GABA_A-current rundown of temporal lobe epilepsy is associated with repetitive activation of GABA_A “phasic” receptors

Abstract

A study was made of the “rundown” of GABA_A receptors, microtransplanted to Xenopus oocytes from surgically resected brain tissues of patients afflicted with drug-resistant human mesial temporal lobe epilepsy (mTLE). Cell membranes, isolated from mTLE neocortex specimens, were injected into frog oocytes that rapidly incorporated functional GABA_A receptors. Upon repetitive activation with GABA (1 mM), “epileptic” GABA_A receptors exhibited a GABA_A-current (I_GABA) rundown that was significantly enhanced by Zn²⁺ (≤250 μM), and practically abolished by the high-affinity GABA_A receptor inverse agonist SR95531 (gabazine; 2.5–25 μM). Conversely, I_GABA generated by “control” GABA_A receptors microtransplanted from nonepileptic temporal lobe, lesional TLE, or authoptic disease-free tissues remained stable during repetitive stimulation, even in oocytes treated with Zn²⁺. We conclude that rundown of mTLE epileptic receptors depends on the presence of “phasic GABA_A receptors” that have low sensitivity to antagonism by Zn²⁺. Additionally, we found that GABA_A receptors, microtransplanted from the cerebral cortex of adult rats exhibiting recurrent seizures, caused by pilocarpine-induced status epilepticus, showed greater rundown than control tissue, an event also occurring in patch-clamped rat pyramidal neurons. Rundown of epileptic rat receptors resembled that of human mTLE receptors, being enhanced by Zn²⁺ (40 μM) and sensitive to the antiepileptic agent levetiracetam, the neurotrophin brain-derived neurotrophic factor, and the phosphatase blocker okadaic acid. Our findings point to the rundown of GABA_A receptors as a hallmark of TLE and suggest that modulating tonic and phasic mTLE GABA_A receptor activity may represent a useful therapeutic approach to the disease.

The GABA_A receptors are presumed to be heteromeric pentamers formed by various combinations of α (α₁ to α₆), β (β₁ to β₃), γ (γ₁ to γ₃), δ, ε, π, θ, and ρ subunits, resulting in a great functional receptor heterogeneity, and different GABA_A receptor isoforms are located at synaptic and extrasynaptic sites, markedly influencing neuronal excitability. In general, synaptic receptor subtypes contain a γ₂ subunit, which is critical for maintaining receptors at synaptic sites, have low affinity for GABA, and mediate fast inhibitory transmission, the so-called phasic (transient) inhibition. In contrast, extrasynaptic receptor subtypes containing α₄, α₅, or α₆ subunits and/or a δ instead of a γ subunit, exhibit high affinity for GABA and mediate tonic (persistent) inhibition (1–6). Tonic currents are caused by GABA that escapes presynaptic uptake and produces slow neuronal inhibition through a persistent activation of extrasynaptic GABA_A receptors.

Impairement of GABA_A receptor-mediated inhibition increases neuronal excitability and plays a critical role during epileptogenesis (7). We have previously described a characteristic I_GABA rundown that occurs during repetitive stimulation of GABA_A receptors microtransplanted from the human mesial temporal lobe epilepsy (mTLE) brain into Xenopus oocytes. The microtransplantation is affected by injecting membranes, isolated from brain tissues resected from TLE patients, into frog oocytes (8). Repeated activation of the microtransplanted epileptic GABA_A receptors generates I_GABA that exhibit a relatively fast rundown followed by a slow recovery (9). These events are regulated by tyrosine kinase B (TrkB) and phosphatase activities (10, 11), and are probably critical for mTLE seizures because the GABA_A-current (I_GABA) rundown is suppressed by the potent antiepileptic drug levetiracetam (LEV; ref. 12).

To provide insights into the relationship between the specific GABA_A receptor isoform composition and the functional impairement of GABAergic inhibitory systems resulting in seizures, we have begun to explore the GABA_A receptor populations that mediate the I_GABA rundown. We took advantage of the fact that GABA_A receptors lacking the γ₂ subunit (extrasynaptic) are more sensitive to GABA and Zn²⁺ than receptors containing the γ₂ subunit, which have little sensitivity to Zn²⁺ (13, 14). For most of the present experiments we examined the I_GABA rundown in mTLE brain-microtransplanted oocytes exposed to external media containing Zn²⁺. In other experiments, we used the high-affinity inverse agonist of GABA_A receptors, gabazine, at concentrations that for rat GABA_A receptors barely block the receptors underlying tonic inhibition but antagonize those mediating phasic inhibition (1). Our central finding was that GABA_A receptors with very low sensitivity to Zn²⁺, which in our microtransplanted oocytes represented the major population of human receptors incorporated, are responsible for the GABA_A receptor rundown. Also, this study was extended to a rat model of epilepsy.

