LMMUNOCYTOCHEMICAL LOCALIZATION OF Y2 SHORT AND Y2 LONG SUBUNITS

Download short form, y2 long form, 'y2 subunit, immunocytochemistry, antibody, rat brain] ... tions of the existence of the yzs and ysL forms ar...
0 downloads 446 Views 4MB Size
Download PDF
Love PNG Images
The Journal

of Neuroscience,

November

1994,

14(11):

7166-7179

lmmunocytochemical Localization of y2 Short and y2 Long Subunits of the GABA, Receptor in the Rat Brain Antonia

Gutitkrrez,

Zafar

U. Khan,

and

Angel

L. De Blas

Division of Molecular Biology and Biochemistry, School of Biological Sciences, University of Missouri-Kansas City, Kansas City, Missouri 641lo-2499

The distribution of the short (r,J and long (y2J subunits of the GABA, receptors in the rat brain has been revealed by light microscopy immunocytochemistry with novel subunitspecific antibodies (anti-y,, and anti-y,,). We have also used other subunit-specific antibodies including anti-y,lL2 (which recognizes both rzS and Yap), anti-a,COOH, and the monoclonal antibody 62-3Gl to & for comparing the regional and cellular distribution of the most abundant GABA, receptor subunits in the rat brain. The distributions of -yzS and yzL immunoreactivities are similar throughout the brain although the relative intensity of both signals varies depending on the brain area and neuronal type. In the hippocampus, cerebral cortex, and olfactory bulb (particularly mitral, periglomerular, and tufted neurons), yzS was more abundant than 72L. In contrast, the inferior colliculus, medulla, and the cerebellar Purkinje cells displayed more yzL than yzS immunolabeling. An important difference in the distribution of the various subunits was found in cerebellum: ySS and ‘yzL were predominantly localized in the molecular layer, whereas (Y, and @,,, were more abundant in the granular layer. In the thalamus, -rSL and rzS were less abundant than either 01, or & subunits. The results showed that there is colocalization of rzS and yzL subunits in some brain areas and neuronal types, as well as areas of mismatch. Colocalization and mismatches were also found among a,, &, and y2, probably resulting from the heterogeneity in the subunit composition of the GABA, receptors through the brain. [Key words: GABA, receptor, benzodiazepine receptor, y2 short form, y2 long form, ‘y2 subunit, immunocytochemistry, antibody, rat brain]

The y-aminobutyric acid,, receptors(GABA,R) are membrane proteins constituted by pseudosymmetric pentameric combinations of subunits. To date, five subunit classesand several isoforms within each class(a,-(~,, PI-&, y,yX, 6, p,, and p2) have beenidentified in mammalianbrain (for reviews, seeOlsen and Tobin, 1990; Burt and Kamatchi, 1991; Liiddens and Wisden, 1991; DeLorey and Olsen, 1992). These subunits are encoded by different genes.Additional subunit diversity is generatedby alternative mRNA splicing of the yz subunit (Whiting

Received Feb. 16, 1994; revised May 23, 1994; accepted May 26, 1994. This research was supported by Grant NSI 7708 from the National Institute of Neurological Disorders and Stroke, and by a postdoctoral fellowship from the Ministeno de Educaci6n y Ciencia, Spain to A.G. Correspondence should be addressed to Angel L. De Blas at the above address. Copyright 0 1994 Society for Neuroscience 0270-6474/94/147168-12$05.00/O

et al., 1990; Kofuji et al., 1991). Both short (yzs)and long (yzL) forms are identical except for the presenceof an eight-aminoacid sequence,encodedby a 24 nucleotide exon, that is present in the large intracellular loop of yzL. The yz subunit is necessary for the binding of benzodiazepines (BZDs) to the GABA,R (Pritchett et al., 1989; Sigel et al., 1990; Moss et al., 1991; Wafford et al., 1991). Nevertheless,all the functional implications of the existenceof the yzsand ysLforms are not yet known. The -rzr octapeptide hasa consensussubstratesequencefor protein kinase C (Kellenberger et al., 1992; Moss et al., 1992)and also can be phosphorylated by a Ca*+/calmodulin-dependent protein kinase II (Machu et al., 1993). In addition, yzL but not yzsconfers ethanol sensitivity to the GABA,R (Wafford et al., 1991; Wafford and Whiting, 1992). We have recently obtained and characterized specificanti-y,, and anti-y,, antibodies (Khan et al., 1994a), and studied the relative distribution in several areasof the rat brain of the yzs and yzL proteins by immunoprecipitation of the GABA,R with anti-y,, and anti-y,, antibodies (Miralles et al., 1994). In addition, we have also studied the distribution of yzs and ylL mRNAs in rat brain by in situ hybridization (Miralles et al., 1994). These studieshave revealed areasof differential distribution of yzs and yzL as well as areasof colocalization. In addition, we have recently shown by immunoprecipitation and immunoblotting experiments that yls and yzL can coexist in somereceptor complexesbut not in others(Khan et al., 1994a,b). In this communication we are addressingthe localization of yzsand yzLsubunits by light microscopy immunocytochemistry with specific anti-y*, and anti-yzL affinity-purified antibodies. Materials and Methods The polyclonalantibodiesthat specificallyrecognizethe short(y2Jor the long (y2J formsof the y2subunitwereobtainedafter synthetic peptide coupled immunizingrabbitswith the corresponding

Antibodies.

to keyhole limpet hemocyanin (KLH) through a cysteine that was added to the carboxy end ofeach peptide. The peptide KKKKNPAPTIDI (yZs) corresponded to amino acids 332-343 (Shivers et al., 1989), while

NPLLRMFSFKAP (y2J corresponded to amino acids 336-347 (Whiting et al., 1990; Kofuji et al., 1991). Thorough characterization of these antibodies has been reported elsewhere (Khan et al., 1994a). Each peptide was immobilized on thiopropyl Sepharose 6B and the corresponding antibody was affinity purified. Each peptide (5 mg) was coupled to 1 gm of activated thiopropyl-Sepharose 6B (Pharmacia LKB) in 0.1 M Tris-HCl. DH 7.4. 0.5 M NaCl. and 1 mM EDTA. Remainina active groups were blocked after incubation with 1 mM fl-mercaptoethanol for 1 hr at room temperature (RT). The gel was equilibrated with phosphatebuffered saline (PBS), and 1 ml of antiserum (diluted fivefold in PBS) was circulated repeatedly through the column overnight at 4°C. After washing with 200 bed volumes of PBS, the antibody was eluted with 50 mM glycine-HCl, pH 2.3, and collected in 1 ml fractions. The pH

