Recent cloning of K+ channel β subunits revealed that these cytoplasmic polypeptides can dramatically alter the kinetics of current inactivation and promote efficient glycosylation and surface expression of the channel-forming α subunits. Here, we examined the expression, distribution, and association of two of these β subunits, Kvβ1 and Kvβ2, in adult rat brain. In situhybridization using cRNA probes revealed that these β-subunit genes are heterogeneously expressed, with high densities of Kvβ1 mRNA in the striatum, CA1 subfield of the hippocampus, and cerebellar Purkinje cells, and high densities of Kvβ2 mRNA in the cerebral cortex, cerebellum, and brainstem. Immunohistochemical staining using subunit-specific monoclonal and affinity-purified polyclonal antibodies revealed that the Kvβ1 and Kvβ2 polypeptides frequently co-localize and are concentrated in neuronal perikarya, dendrites, and terminal fields, and in the juxtaparanodal region of myelinated axons. Immunoblot and reciprocal co-immunoprecipitation analyses indicated that Kvβ2 is the major β subunit present in rat brain membranes, and that most K+ channel complexes containing Kvβ1 also contain Kvβ2. Taken together, these data suggest that Kvβ2 is a component of almost all K+channel complexes containing Kv1 α subunits, and that individual channels may contain two or more biochemically and functionally distinct β-subunit polypeptides.
Voltage-gated K+ channels are composed of pore-forming α subunits and associated cytoplasmic β-subunit polypeptides (Pongs, 1995). These channels are critical for action potential conduction and neurotransmitter release, and are essential to the control of neuronal excitability (Hille, 1992). Expression of α-subunit mRNAs in heterologous cells gives rise to tetrameric channel complexes with electrophysiological characteristics similar to A-type or delayed-rectifier channels (for review, see Pongs, 1992). The large number of α-subunit genes, their ability to assemble into heteromultimers (Ruppersberg et al., 1990; Sheng et al., 1993;Wang et al., 1993; Scott et al., 1994a; Rhodes et al., 1995), and their heterogeneous expression in mammalian brain undoubtedly contributes to the diversity of voltage-gated K+ channelsin situ (Stühmer et al., 1989; Jan and Jan, 1990). However, the recent discovery of auxiliary subunits associated with ion channels in general (for review, see Isom et al., 1994), and with voltage-gated K+ channels in particular (Parcej et al., 1992; Rettig et al., 1994; Scott et al., 1994b), revealed that their electrophysiological and biochemical properties can be dramatically affected by the presence of β subunits (Rettig et al., 1994; Majumder et al., 1995; Morales et al., 1995; Shi et al., 1996). This association of β subunits with K+ channels not only increases the potential for diversity, it also indicates that the functional properties of individual channels are governed by the specific combination of α and β subunits present in the channel complex.
The β subunits of voltage-gated K+ channels were identified as 38–41 kDa polypeptides associated with the dendrotoxin acceptor purified from bovine brain (Parcej and Dolly, 1989, 1992; Scott et al., 1994a,b), and as a 38 kDa polypeptide in immunoprecipitated rat brain K+ channel complexes (Trimmer, 1991). Subsequent cloning of cDNAs encoding a bovine β subunit (Scott et al., 1994b); the related Kvβ1, Kvβ2, and Kvβ3 in rat brain (Rettig et al., 1994; Heinemann et al., 1995); and closely related β subunits in Drosophila (Chouinard et al., 1995), ferret (Morales et al., 1995), and human tissues (England et al., 1995a,b; Majumder et al., 1995; McCormack et al., 1995) indicated that these β subunits are highly conserved and that some can modulate the rate of inactivation of certain α subunits (Rettig et al., 1994;England et al., 1995a,b; Heinemann et al., 1995; Majumder et al., 1995;McCormack et al., 1995; Morales et al., 1995).
We recently reported that Kvβ1 and Kvβ2 associate with α subunits early in their biosynthesis and exert chaperone-like effects on the α subunits, promoting their efficient glycosylation and stable expression in the plasma membrane (Shi et al., 1996). In addition, we reported that Kvβ1 and Kvβ2 associate with all members of theShaker-related (Kv1) α-subunit subfamily upon co-expression in transfected mammalian cells (Nakahira et al., 1996). Here, we used riboprobes and antibodies specific for Kvβ1 and Kvβ2 to examine their expression, subcellular distribution, and co-association in adult rat brain. We observed that these β subunits are widely expressed, and that immunoreactivity for Kvβ1 and Kvβ2 is concentrated in multiple subcellular domains including neuronal somata and dendrites, the paranodal segments of myelinated axons, and in the terminal fields of several cortical and subcortical projection systems. We also observed that Kvβ2 is the predominant β-subunit isoform in rat brain; although Kvβ1 mRNA is widely expressed, the Kvβ1 polypeptide is not a major component of the total rat brain β-subunit pool, and almost all of the Kvβ1 that is present is in K+ channel complexes that also contain Kvβ2. Together, these observations suggest that β subunits are integral components of K+ channel complexes, and that the inclusion of Kvβ1 in Kvβ2-containing complexes may serve to fine tune the electrophysiological properties of channels in specific brain regions.
MATERIALS AND METHODS
Materials. All reagents were molecular biology grade from Sigma (St. Louis, MO) or Boehringer Mannheim (Indianapolis, IN), except where otherwise noted.