Results and Discussion

mTLE GABA_A-Current Rundown.

In agreement with analogous previous experiments (9–11,15), applications of GABA (1 mM) to oocytes injected with membranes from the neocortex of four mTLE patients [nos. 1–4; listed in supporting information (SI) Table 1] elicited inward currents ranging from −30 to −600 nA (depending on both the oocyte and the donor, i.e., frogs or patients). These currents were generated by activation of microtransplanted human ionotropic GABA_A receptors that were blocked by bicuculline and picrotoxin. A considerable rundown of I_GABA occurred after repetitive applications of the neurotransmitter (fall to 40 ± 2% at the sixth GABA application, range: 18–58%; 53 oocytes; 11 frogs, 53/11; patients nos. 1–4; Fig. 1 A and C). This rundown was not accompanied by a significant change in the current decay (T_0.5 = 7.1 ± 0.7 s under control conditions vs. T_0.5 = 7.2 ± 0.4 s at the sixth GABA application; P = 0.26); and recovered partially (≈60%) within 20 min after washout (data not shown; cf. ref. 10). A much smaller I_GABA rundown (fall to 70 ± 1.5%, range: 66–80%; 21/7; pooled data, see ref. 16; Fig. 1 A and B) was detected in oocytes injected with membranes isolated from nonepileptic human brain tissues (patients 5 and 6) and from resected lesional TLE brain tissues (LTLE; patients 7 and 8), again not accompanied by any change in T_0.5 (data not shown). I_GABA rundown from mTLE membrane-injected oocytes was not altered by changing the membrane holding potential (range tested: 0–100 mV; data not shown). Most of these findings have been previously observed and are confirmed here (see refs. 10, 11, 15, and 16).

Fig. 1.

GABA_A-receptor rundown in oocytes injected with human brain membranes. (A) Time course of current rundown in oocytes injected with temporal lobe membranes of three control (○) patients (5, 7, and 8; see SI Table 1; 21/11) and four mTLE patients (nos. 1–4; 45/11) before (●) and after Zn²⁺ (40 μM; □) or gabazine (10 μM; ■). Peak amplitudes of I_GABA were normalized to those elicited by the first neurotransmitter application (I_GABA : ○, 171 ± 19 nA; ●, 225 ± 22 nA; □, 178 ± 23 nA; ■, 24 ± 1 nA). In this and subsequent figures, points refer to means ± SEM. Data fitted to an exponential gave: τ = 41.7 s (○), τ = 43.5 s (●), τ = 42.2 s (□), and τ = 37 s (■). Note that Zn²⁺ and gabazine strongly affect mTLE current rundown without modifying the rate of the decay. (B) Representative currents in response to the first and sixth GABA applications from an oocyte injected with nonepileptic temporal lobe membranes (patient 5; SI Table 1) before (Left) and after 40 μM Zn²⁺ (Center) or 10 μM gabazine (Right) treatment. (C) Representative currents from another oocyte injected with membranes from temporal lobe of a mTLE patient (no. 1; SI Table 1) and treated as shown. Drug concentrations were as in B. In all cases the GABA concentration was 1 mM.

I_GABA Rundown During Selective Inhibition of GABA_A Receptor Populations.

