Analysis of the Presence and Abundance of GABAA Receptors Containing Two Different Types of α Subunits in Murine Brain Using Point-mutated α Subunits*

γ-Aminobutyric acid, type A (GABAA) receptors are pentameric proteins of which the majority is composed of two α subunits, two β subunits and one γ subunit. It is well documented that two different types of α subunits can exist in a singles GABAA receptor complex. However, information on the abundance of such GABAA receptors is rather limited. Here we tested whether mice containing the His to Arg point mutation in the α1, α2, or α3 subunit at positions 101, 101, and 126, respectively, which render the respective subunits insensitive to diazepam, would be suitable to analyze this issue. Immunodepletion studies indicated that the His to Arg point mutation solely rendered those GABAA receptors totally insensitive to diazepam binding that contain two mutated α subunits in the receptor complex, whereas receptors containing one mutated and one heterologous α subunit not carrying the mutation remained sensitive to diazepam binding. This feature permitted a quantitative analysis of native GABAA receptors containing heterologous α subunits by comparing the diazepam-insensitive binding sites in mutant mouse lines containing one mutated α subunit with those present in mouse lines containing two different mutated α subunits. The data indicate that the α1α1-containing receptors with 61% is the most abundant receptor subtype in brain, whereas the α1α2 (13%), α1α3 (15%), α2α2 (12%), α2α3 (2%), and α3α3 combinations (4%) are considerably less expressed. Only within the α1-containing receptor population does the combination of equal α subunits (84% α1α1, 7% α1α2, and 8% α1α3) prevail, whereas in the α2-containing receptor population (46% α2α2, 36% α2α1, and 19% α2α3) and particularly in the α3-containing receptor population (27% α3α3, 56% α3α1, and 19% α3α2), the receptors with two different types of α subunits predominate. This experimental approach provides the basis for a detailed analysis of the abundance of GABAA receptors containing heterologous α subunits on a brain regional level.

From the ‡Institute of Pharmacology and Toxicology, University of Zü rich, Winterthurerstrasse 190, CH-8057 Zü rich, Switzerland and the ¶Department of Chemistry and Applied Biosciences, Federal Institute of Technology, Winterthurerstrasse 190,8057 Zü rich, Switzerland ␥-Aminobutyric acid, type A (GABA A ) receptors are pentameric proteins of which the majority is composed of two ␣ subunits, two ␤ subunits and one ␥ subunit. It is well documented that two different types of ␣ subunits can exist in a singles GABA A receptor complex. However, information on the abundance of such GABA A receptors is rather limited. Here we tested whether mice containing the His to Arg point mutation in the ␣1, ␣2, or ␣3 subunit at positions 101, 101, and 126, respectively, which render the respective subunits insensitive to diazepam, would be suitable to analyze this issue. Immunodepletion studies indicated that the His to Arg point mutation solely rendered those GABA A receptors totally insensitive to diazepam binding that contain two mutated ␣ subunits in the receptor complex, whereas receptors containing one mutated and one heterologous ␣ subunit not carrying the mutation remained sensitive to diazepam binding. This feature permitted a quantitative analysis of native GABA A receptors containing heterologous ␣ subunits by comparing the diazepam-insensitive binding sites in mutant mouse lines containing one mutated ␣ subunit with those present in mouse lines containing two different mutated ␣ subunits. The data indicate that the ␣1␣1-containing receptors with 61% is the most abundant receptor subtype in brain, whereas the ␣1␣2 (13%), ␣1␣3 (15%), ␣2␣2 (12%), ␣2␣3 (2%), and ␣3␣3 combinations (4%) are considerably less expressed. Only within the ␣1-containing receptor population does the combination of equal ␣ subunits (84% ␣1␣1, 7% ␣1␣2, and 8% ␣1␣3) prevail, whereas in the ␣2-containing receptor population (46% ␣2␣2, 36% ␣2␣1, and 19% ␣2␣3) and particularly in the ␣3-containing receptor population (27% ␣3␣3, 56% ␣3␣1, and 19% ␣3␣2), the receptors with two different types of ␣ subunits predominate. This experimental approach provides the basis for a detailed analysis of the abundance of GABA A receptors containing heterologous ␣ subunits on a brain regional level.