Preparation of 35 S-labeled cRNA probes. DNA templates for riboprobe synthesis were prepared by the PCR using plasmid clones containing the full-length rat Kvβ1 or Kvβ2 cDNAs (Nakahira et al., 1996) or a partial Kvβ2 cDNA (K. Nakahira, S. Nawoschik, J. Trimmer, unpublished observations) as PCR templates. Two independent riboprobes targeted to unique, nonoverlapping regions of Kvβ1 or Kvβ2 were used in the present study. The probe sequences were checked versus the Genbank database to ensure that they only recognize the appropriate targets among all deposited sequences. To generate riboprobes for the Kvβ1 subunit, one pair of oligonucleotide primers was designed to amplify a 264 bp region spanning nucleotides −46 to 218 of the rat Kvβ1 cDNA and, in addition, to add the promoter sequences for T3 RNA polymerase. These primers contained the following sequences: 5′-CGGATCCG- CTGTGCTGTGGGGTTCTGAGAGGAC-3′ (forward); 5′-AATTAACC- CTCACTAAAGGGATATTTCATGCCAGTCTGCT-3′ (reverse). The forward primer for this Kvβ1 probe contained eight nucleotides of vector sequence at the 5′ end from the BamHI site used for cloning. A second pair of oligonucleotide primers was designed to amplify a 271 bp sequence spanning nucleotides −321 to −51 of the rat Kvβ1 cDNA and, in addition, to add the promoter sequences for T3 RNA polymerase. These primers had the following sequences: 5′-CCTGACCACATGGATCTGGC-3′ (forward); 5′-AATTAACCCTCACTAAAGGGCGGCAGAGGGTGAGACGTT-3′ (reverse). To generate riboprobes for the Kvβ2 subunit, one pair of primers was designed to amplify a 228 bp fragment from the 3′ untranslated region (nucleotides 1130–1357) of the rat Kvβ2 cDNA and, in addition, to add the promoter sequences for T3 RNA polymerase. These primers had the following sequences: 5′-CCCAGCTCGGACAGTTCCTGGTTCC-3′ (forward); 5′-AATTAACCCTCACTAAAGGGGCATCCAGCGAGG AAG- CGGC-3′ (reverse). A second riboprobe for Kvβ2 was generated using oligonucleotide primers designed to amplify a 300 bp fragment spanning nucleotides 809-1108 of the rat Kvβ2 cDNA and, in addition, to add the promoter sequences for T3 RNA polymerase. These primers contained the following sequences: 5′-ACCAGTGGTTGAAGGACAAG-3′ (forward); 5′-AATTAACCCTCACTAAAGGGTGACTTAGGATCTATAGTCC-3′ (re-verse). All PCR products were gel purified on 1.5% low-melt agarose gels, and bands containing the Kvβ1 or Kvβ2 products were excised, phenol and phenol-chloroform extracted, and ethanol precipitated. The pellet then was dried and resuspended in 1× TE buffer containing (in mm): 10 Tris/HCl, 1 EDTA, pH 7.4. Fifty nanograms of DNA template were used for transcription reactions using35S-CTP (New England Nuclear, Boston, MA) and the Riboprobe Gemini System (Promega, Madison, WI). Probes were examined by Northern analysis of rat brain RNA and by Southern analysis of plasmids containing the Kvβ1 and Kvβ2 cDNAs. Each probe reacted with a single band of appropriate size on Northern blots and with only the appropriate cDNA on the plasmid Southern blots, indicating that these riboprobes specifically recognized Kvβ1 and Kvβ2 transcripts (data not shown).
In situ hybridization. Eight adult male Sprague–Dawley rats were used for analysis of Kvβ1 and Kvβ2 mRNA expression by in situ hybridization histochemistry. Animals were killed by asphyxiation with CO2, and the brains were removed, immediately frozen in a bed of pulverized dry ice, and stored at −70°C. Sections were cut at 10 μm on a Hacker–Brights cryostat and thaw mounted onto chilled (−20°C) slides coated with Vectabond reagent (Vector Labs, Burlingame, CA). All solutions were prepared in dH2O treated with 0.1% (v/v) diethylpyrocarbonate and autoclaved. Sections were fixed by immersion in 4% paraformaldehyde in PBS, pH 7.4, then immersed sequentially in 2× SSC (1× SSC is 0.150 m sodium chloride, 0.015 m sodium citrate), dH2O, and 0.1 mtriethanolamine, pH 8.0. The sections then were acetylated by immersion in 0.1 m triethanolamine containing 0.25% (v/v) acetic anhydride; washed in 0.2× SSC; dehydrated in 50, 70, and 90% ethanol; and rapidly dried. One ml of prehybridization solution containing 0.9 m NaCl, 1 mmEDTA, 5× Denhardt’s solution, 0.25 mg/ml single-stranded herring sperm DNA (Gibco, Gaithersburg, MD), and 50% deionized formamide (EM Sciences, Gibbstown, NJ) in 10 mm Tris, pH 7.6, was pipetted onto each slide, and the slides incubated for 3 hr at 50°C in a humidified box. The sections then were dehydrated by immersion in 50, 70, and 90% ethanol and air dried.
Labeled riboprobes were denatured in a small volume (100 μl) of hybridization solution containing 0.9 m NaCl, 1 mm EDTA, 1× Denhardt’s solution, 0.1 mg/ml yeast tRNA, 0.1 mg/ml single-stranded salmon sperm DNA, dextran sulfate (10%), 0.08% BSA, 10 mm DTT (Boehringer Mannheim, Indianapolis, IN), and 50% deionized formamide in 10 mm Tris, pH 7.6, at 95°C (1 min); placed on ice (5 min); and added at a final concentration of 50,000 cpm/μl to 10 ml of prewarmed (to 55°C) hybridization solution. The hybridization solution then was pipetted onto the sections and allowed to hybridize overnight at 55°C in a humidified chamber. The sections were subsequently washed once for 45 min at 37°C in 2× SSC containing 10 mm DTT, once for 30 min at 37°C in 1× SSC containing 50% formamide, and once for 30 min at 37°C in 2× SSC. Single-stranded and nonspecifically hybridized riboprobe was digested by immersion in 10 mm Tris, pH 8.0, containing bovine pancreas RNase A (Boehringer Mannheim) (40 μg/ml), 0.5 m NaCl, and 1 mm EDTA. The sections then were washed in 2× SSC for 1 hr at 60°C, followed by 0.1× SSC containing 0.5% (w/v) sodium thiosulfate for 2 hr at 60°C. The sections then were dehydrated in 50, 70, and 90% ethanol containing 0.3 m ammonium acetate, and dried.
The slides then were loaded into X-ray cassettes and exposed to Hyperfilm β-Max (Amersham, Arlington Heights, IL) for 3–7 d. Once a satisfactory exposure was obtained, the slides were coated with nuclear-track emulsion (NTB-2) (Eastman Kodak, Rochester, NY) and exposed for 7–21 d at 4°C. The emulsion autoradiograms were developed and fixed according to the manufacturer’s instructions, and the underlying tissue sections were stained with hematoxylin.
To assess nonspecific labeling in the in situ hybridization procedure, a control probe was generated from a template provided in the Riboprobe Gemini System kit (Promega catalog #P2651). This vector was linearized using ScaI, and transcribed using T3 RNA polymerase. The resulting transcription reaction generates two products, a 250 bp and a 1525 bp riboprobe, containing only vector sequence. This control probe mixture was labeled as described above and added to the hybridization solution at a final concentration of 50,000 cpm/μl. No specific hybridization was observed in control sections; i.e., these sections gave a very weak uniform hybridization signal that did not follow neuroanatomical landmarks (data not shown).
Production of synthetic peptides and antibodies. Synthetic peptides corresponding to amino acids 7–28 from the N terminus of the rat Kvβ1 polypeptide (CTEHNLKSRNGEDRLLSKQSST) (Rettig et al., 1994) and amino acids 1–17 of the rat Kvβ2 polypeptide (MYPESTTGSPARLSLRQC) (Rettig et al., 1994) were synthesized (Quality Controlled Biochemicals, Hopkinton, MA) and conjugated to keyhole limpet hemocyanin (KLH) (1 mg peptide/mg carrier protein) using sulfo-m-maleimidobenzoyl-N-hydroxysuccinimide ester (Pierce, Rockford, IL). These peptide/KLH conjugates were injected into rabbits for the production of polyclonal antisera (Pocono Rabbit Farm, Canadensis, PA), and into mice for the production of monoclonal antibodies (mAbs). Production of mAbs and purification of monoclonal immunoglobulins were performed essentially as described previously (Trimmer et al., 1985), and will be described in detail elsewhere (Z. Bekele-Arcuri, M. Matos, J. Trimmer, unpublished observations). For affinity purification of the polyclonal antibodies, the Kvβ1 and Kvβ2 peptides were conjugated to SulfoLink coupling gel (Pierce) via the terminal cysteine residues. Polyclonal antibodies were affinity purified from immune serum by standard procedures (Harlow and Lane, 1988).
