J Neurophysiol 94: 2797–2804, 2005;
10.1152/jn.00445.2005.
Adenosine Down-Regulates Giant Depolarizing Potentials in the Developing
Rat Hippocampus by Exerting a Negative Control on Glutamatergic Inputs
Victoria F. Safiulina, Alexander M. Kasyanov, Rashid Giniatullin, and Enrico Cherubini
Neuroscience Programme, International School for Advanced Studies, Trieste, Italy
Submitted 2 May 2004; accepted in final form 9 July 2005
INTRODUCTION
A particular feature of the developing hippocampus is the
presence of network-driven spontaneous rhythmic activity
called “giant depolarizing potentials” or GDPs (Ben-Ari et al.
1989). GDPs are characterized by recurrent membrane depolarizations with superimposed fast action potentials. They occur at the frequency of 0.03– 0.3 Hz and can be detected both
in vitro (Ben-Ari et al. 1989) and in vivo (Leinekugel et al.
2002). GDPs are generated by the synergistic action of glutamate and ␥-aminobutyric acid (GABA) (Bolea et al. 1999).
Early in postnatal life, GABA depolarizes and excites postsynaptic cells (Ben-Ari 2002; Ben-Ari et al. 1997; Cherubini et al.
1991) due to high intracellular [Cl⫺], which results mainly
from the unbalance of two Cl⫺ cotransporter systems, the
NKCCl and KCC2 that enhance and lower intracellular [Cl⫺],
Address for reprint requests and other correspondence: E. Cherubini, Neuroscience Programme, International School for Advanced Studies, Via Beirut
2-4, 34014 Trieste, Italy (E-mail: [email protected]).
www.jn.org
respectively (Payne et al. 2003; Rivera et al. 1999). The
depolarizing action of GABA during GDPs results in calcium
influx through the activation of voltage-dependent calcium channels and N-methyl-D-aspartate (NMDA) receptors (Leinekugel et
al. 1997). Thus GDPs acting as coincidence detector signals for
enhancing synaptic efficacy (Kasyanov et al. 2004) may contribute to the functional and structural refinement of the hippocampal network.
In previous studies from our and other laboratories it was
shown that GDPs are regulated by a number of neurotransmitter receptors and modulators localized on pre- and postsynaptic
sites of principal cells and interneurons (Avignone and Cherubini 1999; Gaiarsa et al. 1991; Maggi et al. 2001; McLean et al.
1996; Strata et al. 1995; Xie et al. 1994). In turn, neurotransmitters release can be regulated by activity dependent processes like GDPs. Consistent with this, in a recent study
(Safiulina et al. 2005) we demonstrated that adenosine triphosphate (ATP), a widely distributed neurotransmitter and neuromodulator (Burnstock 2004), which regulates network activity
by ionotropic P2X and metabotropic P2Y receptors, can be
released during GDPs. Because ATP is hydrolyzed to adenosine by ectonucleotidases (Cunha et al. 1998; Dunwiddie et al.
1997) the possibility that this neurotransmitter also regulates
frequency of GDPs by activation of adenosine receptors by its
breakdown product adenosine cannot be ruled out.
In the hippocampus adenosine has been shown to act on preand postsynaptic A1 receptors to inhibit glutamate release
(Lupica et al. 1992; Wu and Saggau 1994; Yoon and Rothman
1991) and to enhance potassium conductance (Greene and
Haas 1985), respectively. Altogether these effects would produce a reduction in cell excitability and firing rate (for a review
see Ribeiro et al. 2003b). Here we demonstrate that, during the
first week of postnatal life, endogenous adenosine, by activating A1 receptors localized mainly on glutamatergic terminals
projecting to principal cells or interneurons, regulates the
occurrence of GDPs.
METHODS
Slice preparation
Experiments were performed on hippocampal slices obtained from
postnatal (P) P2–P6 Wistar rats (P0 was taken as the day of birth) as
previously described (Maggi et al. 2001). The procedure was in
accordance with the European Community Council Directive of 24
November 1986 (86/609EEC) and was approved by the local authority veterinary service. Animals were decapitated after being anesthetized with an intraperitoneal injection of urethane (2 g/kg). After
The costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked “advertisement”
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
0022-3077/05 $8.00 Copyright © 2005 The American Physiological Society
2797
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 15, 2017
Safiulina, Victoria F., Alexander M. Kasyanov, Rashid Giniatullin, and Enrico Cherubini. Adenosine down-regulates giant depolarizing potentials in the developing rat hippocampus by exerting a
negative control on glutamatergic inputs. J Neurophysiol 94: 2797–2804,
2005; 10.1152/jn.00445.2005. Adenosine is a widespread neuromodulator that can be directly released in the extracellular space during
sustained network activity or can be generated as the breakdown
product of adenosine triphosphate (ATP). Whole cell patch-clamp
recordings were performed from CA3 principal cells and interneurons
in hippocampal slices obtained from P2–P7 neonatal rats to study the
modulatory effects of adenosine on giant depolarizing potentials
(GDPs) that constitute the hallmark of developmental networks. We
found that GDPs were extremely sensitive to the inhibitory action of
adenosine (IC50 ⫽ 0.52 ␮M). Adenosine also contributed to the
depressant effect of ATP as indicated by DPCPX-sensitive changes of
ATP-induced reduction of GDP frequency. Similarly, adenosine exerted a strong inhibitory action on spontaneous glutamatergic synaptic
events recorded from GABAergic interneurons and on interictal bursts
that developed in CA3 principal cells after blockade of ␥-aminobutyric acid type A (GABAA) receptors with bicuculline. All these
effects were prevented by DPCPX, indicating the involvement of
inhibitory A1 receptors. In contrast, GABAergic synaptic events were
not changed by adenosine. Consistent with the endogenous role of
adenosine on network activity, DPCPX per se increased the frequency
of GDPs, interictal bursts, and spontaneous glutamatergic synaptic
events recorded from GABAergic interneurons. Moreover, the adenosine transport inhibitor NBTI and the adenosine deaminase blocker
EHNA decreased the frequency of GDPs, thus providing further
evidence that endogenous adenosine exerts a powerful control on
GDP generation. We conclude that, in the neonatal rat hippocampus,
the inhibitory action of adenosine on GDPs arises from the negative
control of glutamatergic, but not GABAergic, inputs.
