Cell, Vol. 105, 521–532, May 18, 2001, Copyright 2001 by Cell Press
GABA Itself Promotes the Developmental Switch
of Neuronal GABAergic Responses
from Excitation to Inhibition
tion). Recent work has shown that the developmental
transformation of GABAergic synaptic transmission
from depolarizing to hyperpolarizing is due to a shift in
EGABAtoward a more hyperpolarized potential, which is
likely the result of an ontogenetic decrease in the intra-
cellular Cl?concentration ([Cl?]i; Cherubini et al., 1990;
Luhmann and Prince, 1991; Chen et al., 1996; Owens et
al., 1996). Indeed, changes in the mRNA level for the
K?-coupled Cl?transporter KCC2 have been shown to
correlate with the modification of GABAergic transmis-
sion (Lu et al., 1999; Rivera et al., 1999; Vu et al., 2000).
KCC2 increases the rate of Cl?extrusion, thus leading
to a reduction in [Cl?]iand a consequent shift in EGABA
toward more hyperpolarized potentials (Jarolimek et al.,
1999; Kakazu et al., 1999; Rivera et al., 1999).
This conversion of GABAergic transmission from de-
polarizing to hyperpolarizing is also accompanied by a
change in GABA-mediated biochemical signaling. Only
during this early developmental period, depolarizing
GABAergic potentials activate voltage-dependent Ca2?
channels (VDCCs) and elevate [Ca2?]i (Connor et al.,
1987; Yuste and Katz, 1991; Wang et al., 1994). Such
GABA-induced elevation of [Ca2?]i is likely to play a
critical role in the maturation of the nervous system.
For instance, GABA-mediated increases in [Ca2?]ican
induce BDNF expression (Berninger et al., 1995) and
promote neuronal survival and differentiation (LoTurco
et al., 1995; Marty et al., 1996; Ikeda et al., 1997). GABA-
induced elevation of [Ca2?]i may also be required to
form, stabilize, and strengthen synaptic connections
(Kirsch and Betz, 1998; Caillard et al., 1999; Kneussel
and Betz, 2000).
While the developmental transformation of GABAer-
gic transmission is well documented, little is known
ronal activity is known to increase during development,
we examined in the present study whether synaptic ac-
tivity can regulate the switch of GABAergic transmis-
sion. We found that the change in GABA signaling was
largely prevented by chronic blockade of GABAArecep-
tors, and was accelerated by increased GABA receptor
activation. Changes in the level of KCC2 mRNA tightly
correlated with the observed changes in GABA signal-
ing. In addition, we found that spontaneous GABAergic
activity regulated the activation of voltage-dependent
Ca2?currents. These findings point to GABA as a critical
maturation factor for the switch of the physiological and
biochemical properties of GABA signaling.
Karunesh Ganguly,1,2,4Alejandro F. Schinder,2,4
Scott T. Wong,2,5and Mu-ming Poo1,2,3,5
1Program in Neuroscience
2Section of Neurobiology, Division of Biology
University of California, San Diego
La Jolla, California 92093
GABA is the main inhibitory neurotransmitter in the
synaptic transmission is excitatory and can exert
widespread trophic effects. During the postnatal pe-
riod, GABAergic responses undergo a switch from be-
ing excitatory to inhibitory. Here, we show that the
tors, and accelerated by increased GABAAreceptor
activation. In contrast, blockade of glutamatergic
transmission or action potentials has no effect. Fur-
thermore, GABAergic activity modulated the mRNA
levels of KCC2, a K?-Cl?cotransporter whose expres-
sion correlates with the switch. Finally, we report that
GABA can alter the properties of depolarization-
induced Ca2?influx. Thus, GABA acts as a self-limiting
trophic factor during neural development.
In the adult central nervous system, ?-amino-butyric
acid (GABA) is the primary inhibitory neurotransmitter.
It regulates a neuron’s ability to fire action potentials
either through hyperpolarization of the membrane po-
tential or through shunting of excitatory inputs. During
early postnatal development, however, GABAergic syn-
cellular Ca2?concentration ([Ca2?]i), and even capable
of triggering action potentials (Mueller et al., 1984; Luh-
mann and Prince, 1991; Yuste and Katz, 1991; Reichling
et al., 1994; Wang et al., 1994; Leinekugel et al., 1995;
tal period, in the hippocampus, neocortex, and hypo-
thalamus, as well as other regions of the brain, there is
a switch of the electrophysiological (depolarization to
hyperpolarization) and biochemical (Ca2?-mediated sig-
naling) properties of GABAergic transmission (Mueller
et al., 1984; Ben-Ari et al., 1989; Cherubini et al., 1991;
Luhmann and Prince, 1991; Owens et al., 1996).
The GABAAreceptor channelpredominantly conducts
Cl?ions. Consequently, the nature of GABAergic trans-
mission, excitatory versus inhibitory, is determined pri-
marily by the electrochemical gradient for Cl?, which
depends on the intra- and extracellular concentrations
of Cl?. This electrochemical gradient sets the reversal
potential for GABAergic currents (EGABA; the membrane
Switch of GABAergic Transmission from Depolarizing
To study the change in GABA signaling, we first moni-
tored GABA-induced elevations of [Ca2?]iover develop-
ment. GABA-mediated depolarization was reflected by
an increase in [Ca2?]i. Cultures of hippocampal neurons
were loaded with the Ca2?-sensitive dye Fluo-4 AM and
changes in fluorescence were measured using confocal
4These authors contributed equally to this work.
5Present address: Division of Neurobiology, Department of Molecu-
lar and Cell Biology, University of California, Berkeley, Berkeley
Figure 1. Developmental Changes in GABA-Induced Responses
(A) Pharmacological profile of GABA-induced elevations of [Ca2?]i(BMI ? PTX, 10 and 50 ?M; nimodipine, 10 ?M; thapsigargin ? BHQ, 2 and
10 ?M; and baclofen, 10 ?M). Time course of changes in [Ca2?]i, assessed by Fluo-4 fluorescence intensity. Traces are average fluorescence
intensity (% change shown by the scale) recorded from 30–40 randomly sampled neurons in response to pulses of GABA (open bar, 10 ?M,
15 s). The elevation of [Ca2?]iduring application of thapsigargin ? BHQ was attributed to emptying of intracellular Ca2?stores.
(B) Time course of the GABA switch. The percentage of neurons exhibiting Ca2?elevation in response to GABA is shown for different days
after cell plating. Each point represents mean ? sem (n ? 3 to 21 experiments, 121 in total). Each experiment involved recording from 30–40
neurons. Insets: Sample recordings from 20 randomly selected neurons at the time points indicated by the arrows.
images on the right represent typical [Ca2?]iin neurons before (“baseline”), during (“GABA”), and after (“recovery”) application of a pulse of
GABA (higher intensity represents higher [Ca2?]i). Image field ? 604 ? 604 ?m.
