Enhanced Tonic GABA Current in Normotopic and Hilar Ectopic Dentate Granule Cells After Pilocarpine-Induced Status Epilepticus

Abstract

In temporal lobe epilepsy, loss of inhibitory neurons and circuit changes in the dentate gyrus promote hyperexcitability. This hyperexcitability is compensated to the point that dentate granule cells exhibit normal or even subnormal excitability under some conditions. This study explored the possibility that compensation involves enhanced tonic GABA inhibition. Whole cell patch-clamp recordings were made from normotopic granule cells in hippocampal slices from control rats and from both normotopic and hilar ectopic granule cells in slices from rats subjected to pilocarpine-induced status epilepticus. After status epilepticus, tonic GABA current was an order of magnitude greater than control in normotopic granule cells and was significantly greater in hilar ectopic than in normotopic granule cells. These differences could be observed whether or not the extracellular GABA concentration was increased by adding GABA to the superfusion medium or blocking plasma membrane transport. The enhanced tonic GABA current had both action potential–dependent and action potential–independent components. Pharmacological studies suggested that the small tonic GABA current of granule cells in control rats was mediated largely by high-affinity α₄β_xδ GABA_A receptors but that the much larger current recorded after status epilepticus was mediated largely by the lower-affinity α₅β_xγ₂ GABA_A receptors. A large α₅β_xγ₂-mediated tonic current could be recorded from controls only when the extracellular GABA concentration was increased. Status epilepticus seemed not to impair the control of extracellular GABA concentration by plasma membrane transport substantially. Upregulated tonic GABA inhibition may account for the unexpectedly modest excitability of the dentate gyrus in epileptic brain.

INTRODUCTION

Temporal lobe epilepsy is the most common form of epilepsy in the adult population. Most cases are believed to develop after a lesion of the brain. The most consistent region of neuronal loss is the hilus of the dentate gyrus. Dentate granule cells regulate the propagation of seizures from the highly excitable entorhinal cortex to the hippocampus (Nadler 2003, 2009; Nadler and Zhan 2009). In lesional temporal lobe epilepsy, dentate gyrus circuitry is expected to become more excitable because of the loss of hilar GABA neurons (Buckmaster and Jongen-Rêlo 1999; Obenaus et al. 1993; Sun et al. 2007; Sundstrom et al. 2001) and the marked increase of monosynaptic recurrent excitatory connections (Nadler 2003). In addition, seizures increase the rate of granule cell replication, and some of these newly generated neurons migrate to ectopic locations, most notably the dentate hilus (Parent et al. 1997, 2006; Scharfman et al. 2000). Hilar ectopic granule cells (HEGCs) are more excitable than normotopic granule cells; many even burst spontaneously. Normotopic and ectopic granule cells are synaptically interconnected by recurrent mossy fibers, forming a reverberating network unique to epileptic brain. Formation of recurrent excitatory circuitry in the dentate gyrus is associated with a reduced threshold for granule cell synchronization (Gabriel et al. 2004; Hardison et al. 2000; Masukawa et al. 1992; Okazaki and Nadler 2001; Patrylo and Dudek 1998; Tauck and Nadler 1985). These changes may contribute to a progressive increase in the frequency and duration of spontaneous seizures (Gorter et al. 2001; Zhang et al. 2002).

Contrary to expectation, stimulation of the mossy fibers in vitro (Hardison et al. 2000; Okazaki and Nadler 2001; Patrylo and Dudek 1998; Tauck and Nadler 1985) or the perforant path in vivo (Buckmaster and Dudek 1997) usually does not evoke reverberating excitation in the dentate gyrus of epilepsy models. In kainic acid–treated rats, granule cells are most excitable during the first few days after status epilepticus, coincident with the degeneration of hilar GABA neurons (Sloviter et al. 2006). The initial hyperexcitability reverses over the next 1–2 mo. Thus the loss of dentate gyrus interneurons is compensated eventually, to the point that granule cells exhibit a normal or even subnormal level of excitability under some conditions (Buckmaster and Dudek 1997; Kobayashi and Buckmaster 2003; Sloviter et al. 2006; Uruno et al. 1994; Wilson et al. 1998; Wu and Leung 2001). Mechanisms suggested to explain this progressive loss of excitability include the sprouting of inhibitory axons (Andre et al. 2001; Davenport et al. 1990; Mathern et al. 1995; Wittner et al. 2001) and increased driving of interneurons by mossy fibers (Sloviter et al. 2006).

Increased tonic inhibition could also dampen granule cell excitability. Tonic inhibition involves the persistent activation of perisynaptic and extrasynaptic GABA_A receptors by ambient extracellular GABA (Glykys and Mody 2007a; Walker and Semyanov 2007). Tonic GABA currents have been recorded in many neuronal populations, including dentate granule cells. This study explored the possibility that tonic GABA inhibition increases in granule cells of rats made epileptic with pilocarpine.

METHODS

Pilocarpine-induced status epilepticus

Male Sprague-Dawley rats (150–200 g; Zivic Laboratories, Pittsburgh, PA) were injected intraperitoneally with pilocarpine hydrochloride (340–360 mg/kg) 30 min after pretreatment with scopolamine methyl bromide and terbutaline hemisulfate (both 2 mg/kg, ip). Status epilepticus, defined as a continuous limbic motor seizure of stage 2 or higher (Racine 1972), was allowed to self-terminate after 6–8 h. Rats treated in this way develop extensive and consistent hilar lesions followed by robust mossy fiber sprouting, the accumulation of HEGCs, and spontaneous seizures (Jiao and Nadler 2007; Sloviter et al. 2003). Rats that exhibited only a few brief behavioral seizures, but not status epilepticus, were used as controls to account for any possible action of pilocarpine not mediated by status epilepticus. Histological tests showed no evidence of neuronal degeneration or mossy fiber sprouting in these animals (Okazaki et al. 1999), and their electrophysiological responses were not significantly different from those of age-matched untreated rats (Hardison et al. 2000; Molnár and Nadler 1999; Okazaki and Nadler 2001; Okazaki et al. 1999).

