| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1 Institut de Génomique Fonctionnelle, CNRS UMR5203/INSERM U661/UM1/UM2, Montpellier, France
2
INMED, INSERM U29, Marseille, France
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
(Received 13 July 2005;
accepted after revision 5 August 2005;
first published online 11 August 2005)
Corresponding author P. Tosetti: IGF, 144 rue de la Cardonille, 34094, Montpellier cedex 05, France. Email: ptosetti{at}igf.cnrs.fr
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
-aminobutyric acid (GABA)-receptor mediated synaptic transmission (Chen et al. 1999; Khalilov et al. 2003). Impaired GABAB receptor (GABABR)-mediated signalling has been observed after epileptic activity in the adult hippocampus (Mangan & Lothman, 1996; Wu & Leung, 1997; Haas et al. 1996; Chandler et al. 2003). In the developing hippocampus, however, it is not clear whether GABABR-mediated signalling is functional after seizures. GABABRs play a crucial role in controlling the physiological activity of the immature hippocampal network. GABAA currents, which mediate most of the inhibitory drive in the adult, are excitatory during the first postnatal week of life (Cherubini et al. 1991). Consequently, the synchronous activation of hippocampal interneurones results in excitatory network discharges called giant depolarizing potentials (GDPs) (Ben-Ari et al. 1989). In these peculiar conditions, GABABR activation is essential to limit network excitability. Endogenous GABA released during GDPs activates GABABRs, thus reducing further GABA release from interneurones and leading to GDP termination (McLean et al. 1996). Consistent with this, the pharmacological block of GABABRs progressively increases GDP width, transforming GDPs into epileptiform discharges (McLean et al. 1996; Tosetti et al. 2004). Similar effects can also be observed after the functional impairment of GABABR-mediated presynaptic inhibition of GABA release (GABA auto-inhibition) by a prolonged application of baclofen, a selective GABABR agonist (Tosetti et al. 2004). This observation raises the interesting possibility that a sustained release of endogenous GABA may trigger a functional loss of GABA auto-inhibition, leading to hyperexcitability and pathological network activity. Ictal discharges, during which both GABA and glutamate are continuously released for several seconds (Velazquez & Carlen, 1999; Kohling et al. 2000; Rutecki et al. 1985), are one possible context for such loss to occur.
Here we report that extensive endogenous GABA release during ictal-like discharges impairs GABA auto-inhibition in the newborn rat hippocampus. Such impairment contributes to the network hyperexcitability observed following the epileptiform discharges. These data thus reveal a novel role of GABABR-mediated signalling in acute post-seizure plasticity of the developing hippocampus.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Brains were removed from anaesthetized (350 mg kg–1 chloral hydrate I.P.) male Wistar rats (postnatal day 2–5) and immersed into ice-cold (2–4°C) artificial cerebrospinal fluid (ACSF) of the following composition (mM): NaCl, 126; KCl, 3.5; CaCl2, 2; MgCl2, 1.3; NaH2PO4, 1.2; NaHCO3, 25; and glucose, 11; pH 7.4, when equilibrated with 95% O2 and 5% CO2. All animal experiments were carried out in accordance with the European Communities Council Directive of 24 November 1986 (86/609/EEC). Intact hippocampal formations (IHFs) were dissected out as previously described (Khalilov et al. 1997), and incubated in oxygenated ACSF at room temperature. When needed, 600 µm slices were prepared from IHFs using a McIlwain tissue chopper. Both intact IHFs and slices were equilibrated at room temperature in oxygenated ACSF for at least 60 min prior to electrophysiological recordings.
Recording solutions
ACSF, the extracellular recording solution, was prepared from a 10x stock solution immediately before the experiment, equilibrated with 95% O2 and 5% CO2, heated at 34°C, and delivered to the recording chamber via a pressurized system at a flow rate of 3 ml min–1 (slices), or 10 ml min–1 (IHF).
For whole-cell recordings from hippocampal slices, intracellular solution contained (mM): 150 CsCl2, 2 MgCl2, 0.1 CaCl2, 1 EGTA, 2.5 Na2ATP, 0.4 Na2-GTP, 1.5 Mg-ATP, 10 Hepes, pH = 7.25, 2QX314, osmolarity = 275 mosmol l–1. In these ionic conditions, both ionotropic GABAergic and glutamatergic currents reversed around 0 mV. For patch-clamp recordings of the IHF, the intracellular solution contained (mM): 120 Cs-gluconate, 13.4 CsCl, 10 Hepes, 1.1 EGTA, 0.1 CaCl2, 0.4 Na2GTP, 4 MgATP, pH = 7.4. In these ionic conditions, GABAA currents reversed around –50 mV, while ionotropic glutamatergic currents reversed around +10 mV. For all patch-clamp experiments, 5,6-tetramethylrhodamine biocytin (rhodamine) was dissolved (0.5–1%) in the internal solution to allow post hoc visualization of the recorded neurones. For extracellular field recordings, pipettes were filled with ACSF.