(D) Left: Representative examples of peak GABA-induced currents versus membrane voltages (I-V relationship). The cells were voltage clamped
at ?70 mV and stepped to different potentials (?90 to ?40 mV, 10 mV steps). EGABAwas calculated by fitting the I-V curve to a second-order
polynomial function. Right: Averaged I-V curves (n ? 11 in both cases).
GABA as a Self-Limiting Trophic Factor
Figure 2. Role of Endogenous GABAergic
Transmission in the Developmental Switch
(A) Time course of the switch in GABA signal-
ing for neurons treated chronically with BMI ?
PTX (10 and 50 ?M; n ? 31, 1085 neurons),
KCl (10 mM; n ? 26, 910 neurons), and KCl ?
BMI ? PTX (n ? 14, 490 neurons), compared
to control neurons (n ? 121, 4235 neurons).
In all experiments involving KCl, TTX and
D-APV were included in the medium. Statisti-
cal comparisons were carried out for data
collected for day 8, 10, and 12 neurons.
BMI ? PTX and KCl were significantly differ-
ent from the untreated control and KCl ?
BMI ? PTX (p ? 0.005, single-factor ANOVA
followed by a Bonferroni t test corrected for
multiple comparisons). Insets: I, Representa-
tive recordings of day 14 neurons grown in
results on day 10 neurons grown in the pres-
ence of KCl (10 mM).
(B) Comparison of the effects of chronic
blockade of GABAAand glutamate receptors
at day 13 neurons. Numbers of experiments
for day 7 and day 13 neurons (with or without
chronic treatment with BMI ? PTX). Each
cell. Data points with error bars below denote
mean ? sem (n ? 10 to 11). Data from both
control day 7 and BMI ? PTX day 13 were
significantly different from the control day 13
(p ? 0.001, Kolmogorov-Smirnov test).
to exogenousapplications of GABA(100 ?M, 15s pulse)
with a rapid and reversible increase in [Ca2?]i. As shown
in Figure 1A, these responses were largely blocked by a
combination of GABAAreceptor antagonists (bicuculline
methoiodide, “BMI”, 10 ?M; picrotoxin, “PTX”, 50 ?M)
or by the L-type Ca2?channel antagonist nimodipine
(10 ?M), indicating that elevation of [Ca2?]iis due to
Ca2?influx through L-type Ca2?channels activated by
GABAA-receptor-mediated depolarization. In addition,
application of baclofen, a GABABagonist, had no effect
on [Ca2?]i. Furthermore, depleting intracellular Ca2?
stores with thapsigargin (2 ?M) and 2,5-di-t-butyl-1,4-
benzohydroquinone (BHQ, 10 ?M) did not alter the
GABA-mediated Ca2?transients, indicating that the ele-
vation of [Ca2?]iis a result of Ca2?influx through VDCCs.
The percentage of neurons that responded to GABA
with an elevation of [Ca2?]idecreased with the age of
the culture (Figures 1B and 1C). All neurons displayed
GABA-mediated depolarization at day 4 to 6, while most
neurons exhibited no detectable response by day 13.
Regardless of whether a neuron responded to GABA,
depolarization with a high concentration of extracellular
(Figure 1C), suggesting that the lack of responsiveness
to GABA was due to a decrease in the extent of depolar-
findings from brain slices and cultured neurons (Luh-
mann and Prince, 1991; Chen et al., 1996; Owens et al.,
1996; Wang et al., 1994), that GABA-induced elevations
of [Ca2?]iare developmentally regulated and that cul-
tures of hippocampal neurons provide a useful model
in GABA signaling.
To directly assess the developmental shift in EGABA,
sure the reversal potential of currents elicited by local
application of GABA (10 ?M, 50 ms pulse) in young (day
1D, EGABAwas significantly more hyperpolarized in ma-
ture neurons (?60.7 ? 2.2 mV; mean ? sem, n ? 11)
than in young neurons (?44.5 ? 2.0 mV; n ? 11, p ?
0.001), indicating that the developmental decrease of
EGABAcontributes to the reduction of GABA-induced ele-
vations of [Ca2?]i.
GABA Itself Promotes the Developmental Switch
Based on the hypothesis that the transformation of
GABA signalingdepends onneuronal activity,we exam-
ined the effects of chronic blockade of ionotropic trans-
mitter receptors on the GABA-induced Ca2?transients.
Interestingly, GABAergic, but not glutamatergic activity
was essential for the developmental change (Figures
2A–2C). Chronic blockade of GABAAreceptors with BMI
(10 ?M) and PTX (50 ?M) prevented the transformation
in most neurons (note that similar effects were found
in the presence of BMI alone; Figure 6E). In contrast,
blockade of the N-methyl-D-aspartate (NMDA) sub-
type of glutamate receptors with D-APV (25 ?M), the
non-NMDA receptors with CNQX (15 ?M), or both
(CNQX?APV) did not affect the developmental time
course (Figure 2B). In addition, chronic activation of the
Figure 3. Modulation of KCC2 mRNA Ex-
pression by GABAergic Activity
Levels of KCC2 mRNA assessed by the
RNase protection assay.
(A) Representative experiment displaying a
higher levelof KCC2mRNA in day15 neurons
as compared to day 3 neurons.
(B) Representative experiment showing a de-
crease in KCC2 mRNA expression in day 15
neurons chronically treated with BMI ? PTX,
as compared to untreated controls. The thin
band marked with “*” was due to residual
undigested ?-actin probe, as shown by the
?-actin probe lane alone.
(C) The expression of KCC2 mRNA was in-
creased by chronic depolarization with KCl
(10 mM), and decreased in day 9 neurons
followingthe BMI?PTX treatment.Replicate
samples of results are shown.
(D) Summary of all results on KCC2 mRNA
expression, assayed at day 3, 9, and 12–15.
For each condition, KCC2 mRNA levels were
normalized to that of ?-actin and expressed
as fold induction over the band intensity ob-
served at day 3 in control neurons. “N” ?
“**” denotes p ? 0.005 (t test).
metabotropic glutamate receptor mGluR3 with L-AP4
did not modify the switch (d11; 0.49 ? 0.07, n ? 5, p ?
0.57, t test). Thus, the developmental switch of GABA-
mediated Ca2?signaling is triggered by endogenous
The maturational change in GABA-mediated Ca2?sig-
naling may result directly from a shift in the reversal
whether the ontogenetic modification of EGABAwas also
regulated by GABAergic transmission. As shown in Fig-
ure 2C, chronic blockade of GABAAreceptors prevented
the shift of EGABAtoward hyperpolarized potentials in day
13 neurons. EGABAremained at ?48.4 ? 1.6 mV (n ? 10),
a value that was not significantly different from that of
young (day 7) neurons under control conditions (p ?
