Rochelle D. Schwartz

The GABAa receptor-gated ion channel: Biochemical and pharmacological studies of structure and function

In the past few years, substantial advances have been made in analyzing the structure and function of the GABA receptor-gated Cl- channel. A major goal is to identify the molecular characteristics of the GABAA receptor that are necessary for maintaining normal GABAergic neurotransmission. Future studies will undoubtedly include techniques that have been used successfully to construct a detailed structural and dynamic model of the nACHR-gated ion channel. These include X-ray scattering, single group rotation theory, and genetic homology, deletion and site-directed mutation studies. Such techniques will make it possible to identify the structural defects that give rise to abnormal GABA receptor function and possibly to sleep, anxiety and seizure disorders.

Acute stress enhances the activity of the GABA receptor-gated chloride ion channel in brain

The effect of acute swim stress on the functional activity of the gamma-aminobutyric acid (GABA) receptor/chloride ion channel was studied using an assay to measure 36chloride (36Cl-) uptake into rat brain synaptoneurosomes. Muscimol-stimulated 36Cl- uptake in cerebral cortex, hippocampus, but not cerebellum, was enhanced (increased potency and efficacy) in rats subjected to 10 min of swimming, compared to non-stressed, control rats. The effect of swim stress on the activity of the GABA receptor/Cl ion channel was prevented by adrenalectomy.

Alcohols stimulate γ-aminobutyric acid receptor-mediated chloride uptake in brain vesicles: correlation with intoxication potency

A series of short-chain alcohols, including ethanol, were examined for their abilities to stimulate gamma-aminobutyric acid (GABA) receptor-mediated chloride uptake into isolated brain vesicles. All of the alcohols tested stimulated 36 chloride uptake, at concentrations that occur during acute intoxication, and their potencies in stimulating GABA receptor-mediated chloride uptake were highly correlated with both their intoxication potencies in rats (r = 0.96; P less than 0.0001) and their membrane/buffer partition coefficients (r = 0.91; P less than 0.0005). Thus, the activity of alcohols at the GABA receptor-coupled chloride ion channel appears to be related to their ability to enter hydrophobic regions of the neuronal membrane. These data suggest that the anxiolytic, sedative/hypnotic and intoxicating properties of ethanol may, in part, be mediated via an action at central GABA receptors.

Barbiturate and picrotoxin-sensitive chloride efflux in rat cerebral cortical synaptoneurosomes

The effects of various barbiturates and picrotoxin in modifying the efflux of chloride (36Cl−) was studied in a novel subcellular preparation from rat cerebral cortex, the ‘synaptoneurosome’. Dilution of synaptoneurosomes pre‐loaded with 36Cl− resulted in rapid efflux of 36Cl− that could be measured as early as 10 s following dilution. In the presence of barbiturates such as pentobarbital and hexobarbital there was a significant increase in 36Cl− efflux which was not observed with the pharmacologically‐inactive barbiturate, barbital. The effect of barbiturates in enhancing 36Cl− efflux was also stereospecific [(−)‐DMBB > (+)‐DMBB] and reversed by picrotoxin. By contrast, picrotoxin alone significantly inhibited 36Cl− efflux. These data demonstrate pharmacologically relevant Cl− transport for the first time in a subcellular brain preparation.

Regulation of γ‐Aminobutyric Acid/Barbiturate Receptor‐Gated Chloride Ion Flux in Brain Vesicles by Phospholipase A2: Possible Role of Oxygen Radicals

Preincubation of brain membranes with phospholipase A2 (PLA2) has been shown previously to affect the binding characteristics of various recognition sites associated with the γ‐aminobutyric acid (GABA) receptor complex. In the present study, we have investigated the effects of PLA2 (from Naja naja siamensis venom) on the functional activity of the GABA receptor/chloride ion channel. PLA2 (0.001–0.02 U/mg protein) preincubation decreased pentobarbital‐induced 36Cl‐ efflux and musci‐mol‐induced 36Cl‐ uptake in rat cerebral cortical synaptoneurosomes. The effect of PLA2 was prevented by EGTA and two nonselective PLA2 inhibitors, mepacrine and bromophenacyl bromide. The removal of free fatty acids by addition of bovine serum albumin both prevented and reversed the effect of PLA2. Products of the catalytic activity of PLA2, such as the unsaturated free fatty acids, arachidonic and oleic acids, mimicked the effect of PLA2. However, the saturated fatty acid, palmitic acid, and lysophosphatidyl choline had no effect on pentobarbital‐induced 36Cl‐ efflux. Because unsaturated free fatty acids are highly susceptible to peroxidation by oxygen radicals, the role of oxygen radicals was investigated. Xanthine plus xanthine oxidase, a superoxide radical generating system, mimicked the effect of PLA2, whereas the superoxide radical scavenger, superoxide dismutase, diminished the effects of PLA2 and arachidonic acid on pentobarbital‐induced 36Cl‐ efflux. Similarly, the effect of PLA2 was also inhibited by methanol (1 mM), a scavenger of the hydroxyl radical, and by catalase. These data indicate that exogenously added PLA2 induces alterations in membrane phospholipids, possibly promoting the generation of oxygen radicals and fatty acid peroxides which can ultimately modulate GABA/barbitu‐rate receptor function in brain.