Subsequent experiments were performed in oocytes injected with membranes isolated from the same mTLE patients (nos. 1–4) and treated with Zn²⁺, which efficiently blocks rat GABA_A receptors lacking the γ subunit (see ref. 14 and references therein). We have previously reported that, under identical experimental conditions, the Zn²⁺ IC₅₀ in oocytes injected with mTLE neocortex membranes was ≈500 μM (15). mTLE membrane-injected oocytes, treated with Zn²⁺ at concentrations known to block most of the GABA_A receptors lacking the γ subunit (10–200 μM; refs 13 and 14), exhibited I_GABA peak amplitude values impaired by 10–30% (e.g., Fig. 1 B and C; Zn²⁺ 40 μM, 18 ± 5%; Zn²⁺ 100 μM, 21 ± 6%, 35/9), thus providing an estimate of the proportion of γ-lacking, high Zn²⁺-sensitivity GABA_A receptors incorporated by microtransplanted oocytes. After Zn²⁺ treatment, mTLE GABA_A receptor stability (evaluated by the I_GABA rundown) was strongly decreased compared with Zn²⁺-untreated oocytes (Figs. 1C and 2). The Zn²⁺-induced increase of I_GABA rundown (i.e., decrease of GABA_A receptor stability) was concentration-dependent and reached a plateau at concentrations slightly >40 μM, indicating that most of the (tonic) Zn²⁺-sensitive GABA_A receptors were blocked at that Zn²⁺ concentration (Fig. 2). This increased rundown was not accompanied by changes in either the current decay (T_0.5 = 7.45 ± 0.7) or the GABA apparent affinity (EC₅₀ = 101 μM control). In contrast to mTLE receptors, Zn²⁺ again influenced the amplitude of I_GABA, but Zn²⁺ did not influence the small rundown observed in control (non-mTLE) oocytes (Fig. 1B). Together, these findings indicate that the increase in mTLE I_GABA rundown was associated with the activity of (phasic) GABA_A receptors comparatively insensitive to Zn²⁺.

Fig. 2.

Relationship between Zn²⁺ (Upper) or gabazine (Lower) concentrations and I_GABA rundown. Points represent the percentage of I_GABA rundown (6–35 oocytes for each dose; two frogs) from oocytes injected with mTLE membranes (patients 1–3). Results were fitted with a single exponential. The plateau concentration (rundown: 27% Upper and 87% Lower) was reached with ≈80 μM Zn²⁺ (Upper) and 10 μM gabazine (Lower).

To estimate the contribution of the Zn²⁺-sensitive high-affinity GABA_A receptors to the I_GABA rundown, microtransplanted oocytes were exposed to gabazine, a high-affinity inverse agonist of GABA_A receptors. Under conditions in which the low-affinity (phasic) GABA_A receptors are blocked by gabazine (1), the I_GABA generated by the activation of high-affinity (tonic) GABA_A receptors was picrotoxin-sensitive (data not shown) and exhibited an even greater stability than gabazine-untreated, and similar to gabazine-treated, nonepileptic membrane-injected oocytes (Fig. 1). Specifically, 25 μM gabazine reduced the I_GABA peak amplitude to 3.8 ± 1% (vs. mTLE gabazine-untreated oocytes), and during the rundown the I_GABA fell down by only ≈14% instead of the usual ≈60%. This low level of rundown was also observed at lower concentrations of gabazine (10 μM: peak amplitude ≈14%; 7.5 μM: peak amplitude ≈18%). With 2.5 μM gabazine, the rundown became much closer to that of mTLE gabazine-untreated oocytes (Fig. 2). Futhermore, the stability of I_GABA after gabazine treatment was accompanied by a marked decrease in current decay (T_0.5 >10 s, gabazine 10–25 μM), confirming a slow desensitization of tonic GABA_A receptors (17). The absence of rundown was observed also when GABA was applied at low concentrations, which probably activate only high-affinity GABA_A receptors (10–20 μM: I_GABA ranging from 10 to 150 nA; 9/2; data not shown). Moreover, treating mTLE oocytes for 2–3 min with a mixture of Zn²⁺ (200–250 μM) plus gabazine (10 μM), blocked the I_GABA by >95% (data not shown), supporting the notion that both tonic and phasic mTLE GABA_A receptors are incorporated by the oocyte membrane after microtransplantation. Taken together, these findings indicate that the rundown of mTLE GABA_A receptors is associated with low Zn²⁺-sensitivity, high gabazine-sensitivity phasic GABA receptors rather than Zn²⁺-sensitive tonic receptors.

In experiments aimed at evaluating the functional expression of microtransplanted α₅- containing GABA_A receptors, oocytes injected with human membranes isolated from the mTLE neocortex were exposed to the α₅-GABA_A receptor-specific inverse agonist L-655,708 (18). It was found that L-655,708 (50 μM; 2-min pretreatment) did not significantly modulate either the I_GABA amplitude (fall to 82 ± 2%; 8/2; P = 0.2) or the I_GABA rundown (50 μM; data not shown). We conclude that, under our conditions, the selective modulation of the activity of α₅-GABA_A receptors does not significantly affect I_GABA rundown.

