MolPharm xPharm- The Comprehensive Pharmacology Reference

Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mäkelä, R.
Right arrow Articles by Korpi, E. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mäkelä, R.
Right arrow Articles by Korpi, E. R.

0026-895X/97/030380-09$3.00/0
Copyright © by The American Society for Pharmacology and Experimental Therapeutics
All rights of reproduction in any form reserved.
MOLECULAR PHARMACOLOGY 52:380-388 (1997).

Cerebellar gamma -Aminobutyric Acid Type A Receptors: Pharmacological Subtypes Revealed by Mutant Mouse Lines

Riikka Mäkelä, Mikko Uusi-Oukari, Gregg E. Homanics, Joseph J. Quinlan, Leonard L. Firestone, William Wisden, and Esa R. Korpi

Department of Mental Health and Alcohol Research, National Public Health Institute, Helsinki, Finland (R.M., E.R.K.), Tampere Brain Research Center, University of Tampere Medical School, Tampere, Finland (R.M.), Department of Anesthesiology/Critical Care Medicine, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15261 (G.E.H., J.L.Q., L.L.F.), Medical Research Council Laboratory of Molecular Biology, Medical Research Council Centre, Cambridge CB2 2QH, UK (W.W.), and Department of Pharmacology and Clinical Pharmacology, University of Turku, Turku, Finland (M.U.-O., E.R.K.)

    Summary
Summary
Introduction
Procedures
Results
Discussion
References

The vast molecular heterogeneity of brain gamma -aminobutyric acid type A (GABAA) receptors forms the basis for receptor subtyping. Using autoradiographic techniques, we established the characteristics of cerebellar granule cell GABAA receptors by comparing wild-type mice with those with a targeted disruption of the alpha 6 subunit gene. Cerebellar granule cells of alpha 6-/- animals have severe deficits in high affinity [3H]muscimol and [3H]SR 95531 binding to GABA sites, in agonist-insensitive [3H]Ro 15-4513 binding to benzodiazepine sites, and in furosemide-induced increases in tert-[35S]butylbicyclophosphorothionate binding to picrotoxin-sensitive convulsant sites. These observations agree with the known specific properties of these sites on recombinant alpha 6beta 2/3gamma 2 receptors. In the presence of GABA concentrations that fail to activate alpha 1 subunit-containing receptors, methyl-6,7-dimethoxy-4-ethyl-beta -carboline (30 µM), allopregnanolone (100 nM), and Zn2+ (10 µM) are less efficacious in altering tert-[35S]butylbicyclophosphorothionate binding in the granule cell layer of the alpha 6-/- than alpha 6+/+ animals. These data concur with the deficiency of the cerebellar alpha 6 and delta  subunit-containing receptors in the alpha 6-/- animals and could also account for the decreased affinity of [3H]muscimol binding to alpha 6-/- cerebellar membranes. Predicted additional alterations in the cerebellar receptors of the mutant mice may explain a surplus of methyl-6,7-dimethoxy-4-ethyl-beta -carboline-insensitive receptors in the alpha 6-/- granule cell layer and an increased diazepam-sensitivity in the molecular layer. These changes may be adaptive consequences of altered GABAA receptor subunit expression patterns in response to the loss of two subunits (alpha  and delta ) from granule cells.

    Introduction
Summary
Introduction
Procedures
Results
Discussion
References

Receptors responsible for fast inhibitory neurotransmission in the brain, the GABAA receptors, are extremely heterogeneous. Among the 14 subunits (alpha 1-6, beta 1-3, gamma 1-3, delta , and epsilon ) known to participate in forming pentameric mammalian GABAA receptors (1-3), the alpha 6 subunit displays the most unique features in its cerebellar and cochlear nucleus granule cell-restricted expression (4-6) and benzodiazepine agonist-insensitive pharmacology (7, 8). Furthermore, the alpha 6 subunit imparts high GABA sensitivity (9) and selective furosemide sensitivity (10) to the GABAA receptors of synapses between the GABAergic inhibitory Golgi neurons and glutamatergic excitatory granule cells (11). These features, also detectable in cultured cerebellar granule cells (12, 13), have made the alpha 6 subunit-containing GABAA receptors the most distinctive among the alpha Xbeta 2/3gamma 2/delta GABAA receptors.

Two "alpha 6 knockout" mouse lines have been produced by disrupting the alpha 6 subunit gene through homologous recombination techniques (14, 15). Mutant mice are grossly normal in motor skills and their cerebellar cortical cytoarchitecture, indicating that the alpha 6 subunit is dispensable. However, the homozygous alpha 6-/- animals lack the cerebellar granule cell layer diazepam-insensitive Ro 15-4513 (ethyl-8-azido-5,6-dihydro-5-methyl-6-oxo-4H-imidazo[1,5-a][1,4]benzodiazepine- 3-carboxylate) binding (14, 15), as predicted from earlier work (7, 8, 16). In the experiments by Jones et al. (14), a specific association between alpha 6 and delta  subunits was revealed; the delta  subunit protein was largely absent from the cerebella of alpha 6-/- mice, as demonstrated by immunoprecipitation, immunocytochemistry, and immunoblot analysis with distinct delta  subunit-specific antibodies. A decrease in the affinity of cerebellar GABA sites was also observed (14, 15).