GABA A 1 receptors mediate the majority of fast inhibitory neurotransmission in the mammalian brain by controlling an integral chloride ion channel. Enhancement of neuronal inhibition by positive allosteric modulation of GABA A receptors via the benzodiazepine-binding site underlies the pharmacotherapy of several neurological and psychiatric disorders such as generalized anxiety disorders, sleep disturbances, and seizure disorders. GABA A receptors are pentameric transmembrane proteins build of different classes of subunits (␣1-6, ␤1-3, ␥1-3, 1-3, ␦, ⑀, and ) (1). The vast majority of GABA A receptors are composed of ␣, ␤, and ␥2 subunits, where the ␣ subunit variant determines the ligand binding characteristics of the benzodiazepine site of the various receptor subtypes. GABA A receptors containing the ␣1, ␣2, ␣3, or ␣5 subunit in combination with any ␤ subunit and the ␥2 subunit are the most abundant subtypes in the brain and mediate the modulatory actions of clinically used benzodiazepine site agonists, such as diazepam, flunitrazepam, and clonazepam (2,3). The binding of clinical relevant benzodiazepine site ligands to these socalled diazepam-sensitive GABA A receptor subtypes depends on a single amino acid residue, His at position 101 in ␣1 and ␣2, position 126 in ␣3, or position 105 in ␣5 (4,5).
It is well documented that GABA A receptors exist in brain that contain two different types of subunits in a single receptor complex (6 -15). However, it is currently poorly understood to which degree receptors containing two heterologous ␣ subunits contribute to diazepam-sensitive GABA A receptors. So far, the abundance of ␣1␣2and ␣1␣3-containing receptors in the cerebral cortex and ␣1␣5and ␣2␣5-containing receptors in the hippocampus have been estimated by sequential steps of immunoprecipitation and quantification of precipitated receptors by radioligand binding (6 -8). This methodological approach is a demanding task and because of the amount of tissue required is hardly applicable to all brain regions. In addition, immunopurification studies are limited by the fraction of receptors that can be solubilized from the membranes.
In the present paper an alternative experimental approach was used to overcome these experimental limitations. This method is based on the analysis by ligand binding studies of mutant mouse lines carrying a point mutation in the diazepamsensitive ␣ subunits. Because the diazepam sensitivity of GABA A receptors depends on the presence of a histidine residue in the drug-binding domain of the ␣ subunits ␣1, ␣2, ␣3, and ␣5, they are rendered diazepam-insensitive when this histidine residue is replaced by arginine (4,5). This feature had been previously used to analyze the contribution of GABA A receptor subtypes to the diverse actions of diazepam. By generating mouse lines that contain the His to Arg (H/R) point mutation in the ␣1, ␣2, or ␣3 subunit, it was shown that sedation, anterograde amnesia, and most of the anticonvulsant activities are mediated by ␣1-containing receptors (16), whereas the anxiolytic like activity is mediated by ␣2-containing receptors (17).
Here we show that the different ␣ (H/R) point-mutated mouse lines provide an opportunity to estimate the abundance of receptor populations containing two different types of ␣ subunits in a receptor complex. Immunodepletion studies indicated that in ␣ (H/R) mutant mouse lines only receptors containing two mutated ␣ subunits fail to bind diazepam (e.g. ␣1 (H101R) ␣3 (H126R) ), whereas receptors containing one mutated ␣ subunit and one nonmutated heterologous ␣ subunit (e.g. ␣1 (H101R) ␣3) remain sensitive to diazepam binding. This permitted a detailed analysis on the presence and abundance of receptors containing either two homologous or two heterologous ␣ subunits by comparing the proportion of diazepam-sensitive and diazepam-insensitive binding sites in wild type mice with the mouse lines containing a single type of mutated ␣ subunit (␣1 H101R , ␣2 H101R , or ␣3 H126R ) and newly crossbred mouse lines containing two different types of mutated ␣ subunits (␣1 H101R ␣2 H101R , ␣1 H101R ␣3 H126R , or ␣2 H101R ␣3 H126R ).