2798
SAFIULINA, KASYANOV, GINIATULLIN, AND CHERUBINI
decapitation the brain was removed from the skull and placed in
ice-cold artificial cerebrospinal fluid (ACSF) containing (in mM):
NaCl, 130; KCl, 3.5; NaH2PO4, 1.2; NaHCO3, 25; MgCl2, 1.3; CaCl2,
2, glucose, 25; saturated with 95% O2-5% CO2 (pH 7.3–7.4). Hippocampal slices (500 ␮m thick) were cut with a vibratome and
incubated at room temperature (22–25°C) in oxygenated ACSF.
Registrations of activity were performed from an individual slice in a
submerged chamber continuously perfused at 33–34°C with oxygenated ACSF at a rate of 2–3 ml/min.
Patch-clamp whole cell recordings
Drugs
All substances were prepared as 1,000 ⫻ concentrated stock solutions. ATP, adenosine, picrotoxin, and 6-[(4-nitrobenzyl)thio]-9-␤-Dribofuranosylpurine (NBTI) were purchased from Sigma. Bicuculline
methochloride, 1,3-dipropyl-8-cyclopentylxanthine (DPCPX), 6,7-dinitroquinoxaline-2,3-dione (DNQX), and erythro-9-(2-hydroxy-3nonyl) adenine (EHNA) hydrochloride were purchased from Tocris
Cookson. DPCPX and picrotoxin were dissolved in ethanol (final
concentration of ethanol in solution 0.02%); NBTI and DNQX were
dissolved in dimethylsulfoxide (DMSO). Control experiments using a
solution containing 2.5/1,000 of DMSO did not affect GDPs or sPSPs.
Stock solutions were stored at ⫺20°C. The substances to be tested
were dissolved at their final concentrations in extracellular solution
just before the recording session. All substances were applied through
a three-way tap system. Bath volume exchange was completed within
1.5 min.
Data acquisition and analysis
Data were transferred to a computer after digitization with an A/D
converter (Digidata 1200, Axon Instruments). Data were sampled at
20 kHz and filtered with a cutoff frequency of 1 kHz. Acquisition and
analysis were performed with Clampex 7 (Axon Instruments).
The analysis of sPSPs was performed with Mini Analysis Program
(Synaptosoft, Decatur, GA). General principles and details on the
analysis of synaptic currents can be found elsewhere (Jo et al. 1998;
Poisbeau et al. 1996).
The cumulative amplitude and interevent plots obtained for cell in
controls and after drug application were compared using the Kolmogorov–Smirnov (KS) test.
J Neurophysiol • VOL
RESULTS
Whole cell recordings in current-clamp mode were performed from 148 CA3 pyramidal cells and 12 interneurons
localized in stratum radiatum at P2–P6. All slices exhibited
spontaneous GDPs typical of the immature brain, which lasted
0.5–1 s and occurred at the frequency of 0.05 ⫾ 0.06 Hz (n ⫽
41). In both principal cells and interneurons GDPs were intermingled with small-amplitude action potential– dependent and
–independent spontaneous synaptic potentials (sPSPs). sPSPs
occurred at frequencies of 5.6 ⫾ 0.006 (n ⫽ 98) and 8.5 ⫾ 1.1
Hz (n ⫽ 9), in principal cells and interneurons, respectively.
In CA3 pyramidal cells ATP reduces the frequency of GDPs
by P2 receptors and its breakdown product adenosine
Because the breakdown of ATP in the extracellular space is
one important source of adenosine (Ribeiro et al. 2003b) in a
first set of experiments we tested the effects of ATP on GDPs.
As shown in Fig. 1A, bath application of ATP (50 ␮M)
reversibly blocked GDPs, an effect that was associated with an
increase in frequency of sPSPs. At lower concentrations (2
␮M) ATP induced a reduction of frequency of GDPs from
0.06 ⫾ 0.004 to 0.009 ⫾ 0.001 Hz (n ⫽ 4, P ⬍ 0.01). The
effect of ATP on the frequency of GDPs was dose dependent
with an EC50 of 0.78 ␮M (Fig. 1C). In eight out of 14 cells,
ATP (50 ␮M) induced a membrane hyperpolarization (4.6 ⫾
0.8 mV) associated with an increase in membrane conductance
(9 ⫾ 1%). When ATP (50 ␮M) was applied in the presence of
a high concentration of DPCPX (10 ␮M), which blocks adenosine receptors (Khakh et al. 2003), it produced a biphasic
effect: an initial increase followed by a persistent decrease in
frequency of GDPs (Fig. 1B; see also Safiulina et al. 2005).
The depressant effect was concentration dependent with an
EC50 value of 16.2 ␮M (Fig. 1C). Thus at the concentration of
50 ␮M the reduction in frequency of GDPs produced by ATP
was 43 ⫾ 1.5%. Comparison of the two curves of Fig. 1C
suggests a strong adenosine effect. Interestingly, in 10 ␮M
DPCPX ATP (50 ␮M) was still able to up-regulate the frequency of sPSPs recorded from principal cells, indicating that
this effect was adenosine independent (the frequency of sPSPs
was 10.9 ⫾ 0.9 and 10.8 ⫾ 2.8 Hz in the presence of ATP and
ATP ⫹ DPCPX, respectively; n ⫽ 4). Moreover, it should be
stressed that in the presence of DPCPX, the effects of ATP on
GDPs were never associated with changes in membrane potential and resistance, suggesting that the latter were probably
mediated by postsynaptic A1 adenosine receptors. These data
strongly suggest that endogenous ATP may also regulate the
activity of GDPs by its breakdown product adenosine acting on
A1 receptor subtypes.