The Journal of Neuroscience,

neutralization was done immediately by adding 20 ~1 of 1 M Tris. Antibody-containing fractions were identified by both enzyme-linked immunosorbent assay (ELISA) with immobilized peptide and immunoprecipitation of the GABA,R (Khan et al., 1993). Antibody-containing fractions were pooled and dialyzed in PBS. The anti-y,ILZ antibody to staphylococcal protein A (SPA) fusion protein of the yZs large intracellular loop (IL) that is located between the putative transmembrane segments M3 and M4, and the SPA-nIL fusion protein immunogen were prepared as described elsewhere (Khan et al., 1994a). Anti-r,IL2 recognizes both yZs and yZL. Affinity purification ofanti-y,ILZ was done on immobilizedglutathione-S-transferase (GST)-nIL fusion protein. The preparation and purification of GSTy,IL fusion protein have also been described elsewhere (Khan et al., 1993). For the preparation of the affinity column the GST+L fusion protein (5.4 mg) was incubated with 2 ml of Affigel 10 (Bio-Rad) for 2 hr at RT. Remaining active ester groups were blocked after incubation with 1 M ethanolamine-HCl, pH 6, for 1 hr. The column was washed with 10 bed volumes of 0.1 M sodium acetate, pH 4, followed by 0.1 M NaHCO,, pH 8.5, and PBS. The anti-r,IL2 antibody was affinity purified as described above for anti-y,, and-anti-y,,. The antibody to the carboxy-terminal peptide NREPQLKAPTPHQ of rat (Y, subunit (anti-a,COOH) corresponding to amino acids 4 16428 (Khrestchatisky et al., 1989) was generated by coupling the peptide to KLH using 0.2% alutaraldehvde and immunizina a rabbit with the conjugate. The anti-a:COOH antibody has been characterized elsewhere (Khan et al., 1994b). The antibody to qCOOH was affinity purified on immobilized peptide. For the preparation of the affinity column, 1 gm of activated CH Sepharose 4B (Phannacia LKB) was washed with 200 ml of 1 mM HCI at 4°C followed by 50 ml of buffer A (0.1 M NaHCO,, 0.5 M NaCl, pH 8). Peptide (5 mg in 1 ml of buffer A) was added to gel and incubated for 2 hr at RT. The remaining active groups were blocked by incubating with 0.1 M Tris-HCl, pH 8, 0.5 M NaCl, for 1 hr at RT. The gel was washed with four cycles of alternating pH. Each cycle consisted of a 5 ml washing with 0.1 M sodium acetate, pH 4, and 0.5 M NaCl, followed by a 5 ml washing with 0.1 M Tris-HCl, pH 8, and 0.5 M NaCl. The ael was eauilibrated with 100 ml PBS and affinitv purification of the intibody was done as explained above. The mouse monoclonal antibody (mAb) 62-3Gl recognizes both & and 8, &,,) subunits of the GABA,R. This mAb has been thoroughly characterized elsewhere (De Blas et al., 1988; Vitorica et al., 1988; Park and De Blas, 1991; Park et al., 1991; Ewert et al., 1992). The sheep antibody to rat brain glutamic acid decarboxylase (GAD) (Oertel et al., 198 1a.b) was generously nrovided bv Dr. Irwin J. Kopin from the National Institut& of Heaith, Bethesda. Zmmunoblots. They were done according to De Blas and Cherwinski (1983). Immunocytochemistry. Adult Sprague-Dawley and Fischer-344 rats (Sasco Inc., Omaha, NE) were anesthetized with ketamine-HCl(40 mg/ kg), xylazine (3 mg/kg), and acepromacine maleate (1 mg/kg), and perfused through the ascending aorta with 4% paraformaldehyde, 0.075 M lysine, 0.0 1M sodium periodate in 0.1 M phosphate buffer (PB), pH 7.4 (PLP: McLean and Nakane. 1974). Brains were immersed in this fixative for 3 hr at 4”c, followed by 30% sucrose at 4°C until they sank. Frozen brains were cut in parasagittal sections (25 rm) with a freezing microtome and the sections were stored in PB, pH 7.4, containing 0.02% sodium azide at 4°C. Free-floating sections were processed for immunocytochemistry using the indirect peroxidase-anti-peroxidase (PAP) method described elsewhere (De Blas, 1984; De Blas et al., 1988). Before incubation with the primary antibody, the sections were treated with 3% hydrogen peroxide, 3% methanol in PB for 15 min to eliminate endogenous peroxidases. Sections were incubated with ( 1) 12 mg/ml protein of one of the primary affinity-purified antibodies (anti-y,,, anti-y,,, anti-cu,COOH, and antiy21L2), or the culture medium of the mAb 62-361 diluted l:lOO, or the anti-GAD antiserum diluted 1:20,000, incubations were for 2 d at 4°C with agitation; (2) the appropriate antibody bridge: sheep anti-rabbit IgG (Sigma), rabbit anti-mouse IgG (Sigma), or rabbit anti-sheep IgG (Cauuel) diluted 1: 100 for 1 hr at RT: or (3) the corresponding PAP complex (Sigma): rabbit serum, mouse’ monoclonal, or goat serum diluted 1: 100 for 1 hr at RT. After each incubation sections were rinsed three times with PB, pH 7.4, for 15 min each. All antibodies were diluted in PB containing 0.3% Triton X-100. Reaction product was visualized by incubation with 0.05% 3,3’-diaminobenzidine tetrahydrochloride (DAB from Sigma), 0.005% hydrogen peroxide, 0.03% cobalt chloride, and 0.03% nickel ammonium sulfate in PB for 10-l 5 min at RT. Sec-

12

3.4

November

56

1994, 14(11)