The conservation of expression pattern and sequence of the alpha 6 subunit gene in fish, birds, rodents, and humans (17, 18) suggests an important function for alpha 6 subunit-containing GABAA receptors in the brain. Thus, the lack of any obvious behavioral phenotype of the alpha 6-/- animals (14, 15) might indicate a compensatory rearrangement of GABAA receptor subunit composition in the mutant cerebellum. Here, we used pharmacological binding techniques to compare GABAA receptor fingerprints between wild-type alpha 6+/+ and mutant alpha 6-/- cerebella. Previously, it has been shown that allosteric modulation of the convulsant binding site labeled with [35S]TBPS is an excellent way to probe differences in receptor subunit combinations (9, 19). We examined the modulation of [35S]TBPS binding to receptors in cerebellar cortical layers as affected by the endogenous agonist GABA; the GABA antagonists furosemide, SR 95531 [2'-(3'-carboxy-2',3'-propyl)-3-amino-6-p-methoxyphenylpyrazinium bromide], and Zn2+ ions; the benzodiazepine site agonist diazepam and inverse agonist DMCM; and the neurosteroid agonist allopregnanolone (5alpha -pregnan-3alpha -ol-20-one). The results suggest altered receptor modulations in both the granule cell and molecular layers of the alpha 6-/- mice, consistent with subtle subunit reconfigurations in mutant cerebella.

    Experimental Procedures
Summary
Introduction
Procedures
Results
Discussion
References

Materials. [35S]TBPS, [3H]Ro 15-4513, and [3H]SR 95531 were purchased from Dupont-New England Nuclear (Dreieich, Germany), and [3H]methylamine muscimol was purchased from Amersham (Buckinghamshire, UK). Flumazenil (Ro 15-1788) was donated by F. Hoffmann-La Roche (Basel, Switzerland), diazepam was donated by Orion Pharmaceutica (Espoo, Finland), and zolpidem was donated by Synthelabo Recherché (Bagneux, France). GABA, picrotoxinin, and furosemide were purchased from Sigma Chemical (St. Louis, MO). Unlabeled SR 95531, 5alpha -pregnan-3alpha -ol-20-one (allopregnanolone), and DMCM were obtained from Research Biochemicals (Natick, MA), and ZnCl2 was from Merck (Darmstadt, Germany).

Animals. Two independent 129Sv × C57BL/6 mouse lines, in which the exon 8 of the mouse alpha 6 subunit gene was disrupted at the same site, were created by homologous recombination (14, 15). The brains of 33 homozygous wild-type (alpha 6+/+), 33 homozygous mutant (alpha 6-/-), and six heterozygous mutant (alpha 6+/-) adult (~4-month-old) mice of the F2 and F3 generations were used. Three brains of each genotype for the autoradiography were obtained from the Pittsburgh mice (15), whereas all the other samples originated from the Cambridge line (14).

Preparation of brain membranes and cryostat sections. All mice were killed by decapitation, and the whole brains or cerebella were rapidly dissected and frozen on dry ice. For ligand autoradiography, 14-µM horizontal, coronal, and sagittal serial sections were cut from 13 alpha 6+/+, 13 alpha 6-/-, and six alpha 6+/- brains using a Leitz 1720 cryostat, thaw-mounted onto gelatin-coated object glasses, and stored frozen under desiccant at -20°. To prepare cerebellar membranes, alpha 6+/+ and alpha 6-/- cerebella (from the Cambridge line) were weighed and thawed, and membranes were prepared from them for ligand binding assays as previously described in detail (10). Four membrane pools were prepared from both lines with five cerebella in each. All experiments were carried out in parallel fashion in respect to mouse lines, eliminating any day-to-day variation in receptor assays between the lines.

Ligand autoradiography. The autoradiographic procedures for regional localization of [3H]Ro 15-4513, [3H]muscimol, [3H]SR 95531, and [35S]TBPS binding were as previously described in detail (10, 19-21). Briefly, sections were preincubated in an ice-water bath for 15 min in 50 mM Tris·HCl, pH 7.4, supplemented with 120 mM NaCl in [35S]TBPS and [3H]Ro 15-4513 autoradiographic assays and in 0.31 M Tris-citrate, pH 7.1, in [3H]muscimol and [3H]SR 95531 assays. In some assays, the endogenous GABA, which could interfere with determination of alpha 6 subunit pharmacology (9), was removed by preincubating the sections three times in an ice-water bath for 10 min in 50 mM Tris·HCl supplemented with 1 mM EDTA, pH 7.4.

Final incubations in the preincubation buffer were performed with 6 nM [35S]TBPS at room temperature for 90 min, with 6 nM [3H]muscimol and 20 nM [3H]SR 95531 at 0-4° for 30 min, and with 5 nM [3H]Ro 15-4513 at 0-4° for 60 min. The effects of furosemide, SR 95531, ZnCl2, diazepam, DMCM, and allopregnanolone in the presence or absence of 0.5, 1, 3, or 5 µM GABA were tested on [35S]TBPS binding. Displacement of [3H]Ro 15-4513 binding was studied in the presence of 100 µM diazepam and 100 µM zolpidem. After the incubation, sections were washed three times for 15 sec or twice for 30 sec in an ice-cold incubation buffer in [35S]TBPS and [3H]Ro 15-4513 or in [3H]muscimol and [3H]SR 95531 assays, respectively. Sections were then dipped into distilled water, air-dried under a fan at room temperature, and exposed with plastic [3H] or [14C] standards to Hyperfilm-3H or Hyperfilm-beta max (Amersham), respectively, for 1-6 weeks. Nonspecific binding was determined with 10 µM flumazenil (Ro 15-1788), 10 µM picrotoxinin, and 100 µM GABA in [3H]Ro 15-4513, [35S]TBPS, and [3H]muscimol and [3H]SR 95531 assays, respectively. Substantial background binding was obtained only with [3H]SR 95531 (Fig. 1), due to its binding to monoamine oxidase A in a GABA-insensitive manner (22). Images from representative autoradiography films were produced by scanning the films using Arcus II scanner (Agfa Gevaert, Leverkusen, Germany) and Adobe Photoshop (version 3.0; Adobe Systems, Mountain View, CA) program.