Male C57Bl/6 mice (8–12 wk of age; Charles River, Raleigh, NC) were treated similarly to rats, except that the dose of pilocarpine was 275 mg/kg, ip. Because all mice treated with pilocarpine developed status epilepticus, age-matched untreated mice were used as controls.

All experiments were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved in advance by the Duke University Institutional Animal Care and Use Committee.

Hippocampal slice preparation

Hippocampal slices were prepared 10–40 wk after pilocarpine administration. Animals were decapitated under deep ether anesthesia, and the brain was removed to ice-cold high-Mg²⁺ artificial cerebrospinal fluid (high-Mg²⁺ ACSF; in mM: 112 NaCl, 25 NaHCO₃, 3.1 KCl, 1.8 CaCl₂, 11.2 MgSO₄, 0.4 KH₂PO₄, 1 ascorbic acid, 10 d-glucose, equilibrated with 95% O₂-5% CO₂). Transverse 420-μm-thick slices of the caudal hippocampus were prepared with a vibratome, incubated at 34°C in high-Mg²⁺ ACSF for 30 min, and maintained in standard ACSF (122 mM NaCl, 1.2 mM MgSO₄) at room temperature (22–24°C).

Electrophysiology

A slice was transferred to a submersion-type recording chamber and superfused at ∼3 ml/min with standard ACSF at 22–24 or 34–35°C. Cells were visualized with a Nikon Eclipse E600FN microscope equipped with far infrared-differential interference contrast optics, a CCD camera, and a ×40 water-immersion objective. Criteria for selecting HEGCs were 1) soma located within the hilus and of a size and shape indistinguishable from granule cells in the cell body layer and 2) no more than three dendrites emerged from the soma. Cell identity was confirmed by intracellular dialysis with biocytin and subsequent visualization of cellular morphology. Results obtained from putative HEGCs were included in this study only if cellular morphology appeared identical to previous descriptions of HEGCs (Dashtipour et al. 2001; Scharfman et al. 2000, 2003): small (8–12 μm diam) soma located within the dentate hilus, one to two apical dendrite(s) penetrating into or directed toward the dentate molecular layer, and axon with giant boutons in area CA3 and extensive branches within the hilus. HEGC identity was confirmed in >90% of the experiments.

Patch electrodes pulled from borosilicate glass (1.5 mm OD; 1.1 mm ID) had a tip resistance of 4.5–6.5 MΩ. Recordings were made with an Axopatch 200B amplifier (Molecular Devices, Sunnyvale, CA). Series resistances (<20 MΩ) were compensated 75%. Recordings were rejected if the series resistance varied by >20%. Voltage-clamp recordings were filtered at 2 kHz, digitized at 20 kHz, and stored for analysis off-line with pClamp8.1 (Molecular Devices, Sunnyvale, CA) or MiniAnalysis (Synaptosoft, Decatur, GA) software. Current-clamp recordings were digitized at 50 kHz.

To compare the membrane properties of HEGCs and normotopic granule cells, the electrode was filled with a potassium methylsulfate-based solution [in mM: 125 potassium methylsulfate, 7 KCl, 0.1 EGTA, 2 tris ATP, 0.3 tris GTP, 5 creatine phosphate, 20 U/ml creatine phosphokinase, 1% (wt/vol) biocytin, pH 7.25–7.30, and 294–297 mOsm]. After a tight seal (>2 GΩ) was formed, spontaneous currents were monitored for 5 min in cell-attached mode at a holding potential of −70 mV. Then whole cell access was achieved in current-clamp mode, and the resting membrane potential was determined immediately. Membrane time constant, input resistance, and membrane capacitance were determined from the current response to a 10-mV hyperpolarization from resting V_m applied for 200 ms. The presence or absence of cellular bursts was determined in current-clamp mode at resting V_m and after injecting a series of currents from −0.5 to 1.9 nA in 0.1-nA increments for 1 s each. All membrane potentials were corrected for a liquid junction potential of 9 mV.

To record tonic GABA current and spontaneous inhibitory postsynaptic currents (sIPSCs), the electrode was filled with a CsCl-based solution [in mM: 116 CsCl, 10 QX-314 chloride, 1 MgCl₂, 0.1 CaCl₂, 5 EGTA, 10 HEPES, 2 tris ATP, 5 creatine phosphate, 20 U/ml creatine phosphokinase, 1% (wt/vol) biocytin, pH 7.25–7.30, and 294–297 mOsm]. The liquid junction potential was considered to be 0 when using a chloride-based internal solution. d-2-amino-5-phosphonopentanoate (d-AP5; 50 μM) and 2,3-dihydroxy-6-nitro-7-sulfamoyl-benzo[F]quinoxaline-2,3-dione (NBQX; 10 μM) were added to the superfusion medium to block ionotropic glutamate receptors. After at least a 10-min exposure to these blockers, the baseline holding current was monitored continuously at a holding potential of −70 mV before and during exposure to GABA_A receptor ligands. Tonic GABA current was calculated as the change in baseline current produced by a 2.5-min application of 100 μM 6-imino-3-(4-methoxyphenyl)-1(6H)-pyridazinebutanoic acid hydrobromide (SR95531). The pretreatment baseline current was determined by averaging the current recorded during the 10-s period before addition of SR95531, excluding spontaneous inhibitory postsynaptic currents (sIPSCs). The post-treatment baseline current was determined by averaging the current recorded during a 10-s period that flanked the minimal current recorded during SR95531 application. Tonic current density was calculated by dividing tonic current amplitude by the membrane capacitance. Test compounds were applied for 8 {11,12,13,13a-tetrahydro-7-methoxy-9-oxo-9H-imidazo[1,5-a]pyrrolo[2,1-c][1,4]benzod iazepine-1-carboxylic acid ethyl ester (L-655708)} or 10 [3α,21-dihydroxy-5A-pregnan-20-one (THDOC)] min before the addition of SR95531, and their effect on baseline current was computed similarly to the effect of SR95531. All GABA_A receptor ligands were tested by addition to the superfusion medium.

sIPSCs were counted and their properties were analyzed during a 2.5-min period immediately before exposure to the test compound and again when the effect of test compound was judged to be maximal (7.5–10 min after addition of THDOC or 5.5–8 min after addition of L-655708). Spontaneous events were identified automatically by MiniAnalysis and examined manually to exclude false positives. The amplitude threshold was 5 pA. Peak amplitude, 10–90% rise time, τ, and charge transfer per event were determined for each verified sIPSC and averaged to obtain a single set of values for each recording period examined. Spontaneous events that overlapped were not used for the computation of sIPSC properties. All sIPSCs were abolished by addition of 100 μM SR95531 to the superfusion medium.