Immunofluorescence. Immunofluorescence staining of transiently transfected cells expressing Kvβ1 and Kvβ2 was performed essentially as described previously (Shi et al., 1994). Briefly, green monkey fibroblast COS-1 cells were cultured on glass coverslips that had been previously coated with 25 μg/ml poly-l-lysine. For staining, cells were washed three times in ice-cold PBS, then fixed and permeabilized by treatment for 20 min at 4°C with a freshly prepared ice-cold fixative containing 3% paraformaldehyde/0.1% Triton X-100 in PBS. Cells then were washed three times in TBS containing 0.1% Triton X-100 (TBS-T), and nonspecific protein binding sites were blocked by incubation in TBS-T containing 4.5% w/v nonfat dry milk powder (Blotto-T) (Johnson et al., 1984). Cells then were incubated in primary antibodies diluted in Blotto-T for 1 hr at room temperature, washed three times in Blotto-T for 30 min total, and incubated in the appropriate secondary antibodies for 30 min. After three washes in TBS-T for 15 min total, cells were mounted in PBS medium containing 90% glycerol and 1 mg/mlp-phenylenediamine and viewed on a Zeiss Axiophot microscope using epifluorescence illumination.
Brain membrane preparations. A crude synaptosomal membrane fraction was prepared from freshly dissected adult rat brains, essentially as described previously (Trimmer, 1991, 1993). Briefly, brains were homogenized in 0.3 m sucrose, 10 mm sodium phosphate, pH 7.4, and 10 mm sodium fluoride, containing a protease inhibitor cocktail (1 mm phenylmethyl sulfonyl fluoride, 1 μg/ml leupeptin, 2 μg/ml aprotinin, and 1 μg/ml pepstatin). The resultant homogenate was centrifuged at 3,000 × g for 10 min to remove nuclei and cellular debris. The supernatant then was centrifuged at 45,000 × g for 60 min to pellet the membranes. Aliquots of the membrane preparations were suspended in the homogenization buffer, and protein was determined using the BCA method (Pierce).
SDS/polyacrylamide gels and immunoblotting. For immunoblots, 50 μg of membrane protein was added to SDS sample buffer, boiled, and fractionated on 7.5 or 12% SDS/polyacrylamide gels (Maizel, 1971). Disulfide bonds were reduced by the addition of 20 mm 2-mercaptoethanol to the sample buffer. Lauryl sulfate (Sigma) was the SDS source used for all SDS-PAGE (Shi et al., 1994). After electrophoretic transfer to nitrocellulose paper, the resulting blots were blocked in TBS containing 4% low fat milk (Blotto) (Johnson et al., 1984), incubated in affinity-purified antibody diluted 1:50–1:100 in Blotto for 1 hr or undiluted mAb tissue culture supernatants, and washed three times in Blotto for 30 min total. Blots then were incubated in HRP-conjugated secondary antibody (1:2000 dilution in Blotto) (Cappel, West Chester, PA) for 1 hr and then washed in TBS three times for 30 min total. The blots then were incubated in substrate for enhanced chemiluminescence (ECL) (Amersham, Arlington Heights, IL) for 1 min and autoradiographed on preflashed (to OD545 = 0.15) Kodak X-OMAT LS film.
Immunoprecipitation. Immunoprecipitation reactions were performed using detergent lysates of rat brain membranes. All procedures were performed at 4°C. Membranes (1 mg membrane protein/tube) were solubilized to 1 ml final volume/tube in lysis buffer (1% Triton X-100, 0.15 m NaCl, 1 mm EDTA, 10 mm sodium azide, 10 mm Tris/HCl, pH 8.0, containing a protease inhibitor cocktail (see above). Affinity-purified antibodies were added and the samples incubated for 2 hr on a rotator, followed by addition of 20 μl of a 50% suspension of protein A sepharose and additional incubation for 45 min. After incubation, protein A sepharose was centrifuged at 10,000 × g for 20 sec, and the resulting pellets were washed by resuspension and centrifugation six times with lysis buffer. The final pellets were resuspended in 50 μl reducing sample buffer, and 20 μl electrophoresed on 12% SDS-PAGE and subjected to immunoblotting as described above.
Immunohistochemistry. Twelve adult male Sprague–Dawley rats were deeply anesthetized with sodium pentobarbital (60 mg/kg, i.p.) and perfused through the ascending aorta with 150 ml of 0.9% saline, followed by 500 ml of fixative containing freshly prepared 4% paraformaldehyde in 0.1 mNaPO4 buffer, pH 7.4 (PB). The brains were removed, cryoprotected for 18–48 hr in 20% sucrose in PB, frozen in a bed of pulverized dry ice, and then cut into 35-μm-thick sections on a sliding microtome. Consecutive 1-in-10 series of sections were collected in 0.05 m PB and processed for light microscopic immunohistochemistry as described previously (Rhodes et al., 1995). Affinity-purified rabbit polyclonal anti-Kvβ1 and rabbit anti-Kvβ2 antibodies were used at dilutions of 1:1500 and 1:4000, respectively. Mouse mAbs (purified IgG fractions and tissue culture supernatants) raised against Kvβ1 (clone K9/40) and Kvβ2 (clone K17/70) were used at dilutions of 1:100–1:1500. To verify the specificity of the immunohistochemical reactions, some sections were processed either without addition of the primary antibody or using antibodies incubated previously (1 hr) in vehicle containing an excess of the synthetic peptide/BSA antigen (5–25 μg/ml). No specific staining was observed in these control sections (data not shown).
Analysis of sections processed for in situ hybridization and immunohistochemistry was performed using a Zeiss Axiophot photomicroscope. Low-magnification photographs of immunohistochemically stained sections and autoradiograms generated on film were taken using a Nikon Multiphot macrophotography system.
Localization of Kvβ1 and Kvβ2 mRNA
Analysis of autoradiograms prepared on film indicated that the Kvβ1 and Kvβ2 mRNAs are widely and heterogeneously expressed in adult rat brain (Table 1). The pair of riboprobes for Kvβ1 generated patterns of hybridization signal that were indistinguishable from one another, as did the pair of riboprobes for Kvβ2. Taken together with the result of Northern and Southern blotting using these probes, this result strongly suggests that the riboprobes used in the present study specifically reveal the presence of Kvβ1 or Kvβ2 mRNA.