View larger version (74K):
[in this window]
[in a new window]
  		Fig. 1.   Autoradiographic distribution of GABAA receptor binding sites in alpha 6-/- and wild-type alpha 6+/+ mice. Flumazenil-sensitive benzodiazepine sites were labeled by 5 nM [3H]Ro 15-4513, showing total binding, diazepam-insensitive binding, and zolpidem-insensitive binding. GABA sites were labeled by 6 nM [3H]muscimol, showing total binding, with nonspecific binding in the presence of 100 µM GABA at the film background level. GABA sites were also labeled by 20 nM [3H]SR 95531, showing total binding and nonspecific binding in the presence of 100 µM GABA. OB, olfactory bulb; Ctx, cerebral cortex; Cb, cerebellum; Gr, cerebellar granule cell layer; T, thalamus; Hi, hippocampus.

[3H]Muscimol binding assay. Binding of [3H]muscimol at 10 different concentrations (1-300 nM) was performed in triplicate in a total volume of 250 µl containing ~50-100 µg of cerebellar membrane protein. Incubations were performed for 60 min at 0-4° in 50 mM Tris-citrate buffer, pH 7.1. Nonspecific binding was determined with 100 µM GABA. Incubation was ended by centrifugation at 13,000 × g for 10 min at 0-4°. The pellets were rinsed once with 1 ml of ice-cold water, and the bottom of the tube containing the pellet was cut into a scintillation vial. The pellets were dissolved in 0.5 ml of LUMA Solve (Lumac LSC, Groningen, The Netherlands) overnight at room temperature, after which 4 ml of Optiphase HiSafe 2 scintillation fluid (Wallac, Turku, Finland) was added, and radioactivity was determined in a Wallac model 1410 liquid scintillation counter.

Data analysis. Autoradiography films were quantified using MCID M4 image analysis devices and programs (Imaging Research, St. Catharines, Ontario, Canada) as described in detail by Korpi et al. (19). Binding densities for each brain area were averaged from measurements from one to three sections/brain. The standards exposed simultaneously with brain sections were used as reference, with the resulting binding values given as radioactivity levels estimated for gray matter areas (nCi/mg for 3H and nCi/g for 14C).

Saturation isotherms of [3H]muscimol binding were analyzed for the estimation of Kd and Bmax by nonlinear regression with Prism 2.0 (GraphPAD Software, San Diego, CA).

Statistical significances of the differences between the alpha 6+/+, alpha 6+/-, and alpha 6-/- mice groups and between two population means were assessed with Prism by using one-way analysis of variance followed by Newman-Keuls post hoc test or by using Student's t test, respectively.

    Results
Summary
Introduction
Procedures
Results
Discussion
References

Benzodiazepine agonist-insensitive binding is absent in the mutant alpha 6-/- mice. Total [3H]Ro 15-4513 binding was widespread (14, 15), as expected due to the high affinity of [3H]Ro 15-4513 to all GABAA receptors with benzodiazepine sites (23), and totally displaceable by the benzodiazepine site antagonist flumazenil (Ro 15-1788; data not shown) throughout the alpha 6+/+, alpha 6+/-, and alpha 6-/- brains (Fig. 1). However, there was significantly less [3H]Ro 15-4513 binding to the cerebellar granule cell layer of alpha 6-/- and alpha 6+/- mice compared with binding to that of alpha 6+/+ mice (Table 1).

                              
View this table:
[in this window]
[in a new window]
  		TABLE 1
Benzodiazepine-insensitive [3H]Ro 15-4513 binding in cerebellar cortical layers of alpha 6+/+, alpha 6+/-, and alpha 6-/- mice

Serial brain sections were incubated with 5 nM [3H]Ro 15-4513 in the presence or absence of diazepam or zolpidem at 100 µM. Nonspecific binding was defined in the presence of 10 µM flumazenil. Autoradiographic films were processed and quantified against radioactivity standards. Data are mean ± standard deviation for six animals in each group, with half of each group from the Cambridge line and half from the Pittsburgh line.

Diazepam-insensitive [3H]Ro 15-4513 binding is considered to be the hallmark of alpha 6 subunit-containing GABAA receptors (4, 7, 8, 16). Diazepam (100 µM) only partially displaced [3H]Ro 15-4513 binding from the cerebellar granule cell layer of the alpha 6+/+ and alpha 6+/- mice, whereas the binding from granule cell layer of alpha 6-/- mice was totally displaceable (Table 1), as also demonstrated by Jones et al. (14) and Homanics et al. (15). There was very little diazepam-insensitive binding left in the molecular layer of each mouse genotype (Table 1). The diazepam-insensitive binding in the granule cells is also insensitive to the subtype-selective agonist zolpidem (21), which was confirmed in the alpha 6+/+ and alpha 6+/- mice.

GABA site labeling is reduced in cerebellar granule cell layer of the alpha 6-/- mice. GABA site agonist [3H]muscimol (24) and antagonist [3H]SR 95531 (25) were used as radioligands to determine the regional distribution by autoradiography of GABA binding sites in adult alpha 6+/+, alpha 6+/-, and alpha 6-/- mice. Both [3H]muscimol and [3H]SR 95531 binding was almost completely missing from cerebellar granule cells of alpha 6-/- mice (Fig. 1; Table 2). Identical images were obtained in alpha 6-/- mouse brains from both sources. The amount of [3H]muscimol binding to the cerebellar granule cell layer of alpha 6+/- mice was not significantly reduced compared with that of the alpha 6+/+ mice. The GABA site labeling was thus decreased in alpha 6-/- mice, similar to the benzodiazepine agonist-insensitive site labeling (Table 2). The amount of [3H]muscimol binding to the cerebellar molecular layer of alpha 6-/- mice was slightly (p < 0.05) lower than that to the molecular layer of alpha 6+/+ mice.