The contribution of tonic GABA current to the total baseline GABA current was computed as follows: charge transferred per second by the tonic current/the charge transferred per second by tonic and synaptic currents combined. The synaptic contribution was determined from the mean number of events per second × the mean charge transfer per event.

Grouped data are expressed as means ± SE unless indicated otherwise.

Visualization of cell morphology

After recording, the electrode tip was withdrawn slowly, and the slice was fixed with 5% (wt/vol) paraformaldehyde in 0.1 M sodium phosphate/0.8% (wt/vol) NaCl (PBS). After fixation at 4°C overnight, the slice was cut into 60-μm-thick sections with a vibratome. Sections were incubated in 30% (vol/vol) methanol/2% (vol/vol) H₂O₂ for 90 min to inactivate endogenous peroxidases. After washing with PBS, sections were incubated at 4°C overnight in an avidin-horseradish peroxidase solution (Vectastain Elite ABC Kit, Vector Laboratories, Burlingame, CA) that contained 0.3% (vol/vol) Triton X-100. After washing again with PBS, color was developed with diaminobenzidine/H₂O₂ for 6 min and intensified with nickel ammonium sulfate according to the Vector protocol. Cell morphology was reconstructed from serial sections with use of Neurolucida (MicroBrightField, Williston, VT).

Materials

SR95531, d-AP5, NBQX, TTX, and L-655708 were purchased from Tocris Bioscience (Ellisville, MO). Biocytin, creatine phosphate, creatine phosphokinase, tris ATP, tris GTP, 1-[2-[[diphenylmethylene)imino]oxy]ethyl]-1,2,5,6-tetrahydro-3-pyridinecarboxylic acid hydrochloride (NO-711), and THDOC were purchased from Sigma (St. Louis, MO).

RESULTS

Membrane properties of normotopic and hilar ectopic granule cells

When whole cell patch-clamp recordings were made with a potassium methylsulfate–based internal solution, few differences were noted among HEGCs, normotopic granule cells from rats subjected to pilocarpine-induced status epilepticus (GC-SEs), and normotopic granule cells from control rats (CGCs) (Table 1). HEGCs had a significantly less polarized resting membrane potential than normotopic granule cells. Cell-attached recording showed spontaneous bursting in 31.6% (6/19) of HEGCs but not in any normotopic granule cell. HEGCs found to burst spontaneously had resting membrane potentials between −53 and −77 mV (Fig. 1). The burst rate was greatest in cells with the lowest resting potentials. Of the six spontaneously bursting cells, five had a resting potential below −70 mV. Thus most spontaneously bursting HEGCs had a less polarized membrane potential than the average nonbursting HEGC (−72 ± 1 mV) or normotopic granule cell. Bursting HEGCs did not appear damaged, however. Action potential threshold and amplitude were comparable to those of nonbursting HEGCs, and a stable recording could be maintained for >1 h. Of the 13 nonbursting HEGCs, depolarization evoked bursts in 5. Burst thresholds ranged from −52 to −71 mV. Thus 57.9% of the HEGCs recorded possessed intrinsic burst capability.

TABLE 1.

Comparison of HEGC membrane properties with those of normotopic dentate granule cells

	CGC	GC-SE	HEGC
Resting Vm, mV	−79 ± 1	−78 ± 1	−70 ± 1*
Membrane time constant, ms	8.1 ± 0.5	8.1 ± 0.4	9.7 ± 0.7
Input resistance, MΩ	305 ± 18	216 ± 16	282 ± 32
Membrane capacitance, pF	27.4 ± 2.5	40.8 ± 3.7	41.1 ± 4.9
Action potential threshold, mV	−42 ± 2	−41 ± 2	−42 ± 1
Action potential amplitude, mV	157 ± 7	139 ± 8	147 ± 4
Spontaneous bursting, %	0	0	31.6 (6/19)
Bursting at any Vm, %	0	0	57.9 (11/19)

Values are means ± SE for 10 CGCs, 13 GC-SEs, and 19 HEGCs. Membrane currents associated with spontaneous bursting were monitored during a 5-min cell-attached recording at a holding potential of −70 mV and a temperature of 22–24°C. Seal resistance was >2 GΩ and the leak current was <4 pA. Resting V_m was measured in current-clamp mode immediately on break-in, and the membrane time constant, input resistance, and membrane capacitance were determined from the current response to a 10-mV hyperpolarization from resting V_m applied for 200 ms. The presence or absence of depolarization-evoked bursts was determined by injecting a series of currents from −0.5 to 1.9 nA in 0.1-nA increments for 1 s each. Action potential amplitude was measured from its positive to its negative peak.

^*Significantly less polarized than CGCs or GC-SEs at P < 0.01 (Newman-Keuls test after 1-way ANOVA yielded P < 0.001). CGC, normotopic granule cells from control rats; GC-SE, normotopic granule cells from rats subjected to pilocarpine-induced status epilepticus; HEGC, hilar ectopic granule cells.

FIG. 1.