Endogenous adenosine slows down the frequency of GDPs
To further explore this issue, adenosine was directly applied
to hippocampal slices at concentrations ranging from 0.1 to 10
94 • OCTOBER 2005 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 15, 2017
Electrophysiological experiments were performed from CA3 pyramidal cells and from GABAergic interneurons localized on stratum
radiatum using the whole cell configuration of the patch-clamp technique in current-clamp mode. Neurons were visualized using infrared
Nomarski optics on an upright microscope (Leica DMLFS, Wetzlar,
Germany). GDPs and spontaneous (action potential– dependent and
–independent) ongoing synaptic potentials (sPSPs), were recorded
from a holding potential of ⫺70 ⫾ 5 mV with a standard patch-clamp
amplifier (Axoclamp 2B, Axon Instruments; Foster City, CA). Patch
electrodes, pulled from borosilicate glass capillaries (Hilgenberg,
Malsfeld, Germany), had a resistance of 5–7 M⍀ when filled with an
intracellular solution containing (in mM): KCl, 140; MgCl2, 1; NaCl,
1; EGTA, 1; HEPES, 2; and K2ATP, 2; the pH was adjusted to 7.3
with KOH.
The whole cell capacitance was fully compensated and the series
resistance (10 –20 M⍀) was compensated at 75– 80%. The stability of
the patch was checked by repetitively monitoring the input and series
resistance during the experiment. Cells exhibiting 15–20% changes
were excluded from the analysis. Membrane input resistance was
measured during recordings from the amplitude of the electrotonic
potentials (300-ms duration) evoked by passing hyperpolarizing current steps of 100 –200 pA across the cell membrane.
For each cell, the mean GDPs frequency was calculated in control
conditions and during drug application (starting 3 min after the onset
of drug perfusion). Statistical comparisons were made between mean
values (control vs. drug treatment). The numerical data are given as
means ⫾ SE and compared using the Student’s t-test. The differences
were considered significant when P was ⬍0.05.
ADENOSINE AND GDPS
2799
FIG. 1. Adenosine triphosphate (ATP) also inhibits giant depolarizing potentials (GDPs) by adenosine receptors. A: representative recording from a
CA3 pyramidal neuron (⫺70 mV) demonstrating the reversible inhibitory
action of ATP (50 ␮M) on GDPs. B: recording from another cell showing that
the action of ATP was essentially attenuated in the presence of a high
concentration of 1,3-dipropyl-8-cyclopentylxanthine (DPCPX; 10 ␮M), which
blocks adenosine receptors. Note that in the presence of DPCPX, the depressant effect of ATP was preceded by an early facilitatory action. ATP-induced
increase in spontaneous events (A) was still observed in the presence of
DPCPX (B), indicating that it was independent of the activation of A1
adenosine receptors. Two individual GDPs (marked with a and b) are shown
below in an expanded timescale. C: depressant effects of ATP (open circles,
n ⫽ 35) and ATP ⫹ DPCPX (10 ␮M, closed circles, n ⫽ 21) on the frequency
of GDPs.
␮M. As shown in Fig. 2A, adenosine (10 ␮M) completely
blocked GDPs. This effect was reversible 2–3 min after adenosine was washed out. At a lower concentration (0.5 ␮M)
adenosine reduced the frequency of GDPs from 0.06 ⫾ 0.004
to 0.03 ⫾ 0.002 Hz without modifying their shape (the area
under GDPs was 15.9 ⫾ 1.4 and 16.3 ⫾ 2.2 mV s⫺1 before and
during adenosine application, respectively; P ⬎ 0.05). The
effect of adenosine on GDPs was dose dependent (the EC50
value was 0.52 ␮M; Fig. 2C). The closed circle in Fig. 2C
represents the ratio between the frequency of GDPs obtained in
control conditions and that achieved in the presence of a
saturating concentration of DPCPX (100 nM; see Fig. 3) when
the action of endogenous adenosine is neutralized. This value
corresponds to the inhibitory effect of endogenous adenosine
J Neurophysiol • VOL
FIG. 2. Adenosine acting on A1 receptor subtypes blocks GDPs. A: representative trace from a CA3 pyramidal neuron showing GDPs in control and
during application of adenosine (10 ␮M). At this concentration, adenosine
blocked GDPs without altering the frequency of spontaneous synaptic potentials (sPSPs). B: individual trace from another neuron showing that the
inhibitory effect of adenosine on GDPs was prevented by a low concentration
of DPCPX (100 nM). GDPs marked with a and b are shown below in an
expanded timescale. C: concentration dependency of adenosine effect on
frequency of GDPs (n ⫽ 15). Closed symbol represents the ratio between the
frequency of GDPs obtained in control conditions and that achieved in the
presence of a saturating concentration of DPCPX when the action of endogenous adenosine is neutralized.
94 • OCTOBER 2005 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 15, 2017
on the frequency of GDPs. In four out of seven cells, adenosine
(10 ␮M) produced a membrane hyperpolarization (4.3 ⫾ 0.3
mV) that was associated with an increase in membrane conductance (8 ⫾ 1%). In contrast with ATP, adenosine did not
modify the frequency of sPSPs that at P2–P6 are mainly
GABAA mediated (Hosokawa et al. 1994). In keeping with
this, adenosine (10 ␮M) also failed to modify the frequency
of GABAA-mediated sPSPs when applied in the presence of
the ␣-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid
(AMPA)/kainate receptor antagonist DNQX (10 ␮M). In
DNQX, the frequency of sPSPs was 5.7 ⫾ 1.5 and 6.0 ⫾ 1.7
Hz (n ⫽ 3, P ⬎ 0.05) before and during adenosine application,
respectively. The effects of exogenously applied adenosine on
GDPs, membrane potential, and conductance were prevented
by 100 nM of DPCPX (n ⫽ 4, Fig. 2B), which at this
concentration selectively blocks A1 receptors (Giniatullin and
Sokolova 1998; Ralevic and Burnstock 1998).
2800
SAFIULINA, KASYANOV, GINIATULLIN, AND CHERUBINI
GDPs was from 0.07 ⫾ 0.002 to 0.003 ⫾ 0.002 Hz (n ⫽ 3, P ⬍
0.001; Fig. 4, B and C). As for NBTI, the effect of EHNA was
prevented by DPCPX (100 nM; n ⫽ 4; Fig. 4B). These data
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 15, 2017
FIG. 3. Endogenous adenosine depresses GDPs. A: representative trace
showing GDPs before and after bath application (bar) of DPCPX (100 nM).