                              
View this table:
[in this window]
[in a new window]
  		TABLE 2
The binding of [3H]muscimol, [3H]SR 95531, and [35S]TBPS in cerebellar cortical layers of alpha 6+/+, alpha 6+/-, and alpha 6-/- mice as revealed by quantitative autoradiography

Absorbance values of the autoradiographic films in relation to radioactivity standards are mean ± standard deviation for six animals in [3H]muscimol (6 mM) and [35S]TBPS (6 nM) and for three animals in [3H]SR 95531 (20 nM) experiments in each group, with half of the [3H]muscimol and [35S]TBPS groups from the Cambridge line and half from the Pittsburgh line.

Because the autoradiographic signal of [3H]muscimol picks up only a subpopulation of all cerebellar GABAA receptors (see Ref. 20), possibly corresponding to alpha 6delta variants (14), we extended the analysis by running saturation isotherms of [3H]muscimol binding to cerebellar membranes using a centrifugation assay. The binding values fitted a one-component model, as is evident from the Scatchard transformations shown in Fig. 2. The results indicated a significantly lower affinity (Kd increased to 176% of the wild-type value, p < 0.01) and maximal density (Bmax reduced to 55% of the wild-type value, p < 0.05) in the alpha 6-/- mice (Fig. 2), respectively. This may explain the low granule cell layer labeling by 6 nM (and 20 nM; not shown) [3H]muscimol in alpha 6-/- mice under autoradiographic conditions and might correlate with the loss of alpha 6 and delta  subunit-containing receptors (26).


View larger version (28K):
[in this window]
[in a new window]
  		Fig. 2.   Saturation analysis of [3H]muscimol binding to cerebellar membranes of alpha 6-/- (black-square) and wild-type alpha 6+/+ (square ) mice. Membranes were incubated with various concentrations of [3H]muscimol (1-300 nM). Nonspecific binding was determined in the presence of 100 µM GABA. The specific binding values are mean ± standard error for independent experiments on four alpha 6-/- and four alpha 6+/+ membrane preparations. Binding data were analyzed as described in Experimental Procedures. Kd values were 80.5 ± 13.9 versus 45.7 ± 11.9 nM (mean ± standard deviation, four measurements), and Bmax values were 1.2 ± 0.3 versus 2.2 ± 0.6 pmol/mg of protein for the alpha 6-/- and alpha 6+/+ mice, respectively. Inset, Scatchard plots of the same data illustrating the decreased affinity and maximal binding site density in the alpha 6-/- membranes.

Selective lack of furosemide actions in the alpha 6-/- mice. Furosemide is a loop diuretic that also acts as a selective, noncompetitive antagonist for cerebellar granule cell-specific alpha 6 and beta 2 or beta 3 subunit-containing GABAA receptors (10). In contrast, SR 95531 has been shown to antagonize most GABAA receptor populations throughout the brain (19). As expected, both furosemide (200 µM) and SR 95531 (10 µM) increased the basal [35S]TBPS binding in the absence of exogenous GABA to the granule cell layer of alpha 6+/+ mice cerebella (253 ± 33% and 181 ± 24% of basal values, mean ± standard deviation, 10 animals, respectively) (Fig. 3), whereas in alpha 6-/- granule cells, furosemide did not affect binding (98 ± 10% of basal binding), and SR 95531 reduced binding (63 ± 10% of basal binding). In the cerebellar molecular layer of both alpha 6+/+ and alpha 6-/- mice, furosemide did not affect binding, and SR 95531 tended to decrease basal [35S]TBPS binding (Fig. 3).


View larger version (67K):
[in this window]
[in a new window]
  		Fig. 3.   GABA antagonistic actions of furosemide and SR 95531 on [35S]TBPS binding in alpha 6-/- and wild-type alpha 6+/+ mouse cerebellar sections. Representative autoradiographs of picrotoxinin-sensitive [35S]TBPS binding in serial alpha 6-/- and alpha 6+/+ sections show basal binding, binding in the presence of 200 µM furosemide, 10 µM SR 95531, 5 µM GABA, GABA plus furosemide, and GABA plus SR 95531. Note the selective lack of action of furosemide and the lack of granule cell layer (Gr) antagonism by SR 95531 in the alpha 6-/- mice. Mol, molecular layer.

Exogenous GABA (5 µM) affected [35S]TBPS binding similarly in the granule cell and molecular layers of both alpha 6+/+ and alpha 6-/- mice (24 ± 2% and 19 ± 7% compared with 24 ± 6% and 20 ± 7% of basal binding, respectively). Furosemide (200 µM) was able to reverse inhibition by GABA and elevate the binding over the basal level (145 ± 14% of basal) selectively in the granule cell layer of alpha 6+/+ mice cerebella, which is in agreement with studies on Wistar rats (10). Furosemide had no effect on alpha 6-/- cerebella. On the other hand, 10 µM SR 95531 was able to antagonize the GABA inhibition of [35S]TBPS binding in the granule cells of alpha 6+/+ mice (139 ± 23% of basal) and in those of alpha 6-/- animals (76 ± 14% of basal). SR 95531 reversed the GABA inhibition of [35S]TBPS binding of the molecular layer in a qualitatively similar manner in both alpha 6+/+ and alpha 6-/- mice (Fig. 3). These data are in agreement with complete disappearance of alpha 6 subunit-containing GABAA receptors.