We also tested a positive modulator of the δ-GABA_A receptor-mediated tonic current, the neurosteroid 5α-pregnane-3α,21-diol-20-one (THDOC; refs. 2 and 18). THDOC significantly increased I_GABA by 18 ± 3% (10/2; P = 0.3; THDOC, 50–100 nM; 2- to 60-min pretreatment). However, THDOC did not modify the I_GABA rundown (data not shown). Thus, THDOC, acting preferentially on δ-GABA_A receptors and positively modulating I_GABA, did not modify GABA_A receptor stability.

GABA_A-Current Rundown of Epileptic Rat Receptors.

We wondered whether the I_GABA rundown, observed in human mTLE, was a characteristic peculiar to human epileptic cortex receptors or is a more general adaptive phenomenon linked to the development of epilepsy. In the latter case, the rundown should be present in chronic animal models, and to examine this idea, we used the pilocarpine rat model. In this widely used model of epilepsy, an episode of status epilepticus, induced by the muscarinic agonist pilocarpine, produces, after a latent period of 2–3 weeks, the spontaneous occurrence of seizures (i.e., epilepsy). Therefore, we addressed experiments to investigate whether the I_GABA rundown was present in rats during the “chronic period,” i.e., 4 weeks after pilocarpine-induced status epilepticus. Oocytes injected with membranes isolated from the temporal lobe of two epileptic rats exhibited an obviously increased rundown, compared with two untreated rats (control tissue: fall to 71%, range 57–98%, 21/15; epileptic tissue: fall to 36.5%, range 12–54%, 52/15; Fig. 3A). Here, again, I_GABA rundown was not accompanied by a significant change in I_GABA decay (T_0.5 = 7.7 ± 0.4 s control vs. T_0.5 = 6.5 ± 0.4 s at the sixth GABA application, 52/15, P = 0.1). These findings were confirmed in whole-cell patch clamped temporal lobe pyramidal neurons in brain slices. Specifically, applications of GABA (100 μM) evoked whole-cell currents, with a maximal amplitude of 1.0 ± 0.4 nA (range 0.1–3.8 nA; n = 9) and 2.0 ± 0.8 nA (range 0.1–7.0 nA; n = 10; P > 0.1) for control or pilocarpine-treated animals, respectively. Repetitive GABA applications (1 s, every 15 s) induced a significantly greater current rundown in neurons of the pilocarpine-treated animals vs. controls (Fig. 3B): GABA response decreased to 63 ± 9% in the pilocarpine group (n = 4), vs. 87 ± 6% in controls (n = 6; P < 0.01). Interestingly, oocytes injected with rat membranes isolated from the temporal lobe 24 h after pilocarpine administration (i.e., acute period), exhibited a rundown significantly similar to untreated animals (pilocarpine: 71 ± 2%, 12/2, 2 rats; control: 70 ± 3%, 12/2; 2 rats; P > 0.2).

Fig. 3.

Properties of I_GABA rundown in control rats and rats experiencing recurrent seizures (epileptic). (A) (Upper) Representative currents from two oocytes injected with temporal lobe membranes from control or pilocarpine-treated rats. (Lower) Time course of current rundown (●, control rat; ○, pilocarpine-treated rats). I_GABA was normalized to those elicited by the first application (I_GABA: ○, 121 nA; ●, 135 nA). (B) Rundown of I_GABA from temporal cortex neurons in slices obtained from control and pilocarpine-treated adult rats. (Upper) Representative currents elicited by repetitive GABA (100 μM) applications (horizontal bars) to a neuron in a control temporal slice or in a temporal slice from a pilocarpine-treated rat, before and at the end of the rundown protocol (first and 10th applications, respectively; superimposed traces; holding potential −70 mV). (Lower) Time course averaged from six control neurons (●) or four neurons from pilocarpine-treated rats (○). Time 0 corresponds to the first GABA application of the rundown protocol. Current amplitudes were normalized to the first current amplitude.

In oocytes injected with human mTLE membranes, we have previously shown that the I_GABA rundown was reduced by the antiepileptic drug LEV, the neurotrophin brain-derived neurotrophic factor (BDNF), and the phosphatase blocker okadaic acid (10–12). Under the same experimental protocols used with the human mTLE membranes, all of those drugs antagonized the I_GABA rundown in oocytes injected with epileptic rat temporal lobe membranes (Fig. 4). Finally, experiments were addressed to see whether in epileptic rat membrane-injected oocytes the Zn²⁺-insensitive GABA_A receptor activity was also associated with GABA_A receptor rundown. Oocytes injected with neocortex membranes were pretreated with Zn²⁺ (40 μM) for 2 min before testing I_GABA in the presence of Zn²⁺. The amplitude of the first application of GABA (1 mM) was reduced by 26 ± 3%, whereas, in analogy with human mTLE oocytes, I_GABA rundown was obvious enhanced (Fig. 4).