Note that DPCPX reversibly increased the frequency of GDPs. GDPs marked
with a and b are shown below in an expanded timescale. B: DPCPX increased
the frequency of GDPs in a concentration dependent way (EC50: 11.9 nM).
Each point in the graph represents the mean DPCPX-induced increased in
frequency of GDPs, expressed as a percentage of control, obtained from 3 to
10 CA3 pyramidal cells. Bars represent the SE (sometimes the bars are within
the symbol).
Importantly, DPCPX per se increased the frequency of
GDPs. This effect was concentration dependent and reversible
because it completely recovered after washing out DPCPX.
Thus DPCPX (10 nM) enhanced the frequency of GDPs from
0.08 ⫾ 0.02 to 0.012 ⫾ 0.004 Hz and DPCPX (100 nM), from
0.08 ⫾ 0.02 to 0.21 ⫾ 0.009 Hz (Fig. 3A). This effect occurred
in the absence of any change in the amplitude or shape of
GDPs (the area under GDPs was 7.0 ⫾ 1.0 and 9.1 ⫾ 1.7
mV s⫺1 before and during application of 100 nM of DPCPX,
respectively; n ⫽ 4, P ⬎ 0.05; see insets below Fig. 3A). The
EC50 value for the facilitatory effect of DPCPX on the frequency of GDPs was 11.9 nM (n ⫽ 25; Fig. 3B). These data
indicate that endogenous adenosine by A1 receptor subtypes
contributes to regulation of GDPs.
To further evaluate the contribution of endogenous adenosine to the generation of GDPs we increased the extracellular
level of endogenous adenosine by applying in the bath NBTI,
which inhibits adenosine transport (de Mendonça and Ribeiro
1994; Fredholm et al. 1994; Ribeiro et al. 2003a,b), or EHNA
that blocks adenosine deaminase, the enzyme that converts
adenosine into the inactive metabolite inosine (Ribeiro et al.
2003a,b). NBTI, at the concentration of 20 ␮M, significantly
decreased the frequency of GDPs from 0.07 ⫾ 0.004 to
0.045 ⫾ 0.002 Hz (n ⫽ 5, P ⬍ 0.01; Fig. 4, A and C), in the
absence of any change in GDP amplitude or shape. This effect
was prevented by DPCPX (100 nM; n ⫽ 3), implying that it
was mediated by adenosine acting on A1 receptor subtypes
(Fig. 4A). Similar results were obtained with EHNA (10 ␮M).
The adenosine deaminase inhibitor reduced the frequency of
J Neurophysiol • VOL
FIG. 4. Adenosine transport inhibitor 6-[(4-nitrobenzyl)thio]-9-␤-D-ribofuranosylpurine (NBTI) and the adenosine deaminase blocker erythro-9-(2hydroxy-3-nonyl) adenine (EHNA) reversibly depress the frequency of GDPs.
A, top trace: bath application of NBTI (bar) reversibly reduced frequency of
GDPs. Bottom trace: in the same neuron the effect of NBTI was prevented by
DPCPX (100 nM). B: bath application of EHNA (bar) reversibly blocked
GDPs (top trace), an effect that was prevented by DPCPX (100 nM, bottom
trace). C: each column represents the mean reduction of frequency of GDPs as
percentage of control (n ⫽ 5 for NBTI and n ⫽ 3 for EHNA). **P ⬍ 0.01;
***P ⬍ 0.001.
94 • OCTOBER 2005 •
www.jn.org
ADENOSINE AND GDPS
Endogenous adenosine affects the glutamatergic drive to
GABAergic interneurons
To identify at the network level where the action of adenosine originated from, in additional experiments (n ⫽ 12) direct
recordings from GABAergic interneurons were performed.
Interneurons were localized on stratum radiatum, in close
vicinity to stratum lucidum. They had a resting membrane
potential ranging from ⫺44 to ⫺70 mV (on average ⫺54.7 ⫾
2.4 mV) and input resistance ranging from 129 to 400 M⍀ (on
average 293 ⫾ 23 M⍀). They discharged at high frequency
(63 ⫾ 7 Hz) in response to long depolarizing current pulses
(100 pA, 400 ms; see also Safiulina et al. 2005). In these cells,
GABAergic and glutamatergic activities were studied in isolation in the presence of bicuculline (20 ␮M) and DNQX (20
␮M), respectively. DPCPX, even at high concentration (10
␮M), did not modify the frequency or the amplitude of
GABAergic sPSPs recorded in the presence of DNQX. In three
interneurons the frequency of sPSPs was 8.5 ⫾ 1.3 and 8.2 ⫾
1.2 Hz, whereas the amplitude of sPSPs was 5.9 ⫾ 0.9 and
4.8 ⫾ 0.4 mV, in control and in the presence of DPCPX,
respectively (P ⬎ 0.05).
J Neurophysiol • VOL
In contrast, DPCPX (100 nM) significantly increased the
frequency (but not the amplitude) of glutamatergic sPSPs
recorded in the presence of bicuculline (10 ␮M) from 2.5 ⫾
0.01 to 3.7 ⫾ 0.3 Hz; (n ⫽ 3; P ⬍ 0.05; Fig. 5, A, B, and E).
DPCPX did not modify the membrane potential or the input
resistance of the recorded neurons. Consistent with the expression of presynaptic A1 receptors localized on glutamatergic
terminals making synapses on GABAergic interneurons, exogenous application of adenosine (10 ␮M) in the presence of
bicuculline (10 ␮M) reversibly reduced the frequency (from
1.8 ⫾ 0.06 to 0.9 ⫾ 0.1 Hz; n ⫽ 3, P ⬍ 0.05) but not the
amplitude (1.4 ⫾ 0.3 and 0.9 ⫾ 0.06 mV, before and after
adenosine, respectively; P ⬎ 0.05) of spontaneous glutamatergic events (Fig. 5, C, D, and E). These effects occurred in the
absence of any change in membrane potential and input conductance.