7169

78

1169

80-

50-

5147 45

339

Figure 1. Immunoblots of membranes from rat cerebral cortex with different antibodies. Each strip contains the transfer of 10 pg of membrane protein. All the strips were incubated with 4 &ml of affinitypurified antibodies. Strips I and 2 were incubated with anti-qCOOH, 3 and 4, with anti-y,ILZ; 5 and 6, with anti-y,,; and 7 and 8, with antiyzL. Strips 2, 6, and 8 were incubated with antibodies in the presence of 10 &ml of o(,COOH, yZs, and yZL peptide, respectively. Immunoreaction of strip 4 was similarly blocked with 40 &ml of GST-r,IL fusion protein. The molecular weight (kDa) of the GABA,R subunits was determined by comparison to the relative mobility of markers of known molecular weight. tions were washed twice in PB and mounted on gelatin-coated glass slides. Control sections that were not incubated with the primary antibody showed no immunolabeling. The immunoreaction with each of the different GABA,R subunit antibodies anti-yZs, anti-yz,, anti-cY,COOH, and anti-y,IL2 was completely blocked in all brain areas by incubating each antibody with 60 pg/ml of the corresponding synthetic peptide (yZs, yZL, and cqCOOH) or 180 &ml of GST--y,IL fusion protein.

Results Two novel affinity-purified antibodies have been used for the immunocytochemical localization of yZs and yZL subunits of the GABA,R. Two other affinity-purified antibodies (anti-y,ILZ, which recognizes both yZs and yZL, and anti-cY,COOH, which recognizes a,) as well as the mAb 62-3Gl to & were also used. Immunoblots of membranes from rat cerebral cortex show that the anti-y,, and anti-y,, recognize 45,000 M, and 47,000

7170

Gutihrez

et al. * Distribution

of yzs and

yZL Subunits

in Brain

Figure 2. Rat brain immunocytochemistry with affinity-purified antibodies to yzL (A), yzs (B), a,COOH (C), and the mAb 62-361 to &,, (0) (parasagittal sections). Ctx, cerebral cortex; Cb, cerebellum; CPU, caudate putamen; GP, globus pallidus; H, hippocampus; IC, inferior colliculus; M, medulla; OB, olfactory bulb; SC, superior colliculus; SN, substantia nigra; T/z, thalamus; Rt, reticular thalamic nucleus. Scale bar, 2.3 mm.

M, peptide subunits, respectively (Fig. 1, lanes 5 and 7). The anti-yJL2 recognizesboth 45,000 A4, and 47,000 M, peptides. The immunoreactivity of this antibody (to the IL of yzS) is considerably decreasedwith the addition of the eight-aminoacid peptide that characterizes yzL (Fig. 1, lane 3). The anticqCOOH recognizesa 5 1,000M, peptide (Fig. 1, lane 1). AntiyzLand anti-qCOOH alsoshowedminor reactivity with 79,000 M, and 110,000M, peptides, respectively. The M, of the main peptidescorrespondsto the expected M, of the yzS, yzL, and CY, subunits (Benke et al., 199la,b, Duggan et al., 1992; Khan et al., 1993, 1994a,b).The specificity of the antibodies is remarkable consideringthe heterogeneousmixture of proteins present in the crude membranepreparation and the relatively low content of GABA,R with respectto other proteins. The specificity of the antibodies was also demonstrated by blockage with the corresponding synthetic peptide or fusion protein. The antibodies and their corresponding specificities have been thoroughly characterized elsewhere(Khan et al., 1994a,b). These antibodies efficiently immunoprecipitated the solubilized GABA,/BZD receptors(GABA,R/BZDR). The distribution of yzS, yzL, a,COOH, and Pzn immunolabeling in the rat brain is shown in Figure 2. For all of them, immunostainingwasobservedin cerebralcortex, olfactory bulb, corpusstriatum, thalamus,hippocampus,inferior colliculus, superficial layer of superior colliculus, substantianigra, deep cerebellarnuclei, and cerebellum.Nevertheless,the relative intensity of the staining in theseareaswasdifferent for eachantibody. The yzs and yzL immunolabeling was highest in the molecular

layer of the cerebellum (Fig. 3). The (u,COOH immunolabeling washighest in the olfactory bulb, whereas&, immunolabeling was very intense in all aforementioned areas.The anti-y,,, antiyzL, anti-yJL2, and anti-cY,COOH antibodies showedboth diffuse neuropil labeling and stronger immunoreaction around the surfaceof neuronal cell bodiesand proximal dendritic processes. Immunolabeling was alsofound inside the neuronal cell bodies as shown below. Cerebellum The molecular layer showed the strongestyzL, yzS, and y,IL2 immunolabeling (Fig. 3A-C), whereas (Y,COOH and &, immunoreaction products weremainly located in the granular layer (Fig. 3D,E). Purkinje cell bodieswere labeledstrongly with antiyzL and anti-yJL2 antibodies, less with anti-y,, and antitu,COOH, and very weakly or not at all with anti-&,,. Nevertheless,although small amounts of p2,, subunits are presentin the soma,these subunits are abundant in the dendrites of Purkinje cells(De Blaset al., 1988; Somogyi et al., 1989).The GAD immunoreactivity was presentin molecular, granular, and Purkinje cell layers (Fig. 3F), revealing the presenceof GABAergic synaptic boutons on all the layers. The immunolabeling with anti-y,,, anti-y,,, antiqJL2, and anti-qCOOH affinity-purified antibodies could be completely blocked in all areasof the brain with the correspondingpeptides (60 &ml) or GST-y,IL fusion protein (180 &ml), asillustrated for cerebellum in Figure 4&D.