Postischemic diazepam is neuroprotective in the gerbil hippocampus

In this study, we address the hypothesis that enhancement of gamma-aminobutyric acid (GABA) neurotransmission following an ischemic episode is neuroprotective in the hippocampus. Mongolian gerbils were subjected to transient forebrain ischemia for 5 min by occlusion of the carotid arteries and then administered diazepam (10 mg/kg i.p.) 30 min or 30 and 90 min following ischemia. Diazepam produced a significant decrease in both rectal and brain temperature (4-6 degrees C) in the sham and ischemic gerbils. 1 day following the onset of reperfusion, diazepam substantially reduced the hyperactivity normally induced by the ischemic episode. 7 days later, neuronal viability in the hippocampus was assessed. The single dose of diazepam completely protected the CA1 pyramidal cells of the hippocampus in 62% of the gerbils and the double dose of diazepam completely protected CA1 pyramidal neurons in 67% of the gerbils. There was a significant correlation between the degree of pyramidal cell degeneration in the CA1 area of the hippocampus measured 7 days following ischemia and the degree of hyperactivity measured 1 day following ischemia. Diazepam also prevented the loss of [35S]t-butylbicyclophosphorothionate ([35S]TBPS) binding to GABA-gated chloride channels in the dendritic fields of the CA1 area of the hippocampus. Our findings support the hypothesis that enhancement of GABA neurotransmission following an ischemic event may offset neuronal excitability and prevent neuronal death in specific brain regions. We conclude that GABA-enhancing drugs, such as diazepam, are attractive candidates as neuroprotective agents following ischemic insults.

Inhibition of GABA-gated chloride channel function by arachidonic acid ☆

The effects of arachidonic acid and its metabolites on gamma-aminobutyric acid (GABAA) receptor function were determined in rat cerebral cortical synaptoneurosomes. Incubation of synaptoneurosomes with phospholipase A2 decreased muscimol-induced 36Cl- uptake. Arachidonic acid, the major unsaturated fatty acid released by phospholipase A2, also inhibited muscimol-induced 36Cl uptake. Similar inhibition was obtained with other unsaturated fatty acids (docosahexaenoic, oleic) but not with saturated fatty acids (stearic, palmitic). The effect of arachidonic acid on muscimol responses was inhibited by bovine serum albumin (BSA), and BSA enhanced muscimol responses directly, indicating the generation of endogenous arachidonic acid in the synaptoneurosome preparation. The generation of endogenous arachidonic acid was also indicated by the ability of 2 inhibitors of arachidonic acid metabolism, indomethacin and nordihydroguaiaretic acid (NDGA), to inhibit muscimol-induced 36Cl uptake. We conclude that arachidonic acid probably has both direct and indirect actions on muscimol responses since both enzyme inhibitors inhibited muscimol responses but did not prevent the effect of exogenously added arachidonic acid. In additional experiments, arachidonic acid metabolites generated by cyclooxygenase, prostaglandins D2, E2 and F2 alpha, each decreased muscimol responses; prostaglandins F2 alpha was the most potent inhibitor. Since the unsaturated fatty acids and their metabolites are most susceptible to peroxidation, a generating system of superoxide radicals was tested on muscimol responses. A combination of xanthine and xanthine oxidase inhibited muscimol-induced 36Cl uptake in a concentration-dependent manner. We propose that the inhibition of GABAA neurotransmission by arachidonic acid and its metabolites can lead to increased neuronal excitability. This mechanism may play an important role in the development of neuronal damage following seizures or cerebral ischemia.

