Search

The fastest inhibitory mechanism in the CNS is mediated by ionotropic GABAA receptors and it is known that subunit composition critically determines their properties. While a typical GABAA receptor consists of two α, two β and one γ/δ subunit, there are some exceptions, e.g. αβ receptors. Functional α1γ2 GABAA receptors can be expressed in recombinant model (Verdoorn et al., 1990) and although their role remains unknown, it seems appealing to extend their characterization to further explore the structure-function relationship of GABAA receptors. Intriguingly, this receptor is lacking canonical GABA binding sites but it can be activated by GABA and dose-response relationships for α1β2γ2L and α1γ2L receptors overlap. Deactivation kinetics was similar for both receptors but the percentage of the fast component was smaller in the case of α1γ2L receptors and, consequently, the mean deactivation time constant was slower. The rate and extent of macroscopic desensitization were smaller in the case of α1γ2L receptors but they showed slower recovery. Both receptor types had a similar proton sensitivity showing only subtle but significant differences in pH effects on deactivation. Flurazepam exerted a similar effect on both receptors but the rapid deactivation components were differently affected and an opposite effect was observed on desensitization extent. Rebound currents evoked by pentobarbital were undistinguishable for both receptor types. Taking altogether, although some significant differences were found, α1β2γ2L and α1γ2L receptors showed unforeseen similarity. We propose that functioning of GABAA receptors might rely on subunit-subunit cooperative interactions to a larger extent than believed so far.

Tricyclic antidepressants not only inhibit monoamine reuptake but also modulate Cys-loop receptors. However, it is not understood how this modulation is involved in their therapeutic effects. We analyzed the mechanisms of inhibition of homomeric 5-HT(3A) and alpha7-5HT(3A) receptors by tricyclic antidepressants at the single-channel and macroscopic current levels. These drugs reduce agonist-evoked currents in a noncompetitive and concentration-dependent manner. When they act on the open state, the reduction is similar for both receptors and it is voltage-dependent, thus suggesting an open-channel block process in which the blocked channel can either close or remain stabilized. By acting on the resting state, tricyclic antidepressants reduce the peak current in a voltage-independent manner, with a potency 6-fold higher for 5-HT(3A) than for alpha7-5HT(3A) (IC(50): 6 microM and 1 microM for alpha7-5HT(3A) and 5-HT(3A), respectively). Thus, tricyclic antidepressants may act on closed channels at the unshared extracellular domain from where they inhibit channel opening. Single alpha7-5HT(3A) channels in the continued presence of tricyclic antidepressants show: i) reduced open durations, compatible with open-channel block; ii) reduced burst durations, compatible with closing of blocked channels; and iii) reduced frequency of opening events, compatible with both impaired opening and stabilization of a closed state. In summary, our study reveals that tricyclic antidepressants inhibit homomeric Cys-loop receptors by acting through different mechanisms at open and closed conformational states and probably at two different domains, namely, the pore in the open state and the extracellular domain in the closed state. Fil: Gumilar, Fernanda Andrea. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Bahía Blanca. Instituto de Investigaciones Bioquímicas de Bahía Blanca. Universidad Nacional del Sur. Instituto de Investigaciones Bioquímicas de Bahía Blanca; Argentina Fil: Bouzat, Cecilia Beatriz. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Bahía Blanca. Instituto de Investigaciones Bioquímicas de Bahía Blanca. Universidad Nacional del Sur. Instituto de Investigaciones Bioquímicas de Bahía Blanca; Argentina