Decreased GABA antagonism by zinc in the alpha 6-/- mice. Recombinant GABAA receptor isoforms have been shown to display differential sensitivity to antagonism by the divalent cation Zn2+ (27), with GABA currents of alpha 6beta 3delta receptors being the most sensitive to inhibition by Zn2+ (28). Zn2+ (10 µM) elevated basal [35S]TBPS binding in the granule cell layer of alpha 6+/+ mice but slightly decreased the binding in the alpha 6-/- mice (Fig. 4). This indicates that the ability of Zn2+ ions to antagonize endogenous GABA is diminished in the mutant mice, which is consistent with the lack of alpha 6delta -containing receptors. Zn2+ (10 µM) did not significantly affect [35S]TBPS binding to molecular layer (Fig. 4) or to forebrain areas in either mouse line (data not shown). Both 100 and 300 µM Zn2+ reduced [35S]TBPS binding to cerebellar granule cell and molecular layers and to forebrain regions of alpha 6+/+ and alpha 6-/- mice. Agonism/antagonism of various compounds at the GABAA receptor can be predicted by the convulsant binding assay only in the presence of a relevant GABA concentration (see Ref. 19); therefore, the differential inhibition of the [35S]TBPS binding by higher Zn2+ concentrations (Fig. 4) (see Ref. 29) may be functionally meaningless and just represent another allosteric action of Zn2+, in the same way as other GABAA antagonists do in the absence of GABA (19). However, this Zn2+ inhibition of the convulsant binding to the cerebellar granule cell layer was significantly greater in alpha 6-/- mice than in alpha 6+/+ mice, indicating the presence of different GABAA receptor populations in the alpha 6-/- and alpha 6+/+ granule cells.


View larger version (19K):
[in this window]
[in a new window]
  		Fig. 4.   Effect of Zn2+ on the picrotoxinin-sensitive [35S]TBPS binding in the cerebellar granule cell (black-square, square ) and molecular (bullet , open circle ) layers of alpha 6-/- (black-square, bullet ) and wild-type alpha 6+/+ (square , open circle ) mouse brain sections. The autoradiographic results are expressed as percentages (mean ± standard error of three measurements) of basal [35S]TBPS binding (100%). Values to the left of the gap, obtained in the absence of Zn2+. **, p < 0.01; ***, p < 0.001, statistical significance of the difference from the corresponding value of the alpha 6-/- mice (Student's t test).

Reduced allopregnanolone sensitivity in the alpha 6-/- mice. The expression of the delta  subunit in recombinant receptors together with alpha 1/6beta 3 or alpha 1/6beta 3gamma 2 subunit combinations has been shown to reduce neurosteroid-induced potentiation of GABA-activated currents (13). To determine the effect of alpha 6 subunit, we used a longer preincubation time for the sections (see Experimental Procedures) and an incubation solution with a lower GABA concentration (0.5 µM). Under these conditions, the neurosteroid agonist allopregnanolone (100 nM) decreased (p < 0.01) the binding to 75 ± 13% (mean ± standard deviation, three animals) of the values in the presence of GABA alone (an increase in GABA action) in the granule cell layer of alpha 6+/+ but was ineffective (97 ± 9%) in alpha 6-/- mice. Allopregnanolone at 10 µM reduced [35S]TBPS binding in all brain regions to the background level in both mouse lines and under both preincubation conditions (data not shown). These results are consistent with the presence of alpha 1beta Xgamma 2 and absence of highly GABA-sensitive alpha 6delta receptors in the alpha 6-/- granule cells.

Unexpected interaction of diazepam with cerebellar GABAA receptors of the alpha 6-/- mice. Diazepam (1 and 30 µM) produced a greater enhancement of the [35S]TBPS binding inhibition by 0.5 µM GABA in the molecular layer of the alpha 6-/- than alpha 6+/+ mice (Table 3; Fig. 5), suggesting that the Golgi/granule cell synapses are not the only altered loci in the mutant mice. Diazepam was less efficacious in the granule cell layer than in the molecular layer in both wild-type and mutant mice.

                              
View this table:
[in this window]
[in a new window]
  		TABLE 3
Effects of GABA, diazepam, and DMCM on [35S]TBPS binding in cerebellar granule cell and molecular layers of alpha 6-/- and alpha 6+/+ mice

The statistical significance of the differences from the basal binding in granule cell layer and molecular layer within the alpha 6+/+ and alpha 6-/- mice groups was determined using one-way analysis of variance followed by Newman-Keuls multiple-comparison test. Data are mean ± standard deviation for three animals in each group.


View larger version (78K):
[in this window]
[in a new window]
  		Fig. 5.   Actions of diazepam (DZ) and DMCM (DM) in the presence of a low GABA (G) concentration (0.5 µM) on picrotoxinin-sensitive [35S]TBPS binding in serial alpha 6-/- and wild-type alpha 6+/+ mouse cerebellar sections as revealed by autoradiography. Brain sections were washed extensively before incubation to remove endogenous GABA as described in Experimental Procedures. Concentrations of diazepam and DMCM are given in µM. Gr, granule cell layer; Mol, molecular layer.

Agonistic modulation by DMCM is reduced in the alpha 6-/- mice. The beta -carboline benzodiazepine-site ligand DMCM acts as an inverse agonist at low micromolar concentrations and as an "agonist" through the loreclezole binding site of the beta 2 or beta 3 subunits (30) at higher micromolar concentrations (31). This potentiation is independent of the alpha  subunit and is more pronounced on alpha 6 subunit-containing receptors due to the lack of DMCM inhibition (inverse agonism) via the benzodiazepine site (28, 31). With a thorough preincubation, 30 µM DMCM potentiated the effect of 0.5 µM GABA on [35S]TBPS binding significantly (p < 0.01) more in the granule cell layer of alpha 6+/+ than of alpha 6-/- mice (Fig. 5, Table 3), whereas no such difference was observed in the molecular layer (Table 3).