0.13, t test). These observations indicate that the devel-
opmental increase in GABAergic activity promotes the
transformation of GABA signaling.
slices (Rivera et al., 1999). Since GABAergic activity
drives this shift in EGABA, we examined whether the onto-
genetic increase in the levels of KCC2 mRNA is similarly
modulated. KCC2 mRNA levels were quantified in day
3 and day 12–15 neurons using an RNase protection
assay and normalized to the levels of ?-actin (Lee and
Costlow, 1987). Consistent with previous findings (Ri-
vera et al., 1999), we observed an ?14-fold increase in
the level of KCC2 mRNA in mature neurons (Figures 3A
and 3D). At day 9 (during the onset of the switch), an
?6 fold increase in KCC2 expression was observed,
although only ?20% of neurons failed to respond to
GABA at this time (Figure 1B). This suggests a nonlinear
relationship between the mRNA level of KCC2 and its
functional effects (see Discussion). Furthermore, we ex-
plored whether chronic blockade of GABAergic activity
with BMI ? PTX altered the level of KCC2 mRNA. As
shown in Figures 3B and 3D, KCC2 expression was
decreased by 68 ? 4% (n ? 5, p ? 0.001) in comparison
to age-matched control cultures (n ? 10). To elevate
GABAergic activity, we chronically depolarized the cul-
tures with 10 mM KCl (see Figures 6C and 6D). We
found that KCC2 message levels in day 9 neurons were
GABA Activates the Expression
of the Cl?Transporter KCC2
The hyperpolarizing shift in EGABAhas been correlated
with an increase in KCC2 expression in hippocampal
GABA as a Self-Limiting Trophic Factor
Figure 4. Regulation of Ca2?Channel Properties during Development
(A1) Representative recordings of changes in [Ca2?]iinduced by depolarization with 6, 8, or 10 mM KCl in day 7 and day 13 neurons. (A2)
Summary of normalized changes in Ca2?fluorescence induced by KCl depolarization in young (day 7; n ? 240 neurons), old (day 13; n ? 260)
untreated neurons, and old neurons chronically treated with BMI ? PTX (n ? 111). (“*” and “**” denote p ? 0.05 and p ? 0.001, respectively,
when compared to the control day 13 treatment, single-factor ANOVA followed by a t test).
(B) GABA receptor blockade prevents the developmental shift in the activation of VDCCs. (B1) Representative voltage-clamp recordings of
Ca2?currents. Inward Ca2?currents (downward deflections) were elicited by step depolarizations (holding potential ? ?80 mV, step ? 10
mV). (B2) Normalized peak current versus voltage for day 7 (n ? 12), day 13 (n ? 9), and day 13 BMI ? PTX (n ? 11) neurons. On average,
Ca2?currents measured between ?60 and ?40 mV are 1- to 2-fold larger in younger neurons than in mature neurons (with p ? 0.05 for ?50
and ?40 mV; n ? 9 for 7 day, and n ? 12 for 13 day). (B3) Mean I-V curves for absolute peak current values (same data as in B2). As expected,
error bars were larger in these nonnormalized curves and were omitted for clarity.
enhanced by 69 ? 6% in comparison to control condi-
tions (Figures 3C and 3D), and most neurons in the KCl-
treated day 9 cultures had lost their responsiveness to
GABA (Figure2A). Takentogether, theseresults support
the notion that the expression of KCC2 can be regulated
by GABA-mediated depolarization.
4A). These measurements of KCl-induced changes in
fluorescence intensity most accurately reflect depolar-
ization-induced changes in “free” [Ca2?]i. Therefore,
these results indicate that while a mild depolarization
can robustly increase the free [Ca2?]iin young neurons,
moderate depolarization is required to appreciably in-
crease the levels of [Ca2?]iin mature neurons. We thus
conclude that the switch in GABA-induced Ca2?signal-
ing involves two independent ontogenetic modifica-
tions: a hyperpolarizing shift in EGABAand a reduction in
the depolarization-induced elevation of free [Ca2?]i.
Changes in free [Ca2?]iare likely to depend on two
factors: (1) the activation profile of voltage-dependent
Ca2?currents (ICa), and (2) the Ca2?buffering properties
of a neuron. We conducted voltage-clamp recordings of
ICafrom young and mature neurons. Step depolarization
evoked fast inward currents that displayed slow and
partial inactivation (Figure 4B1). For young neurons, in-
ward currents activated typically between ?60 and ?50
Developmental Regulation of Depolarization-
Induced Elevations of [Ca2?]i
GABA-induced Ca2?transients depend both on the ex-
tent of membrane depolarization as well as the proper-
fiedduringneuronal development.We measuredchanges
larization (induced by KCl) in young (day 7) and mature
(day 13) neurons. While young cells displayed a robust
increase in [Ca2?]iat all levels of depolarization, mature
Figure 5. Spiking Is Not Required for the De-
Chronic blockade of action potentials with
TTX (2 ?M) did not affect the time course of
the GABA signaling switch. I and II: Typical
recordings of Ca2?levels in neurons chroni-
cally treated with TTX, at the indicated time
points. Inset: Membrane currents recorded
from a day 6 neuron before and after applica-
tion of 2 ?M TTX.
mV, reaching a peak at about ?10 mV. In contrast, ma-
ture neurons displayed a shift in the activation profile
toward more hyperpolarized potentials, suggesting that
the kinetics of ICaare developmentally regulated (Figure
4B2). However, the absolute amplitudes of the whole-
polarized potentials (?70 to ?50 mV) were similar (Fig-
ure 4B3). These results suggest that additional factors
underlie the observation that mild depolarization in-
duces large increases in free [Ca2?]ionly in young neu-
rons (Figure 4A). One possible factor is that the cell
surface of mature neurons is typically ?2-fold larger
than that of young neurons. Thus, the current density at
hyperpolarized potentials is likely to be higher in young
cells. Developmental changes in the Ca2?-buffering
properties of neurons are also likely to contribute
(Schierle et al., 1997; Rosenstein et al., 1998; Boukhad-
daoui et al., 2000).
ence of tetrodotoxin (TTX, 2 ?M), which blocks Na?-
dependent action potentials in these neurons (Figure 5).
Interestingly,blocking Na?spikesdid notaffect thetime
course of the transformation, suggesting that spontane-
(see Figure 6B). Spontaneous mGSCs were first de-
tected (at a low frequency) at day 7 (Figure 6A), and the
frequencyincreased steadilytoreacha plateauofabout
0.5 Hz between days 11 and 12. Remarkably, the time
course in which spontaneous GABAergic activity arises
closely parallels that of the switch in GABA signaling
(compare Figures 5 and 6A). It is noteworthy that al-
though substantial glutamatergic activity was also pres-
ent during this period (Figures 6A and 6Bb), this excit-
atory activity did not contribute to the developmental
switch in GABA signaling (Figure 2B).