A very high density of Kvβ1 expression was observed in the striatum, nucleus accumbens, olfactory tubercle, CA1 subfield of the hippocampus, and entorhinal and posterior cingulate cortices, and in several midbrain and brainstem motor nuclei (Fig. 1). Intermediate levels of Kvβ1 expression were observed in the piriform cortex; neocortex; medial septal–diagonal band complex; anterior, mediodorsal, and ventral tier thalamic nuclei; and the cerebellar cortex and deep nuclei. Intermediate levels of Kvβ1 expression also were observed in the laterodorsal, ventral posteromedial (VPM), ventral posterolateral (VPL), and dorsal lateral geniculate, and medial geniculate nuclei of the thalamus. Low levels of Kvβ1 expression were observed in the globus pallidus and hypothalamus. The highest levels of Kvβ2 expression were observed in the piriform cortex, hippocampal formation, and in layer II of the entorhinal cortex, with somewhat lower levels in the neocortex, medial septal–diagonal band complex, and the anterior, VPM, and VPL nuclei of the thalamus. Low and levels of Kvβ2 expression were observed in the remaining thalamic nuclei and in the striatum, globus pallidus, and hypothalamus.
Neocortex and hippocampus
Examination of high-resolution emulsion autoradiograms of sections processed to visualize hybridization to Kvβ1 and Kvβ2 mRNAs indicated that these two transcripts are expressed within cellular profiles with the size and morphology of neurons as opposed to glial cells (Fig. 2). Kvβ1 mRNA was expressed in large pyramidal cells in the deep half of layer III and in layers V and VI. Very high levels of Kvβ1 mRNA were observed in small to medium interneurons in cortical layers II, III, V, and VI. Interestingly, high levels of Kvβ1 mRNA also were observed in small cells juxtaposed to the subcortical white matter (Fig. 1). In virtually all regions of the neocortex, Kvβ2 mRNA was highly expressed in pyramidal cells in layers II, III, V, and VI. In proisocortical areas such as the entorhinal and cingulate cortices, Kvβ1 and Kvβ2 mRNA also were highly expressed in large multipolar neurons in layer II (Fig. 2).
In the hippocampus, Kvβ1 mRNA was expressed in dentate granule cells and in pyramidal cells of all subfields. However, pyramidal cells in the CA1 subfield expressed higher levels of Kvβ1 mRNA than the adjacent CA2 subfield and subiculum (Fig. 1). In addition, Kvβ1 was expressed in large interneurons located in stratum oriens and radiatum of all subfields. The distribution of Kvβ2 mRNA was strikingly similar to that observed for Kvβ1, with the exception that there appeared to be a uniformly high density of Kvβ2 expression across all subfields, and Kvβ2 mRNA did not appear to be expressed at an appreciably greater density in hippocampal interneurons.
Striatum and basal forebrain
The greatest density of Kvβ1 expression was observed in the caudate putamen, nucleus accumbens, and olfactory tubercle. Virtually all neurons in these structures expressed extremely high levels of Kvβ1 mRNA (Figs. 1, 2 E). Kvβ1 mRNA also was highly expressed in the medial septal nuclei and the vertical and horizontal limbs of the diagonal band of Broca in a distribution that closely resembled the distribution of cholinergic neurons in these structures (Fig. 2 G) (Mesulam et al., 1983). In contrast to Kvβ1, there was a low level of Kvβ2 expression in the caudate putamen, nucleus accumbens, and olfactory tubercle (Figs. 1, 2 F). However, there was a high level of Kvβ2 expression in the medial septal and diagonal band nuclei in a pattern that overlapped with, but was somewhat more extensive than, that observed for Kvβ1 (Figs. 1, 2 H).
Thalamus and hypothalamus
A moderate density of Kvβ1 expression was observed throughout the thalamus, with a somewhat greater density in the anterior, mediodorsal, and entopeduncular nuclei (Fig. 1), and a moderate density in the laterodorsal, VPM, and VPL nuclei. A low level of Kvβ1 expression was observed in the hypothalamus, but with a somewhat more intense hybridization signal in the ventromedial hypothalamic nucleus. A low to moderate level of Kvβ2 expression was observed throughout the thalamus, with a somewhat greater density in the anterior, mediodorsal, paracentral, and ventroposterior nuclei. Low levels of Kvβ2 expression also were observed in the hypothalamus, however, as for Kvβ1, there was a somewhat greater density of Kvβ2 expression in the ventromedial nucleus.
Several midbrain motor nuclei, including the red nucleus and all cranial nerve nuclei, contained a very high density of Kvβ1 and Kvβ2 expression (Fig. 1). In addition, cells in the superficial and deep layers of the superior and inferior colliculi, the periaqueductal gray, and the pars compacta and pars reticulata of the substantia nigra expressed intermediate levels of Kvβ1 mRNA. Similarly, large neurons in the deep layers of the superior and inferior colliculi expressed moderate levels of Kvβ2 mRNA, and cells scattered throughout the substantia nigra pars compacta and pars reticulata expressed moderate to low levels of Kvβ2 mRNA (Fig. 1).
Cerebellum and brainstem
In the cerebellar cortex, Purkinje cells and granule cells displayed high and moderate levels of Kvβ1 expression, respectively, and small neurons scattered throughout the molecular layer expressed much lower levels of Kvβ1 mRNA. In contrast, Purkinje cells and granule cells displayed intermediate levels of Kvβ2 expression, and small neurons with a size and distribution similar to basket cells expressed moderate levels of Kvβ2 mRNA. Large neurons in all deep cerebellar nuclei expressed moderate levels of Kvβ1 and Kvβ2 mRNA (Fig. 1).
In the pons and medulla, all cranial nerve sensory and motor nuclei displayed moderate to high levels of Kvβ1 and Kvβ2 expression. In addition, mRNA for both β subunits was highly expressed in large pontine and medullary reticular neurons and in neurons of the superior and inferior olivary complexes.