The neuronal nicotinic acetylcholine receptor alpha4 and beta2 subunits expressed in heterologous expression systems assemble into at least two distinct subunit stoichiometries of alpha4beta2 receptor. The (alpha4)2(beta2)3 stoichiometry is about 100-fold more sensitive to acetylcholine than the (alpha4)3(beta2)2 stoichiometry. In order to investigate if agonists in general distinguish high- and low-affinity alpha4beta2 nicotinic acetylcholine receptors, we have expressed human alpha4 and beta2 nicotinic acetylcholine receptor subunits in two different expression systems. The relative amounts of alpha4beta2 nicotinic acetylcholine receptors with high- and low-affinity for acetylcholine were manipulated by (a) injecting the subunit cDNAs at different alpha:beta ratios into Xenopus oocytes and (b) by culturing HEK-293 cells stably expressing alpha4beta2 nicotinic acetylcholine receptors overnight at different temperatures. The sensitivities of the alpha4beta2 nicotinic acetylcholine receptors to the agonists acetylcholine, 5-I A-85380, and TC-2559 were investigated using the voltage-clamp technique on Xenopus oocytes and using a fluorescent imaging plate reader to measure calcium responses from HEK-293 cells. Like acetylcholine, 5-I A-85380 produced biphasic concentration-response curves and the high-affinity component became larger when the cells were manipulated to produce a greater proportion of (alpha4)2(beta2)3 nicotinic acetylcholine receptors. Interestingly, under all circumstances, TC-2559 produced monophasic concentration-response curves. In oocytes injected with alpha4 and beta2 subunits in the 1:1 ratio the maximum effect of TC-2559 was 28% of that of acetylcholine. The EC50 for TC-2559 was not changed when oocytes were manipulated to express exclusively (alpha4)2(beta2)3 nicotinic acetylcholine receptors, however, the maximum effect of TC-2559 was dramatically enhanced. These results suggest that TC-2559 is a selective agonist of the (alpha4)2(beta2)3 nicotinic acetylcholine receptor stoichiometry.

The distribution and physiological functions of the α6 subunit-containing (α6⁎) nicotinic acetylcholine receptors in the central nervous system make them interesting putative therapeutic targets in several disorders. However, investigations into the receptors have been complicated by their inefficient functional expression in vitro. In the present study we have characterized and compared the pharmacological properties displayed by α6β4 and α6β4β3 nicotinic acetylcholine receptors assembled in tsA201 cells from the classical α6/α3 chimera (C1) and two novel α6/α3 chimeras (C6F223L and C16F223L) identified in a recent study (Jensen et al., 2013). Whereas the Bmax values exhibited by [3H]epibatidine at wild-type α6β4, C1β4, C6F223Lβ4 and C16F223Lβ4 receptors differed substantially, the radioligand and seven orthosteric nicotinic agonists exhibited very similar KD and Ki values at the four receptors. In the FLIPR™ Membrane Potential Blue assay, the agonists exhibited the same rank order of potencies [(±)-epibatidine>sazetidine A>varenicline>(-)-cytisine~(S)-nicotine>acetylcholine>carbachol] at the C1β4, C1β4β3, C6F223Lβ4, C6F223Lβ4β3, C16F223Lβ4 and C16F223Lβ4β3 receptors, albeit the absolute EC50 values differed somewhat between the receptors. Furthermore, four reference antagonists displayed the same rank order of inhibitory potencies at the six receptors [α-conotoxin PIA>2,2,6,6-tetramethylpiperidin-4-yl heptanoate>mecamylamine>dihydro-β-erythroidine]. Although all interpretations based on these results should be made keeping the molecular modifications in the α6 surrogate subunits in mind, this study sheds light on the pharmacological properties of α6β4⁎ nicotinic acetylcholine receptors and demonstrates the applicability of the C6F223L and C16F223L chimeras for studies of these receptors.

We have shown previously that chronic exposure to submicromolar concentrations of nicotine permanently inactivates alpha4beta2 and alpha7 neuronal nicotinic acetylcholine receptors while alpha3beta2 acetylcholine receptors are resistant to inactivation. Phosphorylation of the large cytoplasmic domain has been proposed to mediate functional inactivation. Chimeric subunits consisting of human alpha4 sequence from their N-terminus to either the beginning of the first transmembrane domain or the large cytoplasmic domain and alpha3 sequences thereafter formed acetylcholine receptors with beta2 subunits which were as susceptible to nicotine-induced inactivation as wild-type alpha4 acetylcholine receptors. The converse chimeras, containing the N-terminal parts of the alpha3 subunit and the C-terminal parts of the alpha4 subunit, formed acetylcholine receptors with beta2 subunits which were as resistant to nicotine-induced inactivation as wild-type alpha3beta2 acetylcholine receptors. Thus, inactivation of acetylcholine receptors produced by chronic exposure to nicotine results primarily from effects of the agonist on the extracellular and transmembrane domains of the alpha subunit.