It has been reported that chronic blockade of GABAA
receptors may lead to increased activity even when
GABA is depolarizing (Lamsa et al., 2000). Thus, it is
possible that the effects of chronic blockade of GABAA
receptors on the switch are the result of an increase
in overall neuronal activity. To test this possibility, we
compared the effects of chronic blockade of GABAA
receptors alone (BMI) with those observed in the pres-
ence of BMI ? TTX or BMI ? CNQX in day 13 neurons
(Figure 6E). As no significant differences were found
under these conditions (p ? 0.2, ANOVA), the effects
of chronic BMI treatment were not a consequence of
increased neuronal activity.
The notion that spontaneous GABAergic synaptic ac-
with KCl-induced depolarization (in the presence of
TTX), which enhances the probability of transmitter re-
lease, thus increasing the frequency of mGSCs. As
shown in Figures 6C and 6D, KCl (10 mM) induced a
rapid increase in the mGSC frequency (190 ? 28%, n ?
4, p ? 0.05). Furthermore, chronic treatment with KCl
(10 mM) markedly accelerated the time course of the
transformation of GABA signaling, as more than 90%
of all neurons failed to exhibit GABA-induced [Ca2?]i
GABA Can Promote Developmental Changes
in Depolarization-Induced Ca2?Transients
Since GABA itself promoted the shift in EGABA, we exam-
ined whether the properties of depolarization-induced
Ca2?transients are similarly regulated. As shown in Fig-
ure 4A2, chronic blockade of GABAergic transmission
markedly reduced this developmental modification. In
matureneurons chronicallytreatedwith BMI? PTX(day
that was significantly larger than those in the untreated
parallel cultures. Although the ontogenetic change in
the activation profile for ICawas reduced, the absolute
levels of calcium influx were similar to that seen under
control conditions (Figures 4B2–3). Taken together, these
findings suggest that GABA promotes the develop-
mental decrease in depolarization-induced Ca2?tran-
sients, but this decrease is not fully accounted for by
changes in voltage-dependent Ca2?influx.
The Developmental Switch Does Not Require
To test whether spiking is required for the switch of
GABAergicsignaling, neuronswereculturedin thepres-
GABA as a Self-Limiting Trophic Factor
Figure 6. Developmental Changes in Spontaneous Activity
(A) Developmental increase in the frequency of spontaneous GABAergic and glutamatergic currents, which were recorded in the presence of
TTX ? CNQX/APV and TTX ? BMI, respectively. Representative traces of mGSCs are shown (I and II). Scales: 10 pA, 0.5 s.
(Ba) Spontaneous depolarizing GABAergic potentials recorded in current clamp using gramicidin perforated patch in the presence of CNQX.
Resting potentials were approximately ?60 to ?70 mV. Scales: 10 mV, 0.3 s. (Bb) Voltage clamp recordings (?70 mV) of spontaneous
glutamatergic activity from cultures chronically treated with BMI ? PTX. Scales: 100 pA, 0.3 s.
(C) An example of acute effects of 10 mM KCl on the frequency and amplitude of mGSCs. Each point represents a single mGSC.
(D) Changes in mGSC frequency after KCl treatment (the same experiment as shown in [C]). Inset: average mGSC frequency before and after
KCl treatment (n ? 4). (p ? 0.05, paired t test).
(E) Summary of experiments involving chronic blockade of GABAergic activity only (BMI), GABAergic activity and spiking (BMI ? TTX), and
GABAergic and glutamatergic activities (BMI ? CNQX).
(F) Comparison of chronic activation of GABAAreceptors by muscimol (10–50 ?M; n ? 8) versus control conditions (n ? 21). “*” denotes p ?
0.05 (t test).
neurons did respond to acute depolarization with KCl
with a Ca2?transient. Moreover, the acceleration of the
switch by KCl was reversed in the presence of GABAA
receptor antagonists (KCl ? BMI ? PTX treatment in
Figure 2A). However, this time course was faster than
the one observed for neurons treated only with BMI ?
frequency, KCl had additional effects (see Discussion).
of mGSCs can determine the kinetics of the develop-
mental switch in GABA signaling. Consistent with this
notion, chronic activation of GABAAreceptors with mus-
cimol (10–50 ?M) produced a significant acceleration of
the switch in day 9 neurons (Figure 6F).
Requirement for GABA-Mediated Ca2?Influx
Early in development, GABA-induced depolarization
evokes Ca2?influx through voltage-dependent Ca2?
channels. Electrophysiological recordings from individ-
ual neurons showed that ICawas blocked by ?50% by
the L-type Ca2?channel antagonist nimodipine (10 ?M)
and ?90% by 10 ?M CdCl2(commonly used as a high-
threshold Ca2?channel blocker; Figure 7A). As chronic
blockade of all ICawas found to be deleterious for neu-
ronal survival, a partial block of ICawas performed using
moderate concentrations of nimodipine (0.5–1 ?M). As
shown in Figure 7B, chronic treatment with nimodipine
significantly delayed the developmental switch. To as-
sess whether this observation reflected a true shift in
Figure 7. Role of Ca2?Influx for the Develop-
mental Switch of GABA Function
(n ? 6). Inset: Representative traces of Ca2?
currents elicited by depolarizing pulses in a
day 7 neuron, displaying the effects of nimo-
dipine (10 ?M) and CdCl2(10 ?M).
(B) Effects of chronic blockade of L-type
VDCCs on GABA-induced Ca2?elevation in
cultures ofdifferent ages. (“*” and“**” denote
p ? 0.05 and p ? 0.001, respectively, t test).
(C) Effects of chronic blockade of L-type
VDCCs on the shift in EGABA. Left: Representa-
tive example of a GABA-induced I-V relation-
ship. Cells were voltage clamped at ?70 mV
and stepped to different potentials (?70 to 0
mV, 10 mV steps). Right: Averaged I-V curve
(n ? 9). Note that EGABAat day 11 in the pres-
ence of nimodipine is similar to that at day 7
in untreated cultures (Figure 1).
EGABA, we measured EGABAin cultures chronically treated
with nimodipine (Figure 7C). We found that the shift in
EGABA was prevented at a time in which the effect of
nimodipine on the GABA-induced elevations of [Ca2?]i
was mostsignificant (day 11 neurons).Under these con-
ditions, EGABAwas ?42.5 ? 4.3 mV (n ? 9), similar to that
observed in untreated neurons at day 7 (?44.5 ? 2.0
mV; n ? 11; p ? 0.2). These findings support a model
in which spontaneous GABAergic events elevate [Ca2?]i
and activate signaling pathways which can promote the
tion (Figure 8). This suggests the notion that GABA-
mediated excitation is self-limiting.