Membrane Preparation-Brain tissue from 8 -12-week-old mice was homogenized in 10 volumes of 10 mM Tris-HCl, pH 7.4, 0.32 M sucrose, 5 mM EDTA, 0.1 mM phenylmethylsulfonyl fluoride and centrifuged at 1000 ϫ g for 10 min. Following recentrifugation of the supernatant for 20 min at 12,000 ϫ g, the crude membrane pellet was washed three times with buffer and stored at Ϫ30°C until used.
Solubilization and Immunoprecipitation-For solubilization, the membranes were thawed and washed once with 10 mM Tris-HCl, pH 8, 150 mM NaCl, 1 mM EDTA, 200 mg/liter bacitracin, 0.1 mM phenylmethylsulfonyl fluoride, 2.3 mg/liter aprotinin, 1 mM benzamidine and resuspended in the same buffer to a protein concentration of 5-7 mg/ml followed by addition of sodium deoxycholate to a final concentration of 0.5%. After incubation for 30 min at 4°C, insoluble material was removed by centrifugation for 60 min at 100,000 ϫ g. The supernatant was immediately used for immunoprecipitation experiments.
For immunoprecipitation, aliquots (0.2-0.5 ml) of the deoxycholate extract were incubated with subunit-selective antiserum at concentrations leading to the maximum immunoprecipitation of the corresponding receptors overnight at 4°C, followed by incubation with 50 l of Pansorbin (suspension of 10% Staphylococcus aureus; Calbiochem) for 30 -60 min at room temperature. After extensive washing, the precipitates were resuspended in 10 mM Tris-HCl, pH 8, 150 mM NaCl, 1 mM EDTA, 200 mg/liter bacitracin, 0.1 mM phenylmethylsulfonyl fluoride, 2.3 mg/liter aprotinin, 1 mM benzamidine, 0.2% Triton X-100 for [ 3 H]Ro 15-4513 binding. The specificity of immunoprecipitation was verified in competition experiments using the respective peptide antigen (50 g/ml extract). No specific [ 3 H]Ro 15-4513 binding was observed in the immunoprecipitate after co-incubation of the respective antiserum with the corresponding peptide antigen (data not shown). Diazepam displacement studies using immunoprecipitated GABA A receptors were performed by incubating aliquots of the washed immunoprecipitates with serial dilutions of diazepam at a fixed concentration of [ 3 H]Ro 15-4513 (10 nM). After incubation for 90 min on ice, the reaction was stopped by addition of 4 ml of 50 mM Tris-HCl, pH 7.5, followed immediately by rapid vacuum filtration on Whatman GF/C filters. The filters were washed three times with 4 ml of buffer and subjected to liquid scintillation counting. Nonspecific radioligand binding was determined by including 10 M of flumazenil in parallel incubations. The binding data were analyzed using the program 'KELL for Windows 6.0.5Ј (Biosoft, UK).
To analyze this assumption, the respective ␣1 (H101R) -, ␣2 (H101R) -, and ␣3 (H126R) -containing receptor populations were isolated by immunoprecipitation from deoxycholate extracts of whole brain membranes using subunit-selective antisera. The proportion of diazepam-sensitive and diazepam-insensitive [ 3 H]Ro 15-4513-binding sites in the immunoprecipitated receptor populations was then determined by diazepam displacement of [ 3 H]Ro 15-4513 binding. Interestingly, all of the receptor populations containing the point mutation displayed a high affinity diazepam-binding component corresponding to that of wild type receptors (Fig. 1). Whereas in ␣1 (H101R) -containing receptors only 16 Ϯ 1% remained diazepam-sensitive, 54 Ϯ 10% of ␣2 (H101R) -containing receptors and 73 Ϯ 4% of ␣3 (H126R)containing receptors displayed diazepam-sensitive binding sites ( Fig. 1 and Table II). This result demonstrates that at least some subpopulations of GABA A receptors containing the H/R mutation had retained diazepam-sensitive binding sites.