Fig. 4.

Effect of drugs on I_GABA rundown from oocytes injected with epileptic rat TL membranes. Columns represent current amplitudes at the sixth GABA application (8–20/4) normalized to 98 ± 7 nA. The oocytes were pretreated with LEV (1 μM) or BDNF (500 ng/μl) for 3 hr and Zn²⁺ (40 μM) for 2 min. Okadaic acid (50 nM) was injected into oocytes 8–10 min before recordings. Control refers to untreated oocytes. *, P < 0.01; **, P < 0.001.

Together with previous reports, this work shows that the GABA-receptor rundown, characteristic of oocytes injected with mTLE membranes, is a desensitization process with special features, because it is not coupled to changes in the receptor's affinity for the agonist or altered by changes in membrane potential. Whatever are the mechanisms responsible for the I_GABA rundown, this process may be considered a feature common to epileptic GABA_A receptors because it is present in microtransplanted oocytes and native pyramidal neurons of a rat model of TLE.

We also report that human mTLE I_GABA rundown is generated by the repetitive activation of low-affinity phasic GABA_A receptors, remaining the high-affinity GABA_A receptors stable upon repetitive activation. Noteworthy, at least in an animal model of epilepsy, the tonic current is down-regulated after epileptogenesis (19).

Because the I_GABA rundown occurs also in ex vivo neurons (16), we speculate that development of seizures is favored by the instability of γ-GABA_A receptors. In agreement with current literature, our findings may open a road for therapeutic treatment of seizures by designing neuromodulators acting selectively on tonic or phasic GABA_Aergic populations. Such treatment could help in the mTLE disease by positively modulating GABA_A receptor stability upon seizures. Finally, an open question is whether Zn²⁺ released during seizures could depress GABA_A receptor function (20). It has been ascertained that Zn²⁺ transients, evoked during neurotransmission, may reach an apparent concentration of 10–30 μM (21), which is the same concentration influencing the I_GABA rundown (Fig. 1). We propose that Zn²⁺, coreleased with glutamate from nerve terminals, enhances the depression of the GABAergic inhibition, which is caused by the instability of synaptic γ-GABA_A receptors, through the blockage of tonic epileptic GABA_A receptors.

Materials and Methods

Patients.

Surgical specimens were obtained from the hippocampus and temporal neocortex of four patients afflicted with mTLE (SI Table 1), all operated at the Neuromed Neurosurgery Center for Epilepsy (Pozzilli-Isernia, Italy). The histopathology of all specimens showed the typical neuropathological features of Ammon's horn sclerosis and did not show obvious sclerosis in the temporal lobe. For comparative purposes, when indicated, we used specimens of (i) nontumoral nervous tissue resected from a patient (no. 5; SI Table 1) afflicted with left TL glioblastoma (TLG) who did not suffer epileptic episodes (nonepileptic TLG, grade IV), (ii) neocortical tissue from two patients afflicted with temporal lesion (LTLE; patients 7 and 8, SI Table 1), and (iii) hippocampal tissue from an authoptic disease-free brain (patient 6, SI Table 1). Informed consent was obtained from all of the patients to use part of the biopsy material for our experiments, and the Ethics Committees of Neuromed and the University of Rome “La Sapienza” approved the selection processes and procedures. For more details about patients and screening analysis see ref. 16 and SI Table 1.

Pilocarpine Model.

Male Sprague–Dawley rats (260–280 g; Harlan) were used for all in vivo experiments. Animals were housed under standard conditions: constant temperature (22–24°C) and humidity (55–65%), 12-h dark-light cycle, and free access to food and water. All efforts were made to minimize animal suffering. Procedures involving animals and their care were carried out in accordance with European Community and national laws and policies.

Pilocarpine was administered i.p. (300 mg/kg), and the rat's behavior was observed for several hours thereafter. Within the first hour after injection, all animals developed seizures evolving into recurrent generalized convulsions (status epilepticus; SE). SE was interrupted 2 h after onset by administration of diazepam (10 mg/kg i.p.). In the pilocarpine model of epilepsy, the episode of SE produces, after a latent period of 2–3 weeks, the spontaneous occurrence of seizures (i.e., epilepsy).