The benzodiazepine antagonist flumazenil blocked the inverse agonist action of DMCM in the molecular layer, indicating that this action was mediated by the benzodiazepine site (Fig. 6). Flumazenil failed to antagonize the agonistic effect of DMCM in the granule cell layer. The differential action of 30 µM DMCM between the alpha 6+/+ and alpha 6-/- mice was observed in the presence of a low (0.5 µM) but not of a high (3 µM) concentration of GABA. Thus, the DMCM actions indicate that the mutant granule cell layer has an increased density of sites with low sensitivity to GABA and/or to "agonistic" action of DMCM through the loreclezole site.


View larger version (50K):
[in this window]
[in a new window]
  		Fig. 6.   Effect of the benzodiazepine antagonist flumazenil (FLU) on the action of DMCM in the cerebellar granule cell (Gr) and molecular (Mol) layers of wild-type alpha 6+/+ (A) and alpha 6-/- (B) mice. Data are mean values (three measurements) with standard deviations (not shown) within 25% of the mean values and data demonstrate a flumazenil-insensitive "agonistic" action of 30 µM DMCM in the granule cell layer of alpha 6+/+ mice at low 0.5 µM GABA, action that is absent from the mutant mice. * p < 0.05; ** p < 0.01, statistical significance of the difference from the wild-type values (Student's t test).

    Discussion
Summary
Introduction
Procedures
Results
Discussion
References

By comparing alpha 6-/- mouse lines with wild-type mouse lines, we directly confirmed the specific pharmacological features of the cerebellar granule cell layer generated by alpha 6 subunit-containing GABAA receptors. First, the diazepam-insensitive [3H]Ro 15-4513 binding (7, 8), assumed to represent the benzodiazepine site of alpha 6beta 2/3gamma 2 receptors (4) due to diazepam-sensitive alpha 1, alpha 2, alpha 3, and alpha 5 subunits (each with histidine as the 100th residue) being replaced with the alpha 6 subunit containing Arg100 (32), was absent in the granule cell layer of alpha 6-/- mice. Second, the high GABA sensitivity of cerebellar granule cell receptors and recombinant alpha 6beta 2/3gamma 2 receptors (9, 28, 33) was confirmed; the alpha 6-/- mice exhibited the lack of increase in convulsant ([35S]TBPS) binding in the presence of the GABA site antagonist SR 95531. Third, selective furosemide antagonism of GABA-induced inhibition of [35S]TBPS binding in the cerebellar granule cell layer and recombinant alpha 6 subunit-containing receptors (10) was absent in the alpha 6-/- mice.

The delta  subunit protein is largely lost from the cerebellar granule neurons of the alpha 6-/- mice (14). The alpha 6 subunit-containing receptors are more sensitive to GABA and Zn2+ than alpha 1-containing receptors (9, 28, 33), and both of these properties are further accentuated by the delta  subunit: delta  subunit-containing receptors have an extremely high affinity to GABA agonists and a great sensitivity to functional antagonism by Zn2+ (28). These properties were lost in the granule cell layer of alpha 6-/- mice (diminished [3H]muscimol binding and lower GABA-antagonism by Zn2+). These results suggest that the loss of delta  subunit in the cerebellar granule cell layer of the alpha 6-/- mice, producing a brain-region specific double-subunit inactivation, is clearly detectable pharmacologically. The delta  subunit-containing receptors have reduced sensitivity to neurosteroid agonists (13), but we observed no enhancement by allopregnanolone of the action of 0.5 µM GABA in the mutant delta -deficient granule cell layer. This might be explained by the requirements by delta -deficient receptors of higher GABA concentrations to reveal the allosteric effects of the neurosteroid agonist.

In addition to these gross pharmacological alterations, a few subtle novel properties have emerged in cerebellar GABAA receptors of the alpha 6-/- mice. To avoid the interference of endogenous GABA with the alpha 6 subunit-containing GABAA receptors (9), the brain sections were washed extensively before [35S]TBPS incubations to reveal allosteric interactions. Using a very low exogenous GABA concentration (0.5 µM), we could dissect the allosteric actions on the alpha 1 and alpha 6 subunit-containing receptors. DMCM acts allosterically at low micromolar concentrations as an inverse agonist but as an "agonist" through another binding site dependent on the beta 2 and/or beta 3 subunits (30) at higher micromolar concentrations (31). The latter agonistic action has been shown to be more pronounced on alpha 6 subunit-containing receptors due to the lack of DMCM inhibition (inverse agonism) via the benzodiazepine site (28, 31). Consistent with this, the inverse agonism of DMCM was absent in the granule cell layer of alpha 6+/+ and alpha 6-/- mouse strains, but DMCM was more efficacious in the alpha 6+/+ mouse strain than in the alpha 6-/- mouse strain at a high "agonistic" concentration of 30 µM (Fig. 6). This finding was particularly notable because the actual amount of [35S]TBPS binding remaining unaffected with 30 µM DMCM in the alpha 6-/- cerebellar granule cell layer was greater even in absolute density units than that found in the alpha 6+/+ cerebella (Table 3; Fig. 4). This indicates an increase in the density of receptors with reduced DMCM sensitivity in the alpha 6-/- granule cells. In the presence of 3 µM GABA, 30 µM DMCM potentiated the GABA inhibition of the granule cell layer [35S]TBPS binding down to ~25% of the basal binding in both mouse genotypes (Fig. 6). Thus, in sections incubated with a low GABA concentration, the agonism by DMCM is alpha  subunit dependent. This is primarily due to the differing GABA sensitivity between alpha 6beta 2/3gamma 2 and alpha 1beta 2/3gamma 2 receptors. In addition, the inverse agonist action of DMCM is limited on alpha 6beta 2/3gamma 2 receptors due to the reduction by Arg100 of the affinity of DMCM (4, 31). The key conclusion is that the alpha 6-/- cerebellar granule cell layer has an increased number of receptors (e.g., alpha 1beta 2/3gamma 2 receptors) with such a low GABA sensitivity that 0.5 µM GABA is inefficient to promote the agonistic action of DMCM.