The Journal of Neuroscience,

November

1994, f4(11)

7171

Figure 3. Cerebellum immunocytochemistry with anti-y,, (A), anti-y,, (B), anti-yJL2 (C), anti-Lu,COOH (o), andanti-& (E) GABA,R subunit

antibodiesand GAD antiserum(0. Immunolabeling with anti-y,,, anti-y,,, andantiqJL2 is moreabundantin the molecularlayer,whereasfor anti-a,COOHandanti-&,, it is strongerin the granularlayer. Notice the predominance of y2,.over yzSin the Purkinjecells.G, granularlayer; P, Purkinjecelllayer; M, molecularlayer. Scalebar, 50 pm.

layer, and hilus were immunolabeledwith anti-y,,, anti-y?,, and Hippocampus anti-a,COOH (Fig. 5.4-C). Nevertheless, the relative intensity In the hippocampus(dentate gyrus and hippocampusproper), of the reaction was considerably lower than with the mAb 623Gl to & (De Blas et al., 1988). In CAl, CA2, and CA3 fields the general immunolabeling with anti-y,, was much stronger than with anti-y,, (Figs. 5A,B, 6A,B). In the dentate gyrus, -ySs of the hippocampus proper, the cell bodies of the stratum pyramidale presented the strongest yzL and yzs immunolabeling and yzLimmunoreaction products were located mainly on both (Fig. 6A,B). These were to a large extent located on the surface cytoplasm and external membrane of the granule cells. Antiof the pyramidal cells (Fig. 7,4-C) in a disposition similar to yzs, anti-T2L,and anti-cY,COOH also revealed intracellular imthe one observed with anti-GAD (Figs. 60, 70). Cytoplasmic munolabeling in the basket cells located between the granule labeling, particularly with anti-y,,, was also observed in the cell layer and the plexiform layer, aswell asin hilar interneurons (Fig. 5,4-C). A similar labeling pattern wasobserved with antisomasof the pyramidal cells (Fig. 64. The stratum oriens and & (De Blas et al., 1988) and anti-GAD (Fig. 5D, De Blas et the strata radiatum and lacunosum,wherebasaland apical denal., 1988), which indicates that these GABAergic interneurons drites of pyramidal cells are located, respectively, also showed high labeling with anti-y,,, anti-y,,, and anti-cu,COOH (Fig. 6Aalsohave GABA,R on their surface.Plexiform layer, molecular

Figure 4. GABA,R subunit immunolabeling was blocked with the corresponding peptide or fusion protein. A, Anti-y,, + 60 &ml + peptide. B, Anti-y,, + 60 &ml yzL peptide. C, Anti-cu,COOH + 60 &ml cqCOOH peptide. D, Anti-yJL2 + 180 &ml GST-%IL fusion protem. Cerebellum parasagittal sections. G, granular layer; P, Purkinje cell layer; M, molecular layer. Scale bar, 50 pm.

Figure 5. Dentate gyrus immunolabeling with anti-y,, (A), anti-y,, (B), anti-cY,COOH (C), and anti-GAD (0). All antibodies show immunolabeling around the cell bodies of the granule cells. They also label GABAergic basket cells located at the base of the granule cell layer as well as other GABAergic intemeurons located in the hilus. Notice the higher content of y 2s over y2L in the dentate gyrus. G, granule cell layer; HL, hilus; M, molecular layer. Scale bar, 100 pm.

The Journal of Neuroscience,

November

1994, 14(11)

6. Immunocytochemistry with anti-yll (A), anti-y,, (B), anti-q COOH (C), and anti-GAD (0) of the CA2 region of the stratum pyramidale of the hippocampus. Notice the predominance of yZSover yZLin this structure. SP, stratum pyramidale; SO, stratum oriens; SR, stratum radiatum. Scalebar, 50 pm.

Figure

C), but the relative intensity was lower than with anti-&,, mAb 62-361 (De Blas et al., 1988).

Cerebral cortex yzL and yZS immunoreaction products were distributed through all layers of the cerebral cortex. They were concentrated on the surface of the cell bodies (layers II-VI) as well as in diffuse form through the neuropil (layers I-VI) (Fig. 8A,B). Cytoplasmic immunoreaction was also observed, paiticularly on the pyramidal neurons of layer V. The latter was stronger with anti-y,, than with anti-yzL. Immunolabeling with anti-a,COOH (Fig. 8C) and mAb 62-361 (De Blas et al., 1988) revealed a similar pattern. Figure 9 shows some anti-yzL (A and B) and anti-y,, (C and D) immunoreactive neurons (layers IV-V) at higher magnification. The immunoreaction product has a granular appearance and it is mainly distributed on the somas and proximal dendritic processes, probably on the membrane surface as well as intracellularly.

Olfactory bulb The immunoreactive patterns revealed by anti-y,,, yZL, yJL2, a,COOH, Pm, and GAD antibodies are shown in Figure lOAF. The yZL and yZs immunoreaction products (Fig. 1OA,B) were localized in the glomeruli (Gl), external plexiform layer (EP), mitral cell layer (M), and granule cell layer (GR). Immunolabelie for yzs, yzL, and 7JL2, as well as (Y,COOH and & (Fig. 1OA-E) is mostly located in the EP and GL, which are enriched in proximal and distal dendrites of mitral cells, respectively,

and where synaptic GABA,R are expected to concentrate. Periglomerular and granule cells are GABAergic and make synaptic contact with the distal and proximal dendrites of mitral cells, respectively. Mitral cell bodies showed very intense immunolabeling with anti-y*, and anti-yJL2 (Fig. lOB,C), but considerably less with anti-y,, and anti-cu,COOH (Fig. lOA,D). No reaction in the cell bodies of the mitral cell layer was observed with anti-&, (Fig. 1OE). However, most of the immunolabeling with all the antibodies concentrated in the EP layer containing mitral cell dendrites. Periglomerular and tufted cells also appeared more immunolabeled with anti-y,, and anti-yJL2 than with anti-y,, (Fig. IOA-C). The distribution of the GABA,R subunits is in good correlation with the distribution of GABAergic nerve terminals revealed by anti-GAD (Fig. 100.

Thalamus Anti-r,, and anti-y,, antibodies showed weaker immunoreaction in the thalamus compared to the strong reaction observed with anti-a,COOH and anti& (Fig. 2A-D). A different subunit distribution was observed in the reticular nucleus, which showed strong immunolabeling with anti-y,, and anti-YzL, less intensity with anti-oc,COOH, and almost no labeling with anti& (Fig.

2A-D). Other nuclei Corpus striatum, substantia nigra, and deep cerebellar nuclei were immunolabeled with anti-y,,, anti-TzL, anti-cu,COOH, and anti-&, (Fig. 2A-D). The inferior colliculus as well as the su-

7174 Gutihrez

et al. * Distribution

of ySS and ySL Subunits

in Brain

7. The yZL(A and B) and y2s (C) immunoreaction products are located on the surface of surrounding cell bodies of the CA2 region of the stratum pyramidale. The punctuate (probably synaptic) immunostaining is very similar to the one revealed with anti-GAD (0). Scalebar, 9.2 pm.