Endogenous adenosine affects the release of glutamate from
recurrent collaterals of CA3 principal cells
In a recent work it has been shown that a prolonged exposure
of hippocampal slices from P2–P6 rats to GABAA antagonists
produces low-frequency interictal-like discharges that involve
recurrent glutamatergic collaterals (Khazipov et al. 2004). To
see whether endogenous adenosine can modulate interictal
activity generated by glutamatergic recurrent collaterals in the
absence of GABAA-mediated inhibition, bicuculline (20 ␮M)
was applied for 10 –15 min. In these conditions, interictal
events developed at 0.02 ⫾ 0.004 Hz (n ⫽ 9). Further application of DPCPX (100 nM) increased the frequency of interictal bursts from 0.027 ⫾ 0.001 to 0.04 ⫾ 0.01 Hz (P ⬍ 0.05;
n ⫽ 3; Fig. 6, B and C). In spite of their increase in frequency,
the shape and the area under GDPs was unchanged (9.4 ⫾ 2.1
and 7.1 ⫾ 1.2 mV s⫺1, P ⬎ 0.05 before and during DPCPX
application). This indicates that A1 receptors localized on
recurrent glutamatergic collaterals are involved in adenosine
action. Furthermore, application of adenosine (10 ␮M; n ⫽ 3)
completely blocked bicuculline-induced interictal bursts, an
effect that was associated with a membrane hyperpolarization
(5.5 ⫾ 0.5 mV; Fig. 6A) and an increase in input conductance
(8 ⫾ 2%, P ⬍ 0.05).
DISCUSSION
The present experiments clearly show that early in postnatal
life, endogenous adenosine by A1 receptor subtypes, localized
on glutamatergic terminals projecting mainly to GABAergic
interneurons, exerts a powerful inhibitory action on networkdriven oscillatory events such as GDPs.
GDPs represent a hallmark of developmental networks
(Ben-Ari 2002), which, as already mentioned, are generated by
the neurotransmitter GABA (that at early stages of development is depolarizing and excitatory) and to a lesser extent by
the excitatory transmitter glutamate (Ben-Ari et al. 1989). Our
data suggest that GDPs could be completely blocked by low
micromolar concentrations of adenosine, which specifically
switches off only glutamatergic inputs, indicating a key role of
the latter in the induction of GDPs.
The transient expression of GDPs suggests that these membrane oscillations exert a functional role for only a limited
period of time during postnatal development, acting as coinci-
94 • OCTOBER 2005 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 15, 2017
further support the idea that endogenous adenosine contributes
to control the generation of GDPs.
In CA3 principal cells, NBTI and EHNA did not modify
spontaneous synaptic events that at this developmental stage
are mainly GABAA mediated (see following text). The frequency of sPSPs was 4.3 ⫾ 0.2 and 4.4 ⫾ 0.02 Hz before and
after NBTI, respectively, whereas it was 5.8 ⫾ 0.3 and 5.7 ⫾
0.2 Hz before and after EHNA, respectively (data not shown).
GDPs are generated by the synergistic action of glutamate
and GABA acting on AMPA and GABAA receptors, respectively (Bolea et al. 1999). Therefore they are readily blocked
either by DNQX or bicuculline (Ben-Ari et al. 1989; Bolea et
al. 1999).
In agreement with a previous report (Safiulina et al. 2005)
DPCPX did not modify the frequency or amplitude of sPSPs in
principal cells (the frequency of sPSPs was 5.5 ⫾ 0.7 and
5.4 ⫾ 0.9 Hz, whereas the amplitude was 3.9 ⫾ 0.4 and 3.8 ⫾
0.5 mV before and during DPCPX, respectively; n ⫽ 3). In our
experimental conditions (symmetrical chloride solutions)
sPSPs recorded at ⫺60 mV from principal cells can be attributed to the action of GABA and glutamate acting on GABAA
and AMPA/kainate receptors, respectively. Therefore to see
whether endogenous adenosine regulates glutamate release into
principal cells we blocked GABAA receptors with picrotoxin.
However, in keeping with the late development of glutamatergic connections (Ben-Ari et al. 1989; Hennou et al. 2002;
Hosokawa et al. 1994; Tyzio et al. 1999) in the presence of
picrotoxin (100 ␮M) glutamatergic events were merely detectable, suggesting that during the first postnatal week, spontaneous activity of principal cells is mainly explained by GABA
acting on GABAA receptors. DPCPX applied in the presence
of DNQX (10 ␮M) failed to modify spontaneous GABAAmediated synaptic activity (the frequency of spontaneous
events was 4.7 ⫾ 0.8 and 4.8 ⫾ 0.9 Hz in the presence or
absence of 100 nM DPCPX; n ⫽ 3; P ⬎ 0.05; data not shown).
These data suggest that, in control conditions, endogenous
adenosine exerts its action on GDPs through glutamatergic
terminals localized downstream of principal cells.
2801
2802
SAFIULINA, KASYANOV, GINIATULLIN, AND CHERUBINI
dent detection signals for increasing synaptic efficacy (Kasyanov et al. 2004). Coincidence detection would provide a
combinatorial code by which to achieve outcome that cannot
be generated by isolated stimuli (Spitzer 2004). Therefore early
membrane oscillations such as GDPs may be crucial for setting
the appropriate synaptic connections that lead to the establishment of the adult hippocampal circuit. A number of neurotransmitters and neuromodulators whose extracellular levels
are maintained in an activity-dependent manner regulate network activity. Recently, we have demonstrated that ATP
down-regulates GDPs and differently affects the release of
glutamate and GABA within the CA3 network (Safiulina et al.
2005). In turn, GDPs would control the concentration of
endogenous ATP at synapses. In the present experiments
DPCPX-sensitive changes in ATP-induced reduction of the
frequency of GDPs demonstrate that this nucleotide exerts its
action not only through P2 but also by P1 receptors activated
by the ATP breakdown product adenosine (Cunha et al. 1998;
Dunwiddie et al. 1997). Interestingly, ATP also induced an
increase of sPSPs that was unrelated to the activation of
adenosine receptors because it persisted in the presence of
DPCPX. In accord with a previous report (Safiulina et al. 2005)
ATP-induced increase in the frequency of sPSPs was demonJ Neurophysiol • VOL
strated to be dependent on the activation of presynaptic P2Y1
receptors present on GABAergic terminals.