The Role of GABAergic Activity and Depolarization
We propose that the level of depolarizing GABAergic
activity determines the rate of transformation of neu-
ronal GABAergic responses. This is supported by the
following three lines of evidence: (1) Blockade of GABAA
receptors prevented the transformation; (2) elevating
GABAergic activity or stimulating GABAAreceptors ac-
celerated it; and (3) the developmental increase in spon-
taneous GABA release tightly correlated with the time
course of the transformation. To elevate synaptic activity,
in the presence of TTX. That KCl-depolarization had
indeed elevated GABAergic activity was shown by (1)
the frequency of mGSCs was increased, and (2) the
accelerated rate of transformation was significantly re-
duced when the KCl treatment was combined with block-
ade of GABAAreceptors. However, the transformation
still occurred under conditions of KCl-induced depolar-
ization in the presence of GABAAreceptor antagonists.
As evident from Figure 4A, 10 mM KCl acutely induces
a very robust elevation of [Ca2?]iin immature neurons.
Therefore, it is likely that chronic depolarization with 10
required specificity for GABA-mediated depolarization.
This notion is supported by the fact that physiological
sources ofdepolarization (namelyglutamatergic activity
and action potentials), although present at substantial
levels, are unable to regulate the time course of the
developmental switch (Figures 2B, 6A, and 6B).
Interestingly, in spinal cord neurons, the switch of
GABA signaling is accelerated when neurons are plated
on a higher density of astrocytes (Li et al., 1998). The
mechanism underlying this process remains unknown.
The switch of GABA signaling from excitation to inhibi-
tion occurs during early postnatal brain development,
when synaptic connections are being established,
strengthened, and refined (Katz and Shatz, 1996). Using
a model system of cultured hippocampal neurons, we
tested the hypothesis that this switch is regulated by
neuronal activity. We first characterized two aspects of
GABAergic transmission—the electrophysiological (mem-
brane depolarization) and the biochemical (Ca2?sig-
nals)—which may serve different roles in the developing
nervous system. We found that developmental changes
in these two properties are related to a shift in EGABA
toward hyperpolarized potentials as well as a change in
depolarization-induced elevations of [Ca2?]i. Our results
by GABAergic activity. The expression of the Cl?trans-
porter KCC2 was upregulated by increased spontane-
ous GABAergic activity, and downregulated following
blockade of GABAA receptors, indicating that GABA-
mediated depolarization is itself able to increase the
expression of KCC2 and induce the maturation of inhibi-
GABA as a Self-Limiting Trophic Factor
Figure 8. A Model for the Developmental Switch of GABAergic Transmission
During embryonic development, GABA acts as an excitatory neurotransmitter. GABA-mediated excitation can trigger Ca2?influx through
VDCCs, leading to an increase in the expression of the potassium-chloride cotransporter KCC2. This excitatory GABAergic activity can also
regulate the activation profile of VDCCs. During early postnatal development, the increase in KCC2 activity lowers [Cl?]iand reduces EGABA,
establishing GABA as an inhibitory neurotransmitter.
Recently, astrocytes have been shown to dramatically
in cultured neurons (Ullian et al., 2001). Thus, it is possi-
ble that astrocytes might accelerate the maturation of
GABAergic synaptic transmission, thereby accelerating
the time course of the switch.
that either GABA-mediated elevations of [Ca2?]iare spa-
tially localized, or that additional Ca2?-independent pro-
cesses are required. Consistent with the notion of
GABA-specific signaling is the growing evidence for an
extensive protein network centered around gephyrin, a
protein that may link GABAAreceptors to other signaling
(Kneussel and Betz, 2000).
Specificity of GABA-Mediated Calcium Signaling
Spontaneous GABAergic and glutamatergic activity are
both present during the time window of the develop-
mental switch (Figures 6A and 6B; Hsia et al., 1998; van
den Pol et al., 1998; Lamsa et al., 2000; Palva et al.,
2000). GABAergic events are indeed depolarizing (3–20
mation of unitary events. Such depolarizing responses
are likely to trigger Ca2?influx through VDCCs (Koester
and Sakmann, 1998). Surprisingly, GABAergic but not
glutamatergic activity promotes the transformation of
GABAergic transmission (Figures 2A and 2B).
What is the signal that is specifically activated by
GABA but not glutamate? Nimodipine delayed the
switch (Figures 7B and 7C), indicating that Ca2?influx
through L-type VDCCs participates in the signaling cas-
cade initiated by GABA. Since glutamate-mediated de-
polarization also activates L-type VDCCs (Yuste and
GABA-Dependent Regulation of KCC2 Expression
In hippocampal slices, the temporal expression pattern
for KCC2 coincides with the postnatal switch of GABA-
signaling (Rivera et al., 1999; Vu et al., 2000). In these
cultures, there is a similar developmental increase in the
level of KCC2 mRNA (Figure 3). We found that the level
of KCC2 mRNA can be regulated by GABAergic activity.
The levelof KCC2mRNA wasreduced bychronic block-
transmitterrelease, see Figures 6C and6D).Theobserved
increase in KCC2 mRNA levels may result from en-
hanced transcription or an increase in the stability of
existing mRNAs. Given that the developmental increase
and a reduction of the GABA-mediated depolarization
(Jarolimek et al., 1999; Kakazu et al., 1999; Rivera et
al., 1999), these results further support the notion that
GABA-mediated excitation is self-limited.
While we have monitored the expression of KCC2,
other genes may also contribute to the switch. The ex-
pression of NKCC1, a Na?-K?-2 Cl?transporter, is also
developmentally regulated (Plotkin et al., 1997; Kakazu
et al., 1999; Lu et al., 1999). NKCC1 participates in Cl?
influx, and its expression is reduced during the same
time window in which KCC2 expression is increased,
suggesting a mechanism for reducing Cl?influx as the
Cl?efflux is increased. It may be noted that at day 9,
the expression of KCC2 was 6-fold higher than that at
day 3, while only 20% of the cells had failed to respond
to GABA with Ca2?elevation (Figures 1B and 3D). This
could be attributed to a high rate of Cl?influx due to
NKCC1 activity during this early period of transfor-
activation of NMDA receptors and the maturation of
glutamatergic transmission (Ben-Ari et al., 1997).
GABA-induced elevation of [Ca2?]iis likely to promote
neuronal maturation. In the initial stages of brain devel-
opment, depolarizing GABAergic potentials can prevent
DNA synthesis, leading to the differentiation of cortical
progenitor neurons (LoTurco et al., 1995). Furthermore,
GABA-mediated Ca2?increases can induce BDNF ex-
pression (Berninger et al., 1995) and the differentiation
into specific neuronal phenotypes (Marty et al., 1996).
GABA-induced elevation of [Ca2?]imay also be required
to form, stabilize, and strengthen synaptic connections
(Kirsch and Betz, 1998; Caillard et al., 1999; Kneussel
and Betz, 2000). For glycinergic transmission, there is
a similar developmental transformation from excitation
to inhibition (Reichling et al., 1994; Wang et al., 1994).