Spontaneous and depolarization-evoked bursting of representative hilar ectopic granule cells (HEGCs). Cells were filled with biocytin during whole cell patch-clamp recording, and cell morphology was reconstructed with use of Neurolucida. The apical dendrites of all 3 HEGCs reached the outer edge of the dentate molecular layer, and they all had a short basal dendrite directed into the hilus. The main branch of the mossy fiber reached stratum lucidum of area CA3, and the mossy fiber of HEGC A also sent a recurrent branch into the molecular layer. DG, dentate gyrus. Spontaneous burst-associated currents were monitored for 5 min during voltage-clamp recording in cell-attached mode at 22–24°C. The seal resistance was >2 GΩ, the leak current was <4 pA, and the holding potential was −70 mV. Under these conditions, currents associated with spontaneous bursting were recorded from HEGC B but not from HEGC A or C. During whole cell recording, HEGC B discharged cellular bursts at its resting V_m of −61 mV. HEGC C discharged a burst when depolarized (to −47 mV in this example). Depolarization of HEGC A (to −43 mV in the figure) evoked only a train of action potentials that exhibited spike frequency adaptation, similar to the firing pattern of normotopic granule cells.

Enhanced tonic GABA current in GC-SEs and HEGCs after pilocarpine-induced status epilepticus

Small tonic GABA currents were recorded from dentate granule cells in slices from control rats maintained at 22–24 (2.3 ± 0.3 pA) or 34–35°C (1.4 ± 0.3 pA) (Fig. 2). The size of these currents was undoubtedly limited by the need for rapid superfusion to maintain slice viability (Glykys and Mody 2007a). Regardless of temperature, tonic current amplitude and density were both much greater in GC-SEs and HEGCs than in control granule cells [P < 0.001 in each case by Newman-Keuls test after 2-way ANOVA (temperature × granule cell group) yielded P < 0.001 for effect of granule cell group, P ≈ 0.9 for effect of temperature, and no significant interaction between the variables]. In slices maintained at 34–35°C, for example, tonic current amplitude and density were 16 and 12 times as great, respectively, in GC-SEs and 32 and 25 times as great, respectively, in HEGCs. Tonic current amplitude and density were also significantly greater in HEGCs than in GC-SEs at both temperatures (P = 0.01 for amplitude and P < 0.001 for density by Newman-Keuls test). Absolute values of these measures differed little at the two temperatures for any granule cell group.

FIG. 2.

Tonic GABA current and current density are upregulated in dentate granule cells after status epilepticus. Whole cell recordings were made with a CsCl-based internal solution at a holding potential of −70 mV. Tonic current amplitude was taken as the change in baseline current produced by the addition of 100 μM SR95531 to the superfusion medium. Tonic current density, which normalizes for neuronal surface area, was calculated by dividing tonic current amplitude by the membrane capacitance. Grouped values are means ± SE for 15 normotopic granule cells from control rats (CGCs), 16 normotopic granule cells from rats subjected to pilocarpine-induced status epilepticus (GC-SEs), and 14 HEGCs at 22–24°C and 8 CGCs, 8 GC-SEs, and 7 HEGCs at 34–35°C. P values shown were determined by 1-way ANOVA (P < 0.001) followed by Newman-Keuls test.

There were no consistent differences among dentate granule cells studied at different times after status epilepticus between 10 and 40 wk. Therefore enhanced tonic GABA current in dentate granule cells is a long-lasting, and probably permanent, feature of the pilocarpine model.

The lack of a reduction in tonic GABA current is consistent with data obtained from pilocarpine-treated mice (Zhang et al. 2007). In mice, tonic GABA current was reported to be maintained after status epilepticus but not increased above control. However, the investigators included 5 μM GABA in the medium superfusing their hippocampal slices. In the absence of added GABA, we found tonic GABA current difficult to measure in most granule cells from control mice (0.5 ± 0.2 pA, 0.02 ± 0.01 pA/pF, n = 8). A definite current could be recorded in GC-SEs under the same conditions, however (3.4 ± 0.8 pA, 0.18 ± 0.06 pA/pF, n = 11; P < 0.01 for current amplitude and P < 0.05 for current density compared with controls by Student's t-test). In addition, we measured tonic GABA current in GE-SEs and granule cells from control rats after adding 5 μM GABA to the superfusion medium (Fig. 3). In contrast to the reported results in mice, tonic GABA current (control: 4.3 ± 1.1 pA, GC-SE: 61.0 ± 1.4 pA, n = 6 cells per group; P = 0.003 by Student's t-test) and current density (control: 0.13 ± 0.05 pA/pF, GC-SE: 1.19 ± 0.36 pA/pF, n = 6 cells per group; P < 0.02 by Student's t-test) were many-fold greater in GC-SEs than in controls. Thus increased tonic GABA current could be shown in dentate granule cells after pilocarpine-induced status epilepticus in both mice and rats, and the effect could be shown in rats whether or not GABA was added to the superfusion medium.

FIG. 3.

Tonic GABA current was greater in GC-SEs than in granule cells from control rats when 5 μM GABA was added to the superfusion medium. Whole cell recordings were made at 22–24°C with a CsCl-based internal solution at a holding potential of −70 mV. Traces are from representative experiments.

Tonic GABA current contributed much more to the total baseline GABA current in granule cells after status epilepticus in rats. For example, tonic current contributed 97 ± 1% of the total charge transferred through GABA_A receptors per second in GC-SEs at 34–35°C and >99% in HEGCs, but contributed only 57 ± 8% in controls. Because these findings suggest a greater role for tonic GABA current in regulating granule cell excitability in epileptic brain, we determined the extent to which impaired GABA transport, enhanced GABA neuron firing, and the involvement of different GABA_A receptors contributed to enhancement of the current.