Electrophysiological recordings
CA3 pyramidal neurones were recorded from either hippocampal slices or IHFs (Khalilov et al. 1997). Spontaneous and evoked synaptic activities were acquired with an Axopatch 2B patch-clamp amplifier (Axon Instruments, USA), filtered at 10 kHz, digitized using a Digidata 1200 A/D interface (Axon Instruments, USA) and stored on a Pentium II 450 MHz PC running Axoscope 8.1 (Axon Instruments, USA) for off-line analysis. Field potentials were recorded using a DAM80 Amplifier (World Precision Instruments, USA) with a 1–3 Hz band-pass filter. Capacitance, input and series resistances were measured online with Acquis Software (Biologic, France). Patch-clamp pipettes had a resistance of 3–5 M
, once filled with the internal recording solution. Series resistances ranged from 15 to 30 M. Cells showing series resistances >30 M were discarded. Field recordings were preformed with glass micropipettes of 10–20 M resistance. To reserve the viability of the IHF preparation and avoid ischaemic artefacts, each IHF was recorded for no more than to 2 h.
Impairment of GABA auto-inhibition
In both hippocampal slices and IHFs, functional impairment of GABA auto-inhibition was induced via ictal-like discharge (ILD)-promoted sustained release of endogenous GABA (Fig. 1A). ILDs were triggered by the bath application (9.1 ± 0.9 min, range 8–10 min) of 8.5–10 mM K+. When either 500 µM CGP35348 or 10 µM CNQX +40 µM APV were coapplied with high K+, the cumulative discharge time was monitored to guarantee GABA release consistency among pharmacologically distinct induction phases. In high K+, ILDs represented 41 ± 3% of the overall induction duration. In high K+ + CGP35348, the ILD contribution was not significantly different (39 ± 2%, P > 0.1), although the presence of CGP35348 slightly shortened the ILD duration (high K+: 90 ± 5 s, n = 30; high K+ + CGP35348: 76 ± 3 s, n = 12, P < 0.05) and increased their frequency (high K+: 4.1 ± 0.1 mHz; high K+ + CGP35348: 5.6 ± 0.2 mHz; P < 0.01). In high K+ + CNQX + APV, ILDs were not present, but GABA was still released via short, frequent bursts that constituted 63 ± 4% of the induction phase. When ILDs were induced by 500 µM CGP35348 alone, they represented 44 ± 5% of the induction phase.
|
All recordings were visualized using Axoclamp 8.1 software (Axon Instruments, USA), and analysed using the Acquis (Biologic, France), Minianalysis (Synaptosoft, USA) and Igor software (Wavemetrics, USA).
Paired pulse depression (PPD) of GABAergic synapses impinging onto CA3 pyramidal neurones was measured using pairs of identical stimuli (4–15 V, 0.03 ms each) at 250–350 ms interval, delivered at a frequency of 0.03 Hz with a bipolar tungsten electrode placed in the CA3 stratum radiatum. The resulting pairs of postsynaptic GABAA currents (GABAA-PSCs) were isolated in the presence of 10 µM CNQX and 40 µM D-APV. PPD was measured from the amplitude difference between the first and second GABAA-PSCs, normalized to the amplitude of the first GABAA-PSC. Average PPD values were calculated from 15 paired stimulations delivered at 0.03 Hz starting either 15 min before (control PPD) or 5 min after the different pharmacological treatments. For baclofen experiments, a 10 min ACSF wash was introduced between the end of the baclofen treatment and the restarting of the PPD stimulations, to allow a thorough washout of the drug. Cells showing differences in the amplitude of the first GABAA-PSC before and after treatment were discarded.
Heterosynaptic depression of glutamatergic synapses impinging on CA3 pyramidal neurones was measured by stimulating two independent glutamatergic pathways via two stimulating electrodes positioned in the stratum radiatum on opposite sides of the recording electrode. The electrode position and stimulus intensity were optimized for maximal field EPSP (fEPSP) amplitude. Stimulated pathways were tested for independence by verifying the absence of cross-facilitation. To induce heterosynaptic depression, a single stimulus was given to the first pathway (unconditioned fEPSP). After 30 s, an identical stimulus was given to the same pathway, immediately (300 ms) preceded by a conditioning train (10 stimuli at 50 Hz) to the second pathway. This originated a conditioned fEPSP. The magnitude of heterosynaptic depression was calculated from the amplitude difference between the unconditioned and conditioned fEPSPs, normalized to the amplitude of the unconditioned fEPSP. All recordings were performed in the presence of 20 µM tiagabine, a selective blocker of GABA neuronal transporter 1.
To characterize the physiological activity of the hippocampal network, we measured the amplitude and duration of the GDPs that are spontaneously present in the untreated IHF. GDP duration was measured at 20% of GDP amplitude. To characterize the pathological network activity following the different pharmacological treatment, we measured the amplitude, duration and shape of abnormal events recorded extracellularly from the IHF. Event shape was quantified by the number of phases within each event, where one phase is defined as one change in the slope of the recorded trace. Amplitude and duration of abnormal events were normalized to the respective GDP values from the same cell. Frequency distributions of event amplitude, duration and shape were built for individual cells. These were then averaged to generate mean frequency histograms for each parameter.
Ionotropic GABA and glutamatergic currents underlying network events were isolated by means of their distinct reversal potentials while using Cs-gluconate internal solution. GABA and glutamate charge transfers for each event were then calculated from the areas of the GABAergic and glutamatergic currents underlying the event itself.
Statistical analysis was performed with Kolmogorov–Smirnov (K–S), ANOVA or Student's t-tests, as appropriate. All values were expressed as mean ± standard error (S.E.M.). Differences were considered significant when P < 0.05.