Chicken GABA(A) receptor beta4L and beta4S subunits were expressed in Xenopus oocytes by cRNA injection. Oocytes expressing either beta4 subunit alone or in combination with the chicken alpha1 subunit were studied using the two-electrode voltage-clamp technique. Both the beta4L and beta4S subunits form homomeric GABA-gated Cl- channels with similar efficiencies. In comparison, oocytes expressing either the chicken alpha1 or beta2S polypeptide show no or barely detectable GABA responses, as reported by others for most single-subunit vertebrate GABA(A) receptors. The GABA-gated currents due to the beta4L-subunit homomer were not affected by the presence of actinomycin D during cRNA expression, indicating that nascent oocyte polypeptides are not required for channel formation. The homomeric beta4L-subunit receptors show high affinity for GABA with an EC50 value of 4.3 +/- 0.4 microM and a Hill coefficient of 1.1 +/- 0.1 (n = 6). In response to GABA application at the EC25 value, currents elicited from the beta4L-subunit receptor are enhanced by 50 microM pentobarbital (110 +/- 10%, n = 3) and 10 microM loreclezole (60 +/- 3%, n = 3), inhibited by 10 microM picrotoxinin (93 +/- 3%, n = 3), but not affected by 1 microM diazepam. These properties are similar to those found for oocytes expressing heteromeric chicken alpha1beta4L and alpha1beta2S receptors. Since the beta subunits of GABA(A) receptors provide essential determinants for receptor assembly and subcellular localization, homomeric beta4-subunit receptors are a useful model system for further study of the structure and function of GABA(A) receptors.

We investigated the possible association between delta and gamma2 subunits in native GABA(A) receptors, from different rat brain regions, using subunit-specific anti-delta and anti-gamma2 antibodies. Previous reports have provided somewhat controversial results, indicating both the presence and the absence of association between these two subunits in native receptors. Our results indicate the absence of co-localization between delta and gamma2 subunits. In immunoprecipitation experiments, anti-delta antibody consistently immunoprecipitated [3H]muscimol binding activity (GABA binding sites) from all brain areas tested (10-20% of the total binding). However, under the same conditions, no significant [3H]flumazenil or [3H]ethyl 8-azido-6-dihydro-5-methyl-6-oxo-4H-imidazol[1,5-a]-[1,4]benzodiazepine- 3-carboxylate (Ro15-4513) binding (benzodiazepine binding sites) activity was detected in the immunopellets. These results indicate the absence of association between delta and gamma2 subunits. This question was directly addressed by immunopurification and Western blot experiments. As expected, no gamma2 subunits were detected in anti-delta immunoaffinity purified receptors. Conversely, no delta subunits were identified in anti-gamma2 immunopurified receptors. Thus, these results demonstrate the absence of association between delta and gamma2 subunits in native GABA(A) receptors. Finally, our results also indicate the relevance of the solubilization conditions on the apparent association between different subunits of the native GABA(A) receptor complex.

Studies using recombinant systems indicate that 'diazepam-insensitive' GABA(A) receptors in the central nervous system contain alpha4 and alpha6 subunits while 'diazepam-sensitive' GABA(A) receptors contain alpha1, alpha2, alpha3 and alpha5 subunits. Both native and recombinant diazepam-sensitive GABA(A) receptors typically exhibit large, outwardly rectifying currents. For example, in patch clamp studies, Human Embryonic Kidney (HEK) 293 cells transfected with cDNAs encoding alpha1beta2gamma2 subunits exhibit a rectification ratio (I +60 mV/I -60 mV) of 1.95 +/- 0.21. However, anomalous rectification was observed in recombinant diazepam-insensitive GABA(A) receptors composed of either alpha4beta2gamma2 (rectification ratio, 0.74 +/- 0.09) or alpha6beta2gamma2 (rectification ratio, 0.67 +/- 0.11) subunits. Based on sequence differences between diazepam-sensitive and -insensitive GABA(A) receptor alpha subunits in the vicinity of the putative channel lining, a point mutation was introduced at His273 on the alpha4 subunit. The rectification ratio in cells expressing a mutated alpha4(Asn273)beta2gamma2 receptor increased to 1.92 +/- 0.17. Moreover, mutation of the homologous residue in the alpha1 subunit to histidine reduced the rectification ratio of alpha1(His274)beta2gamma2 to 1.02 +/- 0.12. The affinities of benzodiazepine site ligands at diazepam-sensitive and -insensitive GABA(A) receptors were unaffected by these mutations. Thus, the electrophysiological properties of diazepam-sensitive and -insensitive GABA(A) receptors may be as divergent as their pharmacological characteristics.