Because GABA A receptors exist in brain that contain two different types of ␣ subunits in the same receptor complex (6 -15), the presence of high affinity diazepam binding in the mutated receptor populations is likely to be caused by the subpopulation of GABA A receptors containing a nonmutated heterologous ␣ subunit (which is not ␣4 or ␣6) in the receptor complex that confers high affinity diazepam binding. To verify this hypothesis, brain extracts were immunodepleted in a sequential fashion to initially remove receptors containing the nonmutated ␣ subunit (Fig. 2). For instance, from the brain extract of ␣2 (H101R) mice ␣1-containing receptors were removed by quantitative immunoprecipitation followed by precipitation of the receptor population containing the ␣2 subunit and de-termination of diazepam-sensitive and diazepam-insensitive binding sites. If the diazepam-sensitive binding component in the mutated receptor population is due to a heterologous nonmutated ␣ subunit in the receptor complex (e.g. ␣1 in ␣1␣2 (H101R) ), the proportion of diazepam-insensitive binding sites will be increased after immunodepletion of these receptors because diazepam-sensitive binding sites are removed from the mutated receptor population.

Proportion of Diazepam-insensitive [ 3 H]Ro 15-4513 Binding in Double-mutant Mouse Lines-
The observation that the remaining diazepam-sensitive binding in receptors containing a point-mutated ␣ subunit is due to the presence of a nonmutated heterologous ␣ subunit provides a means to estimate the abundance of receptors containing two different types of ␣ subunits. As a prerequisite, mouse lines were generated by cross-breeding that contained the point mutation in two different ␣ subunits, i.e. ␣1 (H101R) ␣2 (H101R) , ␣1 (H101R) ␣3 (H126R) , and ␣2 (H101R) ␣3 (H126R) . The abundance of receptor subpopulations containing two heterologous ␣ subunits was then estimated by comparing the proportion of diazepam-insensitive binding present in the membrane fractions of brains of double-mutant mice with the sum of diazepam-insensitive binding present in the corresponding singlemutant mice. Because in the double-mutant mice GABA A recep-tors containing the two distinct point-mutated ␣ subunits (e.g. ␣1 (H101R) ␣2 (H101R) ) are rendered in addition insensitive to diazepam binding (compared with the respective single-mutant mice), the amount of diazepam-insensitive binding sites is expected to be greater than the sum calculated from the corresponding single-mutant mice.