Post hoc morphological characterization of recorded cells
After recordings, slices were fixated overnight at 4°C in 0.1 M phosphate buffer saline (PBS) containing 4% paraformaldehyde and 0.9% NaCl. Slices were then rinsed in PBS, mounted on gelatin-coated slides with an aqueous mounting medium (Gel Mount, Biomeda), and coverslipped. Rhodamine-filled cells were identified with an Olympus confocal microscope (Fluoview BX50WI) using a helium/neon laser as the excitation source (543 nm) and an emission filter band-pass (560 nm). Images were taken with a digital camera (ORCA-ER, Hamamatsu). Cells were classified as either pyramidal neurones or interneurones on the basis of their morphology and axonal projections. Electrophysiological data from interneurones were discarded.
Drug delivery
Drugs were diluted from stock solutions into freshly prepared ACSF immediately before the experiment. Drugs used in this study were: Baclofen (Sigma), 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX; Tocris), D(–)2-amino-5-phosphovaleric acid (D-AP5; Tocris), P-3-aminopropyl-P-diethoxymethyl phosphoric acid (CGP35348, Novartis, Basel), tiagabine (Novo Nordisk, Denmark).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Our working hypothesis was that, during epileptiform activity, the sustained exposure of GABAB receptors (GABABRs) to elevated quantities of endogenous GABA could impair GABABR-mediated presynaptic inhibition of GABA release (GABA auto-inhibition). Such functional loss would then increase hyperexcitability and induce spontaneous pathological discharges in the newborn rat hippocampus.
To test this idea, we first investigated whether ictal-like discharges (ILDs) released enough GABA to impair GABA auto-inhibition. ILDs were induced by application of 8.5–10 mM K+ (9.1 ± 0.9 min, range 8–10 min) to either hippocampal slices or the intact hippocampal formation (IHF). We chose this acute model of epilepsy because elevation of [K+]o occurs during seizures and contributes to their generation (Freund et al. 1990; Jensen & Yaari, 1997). High-K+ induced interictal-like activity followed by ILDs lasting 10–20 s in slices, and 40–120 s in the IHF (Fig. 1A). ILDs always started with tonic oscillations at 10–20 Hz and ended with clonic bursts occurring at 1–2 Hz (Fig. 1A). In most cases, an intermediate phase of tonico-clonic bursts at variable frequency could also be observed (Fig. 1A). Paired extracellular field recordings from the stratum radiatum, and whole-cell recordings from CA3 pyramidal neurones demonstrated that GABA was continuously released over tens of seconds before and during ILDs in the IHF (Fig. 1A, upper trace). These are therefore ideal conditions for inducing a functional impairment of GABA auto-inhibition, since GABAB receptors are exposed to a sustained and intense activation during each discharge.
Paired-pulse depression and heterosynaptic depression as indices of efficacy of GABABR-mediated presynaptic inhibition
Paired-pulse depression at GABAergic synapses (PPD) (Davies et al. 1990) and heterosynaptic depression at glutamatergic synapses (Chandler et al. 2003) were used to monitor the efficacy of presynaptic GABABR-mediated inhibition in hippocampal slices.
Both PPD and heterosynaptic depression depended upon GABAB receptor activation, as they were severely reduced by the selective GABABR antagonist, CGP35348. As shown in Fig. 1B, average PPD was reduced from 35 ± 6% (n = 6) in control conditions to 7 ± 2% in the presence of 500 µM CGP35348 (n = 6, P < 0.005). Similarly, heterosynaptic depression was reduced from 35 ± 3% to 12 ± 3% in the presence of the GABABR antagonist (Fig. 1C). Both PPD and heterosynaptic depression were fully restored following CGP35348 washout. These results indicate that both PPD and heterosynaptic depression rely upon activation of presynaptic GABAB receptors in our experimental conditions. The functional impairment of presynaptic GABABR-mediated inhibition should therefore result in a reduction of both PPD and heterosynaptic depression.
In a previous work (Tosetti et al. 2004), we showed that, among the different GABAB receptor-mediated responses, only the presynaptic inhibition of GABA release was impaired after a prolonged application of baclofen. We confirmed this result by measuring PPD and heterosynaptic depression before and after a 5 min application of 100 µM baclofen, a GABABR agonist. While in the bath, baclofen induced a decrease in the probability of GABA release, resulting in a reduction of the first pulse amplitude (Fig. 2A, left panel), and a switch from depression to facilitation (control: 26 ± 5%; during baclofen: –28 ± 16%; n = 6, P < 0.005) (Fig. 2A control, during baclofen). Interestingly, PPD did not recover once baclofen was washed off, and it was still absent (–1 ± 2%) 20 min after removal of the agonist (n = 6, P < 0.005) (Fig. 2A, after baclofen). This could not be accounted for by an incomplete baclofen washout, since the amplitude of the first pulse recovered to pre-baclofen values (Fig. 2A, left panel) and the facilitation disappeared (Fig. 2A, right panel). These results therefore suggest a baclofen-induced loss of GABA auto-inhibition.
|
In conclusion, these results confirm that sustained activation of GABABRs results in a selective decrease in the efficacy of presynaptic GABABR-mediated inhibition of GABA release. These results also show that PPD and heterosynaptic depression can be used as indices of functional impairment of GABABR-mediated presynaptic inhibition in conditions of conserved GABA uptake and release probability (Roepstorff & Lambert, 1994).
