A deficit of functional GABAA receptors in neurons of β3 subunit knockout mice

Abstract

Mice whose γ-aminobutyric acid type A (GABA_A) β₃ subunit gene is inactivated (‘β₃ knockout mice’) have been previously shown to have epilepsy, hypersensitive behavior, cleft palate, and a high incidence of neonatal mortality. In this study, we analyze whole-cell responses to GABA in neurons from β₃^+/+, β₃^+/− and β₃^−/− mice. We demonstrate markedly decreased responses to GABA in both hippocampal and dorsal root ganglion neurons isolated from β₃^−/− mice without major differences in the GABA concentration-response curves. We also utilize the subunit selective pharmacology of Zn²⁺ and the anticonvulsant drug loreclezole to help infer the presence of β₂ and γ subunits in the GABA_A receptors remaining in neurons from β₃^−/− mice.

γ-Aminobutyric acid type A (GABA_A) receptors mediate most of the rapid inhibitory synaptic transmission in the mammalian brain. GABA_A receptors are pentameric complexes consisting of subunits which assemble to form a ligand-gated chloride channel. To date, 16 GABA_A subunits have been identified and cloned (α_1–6, β_1–4, γ_1–4, δ, ε) [1,12,14]. GABA_A receptors are modulated by clinically important drugs such as the benzodiazepines, general anesthetics, and barbiturates [5,11].

The β₃ subunit is a major constituent of neuronal GABA_A receptors and is the predominant β subunit isoform mRNA in a number of brain and spinal cord regions [9,10,20]. β₃ subunit mRNA is particularly abundant in the prenatal and neonatal brain, which suggests a possible role in development [10]. Recently, mice in which the GABA_A β₃ subunit gene was inactivated (‘β₃ knockout mice’) were demonstrated to have epilepsy, hypersensitive behavior, cleft palate, and a high incidence of neonatal mortality [6]. The phenotype of the β₃ knockout mice strengthens the hypothesis that deficits in GABA_A receptor-mediated transmission may underlie certain forms of epilepsy [13] and indicates the importance of GABA_A receptor function in development and behavior. In the present study, we studied the function of GABA_A receptors in the brain and spinal cord of these animals.

The production and phenotypic characterization of β₃ knockout mice has been previously described in detail [6]. For the experiments involving cultured hippocampal neurons, the hippocampus was microdissected, and neurons were dissociated from brains of 1-day-old mouse pups by methods described by Jones and Harrison [7], placed onto a layer of confluent mouse cortical astrocytes on a glass cover slip and maintained in tissue culture for 4–5 days prior to recording. For the experiments involving acutely dissociated dorsal root ganglion neurons (DRGs), DRGs were dissociated from postnatal day 1 mouse pups [18], plated onto poly-D-lysine coated coverslips and used for recordings within 12 h.

For all electrophysiological experiments, coverslips containing neurons were transferred to a chamber which was continuously perfused with extracellular medium containing (in mM): 145 NaCl, 3 KCl, 1.5 CaCl₂, 1 MgCl₂, 5.5 D-glucose, and 10 HEPES/NaOH (pH 7.4, osmolarity 320–330 mosmol). Recordings were made using the whole-cell patch-clamp technique [6,8]. The intracellular solution contained (in mM): 145 N-methyl-D-glucamine hydrochloride, 5 K₂ATP, 5 HEPES/KOH, 2 MgCl₂, 0.1 CaCl₂, and 1.1 EGTA/KOH (pH 7.2, osmolarity 315 mosmol). Neurons were voltage-clamped at −60 mV. Since the intracellular and extracellular solutions contained approximately equal concentrations of chloride ions, the chloride equilibrium potential was approximately 0 mV.