H]Ro 15-4513 Binding in Immuno-isolated GABA A Receptor Populations of Double-mutant
Mice-To estimate the relative abundance of GABA A receptors containing two different types of ␣ subunits within the ␣1-, ␣2-, or ␣3-containing receptor population, GABA A receptors were immunoprecipitated with the respective subunit-selective antisera from deoxycholate extracts of crude brain membranes derived from single-mutant mice (␣1 (H101R) , ␣2 (H101R) , and ␣3 (H126R) ) as well as from doublemutant mice (␣1 (H101R) ␣2 (H101R) , ␣1 (H101R) ␣3 (H126R) , and ␣2 (H101R) ␣3 (H126R) ) and analyzed for the proportion of diazepamsensitive and diazepam-insensitive binding. In all cases the diazepam-insensitive binding sites in the immunoprecipitates from double-mutant mice exceeded the sum of diazepam-insensitive sites in the immunoprecipitates of the corresponding single-mutant mice ( Table II). Immunoprecipitation of ␣1-containing receptors from brain extracts of ␣1 (H101R) ␣2 (H101R) mice yielded 91 Ϯ 7% diazepam-insensitive binding sites, which is 7% larger than the diazepam-insensitive sites from the corresponding single-mutant ␣1 (H101R) (84 Ϯ 1%; Table II). Likewise, immunoprecipitation of ␣2-containing receptors from brain extracts of ␣1 (H101R) ␣2 (H101R) mice resulted in 82 Ϯ 7% diazepam-insensitive binding sites, which is 36% larger than the diazepam-insensitive sites from the corresponding single-mutant ␣2 (H101R) . This result suggests that the ␣1 receptor population comprises 7% ␣1␣2-containing receptors, whereas the ␣2-containing receptor population includes 36% ␣1␣2containing receptors. A similar analysis on ␣1 (H101R) ␣3 (H126R) and ␣2 (H101R) ␣3 (H126R) mice indicated the presence of 8% of ␣1␣3-containing receptors in the ␣1-containing receptor population, whereas the contribution of the ␣1␣3 combination to the ␣3-containing receptor population was 56% (Table II). In addition, the proportion of ␣2␣3-containing receptors was 19% in both the ␣2and ␣3-containing GABA A receptor population. DISCUSSION GABA A receptors are known to contain two ␣ subunits that can be homologous or heterologous. If they are distinct they represent a different class of receptor subtype. In the present paper we studied the presence and abundance of GABA A receptors containing two different types of ␣ subunits based on an analysis of mice containing the H/R point mutation in the ␣1, ␣2, or ␣3 subunit. This point mutation rendered the respective subunits insensitive to diazepam binding.
Basis and Prerequisites-There are two prerequisites for using mouse lines containing the ␣ (H/R) mutation for a quantitative analysis of GABA A receptors containing two different types of ␣ subunits. First, the introduction of the point mutation should not affect the expression levels and the distribution of the mutated subunit as well as those of the nonmutated subunits. It was shown previously that expression and targeting of receptor subunits appear not to be affected by the point mutation. Mice containing the H/R mutation in the ␣1, ␣2, or ␣3 subunit expressed the mutated subunit at normal levels with unaltered expression patterns and did not induce up-or down-regulation of other GABA A receptor subunits to an appreciable extent (16,17).
Second, in receptors containing one mutated and one nonmutated ␣ subunit, the nonmutated ␣ subunit should confer diazepam-sensitive binding to the receptor complex. The first a Percentage of increase refers to the increase of diazepam-insensitive binding sites introduced by the point mutation (i.e. percentage of total mutant mice Ϫ percentage of total of wild type mice).
b Sum of increased diazepam-insensitive binding in the corresponding single-mutant mice. c Fraction of diazepam-insensitive binding in double-mutant mice exceeding the sum of the corresponding single-mutant mice. This fraction likely represents the receptor subpopulation containing the respective heterologous ␣ subunits (e.g. ␣1␣2-containing receptors in the ␣1 (H101R) ␣2 (H101R) mice).
d The data were taken from Ref. 16. e The data were taken from Ref. 17. hint that mouse lines carrying the ␣ (H/R) mutation fulfill this second requirement came from the observation that subfractions in all of the mutated GABA A receptor populations displayed at least some high affinity diazepam binding (Fig. 1). Subsequent immunodepletion studies indicated that the presence of a heterologous ␣ subunit in the receptor complex that is not mutated (e.g. ␣1 in ␣1␣3 (H126R) or ␣2 in ␣2␣3 (H126R) ) conferred sensitivity to diazepam binding (Fig. 2). This feature permitted an assessment of GABA A receptors containing two different ␣ subunits using the various mouse lines carrying the a (H/R) mutation. However, the effect of the presence of a mutated ␣ subunit together with a nonmutated ␣ subunit in a single receptor complex on GABA A receptor function, i.e. pharmacology, is currently unknown. In particular, it is unclear whether the diazepam sensitivity observed in radioligand binding experiments to a mutated ␣ subunit and a nonmutated ␣ subunit reflects the ability of those receptors to functionally respond to diazepam.