ILD-induced release of GABA impairs GABA auto-inhibition
We next investigated whether endogenous GABA released during ILDs could impair GABA auto-inhibition in hippocampal slices. Six ILDs were induced via an 8–10 min application of 10 mM K+ (induction phase), as described above. We measured PPD and heterosynaptic depression before and following the induction phase, and found that only PPD was severely reduced with respect to control (Fig. 3). Average PPD was decreased from 34 ± 4% in control conditions to 8 ± 3% after K+ treatment (n = 8; P < 0.001; Fig. 3A). It should be noted that the amount of residual PPD after K+ treatment is consistent with the amount of GABABR-independent PPD measured in the presence of CGP35348. In contrast, average heterosynaptic depression did not significantly differ before (33 ± 3%; Fig. 3B, control) and after the induction phase (43 ± 9%, n = 7; Fig. 3B, after high K). Thus, in agreement with the results of the baclofen treatment, high-K+ treatment selectively affected GABA auto-inhibition.
|
We next tested the possibility that PPD loss was due to ILDs reducing neuronal GABA uptake. PPD was measured in the presence of 20 µM tiagabine, a selective blocker of GABA neuronal transporter 1 (Andersen et al. 1993), both before and after ILDs. PPD in tiagabine was 48 ± 2% before and 8 ± 4% after high-K+ treatment (n = 9; P < 0.001; data not shown). ILDs therefore significantly reduced PPD even in the absence of functional neuronal GABA transporters, ruling out an impaired neuronal GABA uptake as the cause for PPD loss.
Since neither GABA uptake nor the probability of GABA release were affected by ILDs, functional impairment of presynaptic GABABR signalling is the most likely cause of high-K+-induced PPD loss. To confirm such conclusion, we tested whether PPD loss required the activation of GABABR by coapplying 500 µM CGP35348 during the high-K+ induction phase. Interestingly, no PPD loss was observed after the high-K+ + CGP35348 treatment (Fig. 4A). Average PPD was 27 ± 2%, not significantly different from that measured in control conditions, 27 ± 4% (P > 0.6; n = 9). Thus, these results show that activation of GABABRs by endogenous GABA is required to induce PPD loss.
|
In conclusion, our results show that endogenous GABA released during epileptiform activity can induce the functional impairment of GABA auto-inhibition in the newborn hippocampus. This effect requires a sustained GABA release. Intermittent, short GABAergic discharges mimicking physiological GDPs are ineffective.
ILD-induced functional impairment of GABA auto-inhibition modifies network activity in hippocampal slices
Our next goal was to investigate the acute and long-term consequences of a reduced GABA auto-inhibition on the immature hippocampal network. GDPs, the giant depolarizing potentials, constitute the hallmark of early hippocampal network activity (Ben-Ari et al. 1997). In hippocampal slices, GDP duration was increased by 44 ± 8% following six high-K+-induced ILDs (Fig. 5, after high K). Average GDP width (measured at 20% of GDP amplitude) increased from 300 ± 14 ms in control conditions to 432 ± 20 ms after ILDs (n = 6; P < 0.01). Such increase is consistent with a functional loss of GABABR-mediated presynaptic inhibition, since previous work demonstrated that GABABR activation is required for GDP termination (Tosetti et al. 2004).
|
We finally investigated the effects on GDP duration of a pattern of GABA release unable to impair GABA auto-inhibition. The short, frequent GABAergic bursts generated by the coapplication of high K+ and 10 µM CNQX + 40 µM APV (see previous section and Fig. 4B) did not significantly increase GDP width, which was 312 ± 49 ms in control conditions and 335 ± 44 ms after treatment (n = 8; P > 0.5). Thus, GABABR activation by endogenous GABA could not per se increase GDP width, unless it was long enough to impair GABA auto-inhibition. It is therefore the ILD-induced decrease of GABA auto-inhibition that is responsible for increasing GDP duration.
To further support this conclusion, we tested the effect on GDP width of a 5 min application of 100 µM baclofen, a treatment known to impair GABA auto-inhibition (Fig. 2) (Tosetti et al. 2004). As expected, GDP duration after baclofen treatment was significantly increased by 50 ± 8%, from 275 ± 8 ms to 410 ± 18 ms (n = 8, P < 0.001). Interestingly, this increase is not significantly different from that measured after high-K+ treatment (P > 0.4; Fig. 5A).
In conclusion, our results indicate that ILD-induced GABA release can increase GDP width and modify the spontaneous network activity by impairing GABABR-mediated GABA auto-inhibition in immature hippocampal slices.
ILD-induced impairment of GABA auto-inhibition modifies network activity in the intact hippocampal formation
Since hippocampal slices contain only a fraction of the hippocampal network, they may not be ideal to evaluate network activity. We therefore decided to further investigate the consequence of K+-induced ILDs using the intact hippocampal formation (IHF), a preparation that preserves the hippocampal circuitry (Khalilov et al. 1997).
Network events occurring before and after high-K+-induced ILDs were monitored using extracellular field recordings of CA3 stratum radiatum. Events were then classified according to their amplitude, duration and shape (quantified as number of phases, see Methods). We defined as abnormal all events that significantly differed from GDPs in at least two out of the three above parameters.