Another potential source of endogenous adenosine is cyclic
AMP (cAMP), which can be released from neurons and converted by extracellular phosphodiesterases into AMP and
thereafter by an ecto-5⬘-nucleotidase to adenosine. Evidence
that this pathway is functional in the hippocampus has been
provided (Dunwiddie et al. 1997). In particular, early in postnatal development, cAMP can be produced after activation of
metabotropic glutamate receptors by endogenously released
glutamate (Strata et al. 1995). cAMP can exert opposite effects
on GDPs: it may up-regulate the frequency of GDPs by directly
enhancing GABA release from presynaptic nerve terminals
(Strata et al. 1995) and/or may depress GDPs by indirectly
interfering with glutamate release after having been converted
into adenosine. However, to reach a significant level, multiple
cells must release cAMP over a prolonged period of time
(Brundege et al. 1997).
On the basis of their molecular structures four different
adenosine P1 receptors have been characterized: A1, A2A, A2B,
and A3 (Ralevic and Burnstock 1998). These receptors, which
belong to the G-protein– coupled receptor family, are differentially distributed in the CNS where they exert different phys-
94 • OCTOBER 2005 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 15, 2017
FIG. 5. Endogenous adenosine depresses
the glutamatergic drive to GABAergic interneurons. A: glutamatergic sPSPs recorded
from a GABAergic interneuron in the presence
of bicuculline (20 ␮M) in control (top) and
during application of DPCPX (100 nM, bottom). Note that DPCPX increased the frequency of glutamatergic events. B: cumulative
distributions of interevent intervals (left) and
amplitude (right) of glutamatergic sPSPs before (continuous line) and during (dashed line)
application of DPCPX for the cell shown in A.
Note that DPCPX increased the frequency but
not the amplitude of glutamatergic events. C:
glutamatergic sPSPs in control (top) and during application of adenosine (bottom). D: cumulative distributions of inter event intervals
(left) and amplitude (right) of glutamatergic
sPSPs before (continuous line) and during
(dashed line) application of adenosine for the
cell shown in A. Adenosine produced strong
inhibition of the frequency but did not change
the amplitude of glutamatergic events. E: each
column represents the mean frequency of
sPSPs (as percentage of control), obtained in
DPCPX (100 nM, n ⫽ 3) and adenosine (10
␮M). F: each column represents the mean
amplitude of sPSPs (as percentage of control),
obtained in DPCPX (100 nM, n ⫽ 3) and
adenosine (10 ␮M, n ⫽ 3), **P ⬍ 0.01.
ADENOSINE AND GDPS
iological functions. In the hippocampus, in situ hybridization
and RT-PCR experiments have demonstrated that, whereas
mRNA for A1 receptors is highly expressed, that for A2B and
A3 receptors is below the detection level (Dixon et al. 1996).
Moreover, immunocytochemical experiments have demonstrated the presence of A1 receptors on both pre- and postsynaptic sites (Ribeiro et al. 2003b).
Although at least in the cerebral cortex of immature animals
[3H] DPCPX binding sites represent only 23% of those found
in the adult brain, the present results clearly show that in the
hippocampus A1 receptors are present and already functional
during the first week of postnatal life. However, our data
suggest a prevailing action of adenosine on presynaptic A1
receptors. First adenosine reduced the frequency but not the
amplitude or the shape of GDPs and sPSPs. Second, on the
majority of the cases adenosine did not alter the passive
membrane properties of the patched neurons. Adenosine-induced changes in membrane potential and input conductance
were only rarely observed and, when present, they were produced by relatively high concentrations of adenosine, which
completely blocked GDPs. This is in line with a previous study
from the CA3 region of the immature hippocampus, which has
shown a differential ontogenesis of pre- versus postsynaptic
J Neurophysiol • VOL
GABAB, serotonin, and adenosine G-coupled receptors that
control the same potassium and/or calcium channels (Gaiarsa
et al. 1995). The delayed maturation of postsynaptic responses
to adenosine may involve lack of postsynaptic receptors, absence of G-proteins, or uncoupling between G-proteins and
effectors.
In keeping with previous reports obtained from cultured
neurons (Yoon and Rothman 1991) or adult animals (Lupica et
al. 1992), the present data clearly show that in the immature
hippocampus adenosine affected only glutamatergic synaptic
transmission, as suggested by the reduction in frequency of
AMPA/kainate-mediated sPSPs in GABAergic interneurons.
Therefore by modulating the glutamatergic drive to GABAergic interneurons adenosine exerts a powerful control on the
generation of GDPs. Endogenously released adenosine also
produced a strong inhibitory effect on recurrent glutamatergic
synapses in the CA3 area, thought to be critical for generating
bursting activity (Miles and Wong 1983), as indicated by the
reduction of bicuculline-induced interictal discharges. In line
with the powerful anticonvulsant action of purinoceptor agonists (Dunwiddie 1999), a potent inhibitory role of adenosine
on bicuculline-induced bursting activity in adult hippocampal
slices has been described (Ault and Wang 1986). As for
adenosine-induced down-regulation of AMPA/kainate sPSPs,
adenosine-induced depression of interictal discharges observed
after block of GABAA-mediated inhibition can be attributed to
a reduction of calcium entry through a decrease or an increase
in calcium and potassium conductances present on glutamatergic nerve terminals, respectively (Trussell and Jackson 1987).
The concentration of endogenous adenosine produced during GDPs is unknown. However, the high sensitivity of GDPs
to submicromolar concentrations of exogenous adenosine
(IC50 ⫽ 0.52 ␮M) together with the strong facilitatory effect of
DPCPX suggests that the concentration of endogenous adenosine is probably in the range of the IC50 value and perhaps
could be increased during large depolarizations such as GDPs.