Regardless of the mechanism behind this alteration, it implies that in the absence of alpha 6 and delta  subunits, the remaining granule cell subunits assemble into slightly different combinations. This rearrangement could compensate for the loss of inhibition by alpha 6delta -containing receptors in the alpha 6-/- cerebellar granule cell layer. Because the remaining subunits should be alpha 1, beta 2, beta 3, and gamma 2 (Ref. 6), it is difficult to account for the altered pharmacology unless a substantial proportion of the granule cell GABAA alpha 1beta 2/3gamma 2 receptors are normally pharmacologically "masked" in the wild-type animals, perhaps by the presence of a dominant alpha 6 subunit in the same complex. However, this possibility is controversial (6, 9, 34, 35). An alternate explanation might be an increased amount of beta 1 subunit-containing receptors in the alpha 6-/- cerebella because this beta  subunit is insensitive to the agonistic action of DMCM due to a single amino acid residue change (Ser290 in beta 1 versus Asn290 in beta 2 and beta 3 subunits; Ref. 31). Although the beta 1 subunit mRNA is normally rare in cerebellar granule cells (36, 37), the possible increase in the proportion of beta 1 among the beta  subunit variant proteins should be explored.

Another unexpected finding was the higher efficacy of diazepam in the cerebellar molecular layer of alpha 6-/- mice in the presence of low GABA concentrations. The molecular layer normally has alpha 1beta 2/3gamma 2 receptors on the dendrites of Purkinje and stellate/basket cells (6, 38). At these receptors, diazepam is only a partial agonist (39). Adult Bergman glial cells normally express only alpha 2 and gamma 1 subunits (6), a combination that should be little affected by diazepam (40). Therefore, the novel pharmacology of the molecular layer of the alpha 6-/- animals predicts the presence of other subunit combinations with enhanced GABA and/or benzodiazepine sensitivity.

In conclusion, our results regarding the GABAA receptor alpha 6 subunit gene knockout mouse lines confirm the specific pharmacological features of the alpha 6 and delta  subunit-containing receptors in the cerebellar granule cell layer. In addition, the data revealed several unexpected alterations in the alpha 6-/- cerebella, which could be explained by subtle compensatory subunit reconfigurations in the cerebellar cortex.

    Acknowledgments

We thank P. Johansson (National Public Health Institute, Helsinki, Finland) for skillful technical assistance and A. Jones (Medical Research Council, Cambridge, UK) for expert help with mouse genotyping.

    Footnotes

Received March 26, 1997; Accepted May 29, 1997

   This work was supported in part by the Academy of Finland (E.R.K.), Finnish Foundation for Alcohol Studies (R.M.), National Institutes of Health (G.E.H., L.L.F.), and Medical Research Council (W.W.).

Send reprint requests to: Dr. E. R. Korpi, Department of Pharmacology and Clinical Pharmacology, University of Turku, FIN-20520 Turku, Finland. E-mail: esa.korpi{at}utu.fi

    Abbreviations

GABA, gamma -aminobutyric acid; DMCM, methyl-6,7-dimethoxy-4-ethyl-beta -carboline; TBPS, t-butylbicyclophosphorothionate.