Figure

perficial gray layer of the superior noreactive with anti-y2s, anti-y,,, larly with anti& (Fig. 2A-D). immunoreactivity of the inferior anti-y,, than with anti-y,,.

colliculus were also immuanti-cu,COOH, and particuNoticeable was the lower colliculus and medulla with

Discussion Although the comparative immunocytochemical results presented in this study are qualitative, we feel confident that they also reveal the quantitative relative presence of yzs and yZL in the different brain areas since (1) the sections immunostained with anti-y,, and anti-y,, antibodies were processed together and treated identically, (2) the relative labeling correlates very well with the quantitative immunoprecipitation of the yZs- and y,,-containing receptors in several regions of the brain (Khan et al., 1994a; Miralles et al., 1994), and (3) the intensity of the y2s and yZL immunolabeling correlates with the quantitative Northern and dot blot mRNA assays from several brain regions (Miralles et al., 1994). The qualitative data are also supported by the quantitative computer-assisted densitometric analysis of the immunoreaction products measured in several regions of the brain. The latter was done in connection with studies on the regulation of the expression of the GABA,R subunits during aging (Gutierrez et al., in press).

Specijkity of antibody Two novel affinity-purified antibodies, anti-y,, and anti-yzL, have been used for studying the immunocytochemical localization of the short and long forms of the y2 subunit. To the best of our knowledge, this is the first time that the regional and cellular distribution of the yZs and yZL subunits of the GABA,R in the brain has been described using immunocytochemical techniques. In addition, we have used three other GABA,R subunitspecific antibodies also raised in our laboratory (anti-yJL2, anti-oc,COOH, and mAb 62-361 to p,,,) as well as an anti-GAD antibody for comparative purposes. The subunit specificity of these novel antibodies has been studied in detail elsewhere (Khan et al., 1994a,b) and further supported here by (1) immunoblotting with crude rat brain membranes and (2) the specificity of the immunocytochemical reactions in tissue sections. The immunoblots with brain membranes (Fig. 1) showed the specificity of anti-cr,COOH, anti-yJL2, anti-y,,, and anti-y,, for the corresponding GABA,R subunit (5 1,000 M, for (Y,, 45,000 M, for yZs, and 47,000 M, for y2J. Minor reactivities were also detected for anti-cu,COOH and anti-y,, with 79,000 M, and 110,000 M, peptides, respectively. Nevertheless, the relative strength of the reactive peptides indicates that most of the immunoreactivity corresponds to the appropriate receptor subunit. Two lines of additional evidence strongly support this notion: the relative

The Journal

of Neuroscience,

November

1994,

f4(11)

7175

Figure 8. Cerebralcortex immunolabeling with anti-y,, (A), anti-y2s(B), andanti-y,COOH (C). Scalebar, 90 pm.

immunocytochemical distribution of yzs and yzL in different parts of the brain perfectly correlates with (1) their relative abundancein thesebrain regionsmeasuredby immunoprecipitation of 3H-flunitrazepam and 3H-muscimolbinding with antiyzs and anti-y,, antibodies (Khan et al., 1994a; Miralles et al., 1994), and (2) the relative regional and cellular distribution of yzs and yzL mRNA revealed by in situ hybridization (Miralles et al., 1994).

Cellular localization The subunit-specific antibodies anti-y,,, anti-yzL, and anticu,COOHshow strong membrane immunoreaction around cell bodies and proximal dendrites as well as diffuse neuropil la-

beling. The immunoreaction product that is associatedwith the cell surface is presumably synaptic becauseit often colocalizes with the GAD immunoreactivity present in GABAergic presynaptic terminals. Cytoplasmic labeling is also observed in many instances,which has also been describedby others using different GABA,R subunit antibodies (Richards et al., 1987; Benke et al., 1991b; Zimprich et al., 1991; Thompson et al., 1992). Intracellular localization of (Y,and & subunitshas also been observed by electron microscopy immunocytochemistry. The cytoplasmic labeling probably results from the receptor subunits that are in the process of synthesis, internalization, and/or degradation (Juiz et al., 1989; Somogyi et al., 1989; Yazulla et al., 1989). Nevertheless, anti-y,,, yzL, y,IL2, and

7176

GutiCrrez

et al. * Distribution

of yzS and yZL Subunits

in Brain

Figure 9. Immunolaheling of cerebral cortex neurons (layers IV-V) with anti-y,, (A and B) and anti-y,, (C and D) antibodies. Punctate immunoreaction is associated with the surface of the cell bodies and proximal dendrites. Scale bar, 7 pm. a,COOH antibodies show higher cytoplasmic immunolabeling compared to our anti-&,, mAb 62-361 (De Bias et al., 1988). Similar findings have been reported by others in which anti-y, and anti-a, antibodies (Benke et al., 199 la,b) show prominent cytoplasmic labeling compared to the mAb bd 17 to &,, (Richards et al., 1987). Perhaps the explanation for this cytoplasmic immunolabeling lies in the different localization of the epitopes recognized by the antibodies: the amino-terminal extracellular epitope recognized by mAb 62-361 to & (Ewert et al., 1992) versus the intracellular epitopes of the large IL recognized by our three anti-y, antibodies. The intracellular diffusion of the peroxidase reaction product would be favored when the antibodies recognize intracellular membrane epitopes. The most accepted model for the insertion of the GABA,R subunits in the membrane locates the amino and carboxy ends extracellularly and the loop between M3 and M4 intracellularly (Schofield et al., 1987). The antibodies that recognize extracellular epitopes produce stronger neuropil staining and weaker cytoplasmic reaction, probably due to the restricted localization of the immunoreaction product in the narrow extracellular space existing between membranes of adjacent cells. The membranes may act as diffusion barriers for the accumulated products of the peroxidase reaction. Nevertheless, this hypothesis cannot explain the relatively high cytoplasmic labeling observed with anti-cY,COOH. Presumably, the carboxy end of all GABA,R subunits is located extracellularly. Another possible explanation takes into consideration that the mAb 62-361 recognizes an epitope localized at the amino end of & and p, (Ewert et al., 1992). Therefore, this

antibody may not react with the intracellular precursor proteins that carry the leader peptide, whereas all of the other subunitspecific antibodies used in this study do react with both the intracellular precursor protein and the surface receptor since the epitopes recognized by these antibodies are located far from the amino end. Cytoplasmic immunolabeling may be due to a large extent to the binding of the antibodies to the subunit precursor proteins.