In keeping with this, raising the extracellular concentration of
adenosine with the adenosine transport inhibitor NBTI or with
the adenosine deaminase blocker EHNA (Ribeiro et al. 2003b)
induced a reduction in frequency of GDPs, thus providing
further evidence that endogenous adenosine exerts a powerful
control on the generation of GDPs.
In previous work, using a biochemical assay we noticed that
the concentration of ATP present in the extracellular space
diminished significantly in the presence of TTX, implying that
ATP was released in an activity-dependent manner. Because
adenosine is formed as the breakdown product of ATP from
nerves, this implies that the adenosine concentration is crucially linked to the ongoing neuronal activity. Thus an increase
in the frequency of GDPs may lead to the production of ATP
and adenosine, which in turn may exert a negative feedback on
GDPs by high-affinity presynaptic A1 receptors.
ACKNOWLEDGMENTS
This work was supported by a grant from Ministero Istruzione Universita’ e
Ricerca (COFI 2003) to E. Cherubini, FIRB (Italy), and RFBR (Russia) to
V. F. Safiulina and R. Giniatullin.
GRANTS
This work was supported by a grant from Ministero Istruzione Università e
Ricerca (COFI 2003) to E. Cherubini, Fondo per gli Investimenti della Ricerca
94 • OCTOBER 2005 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 15, 2017
FIG. 6. Endogenous adenosine down-regulates interictal discharges recorded in CA3 principal cells after blockade of ␥-aminobutyric acid type A
(GABAA) receptors with bicuculline. A: representative trace of interictal
bursts recorded from a CA3 pyramidal neuron after application of bicuculline
(20 ␮M) for more than 10 min. Adenosine blocked interictal discharges an
effect that was associated with a membrane hyperpolarization. B: DPCPX (100
nM) increased the frequency of interictal discharges, suggesting activation of
A1 receptors localized on recurrent collaterals of CA3 pyramidal cells. Interictal bursts marked with a and b are shown below in an expanded timescale. C:
each column shows the mean frequency of interictal discharges in control and
in the presence of DPCPX (100 nM, n ⫽ 3), *P ⬍ 0.05.
2803
2804
SAFIULINA, KASYANOV, GINIATULLIN, AND CHERUBINI
di Base (Italy), and Russian Foundation for Basic Research to V. F. Safiulina
and R. Giniatullin.
REFERENCES
J Neurophysiol • VOL
94 • OCTOBER 2005 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 15, 2017
Ault B and Wang CM. Adenosine inhibits epileptiform activity arising in
hippocampal area CA3. Br J Pharmacol 87: 695–703, 1986.
Avignone E and Cherubini E. Muscarinic receptor modulation of GABAmediated giant depolarizing potentials in the neonatal rat hippocampus.
J Physiol 518: 97–107, 1999.
Ben-Ari Y. Excitatory actions of GABA during development: the nature of the
nurture. Nat Rev Neurosci 3: 728 –739, 2002.
Ben-Ari Y, Cherubini E, Corradetti R, and Gaiarsa JL. Giant synaptic
potentials in immature rat CA3 hippocampal neurones. J Physiol 416:
303–325, 1989.
Ben-Ari Y, Khazipov R, Leinekugel X, Caillard O, and Gaiarsa JL.
GABAA, NMDA and AMPA receptors: a developmentally regulated “ménage à trois. ” Trends Neurosci 20: 523–529, 1997.
Bolea S, Avignone E, Berretta N, Sanchez-Andres JV, and Cherubini E.
Glutamate controls the induction of GABA-mediated giant depolarizing
potentials through AMPA receptors in neonatal rat hippocampal slices.
J Neurophysiol 81: 2095–2102, 1999.
Brundege JM, Diao L, Proctor WR, and Dunwiddie TV. The role of cyclic
AMP as a precursor of extracellular adenosine in the rat hippocampus.
Neuropharmacology 36: 1201–1210, 1997.
Burnstock G. Introduction: P2 receptors. Curr Top Med Chem. 4: 793– 803,
2004.
Cherubini E, Gaiarsa JL, and Ben-Ari Y. GABA: an excitatory transmitter
in early postnatal life. Trends Neurosci 14: 515–519, 1991.
Cunha RA, Sebastião AM, and Ribeiro JA. Inhibition by ATP of hippocampal synaptic transmission requires localized extracellular catabolism by
ecto-nucleotidases into adenosine and channeling to adenosine A1 receptors.
J Neurosci 18: 1987–1995, 1998.
de Mendonça A and Ribeiro JA. Endogenous adenosine modulates long-term
potentiation in the hippocampus. Neuroscience 62: 385–390, 1994.
Dixon AK, Gubitz AK, Sirinathsinghji DJ, Richardson PJ, and Freeman
TC. Tissue distribution of adenosine receptor mRNAs in the rat. Br J
Pharmacol 118: 1461–1468, 1996.
Dunwiddie TV. Adenosine and suppression of seizures. Adv Neurol 79:
1001–1010, 1999.
Dunwiddie TV, Diao L, and Proctor WR. Adenine nucleotides undergo
rapid, quantitative conversion to adenosine in the extracellular space in rat
hippocampus. J Neurosci 17: 7673–7682, 1997.
Fredholm BB, Lindstrom K, and Wallman-Johansson A. Propentofylline
and other adenosine transport inhibitors increase the efflux of adenosine
following electrical or metabolic stimulation of rat hippocampal slices.
J Neurochem 62: 563–573, 1994.
Gaiarsa JL, Corradetti R, Cherubini E, and Ben-Ari Y. Modulation of
GABA-mediated synaptic potentials by glutamatergic agonists in neonatal
CA3 rat hippocampal neurons. Eur J Neurosci 3: 301–309, 1991.
Gaiarsa JL, Tseeb V, and Ben-Ari Y. Postnatal development of pre- and
postsynaptic GABAB-mediated inhibitions in the CA3 hippocampal region
of the rat. J Neurophysiol 73: 246 –255, 1995.
Giniatullin RA and Sokolova EM. ATP and adenosine inhibit transmitter
release at the frog neuromuscular junction through distinct presynaptic
receptors. Br J Pharmacol 124: 839 – 844, 1998.