Abundance of GABA A Receptors Containing Two Different ␣ Subunits-The data indicate that the ␣ (H/R) point mutation solely rendered GABA A receptors fully insensitive to diazepam binding that contain two homologous mutated ␣ subunits, whereas receptors containing one mutated ␣ subunit and one nonmutated ␣ subunit retained diazepam binding. This feature permitted an assessment of GABA A receptors containing two different types of ␣ subunits by comparing the extent of diazepam-insensitive binding in the double-mutant mouse lines with that present in the corresponding single-mutant mice. The different receptor subpopulations are calculated by subtracting the sum of diazepam-insensitive binding sites present in the respective single mutated receptors (e.g. ␣1 (H101R) plus ␣2 (H101R) ) from diazepam-insensitive binding sites present in the corresponding double-mutant mice (␣1 (H101R) ␣2 (H101R) ). This difference in diazepam-insensitive sites represents the amount of the respective receptors containing two different ␣ subunits (e.g. ␣1␣2-containing receptors).
Two types of information were obtained from the experimental data: 1) Determination of the proportion of diazepam-sensitive and diazepam-insensitive binding to crude membrane preparations permitted the estimation of the size of GABA A receptor populations in brain containing the different ␣ subunit combinations (Table I). This analysis is restricted to the diazepam-sensitive ␣1-, ␣2-, and ␣3-containing receptors expressed in brain, which, however, represent the vast majority of GABA A receptors (summarized in Table III). 2) Data on the presence of diazepam-sensitive and diazepam-insensitive bind-ing sites in immuno-isolated GABA A receptors allowed an estimate on the abundance of the different ␣ subunit combinations present within a particular GABA A receptor population (e.g. the ␣1␣3 receptor combination in ␣1-containing receptors; Table II). The data obtained from this analysis are summarized in Table IV. Because the experimental data exhibit partially considerable standard deviation (Table II), the data are considered to represent a rough estimate of the receptor populations.
Correspondingly, a similar distribution of the prevalence of receptors containing different types of ␣ subunits in whole mouse brain membranes The abundance of the various receptor subpopulations was calculated from the data presented in Table I. The increase in diazepam-insensitive binding in the single-mutant mice corresponds to the abundance of the receptor populations containing equal ␣ subunits because only receptors containing two mutated ␣ subunits are rendered insensitive to diazepam binding. The receptor populations containing two heterologous ␣ subunits were calculated by subtracting the sum of diazepaminsensitive binding sites present in the respective single-mutant mice from diazepam-insensitive binding sites present in the corresponding double-mutant mice. The difference in the sum of diazepam-insensitive sites present in the single-mutant mice and the diazepam-insensitive sites in the corresponding double-mutant mice represents the amount of the respective receptors containing two different ␣ subunits. For example, the abundance of ␣1␣2 receptors is calculated as follows: 86% (diazepam-insensitive binding in ␣1 (H101R) ␣2 (H101R) mice) minus 73% (the sum of diazepam-insensitive binding in ␣1 (H101R) mice (61%) and ␣2 (H101R) mice (12%)) equals 13%.