Interestingly, Ca2?influx triggered by glycine-mediated
depolarization is required for clustering of postsynaptic
glycine receptors (Kirsch and Betz, 1998), an important
step in synaptogenesis. Similarly, Ca2?influx mediated
byGABA mayalso participatein postsynapticdifferenti-
ing evidence that Ca2?influx through synaptically acti-
vated L-type Ca2?channels is important for activation
of gene expression (Murphy et al., 1991; Berninger et
al., 1995; Deisseroth et al., 1996, 1998).
are transformed. We have demonstrated that GABA it-
self promotes this transformation. Such a self-limiting
trophic action of GABA allows for an activity-dependent
transition of the nervous system from its early depen-
dence on global excitation to the requirement of mature
neural circuits, where inhibition plays a critical role in
its development and function.
GABA-Induced Modifications of ICa
In addition to the developmental transformation of
GABAergic transmission, we found that neuronal matu-
ration involved a shift in the threshold for the activation
of ICa(Figure 4). Pharmacological analysis of Ca2?cur-
ICa seems to reflect primarily the activation of “high-
threshold” VDCCs, with ?50% of those channels corre-
sponding to the L-type (Figure 7A). Our findings suggest
that either the subunit composition of Ca2?channels is
modified or that the expression of low-threshold Ca2?
channels decreases with age. Consistent with this idea,
the expression of Ca2?channel genes appears to be
developmentally regulated (Gruol et al., 1992; Hilaire et
al., 1996; Desmadryl et al., 1998). In fact, the functional
expression of low threshold T-type Ca2?channels has
been shown to decrease in cultured hippocampal neu-
rons between day 4 and day 16, a time window similar
blockade of GABAAreceptors (Figure 4B) reduced the
developmental shift in the threshold for the activation
of ICa, GABAergic transmission can trigger the observed
developmental changes in Ca2?currents.
Cell Culture and Chronic Treatments
Hippocampi from E18 rats were trypsinized (15 min, 37?C), washed,
and gently triturated by passing the tissue through a Pasteur pipette
with a fire-polished tip. Neurons were plated at 500,000 cells/ml on
poly-L-lysine coated coverslips in 35 mm dishes. The plating me-
dium was DMEM (BioWhittaker, Walkersville, MD) containing10%
heat-inactivated fetal bovine serum (Hyclone, Logan, UT) and 10%
Ham’s F12 with glutamine. Twenty four hours after plating, the cul-
ture medium was completely replaced by Neurobasal medium with-
out glutamine (Gibco, Life Technologies) with B-27 supplement
(Gibco). These serum-free conditions supported the growth of neu-
rons but not glial cells. Cultures were fed once a week by adding
?0.4 ml of neurobasal medium. Chronic treatments were started
2 days after plating. BMI, D-APV, TTX, GABA, baclofen, CNQX,
added daily from fresh stocks prepared at 1000-fold the final con-
centrations (indicated in the Results section), except for KCl, which
was added only once. All chronic treatments involving KCl (10 mM)
were done in the presence of TTX and D-APV (2 and 50 ?M).
Two Faces of GABA Signaling: Membrane
Depolarization and Ca2?Elevation
GABA is able to induce electrophysiological (membrane
depolarization) and biochemical (Ca2?signaling) effects,
which may serve different roles in the developing ner-
vous system. While the extent of GABA-mediated depo-
larization depends primarily on EGABA, Ca2?-dependent
signaling through GABA is dependent both on EGABAand
the properties of depolarization-induced elevation of
[Ca2?]i; these two aspects may be independently regu-
lated (Obrietan and van den Pol, 1997). Previous work
has suggested that the GABA-induced depolarization
may serve to either facilitate or inhibit neuronal activity
under different circumstances in the developing brain
(Ben-Ari et al., 1989; Chen et al., 1996; Garaschuk et al.,
2000; Lamsa et al., 2000). A depolarizing GABAergic
potential can result in a shunting inhibition or can facili-
tate excitation, depending on the relative timing and
spatial distribution of GABAergic and glutamatergic in-
puts (Chen et al., 1996). In addition, GABA-mediated
depolarization has been proposed to play a role in the
Neurons were loaded with the Ca2?-sensitive fluorescent dye Fluo-4
(HBS, in mM: NaCl 150, KCl 3, CaCl23, MgCl22, HEPES 10, and
glucose 5; pH 7.4; osmolarity 310 mOsm) containing 15 ?M CNQX
and 10 ?M BMI in the presence of 5 ?M Fluo-4AM dissolved in
DMSO (Molecular Probes, Eugene, OR). HEPES-buffered solution
(in the absence of HCO3?) was used throughout the experiments,
thus consistent with the notion that changes in [Cl?]ican account
GABA as a Self-Limiting Trophic Factor
for the switch in GABA signaling (Staley et al., 1995). All recordings
were carried out at room temperature, using HBS containing 10 ?M
CNQX. Solutions were exchanged using a multibarreled perfusion
laser scanning microscope BioRad MRC1024MP (Biorad, Hemel
Hempstead, England). The scanhead was connected to an Olympus
BX50WI uprightmicroscope and imageswere collected usinga 20?
water-immersion objective. A total of 30–40 neurons were studied
in each experiment. Somatic regions of ?20–50 pixels were chosen
for quantification of fluorescence intensity. Sampling of average
pixel intensitywas performed at0.5 Hz. Transmittedand fluorescent
images reflecting Ca2?dynamics were acquired (picture size of 512 ?
512 pixels) using the Biorad Lasersharp acquisition program. Data
were analyzed using custom made LabView Software (National In-
struments) on a PC. A neuron was considered responsive to GABA
deviations (SD) of the baseline noise. This “six SD” criterion was
developed by repeatedly comparing visual analysis on blind experi-
ments with threshold analysis, until the results matched.
The RatKCC2 clonewas agenerous gift fromJohn Payne.We thank
Xiao-yun Wang for the preparation of the hippocampal cultures and
Benedikt Berninger, Fernanda Ceriani, Miguel Morales, Madhu Rao,
and David Sykes for helpful comments on the manuscript. We also
thank Miguel Morales and Yuki Goda for their invaluable help during
the last stages of this work, and Luca for his support. This work
was supported by grants from the NIH (NS 37831 and NS 36999).
Received September 29, 2000; revised April 18, 2001.
Ben-Ari, Y., Cherubini, E., Corradetti, R., and Gaiarsa, J.L. (1989).
Giant synaptic potentials in immature rat CA3 hippocampal neu-
rones. J. Physiol. 416, 303–325.
Ben-Ari, Y., Khazipov, R., Leinekugel, X., Caillard, O., and Gaiarsa,
J.L. (1997). GABAA, NMDA and AMPA receptors: a developmentally
regulated ‘menage a trois’. Trends Neurosci. 20, 523–529.
