NMDA
receptor-mediated responses were decreased and non-
NMDA receptor-mediated responses increased. It was
previously reported that enhancing CaBP expression in
cultured hippocampal pyramidal neurons with a repli-
cation-defective adenovirus containing the CaBP gene
did not significantly alter the magnitude of glutamate
transmission (Chard et al., 1995). However, these au-
thors reported a marked tendency of non-NMDA recep-
tor-mediated excitatory postsynaptic currents (EPSCs)
tobedecreasedandof NMDA receptor-mediatedEPSCs
tobeincreased(seeTable1 in Chardet al., 1995). These
data match the increased and decreased amplitude of
these respective glutamate receptor-mediated synaptic
responses recorded in CaBP-deficient mice.
Because of the opposite effects of CaBP deficiency on
glutamatergic synaptic responses depending on recep-
tor subtypes, an alteration in glutamate release seems
unlikely to explain these alterations. In addition, the
amplitudeof thepresynaptic fiber volley did not signifi-
cantly differ between wild-type and CaBP-deficient
mice, indicating that the proportion of afferent fibers
recruited for a given stimulation intensity was compa-
rable. In addition, it appeared that the magnitude of
the PPF was not affected in the CaBP-deficient mice.
This suggests that the calcium-dependent mechanisms
controlling glutamate release are not impaired in these
animals. Taken together, these results donot argue for
a significant presynaptic contribution in the impair-
ment of glutamatergic synaptic responses in response
toCaBP deficiency.Several possiblepostsynapticmecha-
nisms must therefore be considered, such as an alter-
ation of receptor regulation, including modifications in
density and/or kinetic properties of the receptors lo-
cated on pyramidal cells.
Evidence indicates the involvement of phosphoryla-
tion/dephosphorylation processes in the regulation of
cell surface ionotropic receptor binding sites (Shaw and
Lanius, 1992; see Pasqualotto and Shaw, 1996, for a
review).Sincetheactivity ofprotein kinasesor phospha-
tases is calcium dependent, a greater stimulus-induced
increase in intracellular calcium concentration (Ca)iin
CaBP-deficient mice could trigger changes in gluta-
mate receptor binding sites. This possible mechanism
has to be considered since the depolarization-induced
increase in (Ca)i was reported to be higher in the
absence of CaBP (Airaksinen et al., 1997). In order to
definitively address the question of receptor density in
the transgenic mice, autoradiographic binding studies
arecurrently under investigation.
Changes in the magnitude of phosphorylation-
dephosphorylation processes in the absence of CaBP
may alsoaffect the receptor-mediated currents through
alterations in single channel properties such as open
probability, mean open time, or desensitization rate
(see references in Raymond et al., 1993; Wang and
Salter, 1994). Moreover, considering the NMDA sub-
type of glutamate receptors, these changes might also
occur without the activation of protein kinases or
phosphatases. In fact, it was recently shown that an
receptor-channels in hippocampal and cerebellar neu-
rons (Legendre et al., 1993; Medina et al., 1994, 1995;
Rosenmund et al., 1995). This response did not appear
to be mediated by phosphorylation-dephosphorylation
processes (Vyklicky, 1993) and correlations were found
between theincreasein (Ca)iand thedegreeof calcium-
induced inactivation of NMDA-activated currents (Me-
dina et al., 1996).As pointedout previously, thedepolar-
ization-induced increase in (Ca)i was found to be of
higher amplitude in neurons lacking CaBP expression
(Airaksinen et al., 1997). Therefore, it could be hypoth-
esizedthat an enhancedinactivation ofNMDA receptor-
channels might occur in CaBP-deficient mice, leading
to the weaker NMDA receptor-mediated synaptic re-
A. J OUVENCEAU ET AL.
sponses. Comparative patch-clamp experiments on iso-
lated NMDA receptor-mediated currents in wild-type
and homozygous mice might be carried out to address
this question. According to this hypothesis, it was
recently reportedin kidney-293 cells cotransfectedwith
cDNAs for CaBP and NMDA receptor subunits that the
calcium binding protein inhibits the onset of calcium-
dependent changes in NMDA-receptor desensitization
(Priceet al., 1997).