Regional distribution We and others have previously shown that (Y,, yz, and &,, are the most abundant GABA,R subunits in the rat brain and that they frequently colocalize in the same receptor complex (Khan et al., 1993, 1994a,b; Mertens et al., 1993; Pollard et al., 1993). Nevertheless, important differences in the distribution of yzs, y2L, yJL2, (I,COOH, and & immunoreaction products were found in several areas of the brain. In the cerebellum, the yzs and y2,.subunits were mainly located in the molecular layer, while (Y, and &, subunits were more abundant in the granular layer. This distribution agrees with other immunocytochemical studies: using an anti-peptide antibody to yz that does not distinguish between yzs and yzL, Benke et al. (199 1b) found predominant immunolabeling in the molecular layer. Richards et al. (1987), using their antibodies to LY, and to /& found predominant immunolabeling in the granular layer. Mismatches in the distribution of 3H-flunitrazepam (BZDR agonist) and ‘H-muscimol (GABA,R agonist) binding sites in the cerebellum have also been reported with radioligand autoradiography (Young and Kuhar, 1979; Palacios et al., 1980;

The Journal

of Neuroscience,

November

1994,

14(11)

7177

Figure 10. Olfactory bulb immunolabeling with anti-Y2,. (A), anti-y,, (B), anti-yJL2 (C), anti-cu,COOH (D), anti-&,, (E), and anti-GAD (F). Notice the higher abundance of yzs over yzL in most cells. EP, external plexiform layer; GL, glomeruli; GR, granule cell layer; M, mitral cell layer. Scale bar, 100 pm.

Sanchezet al., 1991). The yzsand yzL subunits colocalize with the prevalent 3H-flunitrazepam labeling of the molecular layer. On the other hand, (Y, and & colocalize with 3H-muscimol binding, which is predominant in the granular cell layer. These resultsare consistentwith the observation that the presenceof the yz subunit in the GABA,R complex is required for (1) the high-affinity binding of flunitrazepam and other agonist BZDs to the GABAAR, and (2) BZD stimulation ofthe GABA-induced opening of the Cl- channel (Pritchett et al., 1989; Sigel et al., 1990; Moss et al., 1991). The presenceof either yls or yzL in the GABA,R confers BZD potentiation of GABA action (Wafford et al., 1991). Our previous in situ hybridization studieshave revealed that in Purkinje cells the yzL mRNA is present in larger quantities

than y2s mRNA (Miralles et al., 1994). In our immunocytochemistry study we have also dctccted a stronger immunolabeling of the Purkinje cellswith anti-y,, than with anti-y,, (Fig. 3A,B). Purkinje cells were also immunolabeled with anti(u,COOH and anti-&,, antibodies, which is also consistentwith the expressionof yzs, yzL, LY,,&, and 0, mRNAS in thesecells (Zhang et al., 1991; Laurie et al., 1992; Persohn et al., 1992; Wisden et al., 1992; Miralles et al., 1994). In the hippocampus(dentate gyrus and hippocampusproper) and olfactory bulb (particularly mitral cells), yzsimmunolabeling wasmuch more intensethan -yzLimmunolabeling. This result is also in agreementwith our previous studiesshowingthe predominant presenceof yzsover yzL mRNA (particularly in mitral cells) and peptide in these two brain areas,as revealed by in

7178

Gutikrez

et al. * Distribution

of yzs and

7ZL Subunits

in Brain

situ hybridization and GABA,R immunoprecipitation, respectively (Khan et al., 1994a; Miralles et al., 1994). Our immunocytochemical results with the various antibodies are consistent with in situ hybridization studies, which have shown the presence of (Y,, &, &, and y2 mRNAs in granule and pyramidal cells (Zhang et al., 199 1; Persohn et al., 1992; Wisden et al., 1992) as well as yzs and yzL mRNAs (Miralles et al., 1994). The cellular distribution of yzs and y2,. immunoreaction products in the hippocampus proper and dentate gyrus frequently colocalized with GAD immunoreactivity (i.e., in the synaptic boutons surrounding granule cells and the cytoplasm of basket cells and other interneurons). The interneurons also reacted with anti-a, and anti-&,, antibodies in agreement with results previously obtained by us and by others (Richards et al., 1987; De Blas et al., 1988; Houser et al., 1988). These interneurons are also known to be GABAergic (Ribak et al., 1978; Seress and Ribak, 1983; Schmechel et al., 1984; Gamrani et al., 1986) and receive GABAergic innervation (Misgeld and Frotscher, 1986; Schlander et al., 1987). Therefore, the presence of GABA,R subunits in these GABAergic interneurons is consistent with either presynaptic or postsynaptic receptor function or both. Our immunocytochemical results in other brain areas including cerebral cortex, olfactory bulb, thalamus, inferior colliculus, corpus striatum, substantia nigra, deep cerebellar nuclei, and medulla are also consistent with the reported cellular expression of alI, pz, p,, and yz mRNAs by in situ hybridization (Zhang et al., 1990, 199 1; Laurie et al., 1992; Persohn et al., 1992; Wisden et al., 1992) as well as with our previous regional immunoprecipitation studies with anti-y,, and anti-yz, antibodies (Khan et al., 1994a,b; Miralles et al., 1994) and with the regional and cellular expression of yzs and yzL mRNAs revealed by in situ hybridization (Miralles et al., 1994).