Berninger, B., Marty, S., Zafra, F., da Penha Berzaghi, M., Thoenen,
H., and Lindholm, D. (1995). GABAergic stimulation switches from
enhancing to repressing BDNF expression in rat hippocampal neu-
rons during maturation in vitro. Development 121, 2327–2335.
Boukhaddaoui, H., Sieso, V., Scamps, F., Vigues, S., Roig, A., and
Valmier, J.(2000). Q-and L-type calciumchannels controlthe devel-
opment of calbindin phenotype in hippocampal pyramidal neurons
in vitro Eur. J. Neurosci. 12, 2068–2078.
Caillard, O., Ben-Ari, Y., and Gaiarsa, J.L. (1999). Long-term potenti-
ation of GABAergic synaptic transmission in neonatal rat hippocam-
pus. J. Physiol. 518, 109–119.
Chameau, P., Lucas, P., Melliti, K., Bournaud, R., and Shimahara,
T.(1999). Developmentofmultiple calciumchanneltypes incultured
mouse hippocampal neurons. Neuroscience 90, 383–388.
Chen, G., Trombley, P.Q., and van den Pol, A.N. (1996). Excitatory
actions of GABA in developing rat hypothalamic neurones. J. Phys-
iol. 494, 451–464.
Cherubini, E., Rovira, C., Gaiarsa, J.L., Corradetti, R., and Ben Ari, Y.
(1990). GABA mediated excitation in immature rat CA3 hippocampal
neurons. Int. J. Dev. Neurosci. 8, 481–490.
Cherubini, E., Gaiarsa, J.L., and Ben-Ari, Y. (1991). GABA: an excit-
atory transmitter in early postnatal life. Trends Neurosci. 14,
Connor, J.A., Tseng, H.Y., and Hockberger, P.E. (1987). Depolariza-
tion- and transmitter-induced changes in intracellular Ca2?of rat
cerebellar granule cells in explant cultures. J. Neurosci. 7, 1384–
Deisseroth, K., Bito, H., and Tsien, R.W. (1996). Signaling from syn-
apse to nucleus: postsynaptic CREB phosphorylation during multi-
ple forms of hippocampal synaptic plasticity. Neuron 16, 89–101.
Deisseroth, K., Heist, E.K., and Tsien, R.W. (1998). Translocation
of calmodulin to the nucleus supports CREB phosphorylation in
hippocampal neurons. Nature 392, 198–202.
Desmadryl, G., Hilaire, C., Vigues, S., Diochot, S., and Valmier, J.
(1998). Developmental regulation of T-, N- and L-type calcium cur-
rents in mouse embryonic sensory neurones. Eur. J. Neurosci. 10,
Garaschuk, O., Linn, J., Eilers, J., and Konnerth, A. (2000). Large-
scale oscillatory calcium waves in the immature cortex. Nat. Neu-
rosci. 3, 452–459.
in calcium conductances contribute to the physiological maturation
Hilaire, C., Diochot, S., Desmadryl, G., Baldy-Moulinier, M., Richard,
S., and Valmier, J. (1996). Opposite developmental regulation of
mouse sensory neurons. Neuroscience 75, 1219–1229.
Hsia, A.Y., Malenka, R.C., and Nicoll, R.A. (1998). Development of
excitatory circuitry in the hippocampus. J. Neurophysiol. 79, 2013–
All experiments were carried out using the perforated-patch whole
cell recording technique. Neuronswere visualized using phase-con-
trast on a Nikon inverted microscope. Recordings were carried out
using patch clamp amplifiers (Axopatch 200B, Axon Instruments,
Foster City, CA). Signals were filtered at 2 kHz, sampled at 10 kHz
and analyzed using Pclamp 6.0 software (Axon Instruments). Series
resistance was compensated at 80% (lag 30–60 ?s). Micropipettes
were made from borosilicate glass capillaries (KG-33, Garner Glass)
with a resistance of ?2 M?. For recordings of EGABA, pipettes were
tip filled with internal solution and then back filled with internal
solution containing 20 ?g/ml gramicidin A (dissolved in methanol,
Sigma). The internal solution contained (in mM): K-gluconate 154,
NaCl 9, MgCl21, HEPES 10, and EGTA 0.2; pH 7.4; osmolarity 300
mOsm. The bath solution contained HBS with 15 ?M CNQX, and
was constantly perfused at a rate of ?1 ml/min. GABA (100 ?M, 50
ms, 5 psi) was locally applied every 10 s through a micropipette
connected to a Picospritzer (General Valve Corporation). For re-
internal solution containing 200 ?g/ml amphotericin B (dissolved in
DMSO, ICN). The high CsCl2internal solution contained (in mM):
CsCl2 154, NaCl 9, MgCl2 1, HEPES 10, and EGTA 0.2; pH 7.4;
osmolarity 300 mOsm. The bath solution was a modified HBS con-
taining (in mM): NaCl 150, KCl 3, BaCl210, MgCl22, HEPES 10,
glucose 5, andtetraethylammonium 10; TTX 2 ?M.Leak and capaci-
ing pulses. Recordings of miniature synaptic currents were per-
formed using amphotericin B and high CsCl2internal solution. The
bath solution was HBS with CNQX (15 ?M) for GABA-mediated
currents, and with BMI (10 ?M) for glutamate-mediated currents.
Events were filtered at 2 kHz and detected online using the WCP
software (kindly provided by J. Dempster, University of Strathclyde,
Scotland). All experiments were performed at 22–24?C. Experiments
were rejected if the leak current exceeded 100 pA.
Quantitation of KCC2 mRNA in Cultured Neurons
Total RNA was harvested from neurons treated with vehicle, BMI ?
PTX or 10 mM KCl. Cells were lysed with guanidine isothiocyanate
(600 ?l) and purified (RNeasy, Qiagen). An antisense RNA template
was generated using 316 bp of noncoding sequence residing at the
distal 3? end of the mRNA sequence (Payne et al., 1996). An anti-
sense ?-actin probe (245 bp) was mixed with the KCC2 probe and
used as an internal standard in all assays.
were generated using Maxiscript (Ambion), and RNase protection
grams of total RNA were treated with a mixture of the KCC2 and
?-actin antisense probes (5 ? 104cpm each) and annealed at 68?C
for 10 min. The samples were then immediately treated with RNase
T1 (200 U/ml) for 30 min at 37?C. The protected RNA species were
separated over a 5% acrylamide gel containing 8 M urea, and then
dried. Radioactive band intensities were then measured (Molecular
Dynamics Phosphorimager) and quantified (NIH image v1.62,
Cell Download full-text
Ikeda, Y., Nishiyama, N., Saito, H., and Katsuki, H. (1997). GABAA
receptor stimulation promotessurvival of embryonic ratstriatal neu-
rons in culture. Brain Res. Dev. Brain Res. 98, 253–258.