Characterization of GAVAA receptor-mediated 36chloride uptake in rat brain synaptoneurosomes

gamma-Aminobutyric acid (GABA) receptor-mediated 36chloride (36Cl-) uptake was measured in synaptoneurosomes from rat brain. GABA and GABA agonists stimulated 36Cl- uptake in a concentration-dependent manner with the following order of potency: Muscimol greater than GABA greater than piperidine-4-sulfonic acid (P4S)greater than 4,5,6,7-tetrahydroisoxazolo-[5,4-c]pyridin-3-ol (THIP) = 3-aminopropanesulfonic acid (3APS) much greater than taurine. Both P4S and 3APS behaved as partial agonists, while the GABAB agonist, baclofen, was ineffective. The response to muscimol was inhibited by bicuculline and picrotoxin in a mixed competitive/non-competitive manner. Other inhibitors of GABA receptor-opened channels or non-neuronal anion channels such as penicillin, picrate, furosemide and disulfonic acid stilbenes also inhibited the response to muscimol. A regional variation in muscimol-stimulated 36Cl- uptake was observed; the largest responses were observed in the cerebral cortex, cerebellum and hippocampus, moderate responses were obtained in the striatum and hypothalamus and the smallest response was observed in the pons-medulla. GABA receptor-mediated 36Cl- uptake was also dependent on the anion present in the media. The muscimol response varied in media containing the following anions: Br- greater than Cl- greater than or equal to NO3- greater than I- greater than or equal to SCN- much greater than C3H5OO- greater than or equal to ClO4- greater than F-, consistent with the relative anion permeability through GABA receptor-gated anion channels and the enhancement of convulsant binding to the GABA receptor-gated Cl- channel.

Bidirectional modulation of GABA-gated chloride channels by divalent cations: inhibition by Ca2+ and enhancement by Mg2+.

Abstract: The effects of the divalent cations Ca2+, Sr2+, Ba2+, Mg2+, Mn2+, and Cd2+ were studied on γ‐aminobutyric acidA (GABAA) responses in rat cerebral cortical synaptoneurosomes. The divalent cations produced bidirectional modulation of muscimol‐induced 36Cl− uptake consistent with their ability to permeate and block Ca2+channels. The order of potency for inhibition of muscimol responses was Ca2+ > Sr2+ > Ba2+, similar to the order for permeation of Ca2+ channels in neurons. The order of potency for enhancement of muscimol responses was Cd2+> Mn2+ > Mg2+, similar to the order for blockade of Ca2+channels in neurons. Neither Ca2+ nor Mg2+ caused accumulation of GABA in the extravesicular space due to increased GABA release or decreased reuptake of GABA by the synaptoneurosomes. The inhibition of muscimol responses by Ca2+ was most likely via an intracellular site of action because additional inhibition could be obtained in the presence of the Ca2+ ionophore, A23187. This confirms electrophysiologic findings in cultured neurons from several species. In contrast, the effects of Cd2+, Mn2+, and Mg2+ may be mediated via blockade of Ca2+ channels or by intracellular sites, although the results of these studies do not distinguish between the two loci. The effects of Zn2+ were also studied, because this divalent cation is reported to have widely divergent effects on GABAA responses. In contrast to other studies, we demonstrate that Zn2+ inhibits GABAA responses in an adult neuronal preparation. Zn2+ produced a concentration‐dependent inhibition (limited to 40%) of muscimol responses with an EC50 of 60 μM. The inhibition of muscimol‐induced 38Cl− uptake by Zn2+ was noncompetitive. The effect of Zn2+was reduced in the presence of Mg2+ in a competitive or allosteric manner. The portion of GABAA receptors sensitive to Zn2+ may reflect a specific subunit composition in cerebral cortex as previously observed for recombinant GABAA receptors in several expression systems. The modulation of GABAA receptor function by Ca2+ and other divalent cations may play an important role in the development and/or attenuation of neuronal excitability associated with pathologic conditions such as seizure activity and cerebral ischemia.

Optical imaging of intracellular chloride in living brain slices

We developed an optical imaging technique to measure changes in intracellular levels of Cl- in neurons within the living brain slice. After rat brain slices were incubated with the permeant form of the Cl(-)-sensitive dye, 6-methoxy-N-ethylquinolinium chloride (MEQ), neurons could be imaged within the hippocampus, cerebral cortex and cerebellum using fluorescence microscopy. Both soma and dendrites were clearly visible in pyramidal neurons, interneurons, Purkinje cells and cerebellar granule cells. Increased intracellular levels of Cl- were produced by bath application of the inhibitory neurotransmitter, gamma-aminobutyric acid (GABA). Within hippocampal pyramidal neurons and interneurons, GABA produced a concentration-dependent decrease in fluorescence (EC50 = 200 microM). The GABA response was mediated via the GABA receptor since it was blocked by picrotoxin and mimicked by the agonist, muscimol. Muscimol, which is not transported by the GABA re-uptake pump, was approximately 20-fold more potent than GABA. The method developed was also used to image intracellular Cl- levels with UV laser scanning confocal microscopy. Even greater resolution was obtained and deeper structures could be imaged in cerebral cortex and hippocampus. This is the first demonstration of optical imaging to measure intracellular Cl- dynamics in living brain slices using fluorescence microscopy and laser scanning confocal microscopy.