Recombinant human GABAA receptors were investigated in vitro by coexpression of cDNAs coding for α1, β2, and γ2 subunits in the baculovirus/Sf-9 insect cell system. We report that a single amino acid exchange (isoleucine 121 to valine 121) in the N-terminal, extracellular part of the α1 subunit induces a marked decrease in agonist GABAA receptor ligand sensitivity. The potency of muscimol and GABA to inhibit the binding of the GABAA receptor antagonist [3H]SR 95531 (2-(3-carboxypropyl)-3-amino-6-(4-methoxyphenyl)pyridazinium bromide) was higher in receptor complexes of α1(ile 121)β2γ2 than in those of α1(val 121)β2γ2 (IC50 values were 32-fold and 26-fold lower for muscimol and GABA, respectively). The apparent affinity of the GABAA receptor antagonist bicuculline methiodide to inhibit the binding of [3H]SR 95531 did not differ between the two receptor complex variants. Electrophysiological measurements of GABA induced whole-cell Cl− currents showed a ten-fold decrease in the GABAA receptor sensitivity of α1(val 121)β2γ2 as compared to α1(ile 121)β2γ2 receptor complexes. Thus, a relatively small change in the primary structure of the α1 subunit leads to a decrease selective for GABAA receptor sensitivity to agonist ligands, since no changes were observed in a GABAA receptor antagonist affinity and benzodiazepine receptor binding.

The potency of 30 benzodiazepine binding site ligands from 14 different structural classes for inhibition of [ 3 H]Ro 15-4513 (ethyl-8-azido-5,6-dihydro-5-methyl-6-oxo-4 H -imidazo[1,5-a][1,4]benzodiazepine-3-carboxylate) binding to human embryonic kidney (HEK) 293 cells transiently transfected with α 4 β 3 γ 2 S or α 1 β 3 γ 2 S subunits of GABA A receptors was investigated. Most of these compounds were unable to significantly inhibit [ 3 H]Ro 15-4513 binding to α 4 β 3 γ 2 S receptors under conditions where they potently inhibited binding to α 1 β 3 γ 2 S receptors. Nevertheless, compounds from four different structural classes were identified which exhibited a high affinity for α 4 β 3 γ 2 S receptors. Variation of the structure of these compounds could lead to new ligands selectively interacting with α 4 β 3 γ 2 S receptors. Compounds interacting with α 4 β 3 γ 2 S receptors were also able to inhibit [ 3 H]Ro 15-4513 binding to receptors consisting of α 4 γ 2 S subunits with comparable potency. These results support the conclusion that the α subunit is a major determinant of the benzodiazepine binding site properties of GABA A receptors containing α and γ subunits.

An alpha7 nicotinic acetylcholine receptor sequence was cloned from Rhesus monkey (Macaca mulatta). This clone differs from the mature human alpha7 nicotinic acetylcholine receptor in only four amino acids, two of which are in the extracellular domain. The monkey alpha7 nicotinic receptor was characterized in regard to its functional responses to acetylcholine, choline, cytisine, and the experimental alpha7-selective agonists 4OH-GTS-21, TC-1698, and AR-R17779. For all of these agonists, the EC(50) for activation of monkey receptors was uniformly higher than for human receptors. In contrast, the potencies of mecamylamine and MLA for inhibiting monkey and human alpha7 were comparable. Acetylcholine and 4OH-GTS-21 were used to probe the significance of the single point differences in the extracellular domain. Mutants with the two different amino acids in the extracellular domain of the monkey receptor changed to the corresponding sequence of the human receptor had responses to these agonists that were not significantly different in EC(50) from wild-type human alpha7 nicotinic receptors. Monkey alpha7 nicotinic receptors have a serine at residue 171, while the human receptors have an asparagine at this site. Monkey S171N mutants were more like human alpha7 nicotinic receptors, while mutations at the other site (K186R) had relatively little effect. These experiments point toward the basic utility of the monkey receptor as a model for the human alpha7 nicotinic receptor, albeit with the caveat that these receptors will vary in their agonist concentration dependency. They also point to the potential importance of a newly identified sequence element for modeling the specific amino acids involved with receptor activation.