In control conditions, GDPs constituted the totality of network events (n = 24; Fig. 6A, Before high K+) and could be described as stereotyped triphasic waveforms of 21 ± 2 µV amplitude (range 8–43 µV; n = 24), and 484 ± 48 ms duration (range 131–692 ms). Following ILDs, however, GDPs accounted for only 26 ± 4% of total events (n = 24; Fig. 6A, After high K). The remaining 74 ± 4% of network activity was composed of abnormal events that were never recorded in control conditions. They could be divided into three major classes: short oscillations, long oscillations, and interictal-like events (Fig. 6B). Short oscillations were characterized by relatively small amplitude (172 ± 19% of average GDP amplitude; range 35–405%; n = 25), intermediate duration (424 ± 53% of average GDP duration; range 155–1161%) and intermediate number of phases (8.0 ± 0.8 phases; range 2–15; GDPs = 3 phases). Long oscillations not only displayed longer duration (3907 ± 797% of average GDP duration, range 1097–9658%; n = 10), but also larger amplitude (942 ± 334% of average GDP amplitude, range 320–3167%), and always more than 15 phases. Interictal-like events were mostly triphasic (average = 3.2 ± 0.2, range 2–5; n = 14), with intermediate amplitude (478 ± 94% of average GDP amplitude, range 103–1444%), and relatively short duration (196 ± 19% of average GDP duration, range 35–438%). All three types of abnormal events were spontaneously present in the IHF up to at least 1 h 30 min after the induction phase.
|
In addition to releasing endogenous GABA, and impairing the GABABR-dependent response, high-K+-induced ILDs also released large quantities of glutamate (Rutecki et al. 1985). Strong glutamate release could induce several forms of plasticity that could contribute to the observed changes in network activity (Ben-Ari & Gho, 1988). To identify which type of abnormal event depended upon the functional loss of GABA auto-inhibition, we prevented GABABR activation during sustained GABA release with the application of 500 µM CGP35348. This resulted in the recovery of most of the spontaneous GDP activity (70 ± 8% of total events, n = 12; Fig. 6A, After high K + CGP). The remaining 30% of events consisted exclusively of interictal-like waveforms. These results suggest that a decreased GABA auto-inhibition is required for the appearance of both long and short oscillations, but not interictal-like events in the immature hippocampus.
Figure 7 shows frequency distributions of abnormal event properties (duration, shape, and amplitude) measured after high K+ alone or high K+ + 500 µM CGP3534 (n = 24; for histogram values see Table 1). Superposed rectangles highlight properties whose frequency did not significantly differ in the presence or absence of GABABR activation. Interestingly, the duration, shape, and amplitude values that were independent of GABABR activation well matched the properties of interictal-like events. These results confirm that preserving GABABR-mediated signalling prevented the appearance of long and short oscillations.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Endogenous GABA impairs GABA auto-inhibition
We have demonstrated that high-K+-induced epileptiform activity can significantly reduce paired pulse depression (PPD) in the CA3 region of the developing rat hippocampus. PPD loss occurred after as few as six ILDs. Interestingly, a similar decrease in PPD could be observed in the CA1 region and in dentate gyrus granule cells after kindling (Buhl et al. 1996; Wu & Leung, 1997).
How does such functional loss occur? Our data indicate that a sustained release of endogenous GABA during ILDs induces a PPD loss that is the consequence of an impaired GABA auto-inhibition. To support this hypothesis, we showed that activation of GABABRs during high-K+-induced ILDs was required to induce the PPD loss since: (i) the application of the specific GABABR antagonist CGP35348 during the induction phase prevented it; and (ii) a 5 min baclofen application induced a similar PPD loss. We also proved that GABABR activation needed to be long lasting in order to induce the PPD loss. In fact, PPD was not reduced when the pattern of GABABR activation was changed from sustained (during ILDs) to intermittent (during GDP-like activity). Finally, we showed that other mechanisms potentially able to induce PPD loss, like a decreased probability of GABA release or an increased GABA uptake (Roepstorff & Lambert, 1994), were not involved. Altogether, these observations indicate that, in the newborn hippocampus, GABA released during ILDs impairs presynaptic GABABR-mediated inhibition, leading to PPD loss.
The mechanism leading to the functional impairment of presynaptic GABABR-mediated inhibition is presently unknown. The simplest explanation is possibly that the release of endogenous GABA is sufficient to induce a conformational change of the GABAB receptor, leading to a desensitized state insensitive to the agonist. For example, the impaired GABA auto-inhibition could result from uncoupling to the effectors via agonist-induced dephosphorylation of the GABAB receptors (Couve et al. 2002). Alternatively, it could result from phosphorylation by G-protein receptor kinases (GRKs) which, through the binding of ß-arrestins, target the G-protein-coupled receptors for endocytosis (Carman & Benovic, 1998; Ferguson et al. 1998; Tsao & von Zastrow, 2000). Independently of the mechanism involved, the present study provides the first demonstration that endogenous GABA can impair GABABR-mediated inhibition, and reveals one functional context where this occurs: during epileptiform activity.
Functional consequences of ILD-induced impairment of GABA auto-inhibition
We have shown that the ILD-induced functional loss of GABA auto-inhibition increased the duration of physiological GDPs and triggered pathological network oscillations in the intact hippocampus. In fact, procedures that prevented such impairment (coapplication of CGP35348 or CNQX + APV during the high-K+ induction phase) also prevented alterations of network activity. Furthermore, procedures that induced (bath application of baclofen) or simply mimicked impairment (non-saturating concentrations of CGP35348) triggered similar abnormal oscillations and increased GDP duration.
These results are consistent with previous studies reporting the fundamental role of GABABRs in preventing excessive hippocampal excitability during development. In fact, during the first postnatal week, the synchronous activation of GABAergic interneurones triggers depolarizing GABAAR-mediated currents, and contributes to spontaneous network discharges called giant depolarizing potentials (GDPs) (Cherubini et al. 1991). However, once released, GABA also acts on presynaptic GABABRs, limiting further GABA release, and effectively leading to GDP termination (McLean et al. 1996). Impairing GABABR function at this age prevents GDP termination, transforming GDPs into long epileptiform discharges (Tosetti et al. 2004). In agreement with these findings, our present data show that ILDs impair GABA auto-inhibition and decrease the inhibitory control exerted by endogenous GABA on the network activity. As a direct consequence, the hippocampal network develops a situation of hyperexcitability, revealed by the presence of longer GDPs, and abnormal long and short oscillations.