Increased tonic GABA current could not be explained by impaired GABA transport

GABA transport through the plasma membrane regulates the ambient extracellular GABA concentration and thus the magnitude of tonic GABA current. Studies with GABA transport inhibitors assessed the possibility that compromised GABA transport accounted for the enhanced tonic GABA current recorded after status epilepticus. In rodent hippocampus, activity of GAT-1, one of the four plasma membrane GABA transporters, accounts for ∼80% of total GABA transport (Jensen et al. 2003). In dentate granule cells from control rats, tonic GABA current increased by 26 ± 4 pA and the current density by 0.80 ± 0.13 pA/pF during exposure to the GAT-1 inhibitor NO-711 (20 μM; Fig. 4). NO-711 was more effective in raising tonic GABA current than 5 μM GABA, presumably because the extracellular GABA concentration in contact with perisynaptic/extrasynaptic GABA_A receptors was increased to a greater degree. The GAT-3 inhibitor SNAP-5114 had no effect at any concentration from 20 to 240 μM (100 ± 2% of control, n = 4). The current density increase produced by NO-711 was substantially greater in GC-SEs and HEGCs than in control granule cells, and the increase was roughly proportional to the current density before NO-711 was added to the medium. If GABA transport had been impaired substantially after status epilepticus, NO-711 should have had little effect on tonic GABA current (Calcagnotto et al. 2005). Thus the markedly enhanced tonic GABA current after status epilepticus could not be explained by impairment of plasma membrane transport.

FIG. 4.

Inhibition of plasma membrane GABA transport continues to increase tonic GABA current density substantially after status epilepticus. Whole cell recordings were made at 22–24°C with a CsCl-based internal solution at a holding potential of −70 mV. Tonic current density was calculated as described in Fig. 2. Addition of the GAT-1 inhibitor 1-[2-[[diphenylmethylene)imino]oxy]ethyl]-1,2,5,6-tetrahydro-3-pyridinecarboxylic acid hydrochloride (NO-711; 20 μM) to the superfusion medium increased the baseline current density to a greater degree after status epilepticus, which resulted in a larger tonic current. Grouped data are expressed as means ± SE for 8 cells in each group. Effects of NO-711 on tonic current density in different granule cell groups seemed roughly proportional to tonic current density in the absence of NO-711 (cf. Fig. 2, left).

As expected from studies of CA1 pyramidal cells (Bonin et al. 2007), increased tonic GABA current produced by exposure to NO-711 was associated with increased membrane conductance (Fig. 5). More depolarizing current was required to evoke an action potential. Action potential threshold (−44 ± 2 mV with or without NO-711) and peak amplitude (168 ± 6 mV before NO-711 and 170 ± 6 mV in the presence of NO-711) were unchanged.

FIG. 5.

Increasing the extracellular GABA concentration with NO-711 increases membrane conductance in dentate granule cells from control rats. Recordings were made in current-clamp mode at 22–24°C. Results shown in A–D are from a representative cell, whose resting V_m was −75 mV. Similar results were obtained from 5 additional cells. A: a series of depolarizing currents induced membrane potential changes before addition of NO- 711 (20 μM) to the superfusion medium. Note that injection of the 2 largest currents evoked action potential firing. B: the same depolarizing currents induced smaller membrane potential changes in the presence of NO-711. Note that injection of the 2 largest currents now failed to evoke action potential firing. C: the current injection protocol. Stepped currents of 500-ms duration were injected in 0.05-nA increments from 0 to 0.75 nA. D: the relationship between membrane potential change and injected current. Lines were drawn by linear regression. E: the slope of the relationship between membrane potential change and injected current differed significantly by paired t-test (n = 6), indicating increased membrane conductance in the presence of NO-711.

Action potential–dependent and action potential–independent GABA release contribute to the increased tonic current

In area CA1, the mean spontaneous firing rate of surviving GABA interneurons increases after status epilepticus (Cossart et al. 2001; Stief et al. 2007), and indirect evidence suggests that dentate gyrus interneurons may fire more rapidly as well (Shao and Dudek 2005). Studies with TTX tested the hypothesis that increased action potential–driven GABA release accounted for the enhanced tonic GABA current. Application of 1 μM TTX reduced the current by 46 ± 5% in GC-SEs and 43 ± 2% in HEGCs (Fig. 6). Both the TTX-sensitive and TTX-insensitive components of the current were much larger than the tonic GABA current recorded from control granule cells under the same conditions. Sensitivity to TTX implies that enhanced action potential–dependent release of GABA accounts for nearly half the current increase. However, at least as much of the increase depends on the release of GABA by some other mechanism.

FIG. 6.

Action potential–dependent and action potential–independent GABA release contribute to the large tonic GABA currents in GC-SEs and HEGCs. Whole cell recordings were made at 34–35°C with a CsCl-based internal solution at a holding potential of −70 mV. Tonic current amplitude was calculated as described in Fig. 2. TTX (1 μM) was added to the superfusion medium 10 min before application of SR95531 (100 μM). Grouped data are means ± SE for 5 cells.

Different GABA_A receptors mediate tonic current in dentate granule cells after status epilepticus

In dentate granule cells, the tonic current recorded at low extracellular GABA concentrations is produced by activation of perisynaptic (near the outer edges of synapses) and extrasynaptic GABA_A receptors with the composition α₄β_xδ (Stell et al. 2003; Wei et al. 2003). Receptors of this type have a high affinity for GABA (EC₅₀ = 0.5–2.28 μM) (Brown et al. 2002; Stórustovu et al. 2006; Wallner et al. 2003), enabling them to respond to the low extracellular concentrations of GABA present in rodent brain (∼0.1–2.0 μM) (Cavelier et al. 2005), and they exhibit little desensitization. GABA is a low affinity (partial) agonist at these receptors. GABA receptors that include a δ subunit are a prominent target of neurosteroids, brain-derived metabolites of gonadal steroids, and glucocorticoids (Herd et al. 2007). These metabolites increase the efficacy of GABA (Bianchi and Macdonald 2003; Wohlfarth et al. 2002) and thus the magnitude of tonic GABA currents mediated by δ subunit-containing receptors on granule cells (Mtchedlishvili et al. 2001; Stell et al. 2003). The neurosteroid THDOC specifically enhances tonic current mediated by δ subunit-containing GABA_A receptors at concentrations <30 nM.