GABA and other drugs were rapidly applied (>1 ml/min) to neurons by local perfusion using a motor-driven solution exchange device (Bio Logic Rapid Solution Changer RSC-100; Molecular Kinetics, Pullman, WA, USA) as previously described [8]. Responses were low-pass-filtered at 5 kHz and digitized (TL-1–125 interface; Axon Instruments) using pCLAMP5 and stored for off-line analysis. Drugs used in this study were methohexital sodium (Brevital® sodium; Eli Lilly, Indianapolis, IN, USA) and zinc chloride (Sigma, St. Louis, MO, USA). Loreclezole was a kind gift from Janssen Pharmaceutica (Beerse, Belgium).

A marked reduction in the maximal amplitude of GABA-evoked whole cell currents was observed in hippocampal neurons from β₃^−/− mice. Mean maximal GABA responses recorded in hippocampal neurons from β₃ mice were nearly 50% lower than those from neurons of either β₃^+/+ and β₃^+/− mice (Fig. 1). This is very similar to the reported 50% decrease in the B_max for [³H]muscimol binding in whole brain homogenates from β₃^−/− versus β₃^+/+ mice [6]. This suggests that there are fewer GABA_A receptors in neurons from β₃^−/− mice. The amplitudes of GABA-evoked currents in hippocampal neurons from β₃^+/+ or β₃^+/− mice do not differ significantly (Fig. 1C). The GABA concentration-response relationships for neurons from β₃^+/+, β₃^+/−, and β₃^−/− mice are very similar, although the Hill slopes for neurons from β₃^+/− and β₃^−/− mice differ significantly from those from β₃^+/+ mice (P < 0.001 for each of the Hill slopes of β₃^+/− and β₃^−/− mice versus β₃^+/+ mice). Similar small differences in GABA concentration-response relationships have also been noted in recombinant studies of GABA_A receptor subunit combinations containing either the β₂ or β₃ subunit isoform [3].

Fig. 1.

GABA-evoked whole cell currents from cultured hippocampal neurons isolated from newborn β₃^+/+, β₃^+/−, and β₃^−/− mice. (A) Representative currents from individual neurons from each group in response to GABA (10 μM, 30 μM, 100 μM and 1 mM). The bar above the trace represents the duration of GABA application. (B) Pooled GABA concentration-response curves (5 ≤ n ≤ 10) yielding EC₅₀ values of 15.9 ± 0.7 μM, 27.4 ± 0.9 μM, 26.8 ± 3.5 μM for β₃^+/+, β₃^+/− and β₃^−/− mice, respectively. The Hill slopes were 1.7 ± 0.1, 1.1 ± 0.03 and 1.0 ± 0.1. (C) Pooled data for maximal GABA current amplitudes (mean ± SEM) for neurons tested. Numbers above bars indicate number of neurons contributing to mean value. The maximal response to 1 mM GABA for hippocampal neurons from β₃^−/− mice differs significantly from that for both β₃^+/+ (P < 0.002) and β₃^+/− (P < 0.0001) mice (Student’s unpaired t-test). Maximal GABA current amplitudes for hippocampal neurons from β₃^+/+and β₃^+/− mice do not differ significantly from each other.

In marked contrast to the situation in hippocampal neurons, where substantial GABA-evoked currents are observed in neurons from β₃^−/− mice, maximal GABA responses are drastically reduced in DRGs from β₃^−/− mice [6]. In particular, a majority (>80%) of DRGs from β₃^−/− mice (27/36 cells) respond to a maximal GABA concentration (1 mM) with less than 100 pA of current.