References Benke D, Mertens S, Mohler H (199 la) Ubiquitous presence of GABA, receptor containing a,-subunit in rat brain demonstrated by immunoprecipitation and immunohistochemistry. Mol Neuropharmacol 1:103-l 10. Benke D, Mertens S, Trzeciak A, Gillessen D, Miihler H (199 1b) GABA, receptors display association of y,-subunit with (Y,and @Z/3subunits. J Biol Chem 266:4478-4483. Burt D, Kamatchi GL (199 1) GABA, receptor subtypes: from pharmacology to molecular biology. FASEB J 52916-2923. De Blas AL (1984) Monoclonal antibodies to specific astroglial and neuronal antigens reveal the cytoarchitecture of the Bergmann glia fibers in the cerebellum. J Neurosci 4:265-273. De Blas AL. Cherwinski HM (1983) Detection of antigens on nitrocellulose paper immunoblots with monoclonal antibodies. Anal Biothem 133:214-219. De Blas AL, Vitorica J, Friedrich P (1988) Localization of the GABA, receptor in the rat brain with a monoclonal antibody to the 57,000 M, peptide of the GABA, receptor/benzodiazepine receptor/Cll channel complex. J Neurosci 8:602-6 14. DeLorey TM, Olsen RW (1992) y-Aminobutyric acid, receptor structure and function. J Biol Chem 267:16747-16750. Duggan MJ, Pollard S, Stephenson FA (1992) Quantitative immunoprecipitation studies with anti-y-aminobutyric acid, receptor yz l15 Cys antibodies. J Neurochem 58:72-77. Ewert ‘M, De Blas AL, Miihler H, Seeburg P (1992) A prominent epitope on GABA, receptors is recognized by two different monoclonal antibodies. Brain Res 569:57-62. Gamrani H, Onteniente B, Seguela P, Geffard M, Calas A (1986) Gamma-aminobutyric acid-immunoreactivity in the rat hippocampus. A light and electron microscope study with anti-GABA antibodies. Brain Res 364:30-38. Gutierrez A, Khan ZU, Morris SJ, DeBlas AL (in press) Age-related decrease of GABA, receptor subunits and glutamic acid decarboxylase in the rat inferior colliculus. J Neurosci, in press.

Houser CR, Olsen RW, Richards JG, Mohler H (1988) Immunohistochemical localization of benzodiazepine/GABA, receptors in the humanhippocampalformation.J Neurosci8:1370-l383. Juiz JM, Helfert RH, Wenthold RJ, De Blas AL, Altschuler RA (1989) Immunocytochemical localization of the GABA,/benzodiazepine receptor in the guinea pig cochlear nucleus: evidence for receptor localization heterogeneity. Brain Res 504: 173-l 79. Kellenberger S, Malherbe P, Sigel E (1992) Function of the (~,&y~~ r-aminobutyric acid type A receptor is modulated by protein kinase C via multiple phospho-vlation sites. J Biol Chem 26?:25660-25663. Khan ZU. Fernando LP. Escriba P. Busauets X. Mallet J. Miralles CP. Filla M, De Blas AL’ (1993) Antibodies to the y2 subunit of the ;;alminobutyric acid,/benzodiazepine receptor. J Neurochem 60:96 lKhan ZU, Gutitrrez A, De Blas AL (1994a) The short and long form yz subunits of the GABA,/benzodiazepine receptors. J Neurochem, in press. Khan ZU, Gutitrrez A, De Blas AL (1994b) The subunit composition of a GABA,/benzodiazepine receptor from rat cerebellum. J Neurochem 63~371-374. Khrestchatisky M, MacLennan AJ, Chiang MY, Xu W, Jackson MB, Brecha N, Stemini C, Olsen RW, Tobin AJ (1989) A novel 01subunit in rat brain GABA, receptors. Neuron 3~745-753. Kofuji P, Wang JB, Moss SJ, Huganir RL, Burt DR (199 1) Generation of two forms of the y-aminobutyric acid, receptor y, subunit in mice by alternative splicing. J Neurochem 56f7 13-7 15. Laurie DJ. Seebura PH. Wisden W (1992) The distribution of thirteen GABA, receptor subunit mRNAs‘in the rat brain. II. Olfactory bulb and cerebellum. J Neurosci 12: 1063-l 070. Liiddens H, Wisden W (199 1) Function and pharmacology of multiple GABA, receptor subunits. Trends Pharmacol Sci 12:49-5 1. Machu TK, Firestone JA, Browning MD (1993) Ca*+/calmodulindependent protein kinase II and protein C phosphorylate, a synthetic peptide corresponding to a sequence that is specific for the yzL subunit of the GABA, receptor. J Neurochem 61:375-377. McLean IW, Nakane PK (1974) Periodate-lysine-paraformaldehyde fixative. A new fixative for immunoelectron microscopy. J Histochem Cytochem 22:1077-1083. Mertens S, Benke D, Mijhler H (1993) GABA, receptor populations with novel subunit combinations and drug binding profiles identified in brain by c+- and b-subunit-specific immunopurification. J Biol Chem 268:5965-5973. Miralles CP, Gutitrrez A, Khan ZU, Vitorica J, De Blas AL (1994) Differential expression of the short and long forms of the y2 subunit of the GABA,/benzodiazepine receptors. Mol Brain Res, in press. Misgeld U, Frotscher M (1986) Postsynaptic GABAergic inhibition of non-pyramidal neurons in the guinea-pig hippocampus. Neuroscience 19: 193-206. Moss SJ, Ravindran A, Mei L, Wang JB, Kofuji P, Huganir RL, Burt DR (199 1) Characterization of recombinant GABA, receptors produced in transfected cells from murine (Ye,fl,, and y2 subunit cDNAs. Neurosci Lett 123:265-268. Moss SJ, Doherty CA, Huganir R (1992) Identification of the CAMPdependent protein kinase and protein kinase C phosphorylation sites within the major intracellular domains of the @,, yzs, and yzL subunits of the y-aminobutyric acid type A receptor. J Biol Chem 267: 1447014476. Oertel WH, Schmechel DE, Tappaz ML, Kopin IJ (198 1a) Production of a specific antiserum to rat brain glutamic acid decarboxylase by injection of an antigen-antibody complex. Neuroscience 6:2689-2700. Oertel WH, Schmechel DE, Mugnaini E, Tappaz ML, Kopin IJ (198 1b) Immunocytochemical localization of glutamate decarboxylase in rat cerebellum with a new antiserum. Neuroscience 6:2715-2735. Olsen RW, Tobin AJ (1990) Molecular biology of GABA, receptor subtypes: from pharmacology to molecular biology. FASEB J 5:29 162923. Palacios JM, Young WS III, Kuhar MJ (1980) GABA and benzodiazepine receptors in rat and human brain: autoradiographic localization by a novel technique. In: Enzymes and neurotransmitters in mental disease (Usdin E, Sourkes TL, Youdin MBH, eds), pp 573583. New York Wiley. Park D, De Blas AL (199 1) Peptide subunits of y-aminobutyric acid,/ benzodiazepine receptors from bovine cerebral cortex. J Neurochem 56:1972-1979. Park D, Vitorica J, Tous G, De Blas AL (1991) Purification of the