THDOC increased the tonic GABA current in granule cells of control rats by 18 ± 5 pA and the current density by 0.61 ± 0.25 pA/pF (Fig. 7). THDOC also slowed the decay of sIPSCs, but did not affect the mean sIPSC frequency or any other sIPSC property (Fig. 8; Table 2). In GC-SEs, however, THDOC altered neither tonic GABA current nor the rate of sIPSC decay. THDOC did enhance tonic GABA current (by 10 ± 3 pA) and density (by 0.47 ± 0.15 pA/pF) in HEGCs, indicating that δ subunit–containing GABA_A receptors mediate tonic GABA current in hilar ectopic, but not normotopic, granule cells after status epilepticus. The neurosteroid did not, however, affect significantly either the frequency or properties of sIPSCs in these cells. In the absence of any GABA_A receptor ligand, we observed no significant differences among granule cell populations with respect to sIPSC frequency, amplitude, or kinetics, in agreement with a previous study performed under similar experimental conditions (Shao and Dudek 2005).

FIG. 7.

Loss of neurosteroid-induced enhancement of tonic GABA current in GC-SEs, but presence in HEGCs, after status epilepticus. Whole cell recordings were made at 22–24°C with a CsCl-based internal solution at a holding potential of −70 mV. Left: addition of vehicle (2.5 μM ethanol) to the superfusion medium as a control for drift in the recording of baseline holding current. Right: effect of the neurosteroid 3α,21-dihydroxy-5A-pregnan-20-one (THDOC; 20 nM). THDOC enhanced tonic current amplitude in CGCs and, to a lesser degree, in HEGCs, but not in GC-SEs. Bottom: addition of THDOC to the superfusion medium increased the baseline holding current significantly in CGCs and HEGCs but not in GC-SEs (solid bars). Baseline holding current tended to decline slightly over the same time period when vehicle was added to the superfusion medium instead of THDOC (open bars). Grouped values are means ± SE for 5–7 cells. *P < 0.01 compared with vehicle controls (Student's t-test).

FIG. 8.

THDOC prolongs the decay of spontaneous inhibitory postsynaptic currents (sIPSCs) in granule cells from control rats. The traces shown are averaged sIPSCs from representative experiments carried out at 22–24°C. Traces recorded in the presence of the subtype-specific GABA_A receptor ligand are scaled to the peak amplitude of the sIPSC recorded in the absence of ligand. Grouped data from these experiments are presented in Table 2.

TABLE 2.

Effects of subtype-selective GABA_A receptor ligands on frequency and properties of sIPSCs in dentate granule cells

	CGC		GC-SE		HEGC
	Before	+ Ligand	Before	+ Ligand	Before	+ Ligand
THDOC
Frequency, Hz	3.1 ± 0.6	3.8 ± 0.4	3.1 ± 0.7	3.0 ± 0.9	1.1 ± 0.3	1.6 ± 0.7
Amplitude, pA	30.3 ± 4.0	29.1 ± 2.5	27.6 ± 3.5	23.4 ± 4.2	22.8 ± 4.9	25.0 ± 3.1
10–90% rise time, ms	1.7 ± 0.3	2.0 ± 0.2	2.4 ± 0.2	2.4 ± 0.3	2.8 ± 0.9	2.3 ± 0.4
τ, ms	10.6 ± 0.9	13.3 ± 3.2†	10.2 ± 0.8	9.6 ± 1.6	10.8 ± 2.8	10.9 ± 3.1
Charge transfer, fC	301 ± 52	354 ± 44	265 ± 39	221 ± 58	293 ± 88	214 ± 40
L-655708
Frequency, Hz	2.5 ± 0.3	2.2 ± 0.4	2.4 ± 0.9	2.2 ± 0.8	3.0 ± 0.6	2.3 ± 0.5
Amplitude, pA	28.4 ± 3.8	25.7 ± 3.1	27.0 ± 3.6	21.2 ± 5.6	37.5 ± 8.6	25.7 ± 6.6
10–90% rise time, ms	2.8 ± 0.9	2.3 ± 0.4	1.9 ± 0.2	2.1 ± 0.1	1.8 ± 0.2	2.0 ± 0.3
τ, ms	9.5 ± 0.3	7.8 ± 0.4‡	10.9 ± 0.6	9.0 ± 0.9	12.3 ± 1.5	8.0 ± 1.3‡
Charge transfer, fC	234 ± 29	175 ± 20‡	293 ± 46	192 ± 56*	426 ± 95	191 ± 52‡

Values are means ± SE for 5 (CGC/THDOC), 6 (CGC/L-655,708), 6 (GC-SE/THDOC), 7 (GC-SE/L-655,708), 6 (HEGC/THDOC), or 7 (HEGC/L-655,708) cells per group. The number of sIPSCs recorded during a 2.5-min period and the properties of averaged sIPSCs were determined before and during exposure to the test compound at 22–24°C. Where indicated, the ligand altered the sIPSC property significantly at

^*P < 0.05.

^†P < 0.02, or

^‡P < 0.01 (paired t-test). sIPSC, spontaneous inhibitory postsynaptic current. See Table 1 for other abbreviations.