In a recent study, racemic 3-(N,N-dimethylamino)butyl-N,N-dimethylcarbamate (1) was shown to be a potent agonist at neuronal nicotinic acetylcholine receptors with a high selectivity for nicotinic over muscarinic acetylcholine receptors [Mol. Pharmacol. 64 (2003) 865-875]. Here we present the synthesis and pharmacological characterization of a series of analogs of, where the methyl group at C-3 has been replaced by different alkyl substituents. Ring systems have been incorporated into the carbon backbone of some of the molecules, or the amino group has been build into ring systems. Furthermore, the (+)- and (-)-enantiomers of have been separated, and X-ray crystallography has revealed that (-)-1 possesses (S)-configuration. The compounds have been characterized pharmacologically at recombinant nicotinic receptor subtypes. The structure-activity relationship study has provided valuable insight into the mode of interactions of and its analogs with neuronal nicotinic acetylcholine receptors.

Efficient transfection of NG108-15 cells with GABA(B) receptor subunits was achieved using polyethylenimine. Baclofen modulated high voltage-activated Ca(2+) current in differentiated cells transfected with GABA(B1) and GABA(B2) receptor subunits or with the GABA(B2) subunit alone, but not with the GABA(B1) subunit alone. Characteristics of the current modulation were very similar for cells transfected with GABA(B1/2) and GABA(B2) subunits. Using antisense oligonucleotides against GABA(B1) subunits and also western immunoblotting, we are able to show that NG108-15 cells contain endogenous GABA(B1) subunits. Therefore, functional receptors can be formed by the combination of native GABA(B1) subunits with transfected GABA(B2) subunits, in agreement with the proposed heteromeric structure of GABA(B) receptors. Finally, we used selective channel blockers to identify the subtypes of Ca(2+) channels that are modulated by GABA(B) receptors. In fact, in differentiated NG108-15 cells, the recombinant GABA(B) receptors couple only to N-type Ca(2+) channels.

Although the neuronal nicotinic beta3 subunit was cloned several years ago, it has only recently been shown to form heteromeric channels when associated with other nicotinic subunits, and very little information is available concerning its assembly in the native nicotinic receptors of the nervous system. Using subunit-specific antibodies and immunoprecipitation experiments, we have identified the retina as being the chick central nervous system (CNS) area that expresses the highest level of the beta3 subunit. Sequential immunopurification experiments showed that there are at least two populations of beta3-containing receptors in chick retina: in one, the beta3 subunit is associated with the alpha6 and beta4 subunits; in the other more heterogeneous population, the beta3 subunit is associated with the alpha2, alpha3, alpha4, beta2 and beta4 subunits. Both of these receptor populations bind [3H]epibatidine and a number of nicotinic receptor agonists with high affinity (nM) and nicotinic receptor antagonists with a lower affinity (microM). The greatest pharmacological difference between the two populations is the affinity for the alpha-conotoxin MII, which inhibits binding to alpha6-containing receptors and not that to beta3-containing receptors. We also searched for the presence of the beta3 subunit associated with the alpha-bungarotoxin binding subunits alpha7 and/or alpha8 in retina and chick brain. Immunoprecipitation studies using anti-beta3 antibodies did not detect any specific alpha-bungarotoxin labeled receptors, thus, indicating that the beta3 subunit is not present in the alpha-bungarotoxin receptors of these areas.

The properties of wild type and mutant rat nicotinic alpha7 receptors expressed in Xenopus oocytes were investigated using electrophysiology and site-directed mutagenesis. When compared at individual agonist concentrations, neither the normalised nicotinic, nor acetylcholine, responses of the wild type receptors were significantly different from the corresponding responses obtained from a first extracellular domain mutant, phenylalanine(189)tyrosine (P0.05). The dissociation constants (K(D)) of the wild type (4.7 nM) and Phe(189)Tyr mutant (5.2 nM) receptors for alpha-bungarotoxin were estimated by an electrophysiological approach. The similarity of the results suggests that the mutation did not lead to a widespread disruption of structure-function relationships, although a slight change in nicotine sensitivity may have occurred. In contrast, the mutations (Tyr(190)Gln, first extracellular domain), (Glu(261)Ala, M2 region) severely compromised receptor function. An additional mutation was made in a negatively charged motif of the second extracellular domain which is conserved in homomeric nicotinic receptors. This mutation, Asp(268)Ala, also caused a loss of function. Thus the structure-function relationships in nicotinic alpha7 receptors have parallels with heteromeric nicotinic receptors, but there may also be some marked differences.