In the present experimental series, pilocarpine rapidly induced a robust convulsive SE (latency: 19 ± 3 min). Based on behavioral observation (22) and EEG recordings, the severity of SE in the different animals was indistinguishable. Rats underwent 20 days of behavioral monitoring to examine the outcome of the treatment in terms of spontaneous seizures. Spontaneous seizures began to occur 18 ± 2 days after SE. In the week before being killed (21–28 days after pilocarpine administration) animals experienced an average of 15 ± 2 seizures per day, with a severity score of 2.2 ± 0.2 (according to ref. 22). Rats were killed 28 days after SE.

Membrane Preparation, Injection Procedures, and Electrophysiological Recordings from Oocytes.

Membranes were prepared as detailed (8) with tissues from human epileptic brain regions (hippocampus and TL) or rat TL. Preparation of Xenopus laevis oocytes and injection procedures were as detailed (8). From 12 to 48 h after injection, membrane currents were recorded from voltage-clamped oocytes by using two microelectrodes filled with 3 M KCl. The oocytes were placed in a recording chamber (volume, 0.1 ml) perfused continuously (9–10 ml/min) with oocyte Ringer's solution at room temperature (20–22°C). I_GABA rundown was defined as the decrease (in %) of the I_GABA peak amplitude at the sixth GABA jet (current-plateau) after five applications of GABA (1 mM), 10-s duration, at 40-s intervals. The I_GABA desensitization was estimated by the time taken for the current to decay from its peak to half-peak value (T_0.5). In all of the experiments, the holding membrane potential was −60 mV. In the experiments with Zn²⁺ or gabazine, cells were preincubated for 2 min with the drugs and then exposed to GABA (1 mM) plus Zn²⁺ or gabazine at indicated doses.

Whole-Cell Recordings from Rat TL Slices.

Neocortical slices were prepared from adult rats (P36–P50) as described (23). Immediately after surgical resection, transverse slices (250 μm) were cut in sucrose-based artificial cerebrospinal fluid (ACSF) with a vibratome (VT 1000S; Leica Microsystems), placed in a slice incubation chamber at room temperature with oxygenated ACSF, and transferred to a recording chamber within 1–24 h after slice preparation. Whole-cell patch clamp recordings were performed on pyramidal neurons at 24–25°C as described (16). Membrane currents were recorded by using glass electrodes (3–4 MΩ) filled with 140 mM CsCl, 10 mM Hepes, 5 mM 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetate, and 2 mM Mg-ATP (pH 7.3, with CsOH). GABA was delivered to cells by pressure applications (10–20 psi; 1 s; Picospritzer II; General Valve) from glass micropipettes positioned above whole-cell voltage-clamped neurons. In this way we obtained stable whole-cell currents and rapid drug wash before applying the rundown protocol. The current rundown protocol adopted was the following: after current amplitude stabilization with repetitive applications every 120 s, a sequence of 10 GABA (100 μM) applications every 15 s was delivered; the test pulse was resumed at control rate (every 120 s) to monitor recovery of I_GABA. The reduction in peak amplitude current was expressed as percentage amplitude of current at the end of the rundown protocol vs. control (I_tenth/I_first%). For more details see ref. 16.

Chemicals and Solutions.

Oocyte Ringer's solution had the following composition: 82.5 mM NaCl, 2.5 mM KCl, 2.5 mM CaCl₂, 1 mM MgCl₂ 1, and 5 mM Hepes, adjusted to pH 7.4 with NaOH. ACSF had the following composition: 125 mM NaCl, 2.5 mM KCl, 2 mM CaCl₂, 1.25 mM NaH₂PO₄, 1 mM MgCl₂, 26 mM NaHCO₃, and 10 mM glucose (pH 7.35). Sucrose-based ACSF solution contained: 200 mM sucrose, 3 mM KCl, 2 mM CaCl₂, 1.6 mM MgCl₂, 1.25 mM NaH₂PO₄, 25 mM NaHCO₃, 10 mM glucose; 2 mM MgSO₄; and 0.4 mM ascorbic acid (pH 7.35). All drugs were purchased from Sigma with the exception of GABA and L655,708, which were purchased from Tocris. LEV was a gift from Bruno Ferrò (UCB Pharma, Brussels). LEV was dissolved in H₂O, and stored as frozen stock solutions (100 mM; ref. 12). THDOC and L655,708 were dissolved in DMSO (1:1,000 dilution) and stored at −20°C. BDNF and okadaic acid were dissolved and applied as reported (11).

Statistics.

All data represent means ± SEM. Differences among means were analyzed by one-way or two-way ANOVA.