In view of the key role played by glutamate receptors
and particularly the NMDA receptor subtype in neuro-
nal plasticity (see Malenka et al., 1992), one would
predict some alterations in CaBP-deficient mice. Our
results illustrate a specific alteration in the mecha-
nisms of LTP, in agreement with our previous observa-
tions (Molinari et al., 1996). In contrast, STP and LTD
remained unaffected. It has been postulated that a
large influx of calcium through NMDA receptors into
the postsynaptic cells is required for LTP induction,
while a lower threshold is necessary to induce LTD or
STP (Malenka et al., 1992; Lisman, 1989; see Stanton,
1996). Consequently, we may hypothesize that the loss
of LTP in CaBP-deficient mice could be due to the
alteration of NMDA receptor-mediated synaptic re-
sponses recorded in these animals and the consecutive
decreasein Ca entry. STP and LTD would beunaffected
because the increase of intracellular calcium necessary
toinduceSTP and LTD is below thethreshold toinduce
LTP and might bereached in calbindin-deficient mice.
LTP requires the activation of multiple enzyme such
as protein kinase A, protein kinase C, the calcium-
calmodulin kinase II (CaMKII) (Malinow et al., 1988;
Malenka et al., 1989; Fedorov et al., 1995), the tyrosine
kinases (Raymond et al., 1993) and for the later steps,
the synthesis of new proteins (Duffy et al., 1981; Fazeli
et al., 1993). We may suggest that the alteration of
these enzymes might also contribute to the LTP deficit
in theCaBP-deficient mice, since, for instance, calmodu-
lin has been recently found to mediate calcium-
dependent inactivation of NMDA receptors (Hisatsune
et al., 1997; Zhang et al., 1998).
In summary, our results suggest that impairment of
LTP in CaBP-deficient mice mainly results from alter-
ations affecting NMDA receptors. An LTP impairment
was recently reported in the dentate gyrus of calretinin
null mutant mice in relation to some disregulations of
the GABAergic tone (Schurmans et al., 1997). Such a
mechanism is unlikely in CaBP-deficient mice because
no significant alterations in GABA receptor-mediated
inhibitory postsynaptic potentials were found in these
animals (Dutar et al., 1996). Nevertheless, these re-
sults indicate that the mechanisms by which the cal-
cium binding proteins interact with synaptic plasticity
may be different, considering their nature and their
neuronal localization. Supporting this idea, the overex-
pression of another calcium-binding protein, S100?, in
astrocytes alsoaltered hippocampal synaptic plasticity
(Gerlai et al., 1995).
ACK NOWL E DGME NT S
Theauthors thank F. J azat-Poindessous for technical
assistance, R. Rambur for photographic work, and C.
Vaillend for expert advicein statistical analysis.
R E F E R E NCE S
Airaksinen MS, Eilers J , Garaschuck O, Thoenen H, Konnerth A.
1997. Ataxia and altered dendritic calcium signaling in mice carry-
ing a targeted null mutation of the calbindin D28kgene. Proc Natl
Acad Sci USA 94:1488–1493.
Baimbridge KG, Miller J J . 1982. Immunohistochemical localization of
calcium-binding protein in the cerebellum, hippocampal formation
and olfactory bulb of therat. Brain Res 245:223–229.
BaimbridgeKG, CelioMR, Rogers J H. 1992. Calcium-binding proteins
in thenervous system. TINS 15:303–308.
Baudry M, Massicotte G. 1992. Physiological and pharmacological
relationships between long-term potentiation and mammalian
memory. Concepts Neurosci 3:79–98.
Bliss TVP, Collingridge GL. 1993. A synaptic model of memory:
long-term potentiation in thehippocampus. Nature361:31–39.