Dentate granule cells normally express other GABA_A receptor subtypes in perisynaptic and/or extrasynaptic locations, and these subtypes can potentially contribute to tonic GABA current (Glykys et al. 2008; Lindquist and Birnir 2006; Zhang et al. 2007). One such subunit combination is α₅β_xγ₂. Tonic GABA current in hippocampal pyramidal cells is mediated largely by GABA_A receptors of this type (Caraiscos et al. 2004; Glykys and Mody 2006; Glykys et al. 2008; Prenosil et al. 2006). Although dentate granule cells express fewer α₅β_xγ₂ receptors, their activation still contributes ∼30% of the maximal tonic GABA current in mice (Glykys et al. 2008). L-655708 is an inverse agonist highly specific for α₅β_xγ₂ receptors. In hippocampal slices, it retains subtype specificity even at a concentration as high as 50 μM (Caraiscos et al. 2004; Scimeni et al. 2005). Because α₅β_xγ₂ receptors have about an order of magnitude lower affinity for GABA than α₄β_xδ receptors (EC₅₀ = 6–19.4 μM) (Burgard et al. 1996; Caraiscos et al. 2004), they are believed to contribute to tonic GABA current only when the extracellular GABA concentration is near the high end of the normal range. Our results in granule cells from control rats support this view. In these cells, the current was too small to detect a significant reduction with L-655708 (50 μM). Application of L-655708 did reduce the amplitude of tonic GABA current when the extracellular GABA concentration was increased with NO-711 (Fig. 9). In the presence of NO-711, L-655708 reduced the current by 58 ± 2%. However, L-655708 reduced tonic GABA current in GC-SEs and HEGCs to a similar degree under basal conditions; it was unnecessary to increase the extracellular GABA concentration to observe its effect. Thus the activation of α₅β_xγ₂ receptors generated most of the tonic GABA current in dentate granule cells after status epilepticus.

FIG. 9.

Tonic GABA currents become sensitive to L-655708 under baseline conditions after status epilepticus. Whole cell recordings were made at 22–24°C with a CsCl-based internal solution at a holding potential of −70 mV. Tonic current amplitude was calculated as described in Fig. 2. L-655708 (50 μM) reduced tonic GABA current by >50% in GC-SEs and HEGCs but not in CGCs unless the extracellular GABA concentration was increased by inhibiting GABA transport with 20 μM NO-711. Grouped values are means ± SE for 5–7 cells.

Regarding synaptic responses, L-655708 reduced the decay time constant of sIPSCs significantly in both HEGCs and granule cells from control rats, and we observed a similar trend in GC-SEs as well (Fig. 10; Table 2). L-655708 also reduced the charge transfer per event of sIPSCs in all granule cell populations without changing frequency, peak amplitude, or 10–90% rise time significantly.

FIG. 10.

L-655708 reduces sIPSC decay time in granule cells from all groups. The traces shown are averaged sIPSCs from representative experiments carried out at 22–24°C. Traces recorded in the presence of the subtype-specific GABA_A receptor ligand are scaled to the peak amplitude of the sIPSC recorded in the absence of ligand. Grouped data from these experiments are presented in Table 2.

Because solutions that contained THDOC or L-655708 had to be superfused for several minutes before the peak effect of the ligand could be observed, it was possible that drift in the membrane current recordings could have biased the results. We therefore performed some of these experiments with the same protocol, except that vehicle (2.5 μM ethanol) was added to the superfusion medium instead of ligand (Fig. 7). Under these conditions, there was no consistent change in baseline current with time; the mean baseline current was usually found to have declined slightly when measurements were made at times similar to those used to determine effects of ligand. The drift in baseline current density for the different granule cell populations was as follows: control, 0.03 ± 0.05 pA/pF, n = 7; GC-SE, 0.06 ± 0.05 pA/pF, n = 7; HEGC, 0.03 ± 0.05 pA/pF, n = 6. Thus drift in the baseline current little affected our results.

DISCUSSION

Enhanced tonic GABA current after pilocarpine-induced status epilepticus

Our results showed a marked upregulation of tonic GABA current in dentate granule cells after pilocarpine-induced status epilepticus. This change is long-lasting, occurs in both rats and mice, and can be observed both under baseline recording conditions and when the extracellular GABA concentration is increased by adding GABA to the superfusion medium or inhibiting GABA transport. In contrast, upregulation of tonic GABA current in CA1 pyramidal cells could only be shown at an elevated extracellular GABA concentration (Scimeni et al. 2005). Pharmacological studies suggest the enhanced tonic current is mediated largely by α₅β_xγ₂ GABA_A receptors, which contribute significantly in granule cells from control rats only when the extracellular GABA concentration is increased. Enhanced tonic GABA current contrasts with the loss of inhibitory synaptic transmission in granule cells because of degeneration of hilar GABA neurons (Kobayashi and Buckmaster 2003; Okazaki et al. 1999; Shao and Dudek 2005; Sun et al. 2007). Furthermore, electron micrographs suggest that the somata and proximal dendrites of HEGCs receive little inhibitory innervation (Dashtipour et al. 2001). Thus enhanced tonic inhibition compensates for deficient synaptic inhibition in both GC-SEs and HEGCs. The magnitude of this effect suggests it could account for the reversal of granule cell hyperexcitability after status epilepticus.

Possible mechanisms of the tonic current increase

Activation of high-affinity α₄β_xδ receptors accounts for most of the tonic GABA current recorded from dentate granule cells of mice (Glykys et al. 2008), and our studies with GABA_A receptor ligands suggested their activation was also largely responsible for the tonic current recorded from control granule cells in rats. Previous studies showed a downregulation of δ subunit expression in granule cells of epileptic brain (Elliott et al. 2003; Peng et al. 2004; Schwarzer et al. 1997) accompanied by loss of the neurosteroid-induced enhancement of tonic GABA current (Mtchedlishvili et al. 2001; Zhang et al. 2007). We replicated the latter finding. One might therefore expect to observe a smaller tonic GABA current in granule cells of animals that become epileptic after pilocarpine-induced status epilepticus. In contrast, we observed an order of magnitude increase.

Studies with L-655708 suggested that the large tonic GABA currents recorded from both normotopic and hilar ectopic granule cells after status epilepticus in rats mainly reflected the activation of α₅β_xγ₂ GABA_A receptors. The size of the current attributable to activation of these receptors increased by at least a factor of 10, more than replacing the current lost from downregulation of α₄β_xδ receptors. In contrast, Zhang et al. (2007) found that pilocarpine-induced status epilepticus did not change the contribution of α₅β_xγ₂ receptors to the tonic GABA current of dentate granule cells in mice. The difference in our results may be attributable to different granule cell responses to status epilepticus in the two species. However, tonic GABA current increased much less after status epilepticus in mice than in rats. Therefore it might be difficult to detect a significantly increased involvement of α₅β_xγ₂ receptors in mice.