Generation and characterization of Kvβ1- and Kvβ2-specific polyclonal and mAbs
A polyclonal antiserum was produced by the immunization of rabbits with a synthetic Kvβ1 peptide immunogen, corresponding to N-terminal amino acids 7–28 of the deduced rat brain Kvβ1 sequence (Rettig et al., 1994). This sequence is not present in the deduced sequence of the rat brain Kvβ2 β subunit, although some overlap is seen to a recently identified rat brain Kvβ3 β subunit (11/22 positions identical with changes spread throughout the sequences) (Heinemann et al., 1995; Pongs, 1995). Thus, it is likely that antibodies generated to this peptide are specific for Kvβ1. Rabbit polyclonal antibodies raised against this peptide exhibited a high titer against the peptide on ELISA assays and displayed a monospecific reaction on immunoblot assays versus rat brain membranes to a 44 kDa polypeptide (Fig.3). That this peptide was Kvβ1 was supported by analysis of Kvβ1 and Kvβ2 expressed in COS-1 cells, which showed that the expressed recombinant Kvβ1 polypeptide shared immunoreactivity and electrophoretic mobility with the putative rat brain Kvβ1 polypeptide (Fig. 3). No corresponding immunoreactivity was seen to the recombinant 38 kDa Kvβ2 polypeptide expressed in COS-1 cells or present in the crude rat brain membranes on the same immunoblot (Fig. 3). Immunofluorescence staining of transfected COS-1 cells revealed a similar pattern of immunoreactivity in that the anti-Kvβ1 antibody exhibited a strong immunofluorescence staining of Kvβ1- , but not Kvβ2- , transfected cells (Fig.4).
mAbs generated against the same Kvβ1-specific peptide exhibited similar properties. All of the mAbs exhibited strong reactions in ELISA assays against the Kvβ1 peptide immunogen and in immunofluorescence staining of COS-1 cells expressing recombinant Kvβ1 polypeptide (Fig.4). Staining was not observed in nontransfected or Kvβ2-transfected cells (Fig. 4). Immunoblot analysis of crude rat brain membranes revealed that none of the mAbs isolated recognized the SDS-denatured rat brain Kvβ1 polypeptide, although specific immunoreactivity was seen to the recombinant Kvβ1 (but not Kvβ2) polypeptide present in COS-1 cell extracts, perhaps because of the higher expression levels of Kvβ1 in these extracts compared with that observed in rat brain membranes (data not shown). All of these mAbs, and the rabbit polyclonal antibody, selectively immunoprecipitate the recombinant Kvβ1 polypeptide present in 35S-methionine labeled COS-1 cell extracts (data not shown). Additional details on the generation and characterization of these mAbs will be described elsewhere.
A polyclonal antiserum was produced by the immunization of rabbits with a synthetic Kvβ2 peptide immunogen, corresponding to N-terminal amino acids 1–17 of the deduced rat brain Kvβ2 sequence (Rettig et al., 1994). This sequence is not present in the deduced sequence of the rat brain Kvβ1 or Kvβ3 β subunit (Heinemann et al., 1995; Pongs, 1995). Thus, it is likely that antibodies generated to this peptide would be specific for Kvβ2. Rabbit polyclonal antibodies raised against this peptide exhibited a high titer against the Kvβ2 peptide on ELISA assays and displayed a monospecific reaction on immunoblot assays versus rat brain membranes to a 38 kDa polypeptide (Fig. 3). Additional immunoreactivity was exhibited to a minor, 41 kDa polypeptide (Fig. 3) that is also recognized by the pan-β-subunit antibody (Rhodes et al., 1995). The molecular identity of this 41 kDa polypeptide is not known; however, the fact that it exhibits immunoreactivity to both the N-terminally directed anti-Kvβ2 antibody and the C-terminally directed pan-β antibody indicates that it is either a post-translational variant of Kvβ2 or an alternative splice variant of the Kvβ2 gene. That the 38 kDa polypeptide was Kvβ2 was supported by analysis of Kvβ1 and Kvβ2 expressed in COS-1 cells, which showed that the expressed recombinant Kvβ2 polypeptide shared immunoreactivity and electrophoretic mobility (M r = 38 kDa) with the putative rat brain Kvβ2 polypeptide (the 41 kDa variant was not detected). No corresponding immunoreactivity was seen to the recombinant 44 kDa Kvβ1 polypeptide expressed in COS-1 cells or present in the crude rat brain membranes on the same immunoblot (Fig. 3). Immunofluorescence staining of transfected COS-1 cells revealed a similar pattern of immunoreactivity in that the anti-Kvβ2 antibody exhibited a strong immunofluorescence staining of Kvβ2- but not Kvβ1- transfected cells (Fig. 4).
Anti-Kvβ2 mAbs were generated against a glutathione-S-transferase fusion protein containing the entire Kvβ2 polypeptide. All of the N-terminally directed mAbs exhibited monospecific immunofluorescence staining of COS-1 cells expressing recombinant Kvβ2, but not Kvβ1, polypeptide (Fig. 4). No staining to untransfected cells was seen (Fig. 4). Immunoblot analysis of crude rat brain membranes revealed that only the K17/70 mAb recognizes the SDS-denatured rat brain Kvβ2 polypeptide and recombinant Kvβ2 (but not Kvβ1) polypeptide present in COS-1 cell extracts (Fig. 3). Additional details on the generation and characterization of these mAbs will be described elsewhere.
Reciprocal co-immunoprecipitations: subunit association
To determine whether Kvβ1 and Kvβ2 were present in the same rat brain K+ channel complexes, we performed reciprocal co-immunoprecipitation experiments with the anti-Kvβ1 and anti-Kvβ2 antibodies. Detergent lysates were prepared from rat brain membranes under conditions shown previously to preserve α- or β-subunit interactions (Sheng et al., 1993; Wang et al., 1993; Rhodes et al., 1995; Nakahira et al., 1996; Shi et al., 1996). These lysates then were used in immunoprecipitation reactions performed with the anti-Kvβ1 and anti-Kvβ2 polyclonal antibodies and with a pan-β polyclonal antibody that recognizes both Kvβ1 and Kvβ2 (Rhodes et al., 1995). Immunoprecipitations also were performed using a polyclonal antibody against the Kv2.1 α subunit, which should not co-immunoprecipitate either of these β subunits (Rhodes et al., 1995). Immunoprecipitation reactions then were subjected to immunoblot analyses to assay for the presence of Kvβ1 with the anti-Kvβ1 polyclonal antibody, for Kvβ2 with the K17/70 mAb, and for the β-subunit-associated α subunit Kv1.2 with the Kv1.2C polyclonal antibody (Rhodes et al., 1995).
As expected, the anti-Kvβ1 antibody immunoprecipitated the 44 kDa Kvβ1 polypeptide from these rat brain extracts (Fig.5), as did the anti-pan-β polyclonal antibody, whereas the anti-Kv2.1 antibody did not immunoprecipitate detectable amounts of Kvβ1. That the same amount of rabbit IgG was precipitated in each of these reactions is demonstrated by the ∼50 kDa band present in each immunoprecipitation lane that is attributable to immunoreactivity of the rabbit IgG from the immunoprecipitating antibody, with the anti-rabbit secondary antibody used to develop the immunoblots. Surprisingly, levels of Kvβ1 comparable to those observed in the anti-Kvβ1 reaction were co-immunoprecipitated by the anti-Kvβ2 polyclonal antibody. Because these immunoprecipitation reactions were performed under conditions in which antigen is limiting, these data demonstrate that a large proportion of the Kvβ1 found in rat brain is present in complexes that also contain the Kvβ2 β subunit (Fig.5).