Jarolimek, W., Lewen, A., and Misgeld, U. (1999). A furosemide-
sensitive K?-Cl?cotransporter counteracts intracellular Cl?accu-
lationof intracellularchlorideby cotransportersin developinglateral
superior olive neurons. J. Neurosci. 19, 2843–2851.
tion of cortical circuits. Science 274, 1133–1138.
Khazipov, R., Leinekugel, X., Khalilov, I., Gaiarsa, J.L., and Ben-Ari,
Y. (1997). Synchronization of GABAergic interneuronal network in
CA3 subfield of neonatal rat hippocampal slices. J. Physiol. 498,
Kirsch, J., and Betz, H. (1998). Glycine-receptor activation is re-
quired for receptor clustering in spinal neurons. Nature 392,
Kneussel, M., and Betz, H. (2000). Clustering of inhibitory neuro-
transmitter receptors at developing postsynaptic sites: the mem-
brane activation model. Trends Neurosci. 23, 429–435.
Koester, H.J., and Sakmann, B. (1998). Calcium dynamics in single
spines during coincident pre- and postsynaptic activity depend on
old excitatory postsynaptic potentials. Proc. Natl. Acad. Sci. USA
Lamsa, K., Palva, J.M., Ruusuvuori, E., Kaila, K., and Taira, T. (2000).
Synaptic GABA(A) activation inhibits AMPA-kainate receptor-medi-
ated bursting in the newborn (P0–P2) rat hippocampus. J. Neuro-
physiol. 83, 359–366.
Lee, J.J., and Costlow, N.A. (1987). A molecular titration assay to
measure transcript prevalence levels. Meth. Enzymol. 152, 633–648.
Leinekugel, X., Tseeb, V., Ben-Ari, Y., and Bregestovski, P. (1995).
Synaptic GABAAactivation induces Ca2?rise in pyramidal cells and
interneurons from rat neonatal hippocampal slices. J. Physiol. 487,
Li, Y.-x., Schaffner, A., Walton, M.K., and Barker, J.L. (1998).
Astrocytes regulate developmental changes in the chloride ion gra-
dient of embryonic rat ventral spinal cord neurons in culture. J.
Physiol. 509, 847–858.
LoTurco, J.J., Owens, D.F., Heath, M.J., Davis, M.B., and Kriegstein,
and inhibit DNA synthesis. Neuron 15, 1287–1298.
Lu, J., Karadsheh, M., and Delpire, E. (1999). Developmental regula-
tion of the neuronal-specific isoform of K-Cl cotransporter KCC2 in
postnatal rat brains. J. Neurobiol. 39, 558–568.
Luhmann, H.J., and Prince, D.A. (1991). Postnatal maturation of the
GABAergic system in rat neocortex. J. Neurophysiol. 65, 247–263.
Marty, S., Berninger, B., Carroll, P., and Thoenen, H. (1996). GABA-
ergic stimulation regulates the phenotype of hippocampal interneu-
rons through the regulation of brain-derived neurotrophic factor.
Neuron 16, 565–570.
Mueller, A.L., Taube, J.S., and Schwartzkroin, P.A. (1984). Develop-
ment of hyperpolarizing inhibitory postsynaptic potentials and hy-
campus studied in vitro. J. Neurosci. 4, 860–867.
sensitive calcium channels mediate synaptic activation of immedi-
ate early genes. Neuron 7, 625–635.
Obrietan, K., and van den Pol, A.N. (1995). GABA neurotransmission
in the hypothalamus: developmental reversal from Ca2?elevating
to depressing. J. Neurosci. 15, 5065–5077.
Obrietan, K., and van den Pol, A.N. (1997). GABA activity mediating
cytosolic Ca2?rises in developing neurons is modulated by cAMP-
dependent signal transduction. J. Neurosci. 17, 4785–4799.
Owens, D.F., Boyce, L.H., Davis, M.B., and Kriegstein, A.R. (1996).
Excitatory GABA responses in embryonic and neonatal cortical
slices demonstratedby gramicidin perforated-patchrecordings and
calcium imaging. J. Neurosci. 16, 6414–6423.
Palva, J.M., Lamsa, K., Lauri, S.E., Rauvala, H., Kaila, K., and Taira,
T. (2000). Fast network oscillations in the newborn rat hippocampus
in vitro. J. Neurosci. 20, 1170–1178.
Payne, J.A., Stevenson, T.J., and Donaldson, L.F. (1996). Molecular
characterization of a putative K-Cl cotransporter in rat brain. A neu-
ronal-specific isoform. J. Biol. Chem. 271, 16245–16252.
Plotkin, M.D., Snyder, E.Y., Hebert, S.C., and Delpire, E. (1997).
Expression of the Na-K-2Cl cotransporter is developmentally regu-
lated in postnatal rat brains: a possible mechanism underlying
GABA’s excitatory role in immature brain. J. Neurobiol. 33, 781–795.
Reichling, D.B., Kyrozis, A., Wang, J., and MacDermott, A.B. (1994).
Mechanisms of GABA and glycine depolarization-induced calcium
transients in rat dorsal horn neurons. J. Physiol. 476, 411–421.
Rivera, C., Voipio, J., Payne, J.A., Ruusuvuori, E., Lahtinen, H.,
Lamsa, K., Pirvola, U., Saarma, M., and Kaila, K. (1999). The K?/Cl?
maturation. Nature 397, 251–256.
in neocortical transplants. Cell Transplant. 7, 121–129.
Schierle, G.S., Gander, J.C., D’Orlando, C., Ceilo, M.R., and Vogt
Weisenhorn, D.M. (1997). Calretinin-immunoreactivity during post-
natal development of the rat isocortex: a qualitative and quantitative
study Cereb. Cortex. 7, 130–142.
of neuronal excitation by inhibitory GABAAreceptors. Science 269,
Ullian, E.M., Sapperstein, S.K., Christopherson, K.S., and Barres,
B.A. (2001). Control of synapse number by glia. Science 291,
van den Pol, A.N., Gao, X.B., Patrylo, P.R., Ghosh, P.K., and Obrie-
tan, K. (1998). Glutamate inhibits GABA excitatory activity in devel-
oping neurons. J. Neurosci. 18, 10749–10761.
Vu, T.Q., Payne, J.A., and Copenhagen, D.R. (2000). Localization
and developmental expression patterns of the neuronal K-Cl co-
transporter (KCC2) in the rat retina. J. Neurosci. 20, 1414–1423.
Wang, J., Reichling, D.B., Kyrozis, A., and MacDermott, A.B. (1994).
Developmental loss of GABA- and glycine-induced depolarization
and Ca2?transients in embryonic rat dorsal horn neurons in culture.
Eur. J. Neurosci. 6, 1275–1280.
Yuste, R., and Katz, L.C. (1991). Control of postsynaptic Ca2?influx
in developing neocortex by excitatory and inhibitory neurotransmit-
ters. Neuron 6, 334–344.