Subunit combination
Total diazepam-sensitive receptor population a The values are compared to the respective single mutants. % increase refers to the diazepam-insensitive binding sites present in the double-mutant receptor (e.g. ␣1 (H101R) ␣2 (H101R) , precipitated with ␣1 antiserum: 91%) minus the diazepam-insensitive binding sites of the corresponding single-mutant receptors (␣1 (H101R) : 84%). In this example, the difference in diazepam-insensitive binding sites (7%) corresponds to the abundance of ␣1␣2 receptors in the ␣1 receptor population. The K i values of diazepam-sensitive and diazepam-insensitive ͓ 3 H͔Ro 15-4513binding sites were in the range of 30 -100 nM and Ͼ60,000 nM , respectively. The data represent the means Ϯ S.D. of three to six independent experiments. the various ␣ subunit combinations was detected within the ␣1-, ␣2-, and ␣3-containing receptor populations (Table IV). Again, the ␣1␣1 combination (84%) predominated the ␣1␣2 (7%) and ␣1␣3 combinations (8%) in the ␣1-containing receptor population. Within the ␣2 receptor population the ␣2␣2 (46%) and the ␣1␣2 combination (36%) are expressed to a similar extent, whereas the ␣2␣3 combination (19%) is less abundant. Finally, within the ␣3-containing receptor population, the ␣2␣3 (19%) and ␣3␣3 combinations (27%) are clearly dominated by the ␣1␣3 combination (56%). These values are in good agreement with the data derived from the immunodepletion experiments (Table IV). In addition, the data fit nicely to values published so far on the abundance of the ␣1␣2 combination in the ␣2-containing receptor population of rat hippocampus and cerebral cortex (36% versus 36 and 39%, respectively) (6,8) and the ␣1␣3 combination in the ␣3-containing receptor population in the cerebral cortex (56% versus 55%) (6,8). However, for the ␣1␣3 combination in the ␣1-containing receptor population in cerebral cortex, an abundance of 24% was reported (6,8), whereas the estimate of the present report yielded a value of 8% for whole mouse brain extracts. It is currently unclear whether this mismatch is due to experimental reasons, the brain areas used (cerebral cortex versus whole brain), or the species analyzed (rat versus mouse).
Interestingly, only in the ␣1-containing receptor population the ␣1␣1 combination is the most prevalent one, whereas in the ␣2 and particularly in the ␣3 receptor population receptors with two different types of ␣ subunits predominate. Because GABA A receptors containing two different types of ␣ subunits appear to be expressed at considerable levels, it is conceivable that the inclusion of a second type of ␣ subunit may influence GABA A receptor function. However, the consequences of two different types of ␣ subunits within a single receptor complex on GABA A receptor function are largely unknown. In particular, it is not clear how the expression of heterologous ␣ subunits in a receptor complex affects functional benzodiazepine receptor pharmacology. So far, electrophysiological experiments on recombinant ␣1␣3␤2␥2, ␣1␣5␤2␥2, and ␣1␣6␤2␥2 indicate that receptors containing two different ␣ subunit variants display unique kinetics and pharmacological properties (23)(24)(25)(26). All of these studies inherently lack the ultimate certainty that indeed predominantly the subunit combination containing two heterologous ␣ subunits was measured and not a mixture of different subunit combinations containing only a single ␣ subunit variant. However, recent successful approaches linking the subunits covalently together to assess the subunit arrangement within the receptor complex and the functional consequences of subunit position in the receptor pentamer provide now the basis for analyzing GABA A receptor subtypes of predefined subunit arrangement (27)(28)(29). A first study on the functional consequences of the position of ␣1 and ␣6 subunits in the same receptor pentamer clearly demonstrated that only receptors with the ␣1 subunit next to the ␥2 subunit (␥2-␤2-␣6-␤2-␣1) displayed modulation by diazepam, whereas receptors with the ␣6 subunit next to the ␥2 subunit (␥2-␤2-␣1-␤2-␣6) were insensitive to diazepam (28). These data support the results of the present study indicating that only receptors containing two mutated ␣ subunits are insensitive to diazepam binding. In addition, the data suggest that the nonmutated, diazepamsensitive ␣ subunit preferentially assembles with the ␥2 subunit.
In conclusion, the results of this study indicate that the mouse lines containing the ␣ (H/R) mutation represent a valuable model for analyzing the abundance of receptor populations containing two different types of ␣ subunits. This approach allows the analysis to be performed on membrane fractions, tissue homogenates, or even tissue sections without solubilization and immunoprecipitation of the receptors. A detailed analysis of receptor subtypes containing two different types of ␣ subunits on a brain regional level by quantitative receptor autoradiography on brain sections is now feasible.