These same immunoprecipitation reactions then were assayed for the presence of Kvβ2; because these immunoblots were performed with the anti-Kvβ2 mouse mAb K17/70, the rabbit IgG bands typically seen in the immunoprecipitation reactions are not visualized. The rabbit anti-Kvβ2 polyclonal antibody was able to immunoprecipitate the 38 kDa Kvβ2 polypeptide present in these samples. Comparable amounts of Kvβ2 also were observed in the immunoprecipitation reaction performed with the anti-pan-β polyclonal antibody (Fig. 5). Low levels of Kvβ2 also were observed in the reactions performed with the anti-Kvβ1 antibody, confirming the observation above and indicating that individual K+ channel complexes may contain both of these β-subunit polypeptides. As expected, the anti-Kv2.1 antibody also was unable to co-immunoprecipitate detectable amounts of Kvβ2. All of the immunoprecipitation reactions performed with the anti-β-subunit antibodies contained Kv1.2 (Fig. 5), showing that each of these antibodies could effectively recognize and isolate intact K+ channel complexes.
Immunohistochemical localization of Kvβ1 and Kvβ2
Analysis of immunohistochemically stained sections indicated that Kvβ1 and Kvβ2 are present in cell bodies, dendrites, the juxtaparanodal regions of myelinated axons, and terminal fields of several major projection systems. The pattern of immunohistochemical staining observed with mAbs and polyclonal antibodies to the same β subunit was indistinguishable. The areal and laminar distribution of labeled cells suggests and comparisons of sections processed byin situ hybridization and immunohistochemistry suggest that cells expressing Kvβ1 and Kvβ2 mRNA also are immunoreactive for the corresponding proteins. Regions containing moderate to intense immunoreactivity for Kvβ1 and Kvβ2 were the neocortex, hippocampus, piriform cortex, striatum, thalamus, cerebellum, cranial nerve nuclei, and virtually all major white matter pathways. Although it is beyond the scope of this paper to provide a comprehensive account of β-subunit immunoreactivity in every brain region, the salient features of the immunohistochemical staining are described below and summarized in Table 1.
Neocortex and hippocampus
In the neocortex, immunoreactivity for Kvβ1 was concentrated within pyramidal cells and was particularly intense in smaller bipolar and multipolar interneurons in layers II, III, V, and VI (Fig.6 A). In each of these cell populations, immunoreactivity for Kvβ1 was concentrated within the cell body and proximal portions of the dendritic tree. Immunoreactivity for Kvβ2 also was observed in cortical pyramidal cells, with very intense labeling of large pyramidal cells in layer V (Fig. 6 C). However, in contrast to immunoreactivity for Kvβ1, immunoreactivity for Kvβ2 was concentrated throughout the entire apical dendrite of layer V pyramidal cells, including fine branches and apical tufts. Interestingly, the small interneurons that displayed strong immunoreactivity for Kvβ1 were not intensely immunoreactive for Kvβ2. In addition to the somatodendritic staining described above, in many cortical regions, moderate to intense immunoreactivity for Kvβ1 and Kvβ2 was observed in axonal profiles coursing throughout the cortical neuropil and in a pattern consistent with labeling of terminal fields. This pattern of labeling was particularly intense in the piriform, posterior cingulate, retrosplenial, and entorhinal cortices. Myelinated axons in the subcortical white matter as well as other regions also contained immunoreactivity for Kvβ1 and Kvβ2. The staining in these fibers was discontinuous, with a greater density of reaction product in the juxtaparanodal segments at nodes of Ranvier (Shi et al., 1996).
In the hippocampal formation, there was a heterogeneous distribution of Kvβ1 and Kvβ2 immunoreactivity with moderate to intense staining of granule and pyramidal cell somata (data not shown). Immunoreactivity for Kvβ1 and Kvβ2 also was concentrated within interneurons in stratum oriens and stratum radiatum of the CA1–CA3 subfields. In the dentate gyrus, granule cells contained a moderate density of Kvβ1 and Kvβ2 immunoreactivity. In these cells, staining was concentrated in the cell soma and did not extend into dendritic branches, suggesting that this staining is associated with β subunits synthesized within these cells. In the molecular layer of the dentate gyrus, there was a distinct band of Kvβ1 and Kvβ2 immunoreactivity in the middle third and somewhat less intense labeling in the outer third. As described previously (Rhodes et al., 1995), the location of this band of intense immunoreactivity corresponds closely to the termination zone of the medial perforant path. In the CA subfields and subiculum, immunoreactivity for Kvβ1 and Kvβ2 was distributed diffusely throughout the neuropil, with a greater density in stratum radiatum and stratum oriens than in stratum moleculare, and immunoreactivity for Kvβ2, but not Kvβ1, was concentrated throughout the apical dendrites and apical dendritic tufts of hippocampal pyramidal cells. In the CA3 subfield, there was a moderate density of Kvβ1 immunoreactivity in the mossy fiber zone.
Striatum and basal forebrain
In the caudate nucleus and nucleus accumbens, there was a moderate density of immunoreactivity for Kvβ1 and Kvβ2 within neurons of all sizes and morphologies and distributed diffusely throughout the neuropil. Interestingly, there was a higher density of immunoreactivity for Kvβ1 and Kvβ2 in bundles of axons coursing through the ventromedial portion of the caudate nucleus than in similar fiber bundles in the dorsolateral component of this structure. These fibers appeared to originate from cells within the caudate nucleus and could be followed in consecutive sections to their apparent termination throughout the supra- and subcommissural segments of the globus pallidus (Fig. 6 B,D). Additional fibers showing intense immunoreactivity for Kvβ1 and Kvβ2 appeared to continue beyond the globus pallidus and run through the medial forebrain bundle and cerebral peduncle to terminate within the substantia nigra pars reticulata (Fig. 7 A,C). Intense terminal-field labeling for Kvβ1 and Kvβ2 also was observed in the nucleus accumbens and olfactory tubercle, and in the deep half of the molecular layer of pyriform cortex.
In the medial septal and diagonal band nuclei, immunoreactivity for Kvβ1 and Kvβ1 was present in the somatodendritic compartments of magnocellular neurons. These neurons formed a continuous column of cells extending through the medial septum and vertical and horizontal limbs of the diagonal band and into the magnocellular neurons of the nucleus basalis of Meynert. This pattern of Kvβ1 and Kvβ2 immunoreactivity was virtually identical to the distribution of Kvβ1 and Kvβ2 mRNA and is very similar to the distribution and characteristic morphology of basal forebrain cholinergic neurons (Mesulam et al., 1983).
Many cells within various thalamic nuclei were immunoreactive for Kvβ1 and Kvβ2. The distribution of labeled cells corresponded closely to the pattern of expression Kvβ1 and Kvβ2 mRNA. In virtually all of these cells, a large proportion of the immunoreactivity was concentrated within the somatic and dendritic cytoplasm, suggesting that this staining is associated with the β-subunit polypeptides not yet transported to a distinct plasma membrane domain.