In the hippocampus, GABABRs are ubiquitously located at both postsynaptic and presynaptic sites on both GABAergic and glutamatergic neurones. Thus, ILDs can potentially impair the responses of distinct pools of GABABRs, producing complex effects on the hippocampal network. The present results show that GABABRs on glutamatergic terminals, in contrast to those on GABAergic terminals, are not impaired following high-potassium exposure.
We have previously shown that bath application of saturating concentration of baclofen does not impair postsynaptic GABABRs on pyramidal neurones (Tosetti et al. 2004). Considering the low number of ILDs used to induce such functional impairment, we think it unlikely that the postsynaptic GABABR-mediated control of excitability was affected in the present study. However, a recent study (Wetherington & Lambert, 2002) reported that both pre- and postsynaptic GABABRs on glutamatergic neurones desensitized when exposed to GABA agonists for longer periods (2–96 h). It is thus possible that a fully developed status epilepticus, allowing GABABRs to be exposed for longer and to higher concentration of endogenous GABA, could affect additional pools of GABABR-mediated responses. We however, predict that impairing the postsynaptic GABABR-mediated inhibition would contribute to aggravating hyperexcitability and favour the emergence of pathological activity.
It may seem contradictory that the effects of high-potassium exposure on PPD (in slices) and network activity (in the IHF) are mimicked by different concentrations of CGP35348 (500 versus 100 µM). This can be explained knowing that: (i) all (pre- and postsynaptic) GABAB receptors participate in the control of network-driven synaptic activity, including those that do not desensitize following high-K+-induced ILDs; GABA-PPD, however, is mediated mostly by GABABRs on GABAergic terminals; (ii) CGP35348 blocks all GABAB receptors within the network, while high K+ only impairs GABA auto-inhibition. Since high K+ impairs only one group of GABABRs out of the three controlling the network, a partial block of GABAB receptors (using non-saturating concentrations of CGP35348) best mimics the high-K+ effects on the network activity. Higher concentrations of the GABAB antagonist lead to a more dramatic phenotype (i.e. ILDs). In the case of GABA-PPD, however, high K+ impairs precisely the receptor group that is solely responsible for the phenomenon. Thus, an almost complete block of GABABRs on GABAergic terminals (using 500 µM CGP35348) is required to mimic the consequences of high-K+ treatment on GABA-PPD.
Although it is difficult to precisely mimic high-K+ effects on the network activity using a GABAB antagonist, we believe CGP experiments are helpful as they demonstrate that both sustained activation and partial blockade of the GABAB receptors lead to the same consequences on network activity. This observation therefore strengthens the hypothesis that the abnormal activity observed after high-K+ exposure is due to a ‘desensitization-like’ impairment of GABABR-mediated inhibition.
Impairment of GABA auto-inhibition as a possible ictogenic mechanism
Our data demonstrate that ILDs induce the loss of GABABR-mediated control of GABA release in the neonatal hippocampus, thus triggering network hyperexcitability and the appearance of spontaneous pathological discharges. Among the many seizures-induced alterations that characterize the epileptic tissue, some can underlie or facilitate the emergence of epileptiform activity, while others can be seen as compensatory mechanisms to prevent seizures aggravation (Morimoto et al. 2004). Whether GABABR activation is protective or pro-convulsive is still controversial, with effects varying greatly depending on the age, brain region and type of epileptiform activity (Engel., 1995; Bettler et al. 1998; Sutor & Luhmann, 1998; Motalli et al. 1999; Bettler et al. 2004; Tosetti et al. 2004). The protective role of GABABRs during development has been confirmed by the severe epileptic phenotype of GABAB1-deficient mice (Prosser et al. 2001; Schuler et al. 2001). In the adult, however, the activation of GABABRs is generally considered pro-convulsive because it limits the inhibitory action of GABAARs (Engel, 1995). It may thus appear contradictory to evoke a loss of presynaptic inhibition of GABA release as a pro-convulsive effect. It should be considered, though, that in both adult (Kohling et al. 2000; Fujiwara-Tsukamoto et al. 2003) and neonatal rat hippocampus (Dzhala & Staley, 2003), GABAA receptor activation contributes to the emergence of epileptiform activity. In the neonates, GABAA receptors are excitatory in physiological conditions (Cherubini et al. 1991). In the adult system, GABAA receptors are usually inhibitory, but the intracellular accumulation of Cl– ions during sustained GABAAR activation (i.e. ILDs) can switch GABAAR-mediated synaptic responses from hyperpolarizing to depolarizing, favouring network hyperexcitability (Staley et al. 1995; Kaila et al. 1997). Thus, the impairment of GABA auto-inhibition is likely to favour the emergence of pathological activity in both immature and adult animals. Several plastic changes occurring after seizures facilitate the onset of chronic epileptic activity (Morimoto et al. 2004). Our data provide, however, the first evidence that GABABRs may contribute to post-seizure plasticity and remodelling.