    References
Summary
Introduction
Procedures
Results
Discussion
References

1. Sieghart, W. Structure and pharmacology of gamma -aminobutyric acidA receptor subtypes. Pharmacol. Rev.  47:181-234 (1995)[Medline].
2. McKernan, R. M. and P. J. Whiting. Which GABAA receptor subtypes really occur in the brain? Trends Neurosci.  19:139-143 (1996)[Medline].
3. Davies, P. A., M. C. Hanna, T. G. Hales, and E. F. Kirkness. Insensitivity to anaesthetic agents conferred by a class of GABAA receptor subunit. Nature (Lond.)  385:820-823 (1997)[Medline].
4. Lüddens, H., D. B. Pritchett, M. Kohler, I. Killisch, K. Keinänen, H. Monyer, R. Sprengel, and P. H. Seeburg. Cerebellar GABAA receptor selective for a behavioural alcohol antagonist. Nature (Lond.)  346:648-651 (1990)[Medline].
5. Varecka, L., C.-H. Wu, A. Rotter, and A. Frostholm. GABAA/benzodiazepine receptor alpha 6 subunit mRNA in granule cells of the cerebellar cortex and cochlear nuclei: expression in developing and mutant mice. J. Comp. Neurol.  339:341-352 (1994)[Medline].
6. Wisden, W., E. R. Korpi, and S. Bahn. The cerebellum: a model system for studying GABAA receptor diversity. Neuropharmacology  35:1139-1160 (1996)[Medline].
7. Sieghart, W., A. Eichinger, J. G. Richards, and H. Möhler. Photoaffinity labeling of benzodiazepine receptor proteins with the partial inverse agonist [3H]Ro 15-4513: a biochemical and autoradiographic study. J. Neurochem.  48:46-52 (1987)[Medline].
8. Malminiemi, O. and E. R. Korpi. Diazepam-insensitive [3H]Ro 15-4513 binding in intact cultured cerebellar granule cells. Eur. J. Pharmacol.  169:53-60 (1989)[Medline].
9. Korpi, E. R. and H. Lüddens. Regional gamma -aminobutyric acid sensitivity of t-butylbicyclophosphoro[35S]thionate binding depends on gamma -aminobutyric acidA receptor alpha  subunit. Mol. Pharmacol.  44:87-92 (1993)[Abstract].
10. Korpi, E. R., T. Kuner, P. H. Seeburg, and H. Lüddens. Selective antagonist for the cerebellar granule cell-specific gamma -aminobutyric acid type A receptor. Mol. Pharmacol.  47:283-289 (1995)[Abstract].
11. Nusser, Z., W. Sieghart, F. A. Stephenson, and P. Somogyi. The alpha 6 subunit of the GABAA receptor is concentrated in both inhibitory and excitatory synapses on cerebellar granule cells. J. Neurosci.  16:103-114 (1996)[Abstract/Free Full Text].
12. Zhu, W. J., J. F. Wang, S. Vicini, and D. R. Grayson. alpha 6 and gamma 2 subunit antisense oligodeoxynucleotides alter gamma -aminobutyric acid receptor pharmacology in cerebellar neurons. Mol. Pharmacol.  50:23-33 (1996)[Abstract].
13. Zhu, W. J., J. F. Wang, K. E. Krueger, and S. Vicini. delta Subunit inhibits neurosteroid modulation of GABAA receptors. J. Neurosci.  16:6648-6656 (1996)[Abstract/Free Full Text].
14. Jones, A., E. R. Korpi, R. M. McKernan, R. Pelz, Z. Nusser, R. Mäkelä, J. R. Mellor, S. Pollard, S. Bahn, F. A. Stephenson, A. D. Randall, W. Sieghart, P. Somogyi, A. J. H. Smith, and W. Wisden. Ligand-gated ion channel subunit partnerships: GABAA receptor alpha 6 subunit gene inactivation inhibits delta  subunit expression. J. Neurosci.  17:1350-1362 (1997)[Abstract/Free Full Text].
15. Homanics, G. E., C. Ferguson, J. J. Quinlan, J. Daggett, K. Snyder, C. Lagenaur, Z.-P. Mi, X.-H. Wang, D. R. Grayson, and L. L. Firestone. Gene knockout of the alpha 6 subunit of the gamma -aminobutyric acid type A receptor: lack of effect on responses to ethanol, pentobarbital, and general anesthetics. Mol. Pharmacol.  51:588-596 (1997)[Abstract/Free Full Text].
16. Turner, D. M., D. W. Sapp, and R. W. Olsen. The benzodiazepine/alcohol antagonist Ro-15-4513: binding to a GABAA receptor subtype that is insensitive to diazepam. J. Pharmacol. Exp. Ther.  257:1236-1242 (1991)[Abstract/Free Full Text].
17. Bahn, S., R. J. Harvey, M. G. Darlison, and W. Wisden. Conservation of gamma -aminobutyric acid type A receptor alpha 6 subunit gene expression in cerebellar granule cells. J. Neurochem.  66:1810-1818 (1996)[Medline].
18. Hadingham, K. L., E. M. Garrett, K. A. Wafford, C. Bain, R. P. Heavens, D. J. S. Sirinathsinghji, and PJ. Whiting. Cloning of cDNAs encoding the human gamma -aminobutyric acid type A receptor alpha 6 subunit and characterization of the pharmacology of alpha 6-containing receptors. Mol. Pharmacol.  49:253-259 (1996)[Abstract].
19. Korpi, E. R., P. H. Seeburg, and H. Lüddens. Modulation of GABAA receptor tert-[35S]butylbicyclophosphorothionate binding by antagonists: relationship to patterns of subunit expression. J. Neurochem.  66:2179-2187 (1996)[Medline].
20. Olsen, R. W., R. T. McCabe, and J. K. Wamsley. GABAA receptor subtypes: autoradiographic comparison of GABA, benzodiazepine, and convulsant binding sites in the rat central nervous system. J. Chem. Neuroanat.  3:59-76 (1990)[Medline].
21. Wong, G., M. Sarviharju, M. Toropainen, D. Matecka, and E. R. Korpi. Pharmacologic actions of subtype-selective and novel GABAergic ligands in rat lines with differential sensitivity to ethanol. Pharmacol. Biochem. Behav.  53:723-730 (1996)[Medline].
22. Luque, J. M., R. Erat, R. Kettler, A. Cesura, M. Da Prada, and J. G. Richards. Radioautographic evidence that the GABAA receptor antagonist SR 95531 is a substrate inhibitor of MAO-A in the rat and human locus coeruleus. Eur. J. Neurosci.  6:1038-1049 (1994)[Medline].
23. Lüddens, H. and W. Wisden. Function and pharmacology of multiple GABAA receptor subunits. Trends Pharmacol. Sci.  12:49-51 (1991)[Medline].
24. Olsen, R. W. Drug interactions at the GABA receptor-ionophore complex. Annu. Rev. Pharmacol. Toxicol.  22:245-277 (1982)[Medline].
25. Heaulme, M., J.-P. Chambon, R. Leyris, J.-C. Molimard, C. G. Wermuth, and K. Biziere. Biochemical characterization of the interaction of three pyridazinyl-GABA derivatives with the GABAA receptor site. Brain Res.  384:224-231 (1986)[Medline].
26. Quirk, K., P. J. Whiting, C. I. Ragan, and R. M. McKernan. Characterization of delta -subunit containing GABAA receptors from rat brain. Eur. J. Pharmacol.  290:175-181 (1995)[Medline].
27. Draguhn, A., T. A. Verdorn, M. Ewert, P. H. Seeburg, and B. Sakmann. Functional and molecular distinction between recombinant rat GABAA receptor subtypes by Zn2+. Neuron  5:78