CelioMR. 1990. Calbindin D–28k and parvalbumin in the rat nervous
Chard PS, J ordan J , Marcuccilli CJ , Miller RJ , Leiden J M, Roos RP,
Ghadge GD. 1995. Regulation of excitatory transmission at hippo-
campal synapses by calbindin D28k. ProcNatlAcadSci USA 92:5144–
Creager R, Dunwiddie T, Lynch G. 1980. Paired-pulse and frequency
facilitation in the CA1 region of the in vitro rat hippocampus. J
Physiol (Lond) 299:409–424.
Dudek SM, Bear MF. 1992. Homosynaptic long-term depression in
area CA1 of hippocampus and effects of N-methyl-D-aspartate
receptor blockade. Proc Natl Acad Sci USA 89:4363–4367.
Duffy C, Teyler TJ , Shashoua VE. 1981. Long-term potentiation in the
hippocampal slice: evidencefor stimulatedsecretion ofnewly synthe-
sized proteins. Science212:1148–1151.
Dutar P, J ouvenceau A, Billard J M, Lamour Y, Ferrari S. 1996.
Calcium-dependent properties of hippocampal pyramidal neurons
in transgenic mice with reduced calbindin D28k expression. Soc
Neurosci Abstr 22:342.
Fazeli MS, Corbert J , Dunn MJ , Dolphin AC, Bliss TV. 1993. Changes
in protein synthesis accompanying long-term potentiation in the
dentategyrus in vivo. J Neurosci 13:1346–1353.
Fedorov NB, Pasinelli P, Oestreicher AB, DeGraan PNE, Reyman KG.
1995. Antibodies to postsynaptic PKC substrate neurogranin pre-
vent long-term potentiation in hippocampal CA1 neurons. Eur J
Gerlai R, Wojtowicz J M, Marks A, Roder J . 1995. Overexpression of a
calcium-binding protein, S100?, in astrocytes alters synapticplastic-
ity and impairs spatial learning in transgenic mice. Learn Mem
Hess G, Kuhnt U, Voronin LL. 1987. Quantal analysis of paired-pulse
facilitation in guinea pig hippocampal slices. Neurosci Lett 77:187–
HisatsuneC, Umemory H, InoueT, Michikawa T, Kohda K, Mikoshiba
K, Yamamoto T. 1997. Phosphorilation-dependent regulation of
N-methyl-D-aspartate receptors by calmodulin. J Biol Chem 272:
IacopinoAM, QuinteroEM, Miller EK. 1994. Calbindin-D28k: a poten-
tial neuroprotectiveprotein. Neurodegeneration 3:1–20.
Kostyuk PG, Verkhratsky AN. 1995. Calcium signalling in thenervous
system. Chichester, UK: J ohn Wiley & Sons.
Legendre P, Rosenmund C, Westbrook GL. 1993. Inactivation of
NMDA channels in cultured hippocampal neurons by intracellular
calcium. J Neurosci 13:674–684.
Lisman J A. 1989. A mechanism for the Hebb and the anti-Hebb
processes underlying memory and learning. Proc Natl Acad Sci USA
Lynch G, Larson J , Kelso S, Barrionuevo G, Schottler F. 1983.
Intracellular injections of EGTA block induction of hippocampal
long-term potentiation. Nature305:719–721.
Malenka RC, Kauer J A, Zucker RS, Nicoll RA. 1988. Postsynaptic
calcium is sufficient for potentiation of hippocampal synaptic trans-
SYNAPTIC PLASTICITY AND CALBINDIN DEFICIENCY
Malenka RC, Kauer J A, Perkel DA, Mauk MD, Kelly PT, Nicoll RA,
Waxham MN. 1989. An essential role for postsynaptic calmodulin
and protein kinase activity in long term potentiation. Nature
Malenka RC, Lancaster B, Zucker RS. 1992. Temporal limits on the
rise in postsynaptic calcium required for the induction of long-term
potentiation. Neuron 9:121–128.
Malinow R, Madison DV, Tsien RW. 1988. Persistent protein kinase
activity underlying long-term potentiation. Nature335:820–824.
Medina I, Filippova N, Barbin G, Ben-Ari Y, Bregestovski P. 1994.