It is possible that seizure-induced loss of GABA auto-inhibition contributes to the long-term hyperexcitability observed in epileptic tissue. Several studies reported a functional loss of GABABR-mediated inhibition (Haas et al. 1996; Mangan & Lothman, 1996; Wu & Leung, 1997; Chandler et al. 2003) and a decrease in the number of GABABRs after intense epileptic activity in adult models of chronic epilepsy (Kokaia & Kokaia, 2001; Chandler et al. 2003; Straessle et al. 2003). However, the mechanisms underlying the long-term impairment of GABABR-mediated inhibition are not known. Our data raise the interesting possibility that seizure-induced GABA release contributes to the reduced number of GABABRs observed in chronic epileptic tissue. Indeed, the activity-dependent impairment of most G-protein-coupled receptors is mediated by phosphorylation followed by receptor internalization (Tsao & von Zastrow, 2000). It is tempting to speculate that a prolonged GABABR activation, initially induced by ictal activity and later maintained by pathological discharges, leads to receptor internalization and a long-term decrease in the number of functional GABABRs.
Our present results indicate that the functional loss of GABA auto-inhibition is a consequence of ILDs. However, a sustained release of endogenous GABA always precedes the actual development of the ILD, as is well documented in Fig. 1A. It is possible, although speculative, that functional impairment of GABABR-mediated GABA auto-inhibition occurs during such release. If this is the case, the reduced control of GABA release might not be simply a consequence of ILDs, but could play a role in the generation of pathological discharges.
Conclusion
In the neonatal hippocampus, during physiological activity, i.e. GDPs, GABABR signalling is recruited to limit the release of excitatory GABA. Here we show that, during epileptiform activity, an excessive release of GABA impairs such a protective mechanism, leading to aggravation of hyperexcitability and appearance of spontaneous pathological discharges. The ILD-dependent impairment of GABA auto-inhibition may thus contribute to the pro-convulsive plastic changes that characterize chronic epileptic tissues.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Ben-Ari Y, Cherubini E, Corradetti R & Gaïarsa J-L (1989). Giant synaptic potentials in immature rat CA3 hippocampal neurones. J Physiol 416, 303–325.
Ben-Ari Y & Gho M (1988). Long-lasting modification of the synaptic properties of rat CA3 hippocampal neurones induced by kainaic acid. J Physiol 404, 365–384.
Ben-Ari Y, Khazipov R, Leinekugel X, Caillard O & Gaïarsa J-L (1997). GABAA, NMDA and AMPA receptors: a developmentally regulated ‘ménage a trois’. Trends Neurosci 20, 523–529.[CrossRef][Medline]
Bender RA, Soleymani SV, Brewster AL, Nguyen ST, Beck H, Mathern GW & Baram TZ (2003). Enhanced expression of a specific hyperpolarization-activated cyclic nucleotide-gated cation channel (HCN) in surviving dentate gyrus granule cells of human and experimental epileptic hippocampus. J Neurosci 23, 6826–6836.
Bettler B, Kaupmann K & Bowery N (1998). GABAB receptors: drugs meet clones. Curr Opin Neurobiol 8, 345–350.[CrossRef][Medline]
Bettler B, Kaupmann K, Mosbacher J & Gassmann M (2004). Molecular structure and physiological functions of GABA (B) receptors. Physiol Rev 84, 835–867.
Buhl EH, Otis TS & Mody I (1996). Zinc-induced collapse of augmented inhibition by GABA in a temporal lobe epilepsy model. Science 271, 369–373.[Abstract]
Carman CV & Benovic JL (1998). G-protein-coupled receptors: turn-ons and turn-offs. Curr Opin Neurobiol 8, 335–344.[CrossRef][Medline]
Chandler KE, Princivalle AP, Fabian-Fine R, Bowery NG, Kullmann DM & Walker MC (2003). Plasticity of GABA(B) receptor–mediated heterosynaptic interactions at mossy fibers after status epilepticus. J Neurosci 23, 11382–11391.
Chen K, Baram TZ & Soltesz I (1999). Febrile seizures in the developing brain result in persistent modification of neuronal excitability in limbic circuits [see comments]. Nat Med 5, 888–894.[CrossRef][Medline]
Cherubini E, Gaïarsa J-L & Ben-Ari Y (1991). GABA: an excitatory transmitter in early postnatal life. Trends Neurosci 14, 515–519.[CrossRef][Medline]
Couve A, Thomas P, Calver AR, Hirst WD, Pangalos MN, Walsh FS, Smart TG & Moss SJ (2002). Cyclic AMP-dependent protein kinase phosphorylation facilitates GABA(B) receptor-effector coupling. Nat Neurosci 5, 415–424.[Medline]
Davies CH, Davies SN & Collingridge GL (1990). Paired-pulse depression of monosynaptic GABA-mediated inhibitory postsynaptic responses in rat hippocampus. J Physiol 424, 513–531.
Dzhala VI & Staley KJ (2003). Excitatory actions of endogenously released GABA contribute to initiation of ictal epileptiform activity in the developing hippocampus. J Neurosci 23, 1840–1846.
Engel J Jr (1995). Inhibitory mechanisms of epileptic seizure generation. Adv Neurol 67, 157–171.[Medline]
Ferguson SS, Zhang J, Barak LS & Caron MG (1998). Molecular mechanisms of G protein-coupled receptor desensitization and resensitization. Life Sci 62, 1561–1565.[CrossRef][Medline]
Freund TF, Gulyas AI, Acsady L, Gorcs T & Toth K (1990). Serotonergic control of the hippocampus via local inhibitory interneurons. Proc Natl Acad Sci U S A 87, 8501–8505.