The Journal

-r-aminobutyric acid,/benzodiazepine receptor complex by immunoaffinity chromatography. J Neurochem 56: 1962-l 97 1. Persohn E, Malherbe P, Richards JG (1992) Comparative molecular neuroanatomy of cloned GABA, receptor subunits in the rat CNS. J Comp Neuroanat 326: 193-2 16. Pollard S, Duggan MJ, Stephenson FA (1993) Further evidence for the existence-of (Ysubunit heterogeneity within discrete y-aminobutvric acid, receutor subnooulations. J Biol Chem 268:3753-3757. Priichett DB, Sontheimer I?, Shivers BD, Ymer S, Kettenmann H, Schofield PR, Seeburg PH (1989) Importance of a novel GABA, receptor subunit for benzodiazepine pharmacology. Nature 338:582585. Ribak CE, Vaughn JE, Saito K (1978) Immunocytochemical localization of glutamic acid decarboxylase in neuronal somata following colchicine inhibition of axonal transport. Brain Res 140:3 15-332. Richards JG, Schoch P, Hating P, Takacs B, Miihler H (1987) Resolving GABA,/benzodiazepine receptors: cellular and subcellular localization in the CNS with monoclonal antibodies. J Neurosci 7:1866-1886. Sanchez MP, Diet1 MM, De Blas AL, Palacios JM (1991) Mapping of benzodiazepine-like immunoreactivity in the rat brain as revealed by a monoclonal antibody to benzodiazepines. J Chem Neuroanat 4:111-121. Schlander M, Thomalske G, Frotscher M (1987) Fine structure of GABAergic neurons and synapses in the human dentate gyrus. Brain Res 401:185-189. Schmechel DE, Vickrey BG, Fitzpatrick D, Elde RP (1984) GABAergic neurons of mammalian cerebral cortex: widespread subclass defined by somatostatin content. Neurosci Lett 47:227-232. Schofield PR, Darlison MG, Fujita N, Burt DR, Stephenson FA, Rodriguez H, Rhee LM, Ramachandran J, Reale V, Glencorse TA, Seeburg PH, Barnard EA (1987) Sequence and functional expression of the GABA, receptor shows a ligand-gated receptor super-family. Nature 328:221-227. Seress L, Ribak CE (1983) GABAergic cells in the dentate gyrus appear to be local circuit and projection neurons. Exp Brain Res 50: 173182. Shivers BD, Killish I, Sprengel R, Sontheimer H, Kiihler M, Schofield PR, Seeburg PH (1989) Two novel GABA,-receptor subunits exist in distinct neuronal subuouulations. Neuron 3:327-337. Sigel E, Baur R, Trube G, Mihler H, Malherbe P (1990) The subunit composition of rat brain GABA, receptors on channel function. Neuron 5:703-7 11.

of Neuroscience.

November

1994,

74(11)

7179

Somogyi P, Takagi H, Richards JG, Mijhler H (1989) Subcellular localization of benzodiazepine/GABA, receptors in the cerebellum of rat, cat, and monkey using monoclonal antibodies. J Neurosci 9:2197-2209. Thompson CL, Bodewitz G, Stephenson FA, Turner JD (1992) Mapping of GABA, receptor o5 and og subunit-like immunoreactivity in rat brain. Neurosci Lett 144:53-56. Vitorica J, Park D, Chin G, De Blas AL (1988) Monoclonal antibodies and conventional antisera to the GABA,/benzodiazepine receptor/ Cl- channel complex. J Neurosci 8:6 15-622. Wafford KA, Whiting PJ (1992) Potentiation of GABA, receptors requires phosphorylation of the alternative spliced variant of the y2 subunit. FASEB Lett 3 13: 113-l 17. Wafford KA, Burnett DM, Leidenheimer NJ, Burt DR, Wang JB, Kofuji P, Dunwiddie TV, Harris RA, Sikela JM (199 1) Ethanol sensitivity ofthe GABA, receptor expressed in Xenopus oocytes requires 8 amino acids contained in the y2L subunit. Neuron 7:27-33. Whiting P, McKeman RM, Iversen LL (1990) Another mechanism for creating diversity in y-aminobutyric type A receptors: RNA splicing directs expression of two forms of y2 subunit, one of which contains a protein kinase C phosphorylation site. Proc Nat1 Acad Sci USA 87:9966-9970. Wisden W, Laurie DJ, Monyer H, Seeburg PH (1992) The distribution of 13 GABA, receptor subunit mRNAs in the rat brain. I. Telencephalon, diencephalon, mesencephalon. J Neurosci 12: 1040-1062. Yazulla S, Studholme KM, Vitorica J, De Blas AL (1989) Immunocytochemical localization of GABA, receptors in gold-fish and chicken retinas. J Comp Neurol 280: 15-26. Young WS III, Kuhar MJ (1979) Autoradiographic localization of benzodiazepine receptors in the brains ofhumans and animals. Nature 280:393-395. Zhang JH, Sato M, Noguchi K, Tohyama M (1990) The differential expression patterns of the mRNAs encoding p subunits (/3,, B2 and 0,) of GABA, receptors in the olfactory bulb and its related areas in the rat brain. Neurosci Lett 119:257-260. Zhang JH, Sato M, Tohyama M (199 1) Region-specific expression of the mRNAs encoding ,6 subunits (fi,, & and 8,) of GABA, receptor in the rat brain. J Comp Neurol 303:637-657. Zimprich F, Zezula J, Sieghart W, Lassmann H (1991) Immunohistochemical localization of the LY,,(Y*and olj subunits of the GABA, receptor in the rat brain. Neurosci Lett 127: 125-l 28.