In the cerebellum, immunoreactivity for Kvβ1 and Kvβ2 was concentrated within the cell bodies and dendrites of Purkinje cells and was distributed diffusely and with moderate intensity throughout the molecular layer (Fig. 7 B,D). Immunoreactivity for Kvβ2 also was present in the axon terminals of basket cells. These basket-cell terminals form a characteristic plexus surrounding the initial segment of Purkinje cell axons and have been shown previously to contain intense immunoreactivity for Kv1.1 and Kv1.2 α subunits (McNamara et al., 1993; Wang et al., 1993, 1994; Rhodes et al., 1995;Shi et al., 1996). Numerous fine-caliber axons within the granule cell layer also contained immunoreactivity for Kvβ1 and Kvβ2, as did large neurons within the deep cerebellar nuclei. The axons in the granule cell layer appeared to originate from Purkinje cells and frequently appeared to emerge from the core of labeling of the basket-cell terminal plexus. Large-caliber myelinated axons in the cerebellar white matter also contained intense immunoreactivity for Kvβ1 and Kvβ2. As described above, this immunoreactivity was strikingly discontinuous, with intense staining in patches juxtaposed to nodes of Ranvier (Shi et al., 1996).
Midbrain and brainstem
Large motor and sensory neurons within structures such as the red nucleus and all cranial nerve sensory and motor nuclei contained moderate to intense immunoreactivity for Kvβ1 and Kvβ2 (Fig. 6 A,C). In addition, neurons within the substantia nigra pars reticulata and in the deep layers of the superior colliculus were immunoreactive for these two β-subunit polypeptides. Large neurons within the midbrain and pontine reticular formation and inferior olivary complex were moderately to intensely labeled by both anti-β-subunit antibodies, as were axons within midbrain and brainstem white matter pathways and cranial nerve efferents.
Molecular, immunological, and neuroanatomical techniques were used to examine the expression, distribution, and co-association of two voltage-gated K+ channel β subunits in adult rat brain. Although the relative density of Kvβ1 and Kvβ2 expression and immunoreactivity differs considerably across brain regions, the expression of these two subunits shows substantial overlap, and in many cell populations the respective mRNAs and polypeptides probably co-localize. The riboprobes used in the present study reacted with single bands by Northern and Southern analysis, and their specificity is supported by the observation that two independent riboprobes targeted to nonoverlapping sequences of the target mRNA gave identical hybridization patterns. Moreover, the pattern of immunoreactivity revealed by Kvβ1- and Kvβ2-specific antibodies is consistent with localization of the respective polypeptides to the cell bodies, dendrites, or terminals of cells expressing the corresponding mRNA. Antibodies raised against unique N-terminal sequences of Kvβ1 and Kvβ2 reacted with 44 kDa and 38 kDa polypeptides, respectively, in membranes prepared from rat brain and transfected COS cells; in brain, the 38 kDa Kvβ2 polypeptide is far more abundant. On the basis of reciprocal co-immunoprecipitation experiments, it appears that almost all channel complexes containing Kvβ1 also contain Kvβ2.
Using subtype-specific anti-β-subunit antibodies, we unambiguously identified the Kvβ1 and Kvβ2 polypeptides in rat brain. We based this identification on their reaction with antibodies made to unique regions of the deduced amino acid sequence and their co-migration on SDS gels with the recombinant protein expressed in transfected cells. Together, these results indicate that in rat brain, Kvβ1 is a relatively low-abundance polypeptide of M r = 44 kDa, whereas Kvβ2 is an abundant polypeptide ofM r = 38 kDa. Subsequent analyses of theM r = 44 kDa Kvβ1 andM r = 38 kDa Kvβ2 polypeptide pools in rat brain membranes using a polyclonal antibody raised against a C-terminal peptide conserved in both β subunits (“pan-β” antibody) (Rhodes et al., 1995) reveal a striking difference in the relative abundance of these two subunits. Quantitative analysis of immunoblots performed using the pan-β-subunit antibody, which recognizes both Kvβ1 and Kvβ2, shows that Kvβ2 is ∼50-fold more abundant than Kvβ1 in rat brain membranes, supporting the model we initially proposed (Rhodes et al., 1995) that Kvβ2 was by far the more abundant β subunit subtype in brain. In addition, most Kvβ1 appears to be present in complexes that also contain Kvβ2, as evidenced by the fact that Kvβ1-containing complexes are co-immunoprecipitated using a Kvβ2-specific antibody. A small portion of the total Kvβ2 pool in brain is present in these complexes, as evidenced by the small amount of Kvβ2 in immunoprecipitations performed using the anti-Kvβ1 antibody, thus most of the Kvβ2 in brain appears to exist in complexes lacking Kvβ1. Conversely, it appears that the small amount of Kvβ1 present in brain exists in complexes containing Kvβ2; very few “pure” Kvβ1 complexes are present.
Comparison of immunohistochemical staining patterns observed in the present study with those obtained using the pan-β antibody (Rhodes et al., 1995) suggests that the immunoreactivity for Kvβ2 recapitulates virtually the entire pattern and extent of immunoreactivity observed using the pan-β antibody. If we assume that pan-β staining represents the total β-subunit pool associated with Shakerα subunits, then we can assign Kvβ2 as the major component of this pool. In addition, we can assign Kvβ1 as a comparatively minor subcomponent, found predominantly in channel complexes that also contain Kvβ2 (Fig. 8). On the basis of this analysis, we also can speculate that in brain, the distribution of other β-subunit isoforms containing this conserved C-terminal epitope (England et al., 1995a,b; Heinemann et al., 1995; Majumder et al., 1995; Morales et al., 1995) will show considerable overlap with that observed using the pan-β and Kvβ2-specific antibodies, and that these other β subunits will be present in rat brain K+ channel complexes in association with Kvβ2.
The pattern of Kvβ1 expression reported here is virtually identical to that reported for Kvβ1 mRNA by Rettig et al. (1994). Their data and ours indicated that Kvβ1 is highly expressed in the cortical interneurons; basal ganglia; limbic cortical regions such as the hippocampus, posterior cingulate, and entorhinal cortex; and brainstem and cerebellum. In the majority of regions, immunoreactivity for Kvβ1 is concentrated at the terminals of cells expressing Kvβ1 mRNA. The very high levels of Kvβ1 expression in the caudate putamen corresponded well to the high density of Kvβ1 immunoreactivity in terminal fields in the globus pallidus and substantia nigra and, similarly, the high levels of Kvβ1 expression in entorhinal cortex correspond to a band of Kvβ1 immunoreactivity in the molecular layer of the dentate gyrus (data not shown). One clear exception to this relationship is that numerous cortical and hippocampal interneurons contain high levels of Kvβ1 mRNA and also contain a high density of Kvβ1 immunoreactivity at their somata and proximal dendrites.