Fujiwara-Tsukamoto Y, Isomura Y, Nambu A & Takada M (2003). Excitatory GABA input directly drives seizure-like rhythmic synchronization in mature hippocampal CA1 pyramidal cells. Neuroscience 119, 265–275.[CrossRef][Medline]
Haas KZ, Sperber EF, Moshé SL & Stanton PK (1996). Kainic acid-induced seizures enhance dentate gyrus inhibition by downregulation of GABAB receptors. J Neuroscience 16, 4250–4260.
Holmes GL & Ben-Ari Y (1998). Seizures in the developing brain: Perhaps not so begin after all. Neuron 21, 1231–1234.[CrossRef][Medline]
Holmes GL, Gaïarsa J-L, Chevassus-Au-Louis N & Ben-Ari Y (1998). Consequences of neonatal seizures in the rat: morphological and behavioral effects. Ann Neurol 44, 845–857.[CrossRef][Medline]
Jensen MS & Yaari Y (1997). Role of intrinsic burst firing, potassium accumulation, and electrical coupling in the elevated potassium model of hippocampal epilepsy. J Neurophysiol 77, 1224–1233.
Kaila K, Lamsa K, Smirnov S, Taira T & Voipio J (1997). Long-lasting GABA-mediated depolarization evoked by high-frequency stimulation in pyramidal neurons of rat hippocampal slice is attributable to a network-driven, bicarbonate-dependent K+ transient. J Neuroscience 17, 7662–7672.
Khalilov I, Esclapez M, Medina I, Aggoun D, Lamsa K, Leinekugel X, Khazipov R & Ben-Ari Y (1997). A novel in vitro preparation: the intact hippocampal formation. Neuron 19, 743–749.[CrossRef][Medline]
Khalilov I, Holmes GL & Ben Ari Y (2003). In vitro formation of a secondary epileptogenic mirror focus by interhippocampal propagation of seizures. Nat Neurosci 6, 1079–1085.[CrossRef][Medline]
Khazipov R, Leinekugel X, Khalilov I, Gaïarsa J-L & Ben-Ari Y (1997). Synchronization of GABAergic interneuronal network in CA3 subfield of neonatal rat hippocampal slices. J Physiol 498, 763–772.[Medline]
Kohling R, Vreugdenhil M, Bracci E & Jefferys JG (2000). Ictal epileptiform activity is facilitated by hippocampal GABAA receptor-mediated oscillations. J Neurosci 20, 6820–6829.
Kokaia Z & Kokaia M (2001). Changes in GABA (B) receptor immunoreactivity after recurrent seizures in rats. Neurosci Lett 315, 85–88.[CrossRef][Medline]
McLean HA, Caillard O, Khazipov R, Ben-Ari Y & Gaïarsa J-L (1996). Spontaneous release of GABA activates GABAB receptors and controls network activity in the neonatal rat hippocampus. J Neurophysiol 76, 1036–1046.
Mangan PS & Lothman EW (1996). Profound disturbances of pre- and postsynaptic GABAB-receptor- mediated processes in region CA1 in a chronic model of temporal lobe epilepsy. J Neurophysiol 76, 1282–1296.
Morimoto K, Fahnestock M & Racine RJ (2004). Kindling and status epilepticus models of epilepsy: rewiring the brain. Prog Neurobiol 73, 1–60.[CrossRef][Medline]
Motalli R, Louvel J, Tancredi V, Kurcewicz I, Wan-Chow-Wah D, Pumain R & Avoli M (1999). GABAB receptor activation promotes seizure activity in the juvenile rat hippocampus. J Neurophysiol 82, 647.
Prosser HM, Gill CH, Hirst WD, Grau E, Robbins M, Calver A, Soffin EM, Farmer CE, Lanneau C, Gray J, Schenck E, Warmerdam BS, Clapham C, Reavill C, Rogers DC, Stean T, Upton N, Humphreys K, Randall A, Geppert M, Davies CH & Pangalos MN (2001). Epileptogenesis and enhanced prepulse inhibition in GABA (B1) -deficient mice. Mol Cell Neurosci 17, 1059–1070.[CrossRef][Medline]
Roepstorff A & Lambert JDC (1994). Factors Contributing to the Decay of the Stimulus-Evoked IPSC in Rat Hippocampal CA1 Neurons. J Neurophysiol 72, 2911–2926.
Rutecki PA, Lebeda FJ & Johnston D (1985). Epileptiform activity induced by changes in extracellular potassium in hippocampus. J Neurophysiol 54, 1363–1374.
Schuler V, Luscher C, Blanchet C, Klix N, Sansig G, Klebs K, Schmutz M, Heid J, Gentry C, Urban L, Fox A, Spooren W, Jaton AL, Vigouret J, Pozza M, Kelly PH, Mosbacher J, Froestl W, Kaslin E, Korn R, Bischoff S, van der Kaupmann K, PH & Bettler B (2001). Epilepsy, hyperalgesia, impaired memory, and loss of pre- and postsynaptic GABA(B) responses in mice lacking GABA(B(1)). Neuron 31, 47–58.[CrossRef][Medline]
Staley KJ, Soldo BL & Proctor WR (1995). Ionic mechanisms of neuronal excitation by inhibitory GABAA receptors. Science 269, 977–981.
Straessle A, Loup F, Arabadzisz D, Ohning GV & Fritschy JM (2003). Rapid and long-term alterations of hippo