Since in situ hybridization studies have shown that the β₃ subunit is the predominant GABA_A receptor β subunit isoform mRNA in neonatal rat DRGs [10], GABA currents in DRGs can reveal whether there is any replacement of the lost β₃ subunit by another β subunit isoform. A useful pharmacological tool to distinguish between GABA_A β subunit isoforms is the anticonvulsant drug loreclezole, which has been demonstrated to have a much higher apparent affinity at GABA_A receptors containing β₂ or β₃ subunits, relative to those containing β₁ subunits [17,19]. Only a very small minority of DRG neurons from β₃^−/− mice have whole-cell GABA currents large enough to perform potentiation experiments (approximately 1 in every 15 cells tested), but in these cells we determined that submaximal (EC₂₀) GABA currents from β₃^+/+, β₃^+/− and β₃^−/− mice were all potentiated equally by a low concentration (5 μM) of loreclezole, consistent with most or all GABA_A receptors in DRGs from all three genotypes of mice containing β₂ and/or β₃ subunits (Fig. 2). Mean data for potentiation by 5 μM loreclezole were: β₃^+/+(52.3 ± 15.1% potentiation, n = 8), β₃^+/− (34.1 ± 8.3, n =8), and β₃^−/− (56.1 ± 27, n = 3). The barbiturate methohexital also potentiated submaximal GABA currents in neurons from all three mouse genotypes (Fig. 2). Mean data for potentiation by 5 μM methohexital were: β₃^+/+ (82.7 ± 24.7% potentiation, n = 5), β₃^+/− (100.1 ± 31.5, n = 4), and β₃^−/− (134.0 ± 22.2, n = 3). We also studied the effect of Zn²⁺, which has a much higher apparent affinity for GABA_A receptors lacking a γ subunit; in particular, currents at GABA_A αβ receptors can be inhibited completely by low micromolar concentrations of Zn²⁺ [2,16]. There was very little inhibition by a low concentration of zinc chloride (10 μM) for GABA currents from DRGs from either β₃^+/+ (9.7 ± 3.1% inhibition of EC₅₀ GABA current, n = 4) or β₃^−/− mice (3.4 ± 13.4% inhibition, n = 4).

Fig. 2.

Low concentrations of loreclezole (LCZ) and the barbiturate methohexital (MTX) enhance submaximal GABA currents in DRGs from both β₃^+/+ and β₃^−/− mice. Potentiation of approximately EC₂₀ (25 μM) GABA responses by LCZ and MTX at DRGs from both β₃^+/+ mice (A) and β₃^−/− mice (B). Note the much lower amplitude of currents in response to the maximal concentration (1 mM) of GABA in the neuron from the β₃^−/− mouse, which accounts for the noisier appearance of the traces from β₃^−/− cells.

In the present report, we have extended previous observations observations [6] to show a clear reduction in GABA-evoked currents in hippocampal neurons of β₃^−/− mice. The approximately 50% reduction in functional GABA_A receptors in hippocampal neurons from β₃^−/− mice is consistent with evidence from in situ hybridization experiments demonstrating substantial expression of β₃ mRNA in the CA1 neurons of the hippocampus [9]. The GABA currents in the hippocampal neurons from β₃ mice presumably reflect β₂ subunit-containing GABA_A receptors since β₂ mRNA is also detected in appreciable amounts in the hippocampus [9].

Pharmacological experiments in sensory neurons support the idea that there is little or no replacement of the product of the inactivated β₃ subunit gene by another β subunit isoform. Potentiation experiments using loreclezole show that the small GABA currents in a minority of DRG cells from β₃^−/− mice probably result from β₂-containing GABA_A receptors, which are also likely to be present even in β₃^+/+ mice, since β₂ mRNA is detectable at low levels in neonatal animals [10]. This is consistent with the β₁ isoform being predominantly expressed in glial cells [4,15]. In addition, most or all of the GABA_A receptors in DRG cells from β₃^−/− mice still contain the γ subunit (i.e. they are probably α_xβ₂γ receptors) since there is little or no inhibition of GABA currents by micromolar concentrations of Zn²⁺.

The β₃ knockout mouse appears to be a good model to examine the physiological and behavioral consequences of the lack of a significant fraction of the total GABA_A receptor population. The reduced pre- and post-synaptic inhibition resulting from the absence of many functional GABA_A receptors may well result in the hypersensitive behavior and epilepsy seen in β₃−/− mice and may also contribute to the developmental problems (e.g. cleft palate and early mortality) evident in some of these mice.