Enhanced activation of α₅β_xγ₂ GABA_A receptors by ambient extracellular GABA could be explained in three ways: by increased receptor expression/function, a change in receptor localization, or increased extracellular GABA concentration. Dentate granule cells normally express little α₅ mRNA or protein (Brooks-Kayal et al. 1998; Houser and Esclapez 2003; Sotiriou et al. 2005). In rats, α₅ subunit immunoreactivity was reported to increase after pilocarpine-induced status epilepticus, but by <20% in the granule cell body layer and not at all in the dentate molecular layer (Fritschy et al. 1999). In the dentate molecular layer of mice, limited immunohistochemical studies suggested reduced expression of α₅ subunits (Zhang et al. 2007). Thus enhanced expression of α₅β_xγ₂ GABA_A receptors seems an unlikely explanation for the large L-655,708- sensitive tonic GABA current in granule cells after status epilepticus. It is possible that GABA activates these receptors with greater potency after status epilepticus, that the probability of spontaneous channel opening increases or that some of the receptors move from extrasynaptic to perisynaptic locations, where they may be activated more readily by GABA overflow.

Another explanation for enhanced tonic GABA current after status epilepticus is increased ambient extracellular GABA concentration. This hypothesis is supported by the involvement of a relatively low-affinity GABA_A receptor subtype that normally contributes little to the tonic current unless the extracellular GABA concentration is increased. In fact, L-655708 eliminated the same fraction of tonic GABA current in GC-SEs and HEGCs under baseline conditions as it did in controls treated with NO-711 to increase the extracellular GABA concentration. If the ambient extracellular GABA concentration sensed by dentate granule cells does increase after status epilepticus, our results suggest this change could not be explained simply by impaired removal of GABA from the extracellular space. Frahm et al. (2003) reached a similar conclusion based on studies with the GABA transport inhibitor tiagabine. Enhancement of tonic current would presumably involve increased GABA release. Possible sources of the additional extracellular GABA include both vesicular (Bright et al. 2007; Glykys and Mody 2007b) and nonvesicular (Rossi et al. 2003) release mechanisms that may be operational in either interneurons (Shao and Dudek 2005) or the granule cells themselves (Gutiérrez 2005).

Tonic GABA current of HEGCs

HEGCs result from the aberrant migration of some granule cells born after seizures (Parent et al. 1997). In this respect, they differ from most GC-SEs, which are already mature when the seizure occurs. However, both granule cell populations are exposed to the postseizure, epileptic environment. Both HEGCs and GC-SEs have tonic GABA currents much larger than those of control granule cells, but those of HEGCs are the largest. Tonic current accounted for practically all of the total baseline GABA_A receptor-mediated current in these cells. L-655708, NO-711, and TTX altered tonic GABA current similarly in HEGCs and GC-SEs. Thus whatever the mechanism that increases tonic current in dentate granule cells may be, it affects normotopic and ectopic granule cells similarly. The one exception is that δ subunit-containing GABA_A receptors seem to play a role in ectopic cells only. Expression of extrasynaptic high-affinity δ subunit–containing receptors in HEGCs may explain the larger tonic GABA current in these cells than in GC-SEs.

Involvement of α₄β_xδ and α₅β_xγ₂ GABA_A receptors in sIPSCs

THDOC slowed the decay of sIPSCs in control granule cells without affecting these events significantly in any other way. This result is consistent with the activation of perisynaptic α₄β_xδ GABA_A receptors by the low concentrations of GABA that escape the synaptic cleft (Stell et al. 2003; Wei et al. 2003). Their activation by GABA spillover prolonged the synaptic response. In mice, activation of α₄β_xδ receptors slowed the decay of minimal IPSCs evoked in granule cell dendrites but not of those evoked at or near the soma (Wei et al. 2003). In addition, a high neurosteroid concentration was required to detect the slowing of sIPSC decay in somatic whole cell patch-clamp recordings (Stell et al. 2003). Further studies are needed to determine why we could show this effect with a lower neurosteroid concentration in rats.

L-655708 reduced the decay time constant and charge transfer of sIPSCs in all granule cell populations. These findings suggest that many α₅β_xγ₂ GABA_A receptors are localized to the perisynaptic membrane of dentate granule cells, where their activation prolongs spontaneous synaptic responses. The concentration of GABA sensed by these receptors is evidently sufficient to do this even in granule cells from control rats, although their activation contributed substantially to tonic GABA current only after status epilepticus.

Physiological significance of enhanced tonic GABA current

GABA_A receptor–mediated currents produce a shunting type of inhibition in dentate granule cells, because E_Cl is similar or positive to resting V_m (Pathak et al. 2007; Staley and Mody 1992). Tonic shunting inhibition shifts the relationship between output firing frequency and input current to the right, when cell firing is provoked by a constant voltage step. Thus as shown in the present and previous (Bonin et al. 2007) studies, more depolarizing current must be injected into the neuron to reach threshold. Tonic shunting inhibition can also reduce the slope of the input–output relationship during synaptic excitation (Hamann et al. 2002; Stell et al. 2003). The neuron fires fewer action potentials per increase in the mean excitatory conductance. Thus the enhanced tonic GABA current of dentate granule cells would be expected to dampen circuit excitability, compensating for the loss of synaptic inhibition and the formation of novel recurrent excitatory connections. Furthermore, tonic GABA current may also serve as the major inhibitory control over spontaneous and evoked bursting in HEGCs. To the extent that bursting HEGCs enhance synchronous granule cell firing during seizures, the large tonic GABA current may be expected to minimize their influence. Thus enlargement of this inhibitory current may account for the unexpectedly modest excitability of dentate gyrus circuitry in epileptic brain.

GRANTS

This study was supported by National Institute of Neurological Disorders and Stroke Grant NS-38108.