The presence of allosteric binding sites on recombinant GABAA receptors formed after transfection of human embryonic kidney (HEK) 293 cells with alpha 1-, beta 3-, or gamma 2-subunits, or with various combinations of these subunits, was systematically investigated. From all possible subunit combinations, high affinity [3H]muscimol binding sites were induced in cells transfected with alpha 1 beta 3- or alpha 1 beta 3 gamma 2-subunits only. GABAA receptor associated [3H]flunitrazepam binding sites were induced in cells after transfection with alpha 1 gamma 2- or alpha 1 beta 3, gamma 2-subunits, and [35S]r-butylbicyclophosphorothionate (TBPS) binding sites were found in cells transfected with beta 3-, beta 3 gamma 2-, alpha 1 beta 3-, or alpha 1 beta 3 gamma 2-subunits. Binding of [35S]TBPS could be inhibited by pentobarbital, etazolate, (+)-etomidate, alphaxalone, propofol, chlormethiazole, and 4'-chlorodiazepam (Ro 5-4864) with a potency which differed in cells transfected with beta 3-, beta 3 gamma 2-, alpha 1 beta 3-, or alpha 1 beta 3 gamma 2-subunits. Results obtained indicate that receptors with different subunit composition actually can be formed in HEK cells and exhibit distinct pharmacological properties.

The diversity of neuronal glutamate receptors continues to increase with the discovery of multiple subunits and subunit families. The significance of this potential receptor heterogeneity is unknown because pharmacological tools that could clearly distinguish between different structural isoforms have not yet been identified. A novel glutamate receptor antagonist, 5-nitro-6,7,8,9-tetrahydrobenzo[g]indole-2,3-dione-3-oxime (NS-102), has been shown previously to selectively block the low affinity [3H]kainate binding site in rat brain. We have examined the effect of NS-102 on receptors expressed in fibroblasts from either glur6 subunits or a combination of glurB and glurD (glurB/D receptors). NS-102 (3 microM) reduced currents mediated by glur6 receptors and had very little effect on currents mediated by glurB/D receptors. The binding of [3H]kainate to glur6 receptors showed properties similar to those of the brain low affinity [3H]kainate binding site, and NS-102 inhibited specific binding to glur6 receptors with a potency nearly identical to those sites in brain membranes. Our findings suggest that NS-102 will be useful in identifying the functional role of native receptors containing a glur6 subunit.

The binding of the imidazobenzodiazepine, [3H]Ro 15-4513, to cerebellar granule cell-specific GABAA/benzodiazepine receptors is typically insensitive to benzodiazepine receptor agonists such as diazepam. A mutation in the alpha 6 subunit, causing replacement of the arginine at the 100 position by glutamine (Q100), has recently been found in an alcohol- and benzodiazepine-sensitive rat line. The mutant alpha 6(Q100)beta 2 gamma 2 recombinant receptors are sensitive to diazepam. The binding of [3H]Ro 15-4513 to cerebellar diazepam-insensitive receptors is enhanced by GABA, whereas binding to diazepam-sensitive receptors is inhibited. Recombinant receptors consisting of beta 2 and gamma 2 subunits together with the wildtype alpha 6 or mutant alpha 6(Q100) subunit showed positive modulation of [3H]Ro 15-4513 binding by GABA, whereas alpha 1 beta 2 gamma 2 receptors showed negative modulation. The picrotoxin-sensitive binding of a convulsant, t-butylbicyclophosphoro[35S]thionate ([35S]TBPS), was inhibited in the alpha 6 beta 2 gamma 2 and alpha 6(Q100) beta 2 gamma 2 receptors by GABA at concentrations less than one-tenth of those required in the alpha 1 beta 2 gamma 2 receptors. GABA effects on [35S]TBPS binding were only slightly affected by diazepam in the alpha 6(Q100) beta 2 gamma 2 receptors, while profound effects were seen in the alpha 1 beta 2 gamma 2 receptors in the presence of diazepam. The results with the mutant receptor suggest that the alpha 1 and alpha 6 subunits are responsible for differential allosteric actions by GABA on other binding sites, independently of the structures defining the benzodiazepine binding pharmacology.

GABAA receptors with functional benzodiazepine receptors have been reconstituted by coexpression of alpha-, beta- and gamma-subunit cDNAs. In brain, proteins of the alpha- and beta-subunits, but not the gamma 2-subunit have been identified. Using an antipeptide antiserum, we now demonstrate that the gamma 2-subunit is a protein of 43 kDa. It is present in at least 50% of GABAA receptors, as shown by immunoprecipitation.