Overall, the pattern of Kvβ1 expression corresponds closely to that observed for Kv1.4 (Sheng et al., 1992) and, to a lesser extent, Kv1.1 (Wang et al., 1994). In agreement, the high levels of Kvβ1 immunoreactivity in the globus pallidus and pars reticulata of the substantia nigra correspond closely to the patterns reported for Kv1.1 and Kv1.4 in these structures (Sheng et al., 1992; Wang et al., 1994). Similarly, the high levels of Kvβ1 immunoreactivity in cortical interneurons corresponds closely to the patterns of expression and immunoreactivity for Kv1.1, Kv1.2, and Kv1.6 α subunits in these cells (Wang et al., 1994) (M. Monaghan, N. Barrezueta, J. Trimmer, K. Rhodes, unpublished observations). This observation suggests that Kv1.1, Kv1.2, Kv1.6, Kvβ1, and perhaps Kvβ2 may co-localize and co-assemble in these cells. In view of the high spontaneous firing rates and short-duration action potentials reported for cortical (hippocampal) interneurons (Williams et al., 1994), it is tempting to speculate that the high levels of Kvβ1 expression and its subsequent effects on associated α subunits (Rettig et al., 1994) in these cells are related to an intrinsic requirement for rapidly inactivating channels.
Although Rettig et al. (1994) did not examine Kvβ2 expression in their study, the data reported here indicate that Kvβ2 also is widely expressed in rat brain. Cohen and his colleagues (Arai et al., 1992;Cohen et al., 1992) have characterized the expression of a gene, F5 (Sabath et al., 1990), that is highly expressed in activated T lymphocytes and brain. The recent submission of the F5 sequence to the Genbank database identified F5 as the mouse homolog of Kvβ2. The pattern of F5 mRNA expression in brain (Arai et al., 1992; Cohen et al., 1992) is qualitatively similar to that reported here for Kvβ2, but is described more accurately as a composite of Kvβ1 and Kvβ2 mRNA expression. This may be attributable to the probe that was used for hybridization, which was derived from a region, which in rat, is highly conserved between Kvβ1 and Kvβ2 (80% identity at the nucleotide level).
Immunoreactivity for Kvβ2 is concentrated in the juxtaparanodal regions of myelinated axons and in terminal fields of cells expressing Kvβ2 mRNA. This is clearly evident in the cerebellar cortex, where Kvβ2 mRNA is expressed in basket cells and Kvβ2 immunoreactivity concentrated in the basket-cell terminal plexus, and in the hippocampus, where Kvβ2 mRNA is expressed in layer II cells in entorhinal cortex and Kvβ2 immunoreactivity is present in the molecular layer of the dentate gyrus. However, there also are clear examples of Kvβ2 immunoreactivity in somatic and dendritic domains. For example, in cortical and hippocampal pyramidal cells, Kvβ2 immunoreactivity is concentrated in the cell body and throughout the apical dendritic tree, and in cerebellar Purkinje cells, Kvβ2 immunoreactivity is concentrated in somatic and dendritic domains.
In general, the pattern of expression and immunoreactivity for Kvβ2 is best approximated by a composite of that reported for the Kv1.1 and Kv1.2 α subunits (Sheng et al., 1994; Wang et al., 1994; Rhodes et al., 1995). In cerebellar basket-cell terminals, the pattern of Kvβ2 immunoreactivity corresponds precisely to that observed for Kv1.1 and Kv1.2 (McNamara et al., 1993; Wang et al., 1993, 1994; Sheng et al., 1994; Rhodes et al., 1995; Shi et al., 1996), indicating that Kvβ2 is likely to be associated with Kv1.1 and Kv1.2 in these complexes. Similarly, in the juxtaparanodal membrane of myelinated axons, the pattern of Kvβ2 immunoreactivity is identical to that observed for Kv1.1 and Kv1.2 (Wang et al., 1993, 1994; Sheng et al., 1994; Shi et al., 1996). In other regions, such as the somatodendritic membranes of cortical pyramidal cells and the cell bodies of midbrain and brainstem motor neurons, the pattern of Kvβ2 immunoreactivity also corresponds closely to Kv1.1 and Kv1.2. However, in other regions, such as the apical dendritic trees of cortical and hippocampal pyramidal cells, the pattern of Kvβ2 immunoreactivity is matched by that for Kv1.6 (K. Rhodes, M. Monaghan, J. Trimmer, unpublished observations) and Kv4.2 (Sheng et al., 1992), but not by any of the other Kv1 family α subunits. Because Kvβ2 (and Kvβ1) can associate with all Kv1 α subunits, as well as Kv4.2, in heterologous expression systems (Nakahira et al., 1996), in pyramidal cell dendrites, Kvβ2 may be present in channel complexes containing Kv1.6 or Kv4.2. However, this association remains to be confirmed by reciprocal co-immunoprecipitation studies.
The results of the present study indicate that β subunits are fundamental components of voltage-gated K+channels in mammalian brain. Their widespread expression and tight association with α subunits suggests that they play a fundamental role in the function of native channels (Parcej and Dolly, 1989, 1992;Trimmer, 1991; Scott et al., 1994a,b; Rhodes et al., 1995; Shi et al., 1996). As both Kvβ1 and Kvβ2 are found in both presynaptic (axons and terminals) and postsynaptic (soma and dendrites) plasma membrane domains, it may be assumed that they in themselves do not effectively target K+ channels to specific subcellular compartments (Sheng et al., 1992, 1994). Some β subunits, including Kvβ1, do exert dramatic effects on the electrophysiological properties of expressed channels (Rettig et al., 1994; England et al., 1995a,b; Majumder et al., 1995; Morales et al., 1995). However, Kvβ2, the major β subunit in brain, does not, raising questions as to its precise function. We reported recently that Kvβ2 exerts chaperone-like effects on the Kv1.2 α subunit, promoting its efficient glycosylation, surface expression, and stability in the plasma membrane (Shi et al., 1996). Similar effects are mediated by Kvβ1 (Shi et al., 1996). Interestingly, mutations in aDrosophila β subunit result in a reduction in the surface density of K+ channel complexes (Chouinard et al., 1995), supporting the contention that the effects of β subunits on the biosynthetic maturation and stability of α subunits is a fundamental role. Further definition of the sites where β subunits co-localize and co-associate with individual α subunits will further define the role of Kvβ1 and Kvβ2 in determining the properties of K+ channel complexes in mammalian brain.
This work was supported by Wyeth-Ayerst Research and the Center for Biotechnology at Stony Brook, and funded by the New York State Science and Technology Foundation and by National Institutes of Health Grant NS34383 (J.S.T.). This work was done during the tenure of an Established Investigatorship from the American Heart Association (J.S.T.). We thank Drs. James E. Barrett and John A. Moyer for critically reviewing this manuscript.
Correspondence should be addressed to Dr. James S. Trimmer, Department of Biochemistry and Cell Biology, State University of New York, Stony Brook, NY 11794-5215.