Kainate-induced inactivation of NMDA currents via an elevation of
intracellular Ca2?in hippocampal neurons. J Neurophysiol 72:456–
Medina I, Filippova N, Chartron G, Rougeole S, Ben-Ari Y, Khrest-
chatisky M, Bregestovski P. 1995. Calcium-dependent inactivation
ofheteromericNMDA receptor-channels expressedin human embry-
onic kidney cells. J Physiol (Lond) 482:567–573.
Medina I, Filippova N, Bakhramov A, Bregestovski P. 1996. Calcium-
induced inactivation of NMDA receptor-channels evolves indepen-
dently of run-down in cultured rat brain neurones. J Physiol (Lond)
Menetrey D, De Pommery J , Baimbridge KG, Thomasset M. 1992.
Calbindin-D–28k (CaBP28k)-likeimmunoreactivity in ascending pro-
jections: trigeminal nucleus caudalis and dorsal vagal complex
projections. Eur J Neurosci 4:61–69.
Molinari S, Battini R, Ferrari S, Pozzi L, Killcross AS, Robbins TW,
J ouvenceau A, Billard J M, Dutar P, Lamour Y, Baker WA, Cox H,
Emson PC. 1996. Deficits in memory and hippocampal long-term
potentiation in mice with reduced calbindin D28kexpression. Proc
Natl Acad Sci USA 93:8028–8033.
PasqualottoBA, Shaw CA. 1996. Regulation of ionotropic receptors by
protein phosphorylation. Biochem Pharmacol 51:1417–1425.
Pepin MC, Pothier F, Barden N. 1992. Impaired type II glucocorticoid-
receptor function in mice bearing antisense RNA transgene. Nature
Price CJ , Rintoul GL, Baimbridge KG, Raymond LA. 1997. Calbindin-
D28k inhibits the onset of Ca2?-dependent changes in NMDA-
receptor desensitization. Soc Neurosci Abstr 23:941.
Raymond L, Blackstone CD, Huganir RL. 1993. Phosphorylation of
amino acid neurotransmitter receptors in synaptic plasticity. TINS
Rosenmund C, Feltz A, Westbrook G. 1995. Calcium-dependent inacti-
vation of synaptic NMDA receptors in hippocampal neurons. J
Schurmans S, Schiffmann SN, Gurden H, Lemaire M, Lipp HP,
Schwam V, Pochet R, ImperatoA, Bo ¨hme GA, Parmentier M. 1997.
Impaired LTP induction in dentate gyrus of calretinin deficient
mice. Proc Natl Acad Sci USA 94:10415–10420.
Seress L, Gulyas AI, Freund TF. 1992. Pyramidal neurons are
immunoreactive for calbindin D28kin the CA1 subfield of the human
hippocampus. Neurosci Lett 138:257–260.
Shaw CA, Lanius RA. 1992. Reversible kinase and phosphatase
regulation of brain amino acid receptors in postnatal development.
Dev Brain Res 70:153–161.
Silva AJ , Stevens CF, Tonegawa S, Wang Y. 1992. Deficient hippocam-
pal long-term potentiation in alpha-calcium calmodulin kinase II
mutant mice. Science257:201–205.
Stanton PK. 1996. LTD, LTP, and the sliding threshold for long-term
synaptic plasticity. Hippocampus 6:35–42.
Vyklicky L J r. 1993. Calcium-mediated modulation of N-methyl-D-
aspartate(NMDA) responses in cultured rat hippocampal neurones.
J Physiol (Lond) 470:575–600.
Wagner J J , Alger BE. 1995. GABAergic and developmental influences
on homosynaptic LTD and depotentiation in rat hippocampus. J
Wang LY, Salter MW. 1994. Regulation of NMDA receptors by tyrosine
kinases and phosphatases. Nature369:233–235.
Weiss B, Davidkova G, Zhand SP. 1997. Antisense strategies in
neurobiology. Neurochem Int 31:321–348.
Zhang S, Ehlers MD, Bernhardt J P, Su CT, Huganir RL. 1998.
Calmodulin mediates calcium-dependent inactivation of N-methyl-
D-aspartatereceptors. Neuron 21:443–453.
A. J OUVENCEAU ET AL.