Greene RW and Haas HL. Adenosine actions on CA1 pyramidal neurones in
rat hippocampal slices. J Physiol 366: 119 –127, 1985.
Hennou S, Khalilov I, Diabira D, Ben-Ari Y, and Gozlan H. Early
sequential formation of functional GABA(A) and glutamatergic synapses on
CA1 interneurons of the rat foetal hippocampus. Eur J Neurosci 16:
197–208, 2002.
Hosokawa Y, Sciancalepore M, Stratta F, Martina M, and Cherubini E.
Developmental changes in spontaneous GABAA-mediated synaptic events
in rat hippocampal CA3 neurons. Eur J Neurosci 6: 805– 813, 1994.
Jo YH, Stoeckel ME, and Schlichter R. Electrophysiological properties of
cultured neonatal rat dorsal horn neurons containing GABA and metenkephalin-like immunoreactivity. J Neurophysiol 79: 1583–1586, 1998.
Kasyanov AM, Safiulina VF, Voronin LL, and Cherubini E. GABAmediated giant depolarizing potentials as coincidence detectors for enhanc-
ing synaptic efficacy in the developing hippocampus. Proc Natl Acad Sci
USA 101: 3967–3972, 2004.
Khakh BS, Gittermann D, Cockayne DA, and Jones A. ATP modulation of
excitatory synapses onto interneurons. J Neurosci 23: 7426 –7437, 2003.
Khazipov R, Khalilov I, Tyzio R, Morozova E, Ben-Ari Y, and Holmes
GL.Developmental changes in GABAergic actions and seizure susceptibility in the rat hippocampus. Eur J Neurosci 19: 590 – 600, 2004.
Leinekugel X, Khazipov R, Cannon R, Hirase H, Ben-Ari Y, and Buzsaki
G. Correlated bursts of activity in the neonatal hippocampus in vivo. Science
296: 2049 –2052, 2002.
Leinekugel X, Medina I, Khalilov I, Ben-Ari Y, and Khazipov R. Ca2⫹
oscillations mediated by the synergistic excitatory actions of GABAA and
NMDA receptors in the neonatal hippocampus. Neuron 18: 243–255, 1997.
Lupica CR, Proctor WR, and Dunwiddie TV. Presynaptic inhibition of
excitatory synaptic transmission by adenosine in rat hippocampus: analysis
of unitary EPSP variance measured by whole-cell recording. J Neurosci 12:
3753–3764, 1992.
Maggi L, Sher E, and Cherubini E. Regulation of GABA release by nicotinic
acetylcholine receptors in the neonatal rat hippocampus. J Physiol 536:
89 –100, 2001.
McLean HA, Caillard O, Khazipov R, Ben-Ari Y, and Gaiarsa JL.
Spontaneous release of GABA activates GABAB receptors and controls
network activity in the neonatal rat hippocampus. J Neurophysiol 76:
1036 –1046, 1996.
Miles R and Wong RK. Single neurones can initiate synchronized population
discharge in the hippocampus. Nature 306: 371–373, 1983.
Payne JA, Rivera C, Voipio J, and Kaila K. Cation-chloride co-transporters
in neuronal communication, development and trauma. Trends Neurosci 26:
199 –206, 2003.
Poisbeau P, Rene F, Egles C, Felix JM, Feltz P, and Schlichter R.
Characterization of functional GABAergic synapses formed between rat
hypothalamic neurons and pituitary intermediate lobe cells in coculture:
Ca2⫹ dependence of spontaneous IPSCs. J Neurosci 16: 4835– 4845, 1996.
Ralevic V and Burnstock G. Receptors for purines and pyrimidines. Pharmacol Rev 50: 413– 492, 1998.
Ribeiro JA, Lobo MG, and Sebastiao AM. Endogenous adenosine modulation of 22Na uptake by rat brain synaptosomes. Neurochem Res 28:
1591–1595, 2003a.
Ribeiro JA, Sebastiao AM, and de Mendonça A. Adenosine receptors in the
nervous system: pathophysiological implications. Prog Neurobiol 68: 377–
392, 2003b.
Rivera C, Voipio J, Payne JA, Ruusuvuori E, Lahtinen H, Lamsa K,
Pirvola U, Saarma M, and Kaila K. The K⫹/Cl⫺ co-transporter KCC2
renders GABA hyperpolarizing during neuronal maturation. Nature 397:
251–255, 1999.
Safiulina VF, Kasyanov AM, Sokolova EM, Cherubini E, and Giniatullin
RA. ATP contributes to the generation of network-driven giant depolarizing
potentials in the neonatal rat hippocampus. J Physiol 565: 981–992, 2005.
Spitzer NC. Coincidence detection enhances appropriate wiring of the nervous
system. Proc Natl Acad Sci USA 101: 5311–5312, 2004.
Strata F, Sciancalepore M, and Cherubini E. Cyclic AMP-dependent
modulation of giant depolarizing potentials by metabotropic glutamate
receptors in the rat hippocampus. J Physiol 489: 115–125, 1995.
Trussell LO and Jackson MB. Dependence of an adenosine-activated potassium current on a GTP-binding protein in mammalian central neurons.
J Neurosci 7: 3306 –3316, 1987.
Tyzio R, Represa A, Jorquera I, Ben-Ari Y, Gozlan H, and Aniksztejn L.
The establishment of GABAergic and glutamatergic synapses on CA1
pyramidal neurons is sequential and correlates with the development of the
apical dendrite. J Neurosci 19: 10372–10382, 1999.
Wu LG and Saggau P. Adenosine inhibits evoked synaptic transmission
primarily by reducing presynaptic calcium influx in area CA1 of hippocampus. Neuron 12: 1139 –1148, 1994.
Xie X, Hider RC, and Smart TG. Modulation of GABA-mediated synaptic
transmission by endogenous zinc in the immature rat hippocampus in vitro.
J Physiol 478: 75– 86, 1994.
Yoon KW and Rothman SM. Adenosine inhibits excitatory but not inhibitory
synaptic transmission in the hippocampus. J Neurosci 11: 